MAORI

This book provides the most comprehensive linguistic description of Maori currently
available. Today this East Polynesian language of the indigenous people of New Zealand
is under threat and it seems likely that the Maori language of the future will differ quite
considerably from the Maori language of the past. With few inflections, Maori is readily
approached by linguists. It has an elaborate system for the expression of possession, and
while superficially an accusative language, it shows various traces of ergativity. The area
of transitivity in particular provides a number of interesting theoretical problems and
constructions unusual outside the Polynesian family.
Winifred Bauer offers a wide-ranging and detailed description of the structure of the
language, covering syntax, morphology and phonology. There are two main sources of
data: written narrative texts and material from older native-speaking consultants. The
book also includes many comments on current changes in usage.
This descriptive grammar, which is illustrated with a wealth of examples, will be of
interest to both linguistic theoreticians and descriptive linguists, including language
typologists, and provides a resource for the study and resuscitation of the Maori
language.
Winifred Bauer has been involved in linguistic research into the Maori language for
twenty years, and is currently an Honorary Research Fellow in the Linguistics
Department of the Victoria University of Wellington, New Zealand.
 Descriptive Grammars

Series Editor
Bernard Comrie (University of Southern California)
ADVISORY BOARD
W.S.Allen, Cambridge University
J.T.Bendor Samuel, Summer Institute of Linguistics
D.C.Derbyshire, Summer Institute of Linguistics
R.M.W.Dixon, Australian National University
M.E.Krauss, University of Alaska
B.Krishnamurti, Osmania University
Y.Lastra, National Autonomous University of Mexico
S.A.Wurm, Australian National University

ABKHAZ
B.G.Hewitt

KOBON
J.Davies

MANGARAYI
F.Merlan

TAMIL
R.E.Asher

WEST GREENLANDIC
M.Fortescue

JAPANESE
J.Hinds

RUMANIAN
G.Mallison

MODERN GREEK
B.D.Joseph and I.Philippaki-Warburton

AMELE
J.Roberts
 BASQUE
M.Saltarelli

GULF ARABIC
Clive Holes

KANNADA
S.N.Sridhar

FINNISH
H.Sulkala and M.Karjalainen

CATALAN
José Ignacio Hualde

PUNJABI
Tej K.Bhatia
 Maori
Winifred Bauer
with William Parker
and Te Kareongawai Evans

London and New York

 First published 1993 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Reprinted 1996, 1999, 2001 (twice)
Simultaneously published in the USA and Canada by Routledge 270 Madison Ave, New York NY
10016

Routledge is an imprint of the Taylor & Francis Group

This edition published in the Taylor & Francis e-Library, 2005.
“To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of
thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.”
© 1993 Winifred Bauer
All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or
by any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission
in writing from the publishers.
British Library Cataloguing in Publication Data Bauer, Winifred Maori.—(Descriptive
Grammars Series) I. Title II. Series 499
Library of Congress Cataloging in Publication Data Bauer, Winifred Maori/Winifred Bauer. p.
cm.—(Descriptive grammars.) Includes bibliographical references and index. 1. Maori
language—Grammar. I. Title. II. Series. PL6465.B38 1993 499′.4–dc20 92–29383

ISBN 0-203-40372-X Master e-book ISBN

ISBN 0-203-71196-3 (Adobe e-Reader Format)

ISBN 0-415-02254-1 (Print Edition)
 EDITORIAL STATEMENT

Until quite recently, work on theoretical linguistics and work on language description
proceeded almost entirely in isolation from one another. Work on theoretical linguistics,
especially in syntax, concentrated primarily on English, and its results were felt to be
inapplicable to those interested in describing other languages. Work on describing
individual languages was almost deliberately isolationist, with the development of a
different framework and terminology for each language or language group, and no
feeding of the achievements of language description into linguistic theory. Within the last
few years, however, a major rapprochement has taken place between theoretical and
descriptive linguistics. In particular, the rise of language typology and the study of
language universals have produced a large number of theoreticians who require accurate,
well-formulated descriptive data from a wide range of languages, and have shown
descriptive linguists that they can both derive benefit from and contribute to the
development of linguistic theory. Even within generative syntax, long the bastion of
linguistic anglocentrism, there is an increased interest in the relation between syntactic
theory and a wide range of language types.
For a really fruitful interaction between theoretical and descriptive linguistics, it is
essential that descriptions of different languages should be comparable. The
Questionnaire of the present series (originally published as Lingua, vol. 42 (1977), no. 1)
provides a framework for the description of a language that is (a) sufficiently
comprehensive to cover the major structures of any language that are likely to be of
theoretical interest; (b) sufficiently explicit to make cross-language comparisons a
feasible undertaking (in particular, through the detailed numbering key); and (c)
sufficiently flexible to encompass the range of variety that is found in human language.
The volumes that were published in the predecessor to the present series, the Lingua
Descriptive Studies (now available from Routledge), succeeded in bridging the gap
between theory and description: authors include both theoreticians who are also
interested in description (e.g. Peter Cole and Imbabura Quechua) and field-workers with
an interest in theory (e.g. John Davies on Kobon).
The aim of the Descriptive Grammars is thus to provide descriptions of a wide range
of languages according to the format set out in the Questionnaire. Each language will be
covered in a single volume. The first priority of the series is grammars of languages for
which detailed descriptions are not at present available (e.g. Nkore-Kiga). However, the
series will also encompass descriptions of better-known languages with the series
framework providing more detailed descriptions of such languages than are currently
available (as with the monographs on West Greenlandic and Kannada).
Bernard Comrie
 CONTENTS

Acknowledgements ix
Introduction x
Map xv
Abbreviations xvi
Glossary xix

1. SYNTAX 1
2 MORPHOLOGY 259
3 PHONOLOGY 520
4 IDEOPHONES AND INTERJECTIONS 568
5 LEXICON 570

Textual sources 586

Bibliography 588
Index 592
 Acknowledgements

This book has been many years in preparation, and in the course of that time, many have
had an influence on its content. My greatest debt is to the Maori speakers who have given
so generously of their time and their knowledge to make this volume possible. Firstly, I
acknowledge my great debt to the late William Parker MBE, DLitt (Hon.), who
encouraged and supported me from the start of this project until his death. Bill had a
boundless curiosity about his first language, and enjoyed the challenges of the linguist,
taking great pride in finding answers which satisfied him. I hope he would have felt
proud to be associated with the finished product. Secondly, I wish to thank Te
Kareongawai Evans who undertook in 1988 the task of helping me to complete the book.
Kare has also given freely of her time and knowledge, and wrestled valiantly with the
unfamiliar problems posed by academic enquiry, though at times they seemed perverse
and unreasonable. This book is as much their achievement as mine. Between them Bill
and Kare have checked and/or provided all the non-textual examples, and many more
which corroborated the points made. I alone am responsible for such errors of fact or
interpretation as may remain.
I should also like to thank Peter McLean, my first teacher of Maori. He encouraged me
to seek answers to the questions I asked, and sent me to Bill Parker for help. Other
speakers of Maori have also contributed data or their judgements on various occasions
when extra help was sought. I should like to acknowledge the contributions from time to
time of the late Ruka Broughton and Mate Nathan, and of Rameka Cope, Shane Jones,
Joe Malcolm, and Huirangi Waikerepuru. I should also like to thank Dame Joan Metge
and Shane Jones for the extremely important part they played in introducing me to Kare
Evans, and to thank Shane and Ngareta Jones and Kare-Jo and Patrick Cadogan for
making their homes available for working sessions, and at times fitting their lives around
me. He mihi whaanui nooku teenei ki te hunga naana ahau i aawhina. Ahakoa kua huri
eetahi o raatou ki tua, kei te mau tonu oo raatou whakaaro.
I wish to thank the Stout Trust for the award in 1990 of the generously funded J.D.
Stout Research Fellowship, which shortened the gestation time of this book by several
years, and, by enabling me to work uninterrupted on the project, greatly enhanced its
consistency and cohesion. I would also like to thank Jim Collinge, Director of the Stout
Research Centre for providing an academic home for me for a further year, and for his
enthusiastic support during my two years at the Centre.
For their comments on various portions of the manuscript, I should like to thank
Laurie Bauer, Mary Boyce and Chris Lane, with a special vote of thanks to Bernard
Comrie, whose support I have greatly valued, and whose careful reading of the
manuscript has led to many improvements.
Last but not least, I should like to thank my family for their tolerance and
encouragement, especially Laurie who in 1990 took on an increased domestic load to
enable me to benefit from the Stout Fellowship.
 Introduction

Maori is the language of the indigenous people of New Zealand. It belongs to the Eastern
branch of the Polynesian language family. Its closest relation is the language of the Cook
Islands, sometimes called Cook Island Maori, and sometimes called Rarotongan. These
two languages are to a large degree mutually intelligible, given some good will on the
part of the speakers. Further details about the language family can be found in eg. Biggs
(1971).
Today, Maori is a threatened language. A brief history of the language since the first
European contacts follows, to show how this situation arose. In this account I am
indebted to Biggs (1968) and Benton (1981). Early contacts with Europeans were with
whalers and sealers. It appears that this contact took place in some kind of pidgin, and
had little effect on the Maori language. With the advent of the Missionaries in the early
1800’s, the Maori language was written down. An early orthography devised by Prof. Lee
at Cambridge University with two Maori chiefs and the missionary Kendall as
consultants, and later slightly modified, is still used today. The only problem with this
orthography is its failure to mark vowel length. This is still an issue today, and is
discussed later in this introduction. Smyth (1946) contains a detailed account of the
development of the orthography during the early nineteenth century, and McKenzie
(1985) also covers some of the early issues relating to the written language.
The Missionaries produced a Maori translation of the Bible in the early nineteenth
century, and established schools for the “natives”. These were of necessity Maori
medium schools to begin with, and through them, many Maori became literate in their
own language. However, in 1867, it was decreed that schools had to teach in English
(Biggs 1968, 74). For the next hundred years or so, children were punished for using
Maori at school. Maori parents were exhorted to speak English to their children at home
for the sake of their children’s future. During the early twentieth century, those same
policies were pursued by the authorities. Thus by the mid twentieth century, the younger
generation of Maoris were not Maori speakers. Rather too late, Maori was allowed a
place in the curricula of secondary schools and Universities. A survey of Maori speakers
carried out by the NZCER (Council for Educational Research) and completed in 1978
estimated that there were only about 70,000 speakers of Maori in New Zealand, and that
most of those were adults, and many were over the age of 45. There were still a few
isolated rural communities where Maori was the community language, and where young
Maori speakers were still found, but basically, Maori was the language of the elders, used
chiefly on ritual occasions, and on maraes and in the Church. The arrival of radios and
TVs in the homes in isolated rural communities allowed the intrusion of English even in
remote areas. To the most optimistic in the 1960’s, the Maori language seemed doomed
to extinction. However, the 1970’s saw the beginnings of a renaissance among Maori
people which gained impetus in the 1980’s . In a movement led by younger intellectuals,
there has been an attempt to restore Maori pride, Maori values, Maori customs and, not
least, the Maori language. As far as the language is concerned, one of the main thrusts
has been the establishment of the Kohanga Reo, or language nest’, a pre-school
movement using the Maori language as its medium. Run almost exclusively by women, it
has provided a Maori speaking environment not only for children, but also for many
women whose Maori language skills were not great. There were also attempts in
individual schools to introduce bilingual classes in at least some subjects, and Maori
language enrichment programmes. This has been followed by a few Maori medium
primary schools, and attempts to introduce more language programmes into secondary
schools and to increase the bilingual programmes available in at least some schools.
Maori language radio stations have been established, and there are token Maori language
programmes on national television. A Maori Language Commission has been established
with the brief to promote and support the Maori language. For the activists, there is far
too little Government funding for these activities, and the laws relating to the status and
use of the language have little bite; much of the funding has come from the Maori
community itself. It is not yet clear what the future holds for the Maori language. Benton
comments (1981, 43):

The huge rise in the number of people acquainted with Maori has not,
therefore, meant that the rapid decline in the number of native-speakers of
the language has been in any way compensated for

and sums up (1981, 44):

For Maori language teaching in general, it may safely be said that looking
forward from 1960, the progress made in the two following decades has
been so great as to be almost incredible even to the most optimistic
enthusiast. From the standpoint of 1980, however, it is equally obvious
that the task ahead is still of Herculean proportions. The future of Maori
as a living language is far from assured, although there is now no doubt
that it would be greatly honoured as a dead one.

This evaluation still seems to me valid in 1990. That, in itself, is not good news for the
language.
This has led to a linguistic situation of some complexity. The oldest Maoris (those
over the age of about 60) in most cases probably spoke nothing but Maori before they
went to school. However, since their schooling started, the vast majority have used
English more than Maori in their public lives, and in many cases also at home. In general,
only those in rural areas are likely to have continued to speak Maori as their first
language. All are speakers of English of one variety or another, although fluency and
confidence in English varies considerably. Many now confess to feeling more at home
speaking English than Maori. There are a few younger Maoris who, either because they
grew up in an isolated area, or because they were reared by grandparents (or both) are
also fluent speakers of Maori, but they are the exception rather than the rule. Most of the
“middle” generation are at best “semi-speakers” (see eg. Dorian 1989) and at worst know
no more than a few ritual greetings and isolated words. Many of these people feel cut off
from their culture by their lack of linguistic competence. Some of the younger members
of this group have had the opportunity to study Maori in an educational institution, but
the standard they achieve leaves them struggling as second language learners. The
youngest generation are more likely to have had some exposure to Maori at an early age,
through a Kohanga Reo, but are still unlikely to have the opportunity to continue the
exposure to Maori at school. Some children known to me, who are now approaching their
teenage years, have given up speaking Maori, even although they were able to at the age
of 5; their elders report that they respond in English even if spoken to in Maori, due no
doubt to both peer pressure and lack of follow-up Maori language programmes at school.
(Benton (1981, 23) also mentions the peer pressure towards English.) There are still
relatively few homes in which Maori is the normal language of communication, even
those where both parents have the capacity to speak Maori. Similarly, it is not uncommon
to hear elders leave a meeting house where they have been speaking Maori, and to start to
use English immediately, even to each other. One of the reasons such speakers advance
for this is the difficulty of speaking in Maori about many of the activities of today’s
world, through lack of vocabulary. (When too many English words intrude, code-
switching occurs.) This is an area where the Maori Language Commission is
endeavouring to assist, by an extensive programme of word manufacture. However, there
are still the familiar problems of spreading such new vocabulary items through the
community and having them accepted.
It is not surprising that there is a considerable gap between the Maori spoken by older
members of the Maori community and the younger generation. Most of those who speak
Maori and who are under the age of 45 have learned Maori as a second language. Older
speakers are often very critical of the Maori spoken by these young people. The fact that
they are often publicly critical does nothing to boost the confidence of these semi-
speakers, nor to encourage them to continue to try to speak Maori. This in turn becomes
an important problem for the resuscitation of the language. The Maori of these younger
speakers is in general characterized by a very small vocabulary, and the use of large
amounts of English vocabulary, often not even phonetically adapted (note that phonetic
adaptation is normally called ‘transliteration’ in New Zealand): at times it amounts to a
very narrow range of Maori structures, with Maori function particles but English
‘content’ words. The syntax often reflects English patterns (even in gross ways, such as
the lack of a preposition marking the Direct Object, and the stranding of prepositions). In
some instances, the changes can be attributed directly to English, although in others, the
Maori system appears to have been ripe for change. Even in the speech of the more
competent of these semi-speakers, the intonation patterns are frequently audibly English.
Faced with this situation, it was necessary to make a decision about the variety of
Maori to describe in this volume. This book concentrates on the variety of Maori that is
spoken by the older generation, which shows the minimum of English interference. My
consultants have come from that older age group, and are amongst those who spoke
nothing but Maori before the age of seven. However, in those places where it seems to
me that the innovations of the younger speakers are likely to be permanent, I have also
indicated the more recent constructions. Some of these receive more critical attention
from the elders than others, and where I am aware of the attitudes of older speakers, I
have indicated this.
This situation is complicated by the fact that Maori was not homogeneous when the
Europeans arrived: each tribe spoke its own dialect. Maori dialects have not been studied
rigorously, and most of the information is sporadic. However, most commentators are
agreed on the fact that the dialects differ most at the lexical and phonetic levels (see eg.
Harlow, 1979). However, there are no thorough studies of either. A recent survey by
Biggs (Biggs, 1989) is indicative of our lack of knowledge. The lack of standardization
is, I believe, another stumbling block to the revival of the language, while at the same
time, the identification with a particular tribe, and thus with a particular dialect is one of
the strongest motivators in the revival. The differences between dialects have posed
problems in educational institutions ever since Maori was introduced into the curriculum,
with materials written in one dialect being disliked in other areas, with students and
teachers complaining about examiners from one dialect area failing to accept the idioms
of others, etc. The dialects have always been mutually intelligible (apart from occasional
lexical items) to full speakers of Maori. However, dialect differences cause problems for
semi-speakers, and even when differences are due to dialect, they are often attributed to
incompetence nowadays: a speaker regarded as fairly competent by other speakers of that
dialect may be regarded as rather incompetent by speakers from other areas. The lack of a
standard of course raises problems for the linguist: what dialect should be described,
given that it is not possible to make an exhaustive study of them all? I have been forced
by circumstances to answer this pragmatically: I have described the dialect of those
consultants available to me. Unfortunately, I was forced by the death of my first
consultant to change not only consultant but also dialect part way through this volume.
To those familiar with Maori dialects, that shift will be immediately apparent. However, I
do not believe that it is as damaging as might be thought, simply because the chief
differences between dialects are not syntactic. In addition, my consultants have often
been able to indicate places where they are aware of dialectal differences, and I have had
sporadic access to material from other areas. The textual examples also serve to give
some coverage of dialects other than those of my chief consultants. With the section on
the lexicon, while I have given those dialectal variants I am aware of, there are
undoubtedly many places where only one dialect is represented.
The orthography developed by the early missionaries is still used today. Maori has a
small phonemic inventory, and there were only two sounds which did not have a suitable
single symbol in the roman alphabet. These are /ŋ/, which was written with the digraph
ng, and a phoneme whose phonetic realization was almost certainly subject to some
variation even at that time, written wh. This now has various realizations, but was
possibly [M] ~ [hw] at that time. The major problem linguistically with the orthography is
that it failed to mark vowel length, which is phonemic in Maori on one analysis. Later, a
macron was used to mark vowel length, but it was used only sporadically, and not, for
instance, on maps, road signs, etc. Williams’s Dictionary from the 3rd edition onwards
uses the macron on head words only. The text-books used most widely for the teaching of
Maori all use the macron for vowel length. This is despite a long campaign by Bruce
Biggs and colleagues at the University of Auckland to write long vowels as double
vowels. There are advantages both linguistically and pedagogically to doing so, but the
proposal met with no support from the Maori community at large: they feel that their
language looks “clumsy”, “silly”, etc written with double vowels, and prefer the macron.
Many still see no need at all to mark vowel length. The Maori Language Commission has
lent its support to the macron. Despite this, the present work uses double vowels. This
was largely a matter of typographical convenience: fonts with a macron are restricted in
availability, although very recently, moves have been made to rectify this, as more
material is published in Maori. However, at the time the early sections of this work were
drafted, these fonts were not available, and the narrow line spacing required by the
original typographical specifications for the series seemed a good reason to avoid
cluttering up the space with accents. If I were to begin again now, I would probably make
the opposite decision. I have altered the orthography of all my sources to conform to the
double vowel orthography so that readers of the present work will not have to shift from
the one orthography to the other. In many cases, the original did not mark vowel length at
all and in places I may have failed through ignorance to mark length where it occurs,
especially in proper names, which are not always readily checked. There are certain
forms, chiefly grammatical particles, which have two allomorphs, one with a long vowel
and one with a short vowel depending on the phonological environment in which they
occur. In accordance with the practice of Biggs and other users of double vowel
orthography, these forms are written consistently with a single vowel, since the variation
is predictable. Such forms are discussed in the section on syllabics.
There were a number of difficult decisions to be made with respect to the interlinear
glosses. In particular the glossing of possessive forms posed problems, since the Maori
possessive system is rich in distinctions not made in English. The inclusion of all the
information encoded in the Maori in the glosses makes them rather lengthy, and not
necessarily transparent to the reader. Following the suggestion in Language, 1984, 692–
3, the full glosses have been used only in those sections directly concerned with
possession; elsewhere, more transparent glosses are used. In general, the glosses for
particles are under-specified, and the sections discussing these particles need to be
consulted for more detailed information. A further problem arises with respect to Maori
words which are part of NZ English, but not necessarily known to speakers of other
Englishes, a set of words increasing rapidly at the present. In many cases, these cannot be
readily rendered in English. A glossary has been provided of such forms. That glossary
also includes a few English words which have special meanings in the New Zealand
context.
The locations and tribes referred to in the text are marked on the map which follows.
 Abbreviations
Isg/pl first person singular/plural
IIsg/pl second person singular/plural
IIIsg/pl third person singular/plural
Idlincl/excl first person dual inclusive/exclusive
Iplincl/excl first person plural inclusive/exclusive
A A-class possessive
A-E actor-emphatic
actgen actual genitive
addr addressee
adj adjective
adv adverb
ag agentive prefix
C consonant
Canon canonical transitive verb
cls classifying particle
compar comparative
contr contrary to expectations
det aph anaphoric determiner
dist distant from speaker and hearer
dl dual
DO direct object
dup reduplicated segment(s)
eq predicative equative particle
eql relational equative particle
excl exclusive
exp experience verb
fut future tense/time
gen genitive
genIIsg (etc) second person singular (etc) genitive
gp prefix indicating ‘in groups of’
habit habitual marker
id identical (in sense)
imp imperative marker
 incl inclusive
indef indefinite
instr instrumental marker
intens intensifier
intgen intended genitive
IO indirect object
1k linking element
M mood marker
mod modifier
mon monitory mood marker
N noun
neg negator
nom nominalizing suffix
NP noun phrase
NPA agent NP in actor-emphatic
NPP patient NP in actor-emphatic
num numeral particle
num-Q question word for numbers/quantities
O O-class possessive
Obl oblique noun phrase
oblig mood marker for obligation etc
ord ordinal numeral marker
part. particle
pass passive
pers personal article
pl plural
pnum numeral marker for persons
PP prepositional phrase
PPN Proto-Polynesian
pred predicate
prep preposition
pres present tense/time
proxI near speaker
proxII near hearer
pt past tense/time
Q question particle
 S sentence
sg singular
sp specific
spec specifying preposition
Su subject
subj subjunctive
T/A tense/aspect marker
T/A/M tense/aspect/mood marker
top topicalizing particle
V verb; vowel
VP verb phrase
voc vocative marker
↑ raised pitch
extra high pitch
|
stress
| secondary stress
" emphatic stress
≡ is equivalent to
* ungrammatical structure
 Glossary

The following glossary contains brief definitions for those Maori words used in the
translations of Maori sentences because they are untranslatable, and because they are part
of NZ English, although they may not be known to speakers of other varieties of English.
It also contains definitions of some words of NZ English which may not be known to
speakers of other varieties. In most cases the definitions are taken from Orsman, H.W.
(ed) (1989). The Latin names are from Williams’s Dictionary of the Maori Language. In
a few cases, the definitions are my own.

action song: in the Maori context, a song accompanied by stylized actions of hands and
body
bush: native forest, frequently with thick undergrowth
haka: a dance accompanied by a chant, often performed as a challenge to another party
hangi: an earth oven; its contents; a communal meal cooked in an earth oven
hapu: a sub-tribe
hot-spring: a spring of naturally hot water
hot pools: pools of naturally hot water; pools thermally heated iwi: a tribe
kahawai: a common sport fish, arripis trutta
kahikatea: (= white pine) a tall timber tree, podocarpus excelsum
kakapo: a rare dark green native flightless bird, strigops habroptilus
karakia: an incantation
kiekie: a native climbing plant, freycinetia banksii
kit: a basket or bag woven from flax
kiwi: a flightless native bird (several species), apteryx
kohanga reo: lit. ‘language nest’, a type of pre-school where emphasis is placed on
Maori language acquisition
kowhai: a native tree bearing yellow flowers in spring, sophora tetraptera
kumara: the sweet potato as it is known in Polynesia, ipomoea batatas
lolly: a sweet, candy
manuka: a common N.Z. scrub bush or tree, leptospermum
marae: the courtyard of a Maori meeting house, the centre of tribal life
mere(mere): a short flat bone, stone or greenstone hand-club
moki: N.Z. sea fishes of two families, latridopsis ciliaris
pa: a tribal settlement, usually fortified; a Maori village
picture(s): a film; a commercial cinema showing
poi: a light ball on a string swung to a rhythmic beat in Maori songs and dances
pohutukawa: a hardy coastal tree bearing red flowers at Christmas, metrosideros excelsa
puha: a sow thistle boiled as a vegetable, sonchus oleraceus
pukeko: a brightly coloured native swamp hen, porphyrio melanotus
puriri: a large native hardwood tree, the fruit forms in large clusters of red berries, vitex
lucens
raupo: the bulrush, typha angustifolia
rimu: (= red pine) an important native timber tree, dacrydium cupressinum
stick games: games played by throwing short round sticks in patterns between members
of a group, usually to a song accompaniment, with rhythmic percussive use of the
sticks
taiaha: a long wooden weapon, pointed at one end, used for striking and jabbing
tangi: a mourning ceremony, lament
taniwha: a legendary monster, usually dwelling in water tapu: sacred, reserved;
ceremonial restriction
tarakihi: a fine eating fish common in N.Z. waters, dactylopagrus macropterus
titoki: a native tree, alectryon excelsum
tohunga: a wizard, priest or gifted or learned person, a teacher
tuatara: a native reptile, sphenodon punctatus
whanau: an extended family
yellowhead: a native bird, mohona ochrocephala
 1
Syntax

1.1 GENERAL

1.1.1 Sentence types

1.1.1.1 Direct speech and indirect speech

The preferred method of reporting speech is direct quotation. The commonest
introductory verb is mea ‘say’, though others such as karanga ‘call’ can also occur, and
more elaborate formulae are found in some styles, eg. Ko te kupu a Tutaanekai ki a
Hinemoa ‘Tutanekai’s words to Hinemoa were…’. The introductory phrase, of the
general form ‘X said/spoke’ precedes the quoted words. The orthographic convention
varies: sometimes the introductory phrase is treated as an independent sentence, and
sometimes not. In speech, it is always treated as a separate sentence, receiving sentence
final intonation on the final phrase (see 3.3.4.1–2). The following extract from the tale of
Hinemoa will illustrate these points.
(1) Ka karanga atu a Hinemoa, ki taua
T/A call away pers Hinemoa to det aph
taurekareka nei, anoo he reo tane. Moo wai
slave proxI as a voice man intgen who
too wai? Ka mea mai te taurekareka raa.
sggenIIsg water T/A say hither the slave dist
Moo Tutaanekai. Naa, ka mea atu a
intgen Tutanekai then T/A say away pers
Hinemoa, Homai ki ahau.
Hinemoa give to Isg
‘Hinemoa called out to this slave in a man’s voice. “Who is your water for?” That slave
responded/spoke back. “For Tutanekai.” Then Hinemoa spoke. “Give it to me.”’(H, 8)

Indirect speech also occurs. There are many introductory verbs, eg. ui ‘ask’, paatai ‘ask’,
whakahoki ‘reply’, kii ‘say’, mea ‘say’, whakautu ‘respond’. In indirect speech, the
pronouns are changed to accord with the introductory clause (which is always part of the
same sentence), but tenses are not changed. The following examples illustrate. The
corresponding direct speech is given as (b) for comparison.
(2a) Ka whakahoki a Tamahae i te hii ika a
T/A reply pers Tamahae T/A catch fish pers
ia
IIIsg
‘Tamahae replied that he had been fishing’
 Maori 2

(2b) I te hii ika au

T/A catch fish Isg
‘I’ve been fishing’
(3a) Ka paatai a Hata kei whea ngaa ika
T/A ask pers Hata at(pres) where the(pl) fish
‘Hata asked where the fish were’.
(3b) Kei whea ngaa ika?
at(pres) where the(pl) fish
‘Where are the fish?’

The word order normal for Maori (basically VSO) is used in both the introductory and
the quoted sentences, whether the quote is direct or indirect.

1.1.1.2 Interrogative sentences

There are three basic types of question that can be distinguished in Maori, yes-no
questions (1.1.1.2.1), question-word questions (1.1.1.2.2), and echo-questions (1.1.1.2.3).
The first and last of these are distinguished from declaratives principally by intonation.
However, intonation is subject to considerable regional variation. It also appears to vary
with sex. The intonation patterns outlined here are those of a male Ngati Porou speaker
(ie. East Coast, North Island).

1.1.1.2.1 Yes-no questions

The norm for questions of this kind for older speakers is raised pitch throughout.
However, many younger speakers fail to produce raised pitch (indeed some report that
they find it embarrassing to do so), and they tend to rely on other means to mark these
forms as questions. Yes-no questions do not differ from declaratives in word order.

1.1.1.2.1.1 Neutral yes-no questions

Questions of this type are marked by raised pitch throughout the utterance (indicated by
the vertical arrow preceding), and a high rise on the final phrase. Word order is identical
to that of declaratives, eg.
 Syntax 3

To underline the fact that these are questions, it is possible to add raanei ‘or’ following
the verb in (4) and (5) (ie. after ana and haere respectively), and following anoo in (6)
which is non-verbal. This use of raanei is almost certainly increasing, since younger
speakers rely on raanei and final rising pitch to mark such forms as questions.
The intonation of these questions described above presumes unmarked focus. If some
marked constituent is in focus (eg. Hata in (4)), that constituent receives a fall-rise, but
the same general outline is followed for the intonation of the question as a whole.
However, for constituents other than the subject in focus, it is more likely that a different
question will be asked, eg. to focus on the location:
(4a) Hei/Ko te taaone a Hata i te poo nei?
at(fut) the town pers Hata at the night proxI
‘Will Hata be in town tonight?’

If the time adverbial in (4) is in focus, it is likely to be fronted:

(4b) A te poo nei, ka haere a Hata ki te
at(fut) the night proxI T/A move pers Hata to the
taaone?
town
‘Is Hata going to town tonight?’

1.1.1.2.1.2 Leading questions

These have the particle nee sentence finally, in a separate tone group. Otherwise, they
have declarative word order and intonation contours.

1.1.1.2.1.2.1 Expecting the answer ‘yes’

 Maori 4

In these questions, nee receives falling intonation. Raised pitch throughout often occurs,
but is not always used, even by older speakers.

(8) Ko tana whare te mea whero, nee? eq sggenIIIsg house the thing red Q
‘His house is the red one, isn’t it?’

1.1.1.2.1.2.2 Expecting the answer ‘no’

In such questions nee usually receives falling intonation. However, a rise on nee
sometimes occurs, perhaps when the possibility of contradiction is allowed for. While
falling intonation is probably normal for the phrase preceding nee, various modifications
occur, particularly if the speaker is pleading. The commonest modification appears to be
a fall-rise. Whether this is just a sandhi effect I am unsure. The fall on nee often appears
to start higher when the answer ‘no’ rather than ‘yes’ is expected.

(10) Kaaore i maka·ia atu e koe te reta, nee?

not T/A throw·pass. away by IIsg the letter Q
‘You didn’t throw away the letter, did you?’

1.1.1.2.1.3 Alternative questions

These consist of a positive statement first, with rising intonation on the final phrase,
followed by a phrase of the general form ‘negative word+raanei’, with falling intonation.
The negator used is that appropriate to the full negation of the declarative (see 1.4).

(12) He tangata kai paipa koe, eehara raanei?

cls man eat pipe IIsg neg or
‘Are you a smoker, or not?’
 Syntax 5

1.1.1.2.2 Question-word questions

These are marked by a rise on the question word, and a fall sentence-finally. However, if
the question-word falls in the final phrase, either falling or rising intonation may occur
sentence finally. I have been unable to discover any distinction dependent on this choice.
Many such questions, but not all, also involve changes in sentence structure. These are
dependent on the grammatical function of the constituent in which the question-word
occurs.

1.1.1.2.2.1 Questioning sentence elements

1.1.1.2.2.1.1 Main clause constituents that can be questioned

Any constituent of the main clause can be questioned, but the construction used varies
according to the grammatical relation of the questioned element, the type of sentence, and
in some instances, the tense of the sentence. The predominant construction involves the
insertion of the question word in the normal position of the questioned element.

The question-words are as follows:

wai is used for persons, and also in many cases for names
aha is used for non-persons and actions
(w)hea is used for locations (temporal and spatial)
hia is used for numbers and mass nouns (prefixed by toko for persons,
and otherwise preceded by e)
pee(w)hea is used for ‘like what’ and some other rather diverse
adverbial functions
(t)ee(w)hea is used for a choice from a limited set of objects, the t-form
being the singular (see 2.1.1.8.3, 2.1.2.6.1.2).

1.1.1.2.2.1.1.1 Question-word questions in non-verbal sentences

In equative sentences, the ko-marked predicate may be questioned, eg.

It appears that the subject in such sentences cannot be questioned (but see Bauer, 1991
for some discussion of the difficulties of distinguishing subject from predicate in such
sentences):
(14) *Ko wai ko Hata?
top. who eq Hata
‘Who is Hata?’
 Maori 6

(15) *Ko Hata a wai?

eq Hata pers who
‘Who is Hata?’

Note, however, the following, which appears to be two successive questions:

In classifying sentences, the he-marked predicate may be questioned, eg.

The subject in such sentences cannot be questioned.

(19) *Ko te aha he whero?
top. the what cls red
‘What is red?’

The sentence must be restructured so that the questioned element is a predicate:

In prepositional sentences, the predicate may be questioned, eg.

 Syntax 7

The subject may also be questioned, using subject fronting, eg.

If the subject is questioned with (t)ee(w)hea or wai subject fronting with ko is required:
(23) Ko wai kei roto i te kaapata raa?
top. who at(pres) inside at the cupboard dist
‘Who is in that cupboard?’
(24) Ko eewhea kei roto i te kaapata raa
top. which(pl) at(pres) inside at the cupboard dist
‘Which ones are in that cupboard?’

Questions involving numbers take the following form:

(26) E hia ngaa waka i runga i te roto

num how many the(pl) canoe at top at the lake
inanahi?
yesterday
‘How many canoes were on the lake yesterday?’

For questions involving numbers of people, hia is prefixed by toko-, which appears to be
akin to a classifier for persons:

However, it is not at all uncommon to hear younger speakers use e hia for questions
involving people.
E hia is also used for mass nouns, eg.
 Maori 8

A question of this kind can be answered by expressions like Ka nui ‘A lot’ (more lit. ‘It is
big’), or Torutoru noa ‘Not much’.

1.1.1.2.2.1.1.2 Question-word questions in verbal sentences

1.1.1.2.2.1.1.2.1 Questioning subjects

Subjects of verbal sentences may be questioned in general using ko-fronting if definite, or
indefinite subject fronting. This is true of subjects of intransitives (29), neuter verbs (30),
and experience verbs (31) (for neuter verbs see 2.1.3.1.5, and for experience verbs see
1.2.1.2.2):

(30) He aha kua mahue i te tamaiti?

a what T/A leave behind cause the child
‘What has the child left behind?’
(31) Ko wai i piirangi ki te kaakahu?
top. who T/A want to the garment
‘Who wanted the dress?’

With canonical transitive verbs, the actor-emphatic construction (see 1.11.2.1.7) is used
to front the subject if the sentence is past (32) or future (33), but if it is present, there is
no corresponding actoremphatic, and ko-fronting is used (34):

(33) Maa wai e whaangai ngaa kaawhe?

intgen who T/A feed the(pl) calf
‘Who will feed the calves?’
(34) Ko wai kei te whaangai i ngaa kurii?
top. who T/A feed DO the(pl) dog
‘Who is feeding the dogs?’

Questions corresponding to (32) and (33) using ko-fronting are syntactically possible, but
are judged very odd.
Questioning the subject without fronting produces echo questions.
 Syntax 9

1.1.1.2.2.1.1.2.2 Questioning DOs

DOs can be questioned, but there is some variety of methods used. The construction
preferred by older speakers involves a possessive relative clause (see 1.1.2.3.4.2.3),
which functions as the subject in a non-verbal sentence. The questioned element appears
as a fronted subject. Some examples should help to clarify the construction. A
corresponding declarative is provided as (b) to facilitate comparison.

(35b) Kei te horoi teeraa wahine i te kete

T/A clean that woman DO the kit
‘That woman is cleaning the kit’

(English cannot readily mark that a plural is implied, as the Maori does.)
(36b) I aawhina a Hata i ngaa tamariki iti
T/A help pers Hata DO the(pl) children small
‘Hata helped the small children’

It is also possible, but on at least some occasions deemed less natural, to use a question-
word in the DO slot in a declarative. (This does not automatically produce an echo-
question in Maori, as it does in English.) Thus, alongside (35a), it is possible to ask

In the case of an example like (36a), this construction is preferred by some speakers, who
find the plural form above awkward. They prefer:
 Maori 10

Notice that, as the examples show, the possessive-relative construction is normal whether
the DO questioned is animate or inanimate.
However, with the second argument of experience verbs (see 1.2.1.2.2), somewhat
different preferences are found. With nonpersonal arguments, the substitution pattern was
often judged equally good, so that (37c) is as likely as (37a):

(37b) I piirangi a Hata ki te mere

T/A want pers Hata to the mere
‘Hata wanted the mere’

However, when arguments with personal referents are questioned the substitution pattern
is more likely to be elicited, and is sometimes judged considerably better than the
possessive-relative pattern. Thus (38c) is commonly preferred to (38a):

(38b) I kite a Rewi i a Inia te Wiata

T/A see pers Rewi DO pers Inia te Wiata
‘Rewi saw Inia te Wiata’

The possessive-relative construction appears to be limited to DOs and verbs, see

1.1.1.2.2.1.1.2.4. It also appears that younger speakers, in general, do not use this pattern
(and in some cases do not even recognise it or understand it), but choose the substitution
pattern as the norm for questioning DOs, or opt to promote the DO to subject via the
passive, and to question it using the possibilities for questioning subjects. Note that here,
 Syntax 11

as in several other places in the syntax of Maori, the DO is not as readily subjected to the
syntactic process as oblique NPs, for instance. One may speculate whether this happens
because the DO is a relatively new syntactic relation in a system formerly ergative. In
many of the places where this phenomenon occurs, older speakers use the possessive-
relative construction, but younger speakers often extend the oblique pattern to DOs.

1.1.1.2.2.1.1.2.3 Questioning IOs and other oblique NPs

The substitution of a question-word in the appropriate slot is used in almost all instances,
with one notable exception, namely time questions. The following examples illustrate the
predominant pattern:

(40) I haere atu a Rewi i Rotorua ki hea?

T/A move away pers Rewi from Rotorua to where
‘Where did Rewi go to from Rotorua?’
(41) I waiata a Hata moo wai?
T/A sing pers Hata intgen who
‘Who did Hata sing about?’

However, it appears that the passive agent is not usually questioned, and the normal
interpretation of such questions is that they are echoquestions, eg.

With the cause NP in sentences containing neuter verbs or stative intransitives, there are
two possibilities: besides using the above-mentioned substitution pattern, the i-marked
NP can also be questioned in past and future tenses using what appears to be the
actoremphatic construction, but with ai or nei inserted post-verbally according to tense.
Even more frequently, however, a related transitive is used. Thus relating to a declarative
like
(43a) I riri a Hata i a Rewi
T/A angry pers Hata cause pers Rewi
‘Hata was angry because of Rewi’

there are three possible questions: with substitution (43b), with the actor-emphatic (43c),
or with a related transitive verb (43d):
 Maori 12

A parallel set of examples with the neuter verb pau is:

(44a) Kua pau ngaa kai i ngaa kurii
T/A exhausted the(pl) food cause the(pl) dog
‘The dogs have eaten up all the food’
(44b) Kua pau ngaa kai i te aha?
T/A exhausted the(pl) food cause the what
‘What ate up the food?’
(44c) Naa te aha i pau ai ngaa kai?
actgen the what T/A exhausted part. the(pl) food
‘What ate up the food?’
(44d) Naa te aha i whaka·pau ngaa kai?
actgen the what T/A cause·exhausted the(pl) food
‘What ate up the food?’

Time questions require the fronting of the question phrase, accompanied normally by the
fronting of the subject, and in most cases, ai must be inserted post-verbally. While past
time phrases are most commonly introduced by the preposition i ‘at(pt)’, and under
special circumstances by noo ‘actgen’, past time questions are always introduced by noo.
Future time questions are introduced by a ‘at(fut)’:

(46) A whea a Hata haere mai ai?

at(fut) when pers Hata move hither part.
‘When will Hata come?’

As an alternative, in (46), haere mai ai may be replaced by ka haere mai ‘T/A move
hither’. In this case, ai is not used. Note also the absence of a tense/aspect marker in (46).
It is not clear whether this is an ellipsis but it may well arise from the general demise of e
as a tense marker in Maori. The possibility that such sentences involve a more radical
departure from the syntax of the corresponding declarative must be at least mentioned. It
 Syntax 13

would be possible to analyse them as involving a non-verbal predication of the

prepositional type (see 1.2.1.1.3.2) with a sentential subject, from which raising of the
subject occurs. Such an analysis is difficult to motivate conclusively, but at the same
time, it is difficult to reject, as other similar structures must be recognized for Maori.
Some ‘why’ questions are formulated as questions on oblique NPs, eg.
(47) Tangi ana a Tohe ki te aha?
cry T/A pers Tohe to the what
‘Why is Tohe crying?’

However, these are also frequently formulated as non-verbal questions, using a relative
clause on te take ‘the reason’, eg.
(48) He aha te take i reti ai e Maaui te
cls what the reason T/A lassoo part. by Maui the
raa?
sun
‘Why did Maui catch the sun?’

This construction using a relative clause in a non-verbal sentence is also a frequent

alternative to questioning an oblique NP for questions involving ‘how’, eg.
(49) He aha te mea i poro ai a Hata?
cls what the thing T/A cut part. pers Hata
‘How did Hata cut himself?’
‘What is the thing Hata cut himself on?’
(50) He aha te mea i patu ai i a Hata?
cls what the thing T/A kill part. DO pers Hata
‘How did Hata get killed?’

1.1.1.2.2.1.1.2.4 Questioning the verb

The verb can be questioned by direct substitution of aha in the verb slot. (Like many
bases in Maori, aha occurs in both nominal and verbal phrases without change of form,
see 1.16.) Eg.

In addition, with transitive verbs, two other possibilities are found, both involving
mahi ‘work, activity, do’. The first involves a possessive-relative clause, with mahi as the
verb of the relative clause:
 Maori 14

The second involves the use of mahi as a noun:

(53) He aha te mahi a Hata ki teeraa raakau?
cls what the deed gen Hata to that tree
‘What did Hata do to that tree?’
‘What has Hata been doing to that tree?’

More general questions, which correspond frequently to English questions with happen
are also possible, and involve the question-word being substituted in the verb slot, eg.

This form is frequently used if there is an implication that another person was involved,
and on occasions appears to correspond to ‘What was done to that child?’ A further
general question is
(55) I peewhea·tia te tamaiti raa?
T/A how·pass. the child dist
‘What happened to that child?’
(55), unlike (54), can be used when no specific action is implied.

1.1.1.2.2.1.2 Questions on constituents of subordinate clauses

There is a strong tendency to avoid questioning constituents of subordinate clauses.
However, many types of constituent can be questioned in many types of subordinate
clause. Question-word questions are possible in many types of noun clause, but in
relative clauses and adverbial clauses, it appears that only echo-questions are possible.
Because of the difficulty of eliciting data in this area, given the likelihood of avoidance of
these constructions, and because of the uncertainty of the judgements of grammaticality
in some instances, the following account may not represent the complete picture.
In sentential DO clauses, time and place adverbials can be questioned fairly freely:
(56) E moohio ana a Hata i maarena·tia a
T/A know T/A pers Hata T/A marry·pass. pers
 Syntax 15

As the examples show, to question place involves merely the substitution of the question
word in the appropriate slot, together with question intonation. Time questions, however,
involve the fronting of the time phrase and optionally the subject, as well as substitution
of the appropriate question word, question intonation, and the insertion of the particle ai
following the verb. It will thus be seen that the rules for main clause questions apply also
in the case of these subordinate clauses.
Subjects can be questioned using the possessive-relative strategy which is reserved in
main clauses for DOs, but this involves changing the sentential complement to a non-
finite form in some instances:

(59) Ko wai taa Hata e moohio ana hei

eq who sggen Hata T/A know T/A for(fut)
aawhina i a taaua?
help DO pers Idlincl
‘Who does Hata know will help us?’

It appears that it is tense which determines which construction is used: the first
construction is available only for i and e, the tense markers of the actor-emphatic.
Direct questioning of subjects, using the main clause subject patterns produces echo-
questions only. (58a) and (59a) correspond to (58) and (59) above:
 Maori 16

It is marginal whether DOs in sentential DO clauses can be questioned, although

sometimes the effect is achieved by promoting DOs to Su, eg.

Echo-questions involving the DO are possible by direct substitution, although even in

echo-questions, the actor-emphatic or passive are often preferred:

The verbal element can be questioned by substitution, thus:

With sentential clauses functioning as subject, there are two types of construction to
consider, negatives and the actor-emphatic (see 1.11.2.1.7—not all scholars agree that
these have clausal subjects). In negatives, the subordinate clause subject is normally
raised to the higher predication (see Negation, 1.4). This is then questioned with subject-
fronting, eg.
 Syntax 17

DOs are questioned by promoting to subject, eg.

Direct questioning of the verbal element produces an echo-question only, eg.

The effect of an information request can be obtained using mahi ‘do’, eg.

The only obliques which can readily be questioned in negatives appear to be time and
place. Place is questioned by direct substitution, eg.

Time questions involve fronting, as in main clauses:

 Maori 18

In the actor-emphatic construction, the questioning of the subject of the subordinate

clause is possible only as an echo-question:

(This is also true if the subject of the subordinate clause is fronted with ko or indefinite
subject fronting:
(69a) Ko te aha naa Hata i here?
top. the what actgen Hata T/A tie
‘What did Hata tie up?’
(69b) He aha naa Hata i here?
cls what actgen Hata T/A tie
‘What was it Hata tied up?’)

The verb can be questioned directly:

Obliques such as those expressing place are also directly questioned:

In both the latter cases, while not in doubt about the possibility of these question forms,
my consultant felt that non-actor-emphatic alternatives were more likely. Presumably
there is some conflict between the stress on the actor provided by this construction, and
focussing the question on some other constituent.
In kia noun clauses, the subject is questioned with the possessive-relative construction:
 Syntax 19

The underlying DO is also questioned thus.

The verb can be directly questioned, eg.

However, this is not the preferred form, which uses the possessive-relative construction
as for main clauses:

Oblique NPs, including time NPs, can be questioned directly:

(74) E piirangi ana a Hata kia maarena a
T/A want T/A pers Hata subj marry pers

In ki te clauses, there is no subject constituent. DOs are questioned using the possessive-
relative strategy, as for main clauses:

The verbal element may be questioned by direct substitution:

 Maori 20

Place NPs are also questioned by direct substitution:

In me clauses, all elements can be questioned using the same structures as for main
clauses, eg.

(Forms like those for T/A-marked clauses are possible, but not preferred.)

In hei nominalizations, it appears that only the oblique NP (with maa) can be
questioned:

In non-introduced nominalizations, the DO can be questioned using the possessive-

relative strategy:
 Syntax 21

Obliques of place may also be questioned in this construction, but the questioned NP is
fronted thus:

The nominalization itself may be questioned by direct substitution:

In relative clauses, direct questions appear to be impossible. Echo-questions can,

however, be formed on any element by direct substitution. The following will serve to
illustrate:
(86) E haere mai ana te tangata e moohio ana
T/A move hither T/A the man T/A know T/A
a wai?
pers who
‘The man that who knows is coming?’
(87) E haere mai ana te tangata i aha·tia e
T/A move hither T/A the man T/A what·pass. by
Hata?
Hata
‘The man that Hata did what to is coming?’
(88) E haere mai ana te tangata e tuhituhi ana
T/A move hither T/A the man T/A write T/A
i te aha?
DO the what
‘The man who is writing what is coming?’

In all types of adverbial clause construction, it appears that only echo-questions are
possible. These are achieved by direct substitution.
 Maori 22

The following gives some idea of the range found.

(89) Ka kati·a e koe te wini inaa haere
T/A close·pass. by IIsg the window when move
mai a wai?
hither pers who
‘You will close the window when who comes?’
(90) I kati·a e koe te wini i te haere·nga
T/A close·pass. by IIsg the window at the move·nom
mai o wai?
hither gen who
‘You closed the window when who came?’
(91) Mehemea ki te wareware i a wai te
if to the forget cause pers who the
whaangai te ngeru ka riri·a e koe a wai?
feed the cat T/A angry·pass. by IIsg pers who
‘If who forgets to feed the cat, you’ll be angry with them?’
(92) Kua oti i a Hata te kuti te hipi,
T/A finished cause pers Hata the shear the sheep
ahakoa kaaore a wai i aawhina i a
although neg pers who T/A help DO pers
ia?
IIIsg
‘Hata has finished shearing the sheep although who didn’t
help him?’
(93) Kua oti i a Hata te kuti te hipi,
T/A finished cause pers Hata the shear the sheep
ahakoa i aha·tia e Rewi?
although T/A what·pass. by Rewi
‘Hata has finished shearing the sheep although Rewi did what?’

Because of the complexity of the data in the area of question formation, an attempt is
made in Table 1 to summarize the possibilities. There appear to be very few
generalizations that can be made about this data.
Table 1: Summary of c uestioning strategies
Constituent Type
Construction Su DO V Place Time
Type ↓
Main clause Front Poss-rel or Subs Subs Front
Subs
T/A-marked DO noun-clause Poss- Prom Subs Subs Front in cl
rel
Negative Front Prom Echo or Subs Front
mahi
 Syntax 23

Actor-emphatic Echo NA Subs Subs Front

(fut.)
Kia DO clause Poss- Prom Subs Subs Subs
rel
Ki te DO cl NA Poss-rel Subs Subs Front in cl
Me clause Front Poss-rel Subs Subs ?
Hei nom NA NA (see OBL) * OBL ?
Front
Plain nom NA Poss-rel Subs Front Front
Adj clause & Adv. clause Subs Subs Subs Subs Subs
(echo)

Key: cl=clause; Front=fronting; (fut.)=in future tense only; NA= constituent not found in
that type; nom=nominalization; Poss-rel= possessive-relative construction;
Prom=promotion; Subs= substitution; ?=probably not possible;
*=ungrammatical/impossible.

1.1.1.2.2.1.3 Constituents of NPs that can be questioned

Questioning constituents of noun phrases directly produces only echo-questions.
Possessives, modifiers and heads, as well as constituents of relative clauses may be echo-
questioned by substitution of an appropriate question-word, eg.
(94) Ko teenei te kurii a wai?
eq this the dog gen who
‘This is whose dog?’
(95) He whare aha teenei?
cls house what this
‘This is a what sort of house?’

(Possible answers include he whare whero ‘a red house’, he whare kai ‘an eating house’.)
(96) too tama peehea te roa?
sggenIIsg son how the height
‘your how tall son?’

(answered by eg. too tama e ono putu te roa ‘your 6ft tall son’)
(97) te aha a/o Tuu?
the what gen Tu
‘Tu’s what?’
(98) I kite koe i te nui aha?
T/A see IIsg DO the big what
‘You saw the big what?’
 Maori 24

(Notice in (98) the unusual order adj—head; the answer, however, uses normal head—adj
order: te taraka nui ‘the big truck’.)
Non-echo questions relating to NP constituents are achieved using non-verbal
sentences. Equivalents for the above are given here:
(99) Naa wai te kurii?
actgen who the dog
‘Whose dog is it?’
(100) He aha te kara o teeraa whare?
cls what the colour gen that house
‘What is the colour of that house?’
(101) Hei aha teeraa whare?
for what that house
‘What is that house for?’
(102) Peehea te roa o too tama?
how the height gen sggenIIsg son
‘How tall is your son?’
(103) He aha naa Tuu? OR He aha aa/oo Tuu?
a what actgen Tu a what gen Tu
‘What belongs to Tu?’
(104) He aha te mea nui i kite naa koe?
cls what the thing big T/A see proxII IIsg
‘What is the big thing that you saw?’

1.1.1.2.2.1.4 Elements of prepositional phrases that can be questioned

In a prepositional phrase, only the head noun can be questioned directly, eg.
(105) I hoatu e koe te pukapuka ki a wai?
T/A give by IIsg the book to pers who
‘To whom did you give the book?’

It is possible to question other elements of the noun phrase using nonverbal sentences as
in 1.1.1.2.2.1.3.
The preposition itself cannot be wh-questioned.

1.1.1.2.2.1.5 Elements of coordinate structures that can be questioned

The coordination of personal nouns (including pronouns) involves the use of a collective
pronoun and ko as conjoining items (see 1.3 for details). In such conjoined personal
phrases, it is possible to question the second conjunct; the known party must occur first,
eg.
(106) Naa Hata raaua ko wai i horoi ngaa riihi?
actgen Hata IIIdl spec who T/A clean the(pl) dishes
‘Hata and who washed the dishes?’
(107) *Naa wai raaua ko Hata i horoi ngaa riihi?
 Syntax 25

actgen who IIIdl spec Hata T/A clean the(pl) dishes

‘Who and Hata washed the dishes?’

Non-personal NPs without a preposition are conjoined with me, a comitative marker. The
me NP may be questioned:
(108) Kua ngaro i a koe too pukapuka me
T/A lost cause pers IIsg sggenIIsg book with
too aha?
sggenIIsg what
‘You’ve lost your book and what?’

Prepositional phrases are coordinated by juxtaposition of the prepositional phrases. The

NP of the second phrase may be questioned, eg.
(109) Naa ngaa kau, naa ngaa aha i kai
actgen the(pl) cow actgen the(pl) what T/A eat
ngaa kuumara?
the(pl) kumara
The cows and what ate the kumaras?’

(Me coordination is also possible here, but is considered less acceptable by some speakers
than repeating the preposition.) This juxtaposition is also used for ‘or’ coordination:
(110) E piirangi ana koe ki a Tuu, ki a wai
T/A want T/A IIsg to pers Tu to pers who
raanei?
or
‘You want Tu or who?’

(‘You want Tu and who?’ would use the pronoun+ko construction.)

Verb phrases are coordinated by juxtaposition. The second verb phrase may be
questioned with the addition of hoki for ‘and’ coordination, and raanei for ‘or’
coordination:
(111) I whaangai·a, i aha·tia hoki ngaa kau?
T/A feed·pass. T/A what·pass. also the(pl) cow
‘He fed and what the cows?’
(112) I whaangai·a, i aha·tia raanei ngaa kau?
T/A feed·pass. T/A what·pass. or the(pl) cow
‘He fed or what the cows?’

Modifiers in noun phrases are coordinated by juxtaposing the second modifier+hoki

‘also’. The second modifier may be questioned, but atu ‘away’ is usually required before
hoki:
(113) te tangata moomona, aha atu hoki?
the man fat what away also
‘the fat, what man?’
 Maori 26

Clauses are coordinated by juxtaposition with no marker. Any feature of either clause
may be questioned, subject to the rules and constraints for questioning elements of non-
conjoined structures (see 1.1.1.2.2.1.1). However, such structures are not common.
(114) I hea a Rewi, ka tuutuki too·na motokaa
at where pers Rewi T/A collide sggen·IIIsg car
ki te tereina?
to the train
‘Where was Rewi when his car hit the train?’
(lit. ‘Where was Rewi and his car hit the train?’)

1.1.1.2.2.1.6 Questioning more than one thing in a sentence

Most combinations of items can be questioned together. There is a tendency to use the
question-word substituted in the declarative slot for all constituents questioned, although
the subject NP frequently undergoes one of the fronting strategies discussed in
1.1.1.2.2.1.1.2.1. The following examples illustrate some of the range of possibilities:
(115) Naa wai i aha?
actgen who T/A what
‘Who did what?’
(116) Naa wai i patu a wai?
actgen who T/A beat pers who
‘Who hit whom?’
(117) Ko wai kei te patu i a wai?
top. who T/A beat DO pers who
‘Who is hitting whom?’
(118) I patu·a a wai e Hata ki te aha?
T/A beat·pass. pers who by Hata with the what
‘Hata hit whom with what?’
(note that the Su is not fronted)
(119) Ko wai e aha ana i a wai i whea?
top. who T/A what T/A DO pers who at where
‘Who is doing what to whom where?’
(120) Ko wai taa Hata e patu naa ki te aha?
top. who sggen Hata T/A beat proxII with the what
‘Whom is Hata hitting with what?’

1.1.1.2.2.2 Changes to the questioned element

1.1.1.2.2.2.1 No change
Not applicable.
 Syntax 27

1.1.1.2.2.2.2 Moved to initial position

Questioned subjects are normally in initial position, but only with indefinite subjects does
this appear to be merely a movement transformation. (See 1.1.1.2.2.1.1.1.)
When the actor-emphatic is used to question the subject in transitive past and future
tense examples, this has the effect of placing the subject in initial position, but is not a
pure movement transformation. It is not entirely clear whether the fronting of time
phrases and of place phrases in certain subordinate constructions involves fronting or
some other process (see 1.1.1.2.2.1.1.2.3).

1.1.1.2.2.2.3 Moved to pre-verbal position

This is equivalent to initial position in Maori.

1.1.1.2.2.2.4 Clefted
Ko-fronting of subjects is probably a kind of clefting (see 1.11.2.1.4 and Bauer, 1991).
The possessive-relative question pattern for DOs would also appear to be pseudo-clefting.
Clefting thus accounts for the movements of most questioned elements. Whether the
actor-emphatic is a kind of clefting is not clear to me (see 1.11.2.1.7).

1.1.1.2.2.2.5 Forms intonation nucleus

This appears to be true of all questioned elements. If they are initial, they usually take
rising intonation, if final, there appears to be a choice between a rise and a fall, though
the fall appears to be the unmarked choice.

1.1.1.2.2.2.6 Other
The treatment of time phrases may come under this head. In general, there appears to be a
tendency to choose syntactic constructions in which the questioned element is naturally
first. Thus the actor-emphatic construction puts the underlying subject first; the passive
promotes the DO to subject, when it is readily placed in first position by ko-fronting;
causers with neuter verbs can be the subject of a corresponding transitive with whaka-,
etc. Questioning time phrases by means of a non-verbal sentence is another example of
this tendency if that proves to be the appropriate analysis.

1.1.1.2.2.2.7 Other elements moved with the Q-element

If a questioned element is moved, its phrase-type markers will accompany it, but no other
clause elements move obligatorily. However, when time elements are questioned, the
subject is very frequently moved to pre-verbal position as well, ie. is fronted or raised,
depending on the analysis.
 Maori 28

1.1.1.2.2.2.8 More than one element questioned

If the subject is questioned under these circumstances, it is normally fronted by one or
other available strategy. Other elements use the substitution-in-place strategy, although
the DO may use the possessive-relative strategy. The intonation nucleus falls on the
subject if questioned, otherwise on the final questioned item.

1.1.1.2.3 Echo questions

While a considerable variety of echo-questions is possible in Maori, they are not widely
used in natural discourse, and some speakers find them somewhat discourteous. These
speakers prefer to wait and hope to find out the information they missed from the
remainder of the discourse.

1.1.1.2.3.1 Yes-no echo questions

Yes-no echo questions may repeat the entire statement, but more frequently repeat the
questioned constituent. The question receives extra-high pitch (indicated by ), and
rising intonation on the questioned constituent.

(122) A. I haere au ki te whare pukapuka

T/A move Isg to the house book
‘I went to the library’
 Syntax 29

1.1.1.2.3.2 Question-word echo questions

The appropriate question word appears in the place of the questioned element in its
constituent. The entire sentence or a constituent containing the questioned element may
be, but is not usually, repeated. The echo-question has extra-high pitch, and the
questioned constituent has a rise.

1.1.1.2.3.3 Yes-no question echo questions

These occur, and follow the form and intonation contour of the original question, with
appropriate change of pronoun. Sometimes they receive extra-high pitch, but sometimes
the pitch appears to be raised only to the same extent as for non-echo yes-no questions.
 Maori 30

1.1.1.2.3.4 Question-word question echo-questions

These occur, and follow the form and intonation of the original question, with appropriate
change of pronoun. Sometimes, but not always, they receive extra high pitch.

1.1.1.2.3.5 Constituents that can be echo-questioned

It appears that any sentence element can be echo-questioned, including the passive agent
NP (with e), which is not commonly subject to direct question. For a discussion of echo-
questions in subordinate clauses, see 1.1.1.2.2.1.2.

1.1.1.2.3.6 Echo-questioning of more than one constituent

It is possible for several elements to be echo-questioned at once. The question-forms used
are indistinguishable from direct question forms (see 1.1.1.2.2.1.6). I am not aware of any
restrictions on combinations.

1.1.1.2.3.7 Questioning of different word-types

See 1.1.1.2.2.1.1.
 Syntax 31

1.1.1.2.4 Answers

1.1.1.2.4.1 Answers as distinct speech acts

They are not distinct speech acts in Maori.

1.1.1.2.4.2 Answers in the form of incomplete sentences

An incomplete sentence is the normal answer form for questions in Maori.

1.1.1.2.4.2.1 Yes-no questions

1.1.1.2.4.2.1.1 Minimum answers

The minimum answers to yes-no questions are: aae ‘yes’; kaaoo (sometimes kaaore or
kaahore) ‘no’; eehara (kee) ‘no’ (see Negation 1.4); pea ‘perhaps’; taihoa ‘wait’ (used to
indicate postponement of the answer).

1.1.1.2.4.2.1.2 Replies to leading questions

To leading questions expecting the answer ‘yes’, the normal reply is aae ‘yes’. To
contradict the expectation, kaaoo/eehara ‘no’ plus a positive statement of what is correct
is needed.
To leading questions expecting the answer ‘no’, the expected reply is kaaoo/eehara
‘no’. To contradict the expectation, aae ‘yes’ plus a positive statement of the correct
proposition is used.
In alternative questions, the answer aae ‘yes’ agrees with the positive statement. The
answer kaaoo/eehara ‘no’ agrees with the negative. The minimum answers may be
supplemented by a confirmatory statement, but this is not necessary.

1.1.1.2.4.2.2 Answers to question-word questions

The minimum answer involves the constituent containing the required information. Thus
possible answers to some of the questions in 1.1.1.2.2.1.1.1 are
(128) (i) Ko Rangi
eq Rangi
‘Rangi’ (to (13))
(ii) He whero
cls red
‘Red’ (to (17))
(iii) He tere ki te oma
cls fast to the run
‘A fast runner’ (to (18))
 Maori 32

(iv) He kara
cls flag
‘A flag’ (to (20))
(v) Noo Pooneke
actgen Wellington
‘From Wellington’ (to (21))
(vi) He pene
a pen
‘A pen’ (to (22))
(vii) Ko Rewi
top. Rewi
‘Rewi’ (to (23))

1.1.1.3 Imperatives

1.1.1.3.1 Positive imperative form

Different classes of verb require different structures to form the imperative. Most—
possibly all—of these constructions, however, are marked as distinct from declarative
constructions by the absence of a tense/aspect marker initiating the verb phrase (but see
the discussion of e below). All forms have a special intonation pattern, which consists of
a fall from the first mora. If this is the marker e, the starting pitch is usually rather high. If
the imperative is longer than two phrases, the fall is usually not sustained; declarative
patterns follow. (For a motivation of the term ‘mora’ as applied to Maori, see Bauer,
1981b. Briefly, a mora is a unit of the form (C)V, where V is a short vowel.)
Four different constructions must be discussed. The first applies to intransitive verbs,
and has two variants depending on the number of morae in the verb stem. If the stem
contains two morae or fewer, then the imperative form consists of the particle e+verb
stem (+particles), eg.
(129) E tuu!
imp. stand
‘Stand!’

If the verb stem has three morae or more, e is not used, eg.
(130) Haere atu!
move away
‘Go!’

(Tuu has two morae, haere has three.) Directional post-verbal particles apparently count
as part of the verb stem for the purposes of this rule, (see eg. Biggs, 1969,60; Head,
1989,69). Thus compare:
(131) Kake mai!
climb hither
‘Climb up here’ (speaker is already up)
 Syntax 33

(132) E kake!
imp climb
‘Climb up!’ (speaker is not up)

However, there seems to be some native speaker variation in this area.

The form e has been glossed ‘imperative marker’ here, though there is some doubt
whether this is the most appropriate description of its function. The same form, e, occurs
with vocatives under the same phonological condition, ie. with names containing two
morae or fewer, and with numerals 1–10 (all of which have two morae). It also occurs at
the end of lines of poetry, where its function appears to be to preserve the rhythm by
providing extra syllables. Waite (1990, 403) claims that e before imperatives and
numerals is the tense marker e which is found in the actor-emphatic future and before
negatives with kore, for instance. However, I do not regard this as beyond question.
Firstly, in the actor-emphatic (and in subordinate clauses of various kinds), the
phonological restriction on e does not apply: e occurs regardless of the number of morae
in the stem. Secondly, e can be demonstrated in those cases to be in paradigmatic
relationship with other T/A markers, such as i. Thirdly e is clearly non-past in
uncontroversial cases. With the imperative, there is phonological restriction. With
numerals, e does not occur with tekau ‘10’, which has three morae, suggesting that the
phonological restriction also applies there, and its sense is not non-past: e hia ‘how
many?’ is an appropriate form for a question about numbers even in past-time contexts
(see eg. (26) in 1.1.1.2.2.1.1). One cannot say *I hia…, for instance. I suspect that the e in
imperatives and numerals is the same form as the e with vocatives, not a tense marker,
but used to maintain the basic phonological structure of the Maori phrase: no base in
Maori has fewer than two morae, and the basic phrase structure of ‘phrase-type
marker+base’ ensures a minimum of 3 morae. Where no specific phrase-type marker is
required, if the phrase has fewer than three morae, e fills the empty slot, thus preserving
the minimum phonological form. It thus appears probable that it has the same function in
all these instances—it is a rhythmic filler. Because of this doubt, the glosses here remain
the more specific, conservative choice: they are glossed as specific phrase-type markers.
If the addressee is included in this imperative construction, it is normally in a separate
tone group, and takes the appropriate vocative construction if it is a proper name, eg.
(133) E tuu, koe!
imp. stand IIsg
‘Stand, you!’
(134) Haere atu, e Mere!
move away voc Mary
‘Go, Mary!’
(135) E oho, Tamahae!
imp. wake Tamahae
‘Wake up, Tamahae!’

The addressee may also precede, in which case some speakers require e koe for ‘you’.
There is also some native speaker variation in applying the rule for the insertion of e
before vocatives, eg. the text of the story Hinemoa begins Titiro, e Kawana ‘Look,
 Maori 34

Governor’, where Kawana has three morae. In an example like (133), koe has only two
morae, but does not usually take e. However, my consultant said that it would be possible
to have e koe to be emphatic, and that it would be used, for instance, if the addressee
wasn’t taking any notice. Alternatively, a pronoun can be used without a separate tone
group, eg.
(133a) E tuu koe
imp stand IIsg
‘You stand’

Under these circumstances, the pronoun is never accompanied by e, and appears to be a

subject rather than a vocative NP.
The second construction is used for canonical transitive verbs. The imperative of a
transitive verb requires the passive verb form, and the underlying DO promoted to
subject:
(136) Patu·a te kurii raa!
beat·pass. the dog dist
‘Beat that dog!’

If a non-pronominal addressee is included, it takes the vocative form. Non-pronominal

addressees normally follow, but may precede, eg.
(137) Patu·a te kurii raa, Tamahae!
beat·pass. the dog dist Tamahae
‘Tamahae, beat that dog!’

Note that the addressee is not in the form of the passive agent. For pronominal
addressees, see 1.1.2.1.1.
There are many verbs which appear with both the transitive and the intransitive
pattern, apparently with no change in meaning, and a single speaker can vary freely
between the two, eg. whio ‘whistle’, eg.
(138) E whio ki ngaa tamariki kia haere mai
imp. whistle to the(pl) children subj move hither
‘Whistle out to the children to come!’
(139) Whio·nga atu too kurii kia hoki mai
whistle·pass. away sggenIIsg dog subj return hither
‘Whistle out to your dog to return!’

Head (1989, 67–8) suggests that the two constructions differ in that the intransitive
pattern is action oriented, and the transitive pattern is goal-oriented. Note, however, that
it is not always the case, as might be expected, that the intransitive form is used when no
object is expressed, and the transitive form is used when an object is expressed. All four
possible combinations are attested.
The third construction is used for statives (both adjectives and experience verbs see
1.2.1.2.2). In many cases, it is clear that these are not true imperatives, and this is
probably true of them all. The construction has the subjunctive marker kia in the
tense/aspect slot; no further modifications to the structure are made. While the stative
 Syntax 35

adjective imperative normally uses the fall characteristic of the imperative, examples with
experience verbs usually use a contour which differs from the declarative only in having
a jump up on the stressed syllable of the verb phrase.
(140) Kia tere!
subj fast
‘Hurry!’
(141) Kia maumahara koe ki te raa whaanau o
subj remember IIsg to the day birth gen
too whaea
sggenIIsg mother
‘Remember your mother’s birthday!’

The conventional greeting, Kia ora lit. ‘Be well’, is a stative adjective imperative. It
appears that neuter verbs do not even occur as pseudoimperatives. Thus in the appropriate
sense, (142) was rejected:
(142) *Kia mutu te mahi naa!
subj finished the work proxII
‘Finish that job!’

A transitive with whakamutu was required.

Imperatives involving actions with body parts use a fourth construction. The verb is
active in form, and takes no imperative marker, and the body part is expressed in a
subject phrase, eg.
(143) Ruru te mahunga!
shake the head
‘Shake your head!’

It must also be noted here that under certain circumstances, the second of two
coordinated imperatives does not have imperative form, but has declarative form with the
particle ai post-verbally. For full discussion and illustration, see 1.3.1.1.1.

1.1.1.3.1.1 Person-number combinations in imperatives

Imperatives using the above constructions are possible in first person inclusive dual and
plural, and all numbers (ie. singular, dual and plural) of the second person. Thus
(144) Haere taaua/taatou//koe/koorua/koutou
move Idlinc/Iplinc/ /IIsg/IIdl/IIpl
‘Let’s go!//Go, you!’

Pronominal addressee forms and positions are somewhat complex in transitive

imperatives. The only pronoun to precede is e koe (voc IIsg) (though non-pronouns are
possible in this position). Following the verb is the normal position for pronominal
addressees, and in this position, all take e. It is not entirely clear whether this is the
passive agent marker e, but this seems likely, since four of the five possible pronouns
have too many morae to require the vocative e. (Note that proper names are not possible
 Maori 36

in this medial position.) Of the pronouns, only e koe can occur finally (but non-
pronominal addressees regularly occur in this position). The other pronouns are not
possible in this position with or without e. The following tabulates this information.
Three addressee slots are identified in the example sentence, and the possible forms for
each slot are listed in the table underneath:
ADDR 1, puuhi·a e ADDR 2 te manu, ADDR 3
ADDR 1 shoot·pass. by ADDR 2 the bird ADDR 3
‘Shoot the bird, ADDRESSEE!’
ADDR 1 ADDR 2 ADDR 3
E koe koe e koe
*Koorua koorua *[(e) koorua]
*Koutou koutou *[(e) koutou]
*Taaua taaua *[(e) taaua]
*Taatou taatou *[(e) taatou]

The three slots identified are, of course, mutually exclusive.

1.1.1.3.1.2 Degrees of imperative

Maori has what is usually described as a ‘weak imperative’ which uses the marker me
‘obligation’ in the T/A slot. (Note that me was formerly never followed by a verb form
with passive ending, though the remainder of the construction may be identical to the
passive construction; and note the superficial ergativity of this marking. However, it is
increasingly common among younger speakers to use passive terminations on verbs after
me.) Compare the following pairs:
(145a) Haere atu koe!
move away IIsg
‘Go, you!’
(145b) Me haere atu koe!
oblig move away IIsg
‘You should go’
(146a) Whai·a e koe te maatauranga!
chase·pass. by IIsg the knowledge
‘Pursue knowledge, you!’
(146b) Me whai e koe te maatauranga
oblig chase by IIsg the knowledge
‘You should pursue knowledge’
and (141) above with
(147) Me maumahara koe ki te raa whaanau o
oblig remember IIsg to the day birth gen
too whaea
sggenIIsg mother
 Syntax 37

‘You should remember your mother’s birthday’

It seems clear that these forms with me are not true imperatives, but modal statements.
Nonetheless, they can have the effect of a command if they come from a source with
sufficient authority.
The commonest means of increasing the force of an imperative is by tone of voice, in
which volume plays a substantial part.

1.1.1.3.2 Negative imperatives

Negative imperatives are formed with kaua, kauaka or aua (dialectal variants) as the
negator. The negative marker is followed by a declarative with the marker e ‘non-past’.
The negative imperatives corresponding to some of the positive imperatives above are
(148) Kaua e tuu!
neg imp T/A stand
‘Don’t stand!’
(149) Kaua e haere atu!
neg imp T/A move away
‘Don’t go!’
(150a) Kaua e patu i te kurii!
neg imp T/A beat DO the dog
‘Don’t beat the dog!’

Note that this negative imperative is active, although the positive cannot be. It is also
possible for the negative imperative to be passive:
(150b) Kaua e patu·a te kurii!
neg imp T/A beat·pass. the dog
‘Don’t beat the dog!’
(151) Kaua e tere rawa te haere
neg imp T/A fast intens the move
‘Don’t hurry/go so fast!’

(Note that the intensifier appears to be required here, but not in the positive. My
consultant much preferred a positive version of this, eg.
(152) Kia aata haere!
subj slowly move
‘Go slowly!’)

The addressee system is again rather complex. Proper names follow or precede in the
appropriate vocative form. Pronouns may also precede, in which case the predicted
vocative forms of the pronouns occur. However, the commonest position for pronouns is
as raised subjects (so none of them take e), following the negative imperative form,
except in the passive construction, (150b), where subject position is already filled. In this
case, pronouns may follow the verb as passive agents. Pronouns do not occur in final
 Maori 38

position (unless this is equivalent to post-verbal position). The following tabulations may
make this clearer:
ADDR 1, kaua ADDR 2 e haere, ADDR 3
ADDR 1 neg ADDR 2 T/A move ADDR 3
‘Don’t go, ADDRESSEE!’
ADDR 1 ADDR 2 ADDR 3
Pania koe Pania
E Mere koorua E Mere
E koe koutou *[(e) koe]
Koorua taaua *[(e) koorua]
Koutou taatou *[(e) koutou]
Taaua *Pania *[(e) taaua]
Taatou *e Mere *[(e) taatou]
ADDR 1, kaua ADDR 2 e patu·a e ADDR 3 te
ADDR 1 neg ADDR 2 T/A beat·pass. by ADDR 3 the
kurii, ADDR 4
dog ADDR 4
‘Don’t beat the dog, ADDRESSEE!’
ADDR 1 ADDR 2 ADDR 3 ADDR 4
Pania *Pania *Pania Pania
E Mere *e Mere *Mere e Mere
E koe *e koe koe *e koe
Koorua *koorua koorua *koorua
Koutou *koutou koutou *koutou
Taaua *taaua taaua *taaua
Taatou *taatou taatou *taatou

1.1.1.3.2.1 Person-number in negative imperatives

Negative imperatives occur with the same person-number combinations as positive
imperatives. The position and form of pronominal addressees was outlined in the
previous section. The following examples illustrate:
(153) Kaua taaua e haere!
neg imp Idlincl T/A move
‘Don’t let’s go!’
(154) Kaua taatou e kai i teenaa aaporo
neg imp Iplincl T/A eat DO that apple
‘Don’t let’s eat that apple!’
 Syntax 39

but
(155) Kaua e kai·nga e koutou teenaa aaporo
neg T/A eat·pass. by IIpl that apple
‘Don’t you eat that apple!’

1.1.1.3.2.2 Degrees of negative imperative

A mild negative command can be produced by using a negative declarative, eg.
(156) Kaaore kee te pere o Mikaere e tangi
neg contr the bell of Michael T/A sound
‘The bell of St. Michael’s is not to sound!’

In some cases, a stronger negative command is possible using kei noho, eg.
(157) Kei noho koe ka koorero parau
mon? never IIsg T/A speak false
‘You must never tell lies!’

Such a usage contains a threat, as well as a strong admonishment. Noho has not been
traced in this sense in other constructions, and doubt remains as to the appropriate gloss
of both noho and kei here.
Not all commands with kei are as strong as this, however, eg.
(158) Kei patu·a e koe te tangata raa moo te
mon? beat·pass. by IIsg the man dist intgen the
kore take noa iho
not matter unlimited down
‘Don’t you beat that man for no reason at all!’

It is not clear whether kei in this construction is the same morpheme as kei ‘monitory’. I
have been unable to find examples which native speakers find ambiguous.

1.1.1.3.3 Other means of expressing these imperatives

Any negative sentence can have the function of a negative command if the speaker has
absolute authority. There are also a few cases where a declarative can have the force of a
weak command, eg.
(159) Maa·u tonu e koorero atu
intgen·IIsg still T/A speak away
‘You tell him!’

Since this uses the actor-emphatic construction, it is available only for transitive
sentences. It is not uncommon under these circumstances, particularly from female
speakers, I suspect, although more research would be needed to establish the fact.
 Maori 40

1.1.1.4 Exclamations
Exclamations may constitute a separate sentence type in some instances, though they do
not always. Sometimes they consist simply of a noun phrase, with a special prosodic
pattern, of which the predominant characteristic is a lento, emphatic pronunciation of the
first phrase, often with somewhat breathy voice, which results in ‘stress’ on te. The basic
intonation contour is a fall, although the degree of wonderment determines whether this
starts high or not, eg.
(160) Te aataahua hoki o te ngahere!
the beauty indeed of the bush
‘How beautiful the bush is!’
(161) Te toko·maha hoki o te tangata!
the pnum·many indeed gen the man
‘What a lot of people!’

However, exclamations do not have to follow this syntactic pattern. They may take the
form of a declarative sentence, although they are characterized by the same prosodic
features—lento, emphatic, and often breathy articulation of the initial phrase. Compare
(162) Ka mutu te rangi aataahua!
T/A finished the day beauty
‘What a beautiful day!’
(163) Ka mutu te rangi aataahua.
T/A finished the day beauty
‘The beautiful day ended’

The ka in the exclamation is considerably longer, and the tempo of the phrase ka mutu
considerably slower than in the declarative. There is not necessarily any observable
difference in the pitch patterns.

1.1.1.5 Sentence types regularly used in functions other than their

normal ones
This does not appear to be a particularly widespread feature of Maori. There are three
main ones of which I am aware.
Imperative forms with rising intonation are frequently used to make polite
suggestions, eg.
(164) Tango·hia raa too koti
take off·pass. dist sggenIIsg coat
‘Do take your coat off’

Requests may also be made by statements, usually with rising intonation sentence finally,
eg.
(165) Maa·u e whakakapi atu te kuuaha
intgen·IIsg T/A close away the door
 Syntax 41

‘You will clpse this door?’

Negative statements, and sometimes positive ones, especially the actor-emphatic

illustrated in example (159) above may be commands (see 1.1.1.3.2.2–1.1.1.3.3).

1.1.2 Subordination

1.1.2.1 General markers of subordination

Maori presents a vast array of different types of subordination with no general markers.
There are some forms which are frequently associated with subordinate constructions, but
which are also used in non-subordinate constructions, eg. kia, ki te, ai.
Intonation does not appear to be a main source of marking for subordination:
sequences of two sentences are frequently indistinguishable from main-clause plus
subordinate-clause in terms of intonation.
There are some restrictions on T/A markers in subordinate clauses, which in some
instances help to signal subordination, such as the use of e ‘non-past’ primarily in
subordinate structures, but in many instances, the T/A markers provide no evidence for
subordination.
It appears that, in general, adverbials cannot occur in pre-verbal position in
subordinate structures, but there is not a total ban on the fronting of material to this
position: subjects, for instance, may be fronted with ko.

1.1.2.2 Noun Clauses

1.1.2.2.1 General marking of noun clauses

There is no marking distinctive of noun clauses, which may function as either subject or
DO or occasionally as the complement of a preposition. They occur in the position usual
for their function. DO clauses, if finite, do not, however, occur with the DO preposition (i
or ki).

1.1.2.2.2 Types of noun clause

Noun clauses may take the form of an unmodified sentence. This is characteristic only of
DO clauses and subjects of passive verbs. These may be introduced by any of the T/A/M
markers, although there are a few restrictions, see 1.1.2.5.1.
If noun clauses refer to hypothetical or unrealized events, they are introduced by the
subjunctive kia, or by ki te, depending on the type of verb and the identity of the
participants in the clauses.
They may take the form of nominalized verb constructions, or verb stem constructions
accompanied by morphology typical of nominal constructions.

1.1.2.2.2.1 Sentential noun clauses

 Maori 42

These are used with verbs which take statements of (assumed) fact or actual events as
DOs. The first example below shows a subordinate verbal clause, and the second a
subordinate non-verbal clause with ko-fronting of the subject The third shows a sentential
subject clause.
(166) E maumahara ana ahau i whiu·a koe
T/A remember T/A Isg T/A punish·pass. IIsg
moo te haehae i roto i too pukapuka
intgen the scribble at inside at sggenIIsg book
‘I remember that you were punished for scribbling in your book’
(167) Kua moohio kee mai ia ko te paa i
T/A know contr hither IIIsg top. the pa at
runga i te aromaunga, too Te Aotakii
top at the mountain·face sggen Te Aotaki
‘But he already knew that the pa above on the mountain-face was that of Te Aotaki’ (TWh,
19)
(168) Kua riro maa raatou e tango ngaa
T/A come about intgen IIIpl T/A take the(pl)
ika i ngaa taumanu
fish from the(pl) thwart
‘It had come to be the habit for them to take the fish (themselves) from the thwarts’ (TWh,
17)

Negation also involves noun clauses as subjects of stative verbs, see 1.4.
The two mood markers, me ‘obligation’ and kei ‘monitory’ also introduce full
sentences as complements. Their place in the system appears to be somewhat different
from that of kia, also a mood marker, which is accordingly treated separately, see
1.1.2.2.2.3.
(169) Kua whakarite·a me tae raatou ki te mira
T/A arrange·pass. oblig arrive IIIpl to the mill
o Kawerau i te iwa karaka
gen Kawerau at the nine clock
‘It was arranged that they should arrive at the Kawerau mill at nine o’clock’ (TR2, 99)
(170) I mataku au kei hoki ia ki te kaainga
T/A fear Isg mon return IIIsg to the home
‘I feared that he might return home’

1.1.2.2.2.2 Prepositional complements

There are a number of instances where Maori appears to have an underlying sentential
noun clause as the complement of a preposition. However, subject raising is compulsory
in this type of construction. Sometimes the raised subject takes the form of a possessive,
and sometimes it is a simple NP. The two treatments of the subject are not
interchangeable in particular examples, but I do not have sufficient data to be sure of the
appropriate generalization. Consider the following:
(171) Kei te mea hoki ia ki toona kore i paatai
T/A say also IIIsg to sggenIIIsg neg T/A ask
 Syntax 43

ki ngaa tamariki raa ko teewhea a Ruataupare

to the(pl) children dist eq which pers Ruataupare
‘He was asking himself why he hadn’t questioned the children as to which was Ruataupare’
(TWh, 19)

We are concerned here with the complement of the preposition ki in bold. The underlying
sentence must be of the form
(171a) (I) kore ia i paatai…
T/A neg IIIsg T/A ask
‘He didn’t ask…’

where ia is already raised from the paatai predication (see Negation 1.4). When this is
embedded under ki, the underlying ia surfaces as a genitive, toona.
The following quite common construction must, I think, also be analyzed as involving
a sentential complement to a preposition:
(172) I a ia e haere ana i te ngahere, ka
at pers IIIsg T/A move T/A at the bush T/A
taka·hia noa·tia e ia teetahi ngaarara
tread·pass. sudden·pass. by IIIsg a(sp) lizard
‘While he was walking in the bush, he suddenly stepped on a lizard’

I think the subordinate construction here is to be analyzed as

(172a) I [e haere ana ia i te ngahere]s
at T/A move T/A IIIsg at the bush

ie. as involving a sentential prepositional complement, with obligatory raising of the

clausal subject to prepositional complement position. Whether the critical factor in
determining the difference in the form of the raised subject in these examples is how
many times it has undergone raising I am unsure.

1.1.2.2.2.3 Kia clauses

Kia and ki te clauses (see 1.1.2.2.2.4) are both used for irrealis clauses, in complementary
distribution. However, some details of their distribution remain unclear. If the verb in the
subordinate clause is an experience verb, a neuter verb (see 1.2.1.2.2), a negative, or a
passive, the clause is introduced by kia. If the subordinate clause verb does not fall into
one of these categories, the distribution of kia and ki te depends primarily on the identity
or otherwise of the subjects of the main and subordinate clause. Kia is used for subjects
not identical, and ki te for identical subjects. If the main clause verb is bivalent, these
rules appear to predict accurately whether kia or ki te is required. If, however, the main
clause verb is trivalent, these rules fall short, and I have failed to find what further rules
apply, or whether there is variation between speakers. Examples illustrating this problem
are given at the end of 1.1.2.2.2.4.
Under the same conditions, kia and ki te clauses occur as subjects of passives and non-
verbal sentences, and as appositional noun clauses.
The following examples illustrate kia clauses:
 Maori 44

(173) Kii atu, kia mau·ria mai e koutou

say away subj bring·pass. hither by IIpl
oo maaua kaakahu
plgenIdlexcl clothes
‘Tell him that you must bring us our clothes’ (TWh, 19)
(174) E piirangi ana maatou kia haere raatou
T/A want T/A Iplexcl subj move IIIpl
‘We want them to go’
(175) Ka piirangi ia kia moohio ki te kootiro raa
T/A want IIIsg subj know to the girl dist
‘He wanted to know that girl’
(176) Ka mea atu ia kia raatou kaua e hoki
T/A say away IIIsg subj IIIpl neg T/A return
ki te kaainga
to the home
‘He told them that they should not return home’

When the embedded clause has a subject identical to that of the main clause, it is normal,
as in (175) to delete the identical subject. However, it may also be pronominalized:
(177) I inoi a Hone kia whakaora·ngia ia i
T/A pray pers John subj cure·pass. IIIsg from
tana maauiui
sggenIIIsg illness
‘John prayed that he might be cured of his illness’

Subject in non-verbal sentence:

(178) Ko te tuumanako, kia tae mai koutou ki
eq the wish subj arrive hither IIpl to
taa taatou hui
sggenIplincl meeting
‘It is [our] wish that you should come to our meeting’

Appositional noun clause:

(179) Kaatahi raatou ka whakarite ki te tautoko i te
then IIIpl T/A arrange to the support DO the
take a Te Arawa kia tua·ina ngaa raakau paina
case gen Te Arawa subj fell·pass. the(pl) tree pine
‘Then they arranged to support the Te Arawa case that the pine trees should be cut down’

(Note that, in addition to the rules given above for the use of kia and ki te, there appear to
be certain verbs which never occur with ki te complementation, even if the like-subject
constraint is met, eg. inoi ‘pray, beg’.)
 Syntax 45

1.1.2.2.2.4 Ki te clauses
These structures are not entirely verbal, in contrast to those discussed in the previous two
sections: they have a number of nominal characteristics. There is no tense/aspect/mood
marker; that slot is filled by ki te (lit. ‘to the’). Though written as two words, it acts as a
unit: ki cannot be replaced by other prepositions, and te cannot be replaced by other
determiners. Ki te can only be used when the verb in the subordinate clause is active
transitive or intransitive, and only when the subject of the main verb and subordinate verb
are identical. The subject must be deleted from the subordinate clause.
(180) E whakaaro ana maaua ki te haere ki Tonga
T/A decide T/A Idlexcl to the move to Tonga
‘We are thinking of going to Tonga’
(181) Ka tono·a e ia ngaa heepara ki te
T/A order·pass. by IIIsg the(pl) shepherd to the
whiu mai i ngaa hipi ki te wuuruhete
chase hither DO the(pl) sheep to the woolshed
‘The shepherds were ordered by him to chase the sheep to the woolshed’

The latter example also shows that it is the surface subject in the main clause, and not the
underlying subject, which determines the likesubject constraint
As mentioned in 1.1.2.2.2.3, there are problems with the distribution of kia and ki te
when the main verb is trivalent. Thus while the second example above appears to follow
the ‘like-subject constraint, the following examples also occur:
(182) Karanga·tia atu a ia kia noho ki raro
call·pass. away pers IIIsg subj sit to underneath
‘Call to him to sit down’

but
(183) Karanga·tia a ia ki te koorero
call·pass. pers IIIsg to the speak
‘Call to him to speak’

and
(184) Ka tono a Hata i tana kurii ki te
T/A order pers Hata DO sggenIIsg dog to the
whiu i ngaa hipi
chase DO the(pl) sheep
‘Hata ordered his dog to chase the sheep’

but
(185) Ka tono a Hata ki a koe kia noho ki
T/A order pers Hata to pers IIsg subj sit to
raro
underneath
‘Hata ordered you to sit down’
 Maori 46

While the verbs embedded under kia tend to be intransitive, and are frequently verbs of
motion, intransitive uses of verbs also appear under ki te. It is not clear whether
transitivity is the factor influencing the choice. Hohepa (1970, 23) suggests that ki te does
not occur as the complementizer if the matrix verb has an object, but this is contradicted
by (178); he also specifies that both matrix and complement verb must be active for ki te
to be used, but this is contradicted by (175), for instance. Thus his rule does not account
for all the data I have collected either.

1.1.2.2.3 Indirect statements

A number of examples of indirect statements have already been given in 1.1.1.1. In
addition to the types of indirect statement illustrated there, there are also those which
correspond to headless relative clauses in English. In Maori, these have forms which in
all cases but one correspond to questions, although no question is being asked. In (186)—
(188), the subordinate indirect statement is identical to the corresponding interrogative:
With the subject “questioned”:
(186) I paanui·tia ko wai i wini
T/A announce·pass. top. who T/A win
‘[He] announced who had won’

With a place Obl “questioned”:

(187) I koorero mai ia e haere ana ia
T/A say hither pers IIIsg T/A move T/A IIIsg
ki hea
to where
‘He said where he was going’

With a time Obl “questioned”:

(188) I koorero mai a ia a whea a
T/A say hither pers IIIsg at(fut) when pers
ia ka haere
IIIsg T/A move
haere ai
move part.

‘He said when he was going’

However, with a “questioned” DO, there is no interrogative word, but a possessive-

relative structure subordinated to a definite possessive head, ie. a structure corresponding
to ‘He said that which I told him’, eg.
(189) I koorero mai a ia i taana i
T/A say hither pers IIIsg DO sggenIIIsg T/A
moohio ai
 Syntax 47

know part.
‘He told me what he knew’

Note the occurrence of the DO marker, i, and compare the remarks on DOs in
1.1.1.2.2.1.1.2.2.
Structures like these also occur as the DO noun clauses following such constructions
as ‘don’t know’ and ‘don’t remember’. However, in these instances, when the DO is
“questioned”, the interrogative form is used. Illustrations are given here:
With subject “questioned”:
(190) Kaaore au i te maumahara ko teehea tamaiti
neg Isg T/A remember top. which(sg) child
e pupuri ana i te paaoro
T/A hold T/A DO the ball
‘I don’t remember which child was holding the ball’
(191) Kaaore au i te maumahara naa wai teenaa i
neg Isg T/A remember actgen who proxII T/A
koorero mai ki a au
say hither to pers Isg
‘I don’t remember who told me that’

With DO “questioned”:
(192) Kaaore au i te maumahara he aha taana
neg Isg T/A remember cls what sggenIIIsg
i koorero mai ai ki ahau
T/A say hither part. to Isg
‘I don’t remember what he told me’
(193) Kaaore au i te maumahara he aha aa taaua
neg Isg T/A remember cls what plgenIdlincl
i hoko mai ai
T/A barter hither part.
‘I don’t remember what we bought’

Note the absence of the DO marker i before these sentential structures.

With place Obl “questioned”:
(194) Kaaore au i te maumahara i waiho e au
not Isg T/A remember T/A leave by Isg
taku koti ki whea
sggenIsg coat to where
‘I can’t remember where I left my coat’ oat’
(195) Kaaore au i te maumahara oo whea a ia
neg Isg T/A remember ctgen when pers IIIsg
i tae mai ai
T/A arrive hither part
‘I don’t remember when he arrived’
 Maori 48

1.1.2.2.4 Indirect questions

See also 1.1.1.1. Yes-no indirect questions, including alternativequestions, are introduced
by mehemea ‘if’, eg.
(196) I paatai mai a ia mehemea e haramai
T/A ask hither pers IIIsg if T/A come
ana ahau
T/A Isg
‘He asked if I was coming’
(197) I paatai mai a ia mehemea toomuri a
T/A ask hither pers IIIsg if late pers
ia, kaaore raanei
IIIsg neg or
‘He asked whether he was late or not’

Question-word indirect questions in all cases embed the appropriate direct question with
pronominal changes, eg.
(198) I paatai mai a ia kei whea ngaa
T/A ask hither pers IIIsg at(pres) where the(pl)
pene
pen
‘He asked where the pens were’
(199) I paatai mai a ia ko wai e haere
T/A ask hither pers IIIsg top. who T/A move
mai ana
hither T/A
‘He asked who would be coming’
(200) I paatai mai a ia he aha taa Rewi
T/A ask hither pers IIIsg cls what sggen Rewi
e whaangai ana
T/A feed T/A
‘He asked what Rewi was going to feed’
(201) I paatai a ia e haere ana maatou ki
T/A ask pers IIIsg T/A move T/A Iplexcl to
 Syntax 49

1.1.2.2.5 Indirect commands

Indirect commands take the form of kia or ki te noun clauses, following the usual rules
for choice between the two (see 1.1.2.2.2.2.). There is a variety of introductory verbs,
including tono ‘order’, whakahau ‘order’, kii ‘say’, koorero ‘speak’. The following
examples illustrate:
(203) I koorero mai a ia kia whakatika ahau
T/A speak hither pers IIIsg subj stand up Isg
‘He told me to stand up’
(204) I whakahau a ia i ahau ki te pupuhi
T/A order pers IIIsg DO Isg to the shoot
i te manu
DO the bird
‘He ordered me to shoot the bird’

Thus these constructions bear no direct relationship to the corresponding imperatives.

Note the preposition accompanying the raised subject in (204).

1.1.2.2.6 Non-finite noun clauses

It is questionable whether the distinction between finite and non-finite verbs is relevant to
Maori (see 2.1.3.5, where it is suggested that the relevant distinction is between verbal
and nominal constructions). Nevertheless, the term is retained here for expository
purposes.
The ki te construction has already been discussed and illustrated. It is the most clearly
clausal of the non-finite constructions with nominal function. The following
characteristics are found:
– the T/A/M markers do not occur, and the verb cannot be passive in form
– the verb form is the stem; ki te is formally indistinguishable from ‘to the’
– the subject is obligatorily deleted under identity
– the remaining arguments are unchanged
– there are no morphological insertions
 Maori 50

– adverbials retain their normal form and position.

There are two other types of construction which must be discussed.
Firstly, there are nominalizations of statements. These have the following
characteristics:
– they are introduced by the preposition normal for non-sentential arguments of the verb
– there is accordingly no T/A/M marker, and the active/passive distinction is not made in
the verb
– the verb stem takes the suffix -(Ca)nga (see 2.2.1)
– it is possible to include any argument from a corresponding finite form
– if a subject is included, it appears as a genitive; the selection of the appropriate form is
discussed in 1.10.6 and 2.1.1.3.3–4
– the definite article, te precedes the nominalization unless a pronominal (genitive)
subject occurs
– if manner adverbials are present, they are nominalized with -tanga, with the exception
of the small class of adverbials which regularly precede the verb stem
– no adverbials precede the introductory preposition.
The following examples illustrate these points.
(205) E maumahara ana au ki te whiu·nga
T/A remember T/A Isg to the punish·nom
paakaha·tanga i a koe
severe·nom DO pers IIsg
‘I remember you being severely punished’
(206) E maumahara ana au ki taku tino
T/A remember T/A Isg to sggenIsg very
maataku·tanga
frighten·nom
‘I remember being very frightened’
(207) I rongo au i te tangi·hanga hotuhotu·tanga
T/A hear Isg DO the cry·nom sob·nom
o ngaa taangata i te tangihanga ki a
gen the(pl) people at the mourning to pers
Maui Poomare
Maui Pomare
‘I heard the people sobbing at Maui Pomare’s tangi’
(208) E whakahee ana au ki te mate·kai·tanga
T/A deplore T/A Isg to the lack·food·nom
rawa·tanga o ngaa iwi i teenei
extreme·nom gen the(pl) tribe cause this
kawanatanga mai anoo i too raatou whiwhinga
government hither indeed from sggenIIIpl accession
ki te mana whakahaere
to the position manage
‘I deplore the reduction to the breadline of the people caused by this government right
from their accession to power’.
 Syntax 51

A second type of nominalization can be used as the non-finite equivalent of the indirect
statements discussed in 1.1.2.2.3, under negations of verbs like moohio ‘know’ and
maumahara ‘remember’. These nominalizations have the following characteristics:
– the nominalization is introduced by hei; whether this is a nominal or prepositional
particle remains uncertain
– there is no T/A/M marker, and the active/passive distinction in the verb disappears
– the nominalization is formed with the suffix -(Ca)nga
– no arguments are omitted
– the subject appears in the form maa/moo+NP, following the nominalization; one NP
always precedes the hei phrase—the NP in question form
– single-word manner adverbials are nominalized if present
– the word order is questioned NP+hei+nominalization +maa/moo+subject+…
The construction is illustrated below.
(209) Kaaore au i te moohio he aha hei hanga·nga
neg Isg T/A know cls what for? build·nom
maa·ku
intgen·Isg
‘I don’t know what to build’
(210) Kaaore au i te moohio ko teewhea tamaiti hei
neg Isg T/A know top. which(sg) child for?
whiu·nga paakaha·tanga maa·ku
punish·nom severe·nom intgen·Isg
‘I don’t know which child to punish severely’

A similar pattern is available for a few adverbials involving movement towards a goal,
with the addition of he waahi ‘a place’ preceding hei:
(211) Kaaore au i te moohio kei whea he waahi
neg Isg T/A know at(pres) where a place
hei piki·nga atu moo·ku
for? climb·nom away intgen·Isg
‘I don’t know where to climb to’

(Note that the intransitive piki requires moo for the agent, see further 2.1.1.3.3–4.)

1.1.2.3 Adjective clauses

1.1.2.3.2 The marking of adjective clauses

There is a variety of strategies for forming relative clauses, and the marking differs in
each case. Their intonation pattern is indistinguishable from that of a sequence of two
independent sentences.
There are some restrictions on the T/A markers which can occur in such clauses, see
1.1.2.5.2, but these do not unambiguously mark subordination.
 Maori 52

It is possible that ai and the deictic particles serve to mark the subordination of such
clauses when they occur: they are sometimes obligatory and sometimes optional. For
some discussion, see Bauer, 1982, esp. pp 337ff.)

1.1.2.3.2 Restrictive vs. non-restrictive

Maori does not make this distinction. Native speakers of Maori who are virtually
bilingual are often unaware of this distinction in English, and do not respond to it. It is
difficult to determine whether relative clauses in Maori are interpreted restrictively or
non-restrictively. (There is no difference in meaning obtainable by having a separate tone
group for the relative clause.) Given eg.
(212) Ko taku iraamutu e noho mai raa i
eq sggenIsg nephew T/A stay hither dist at
Oohope, he kaiako
Ohope cls teacher
‘My nephew who lives in Ohope is a teacher’

and asked whether they could deduce how many nieces/nephews the speaker has,
consultants could only say that only one was mentioned, and that no implications could
be drawn about the existence of others.
When I asked my consultant to translate ‘(I have three nephews.) John, who is coming
to stay, is the eldest’, the first response was a non-relative, thus:
(213) Ko Hone te mataamua, ka haere mai, noho
top. John the eldest T/A move hither stay
mai
hither
‘John is the eldest, and [he] is coming to stay’

However, the following relative clause:

(214) ?Ko Hone kei te haere mai ki te noho, ko te
top. John T/A move hither to the stay eq the
mataamua
eldest
‘John, who is coming to stay, is the eldest’

was proffered and accepted with some diffidence. This might suggest that relative clauses
are basically restrictive, and that the problem with (214) is that it has to be non-
restrictive. However, it is possible to have proper names as antecedents in relative
clauses, and many such instances have non-restrictive readings, eg.
(215) Ka hoki te koorero ki a Hotu i ruku raa
T/A return the story to pers Hotu T/A dive dist
i te punga o too raaua waka
DO the anchor gen sggenIIIdl canoe
‘The story returns to Hotu, who had dived for the anchor of their canoe’ (KH, 3)
 Syntax 53

There is only one Hotu in the story; the relative clause recapitulates the action at the point
where the narrator pursued events relating to the other main protagonist.
When I explained the difference in English between pairs like The trees, which we cut
down, were all small’ and The trees we cut down were all small’, and asked how to
capture the difference in Maori, my consultant consistently modified the antecedent NP
of the relative clause to match the restrictive sense, and rendered the non-restrictive
reading with a less specific structure, eg.
(216) Ko ngaa raakau i tua·ina e maatou he
top. the(pl) tree T/A fell·pass. by Iplexcl cls
ri·riki katoa
dup·small all
‘The trees, which we cut down, were all small’
(217) Ko aua raakau i tua·ina raa e maatou
top. pldet aph tree T/A fell·pass. dist by Iplexcl
he ri·riki katoa
cls dup·small all
‘The trees we cut down were all small’

This suggests that the basic reading is non-restrictive. The evidence is thus contradictory,
and it is not at all clear what conclusion can be drawn, except perhaps that this distinction
is irrelevant for Maori.

1.1.2.3.3 The position of the head noun

The head noun always precedes the adjective clause. On rare occasions, another (usually
one-word) phrase may intervene. See human locative genitive, under 1.1.2.3.4.2.1.

1.1.2.3.4 Treatment of the relativized element

1.1.2.3.4.1 Preservation in full

The relativized element is never preserved in full.

1.1.2.3.4.2.1 Replacement by personal pronoun

This treatment of the relativized element is obligatory for relativizing on the maa/naa-NP
of the actor-emphatic construction. These two prepositions take special cliticized forms
of personal pronouns (see 2.1.2.4.3; 2.1.2.4.10), eg.
(218) Kua tae mai te kootiro naa·na i hoko
T/A arrive hither the girl intgen·IIIsg T/A barter
mai ngaa whurutu
hither the(pl) fruit
‘The girl who bought the fruit has arrived’

Replacement by a personal pronoun is also compulsory for relativizing on objects of

comparison. (Sentences of this kind raise a smile, and should probably be considered
 Maori 54

marginal, though there does not appear to be any doubt about the grammaticality or
comprehensibility of the result)
(219) Ko te tamaiti he iti iho a Rewi i
top. the child cls small down pers Rewi compar
a ia he koretake
pers IIIsg cls good-for-nothing
‘The youth that Rewi is smaller than is a good-for-nothing’

Such examples are only possible, however, if the object of comparison is human, since
personal pronouns in Maori are, indeed, personal.
Pronoun replacement is also obligatory for human NPs in locative genitives (ie. as the
complements of the adnominal prepositions in complex locational expressions of the
form prep+local noun+prep+NP, eg. i raro i… ‘underneath…’), eg.
(220) Ko te tangata teenaa naa·ku i waiho te
eq the man that actgen·Isg T/A leave the
panana ki raro i a ia
banana to underneath at pers IIIsg
‘That’s the man I put the banana under’

It is possible, though not obligatory to retain a personal pronoun copy of the relativized
element when relativizing on the following grammatical relations: notional IOs, oblique
NPs, and genitives, provided that the head noun has a personal referent. This is not the
commonest treatment for these categories, however.
With a notional IO:
(221) Ko teenei te tangata naaku i hoatu ki a
eq this the man actgenIsg T/A give to pers
ia te pukapuka
IIIsg the book
‘This is the man to whom I gave the book’

With an oblique NP (in this case the cause NP with a neuter verb, see 1.2.1.2.2 and
2.1.3.1.5):
(222) Ko Tamahae te tamaiti i mau nei i
eq Tamahae the child T/A caught proxI cause
a ia te tarakihi
pers IIIsg the tarakihi
‘Tamahae is the child by whom the tarakihi was caught’

With a genitive NP:

(223) I kaute ia i ngaa tamariki kua eke
T/A count IIIsg DO the(pl) children T/A reach
oo raatou tau ki te tekau maa tahi
plgenIIIpl year to the ten and one
‘He counted the children whose age had reached 11’
 Syntax 55

1.1.2.3.4.2.2 Replacement by a relative pronoun

There are no clear relative pronouns in Maori, although the particles discussed in
1.1.2.3.4.2.3 below have sometimes been described as substitutes for them.

1.1.2.3.4.2.3 Replacement by an invariant particle

A good deal of uncertainty surrounds the function of ai and the deictic particles (nei, naa,
raa) in those relative clauses in which they occur in Maori. However, it is possible that
they function as pro-forms for relativized elements. For a discussion of the problems with
this analysis, and some alternative suggestions, see Bauer 1982, esp. pp329ff.
Either ai or a deictic particle is obligatory when the relativized element is a non-
human oblique NP. It is also obligatory for non-human locative genitives, and is normal
for notional IOs. It is also obligatory in the possessive construction which is one of three
alternatives when the relativized element is a DO. For (usually) younger speakers, who
use the construction formerly limited to oblique NPs for DOs, the use of ai or a deictic is
also compulsory for this pattern of DO relativization.
These forms also occur optionally when the relativized element is a subject, or the DO
of an experience verb. At least some scholars who appear to regard these particles as pro-
forms in the cases covered by the list of compulsory environments regard them as having
a different function in these optional environments (see Chung & Seiter, 1980, 631).
The examples below illustrate these points. In addition, there are certain tense
restrictions on these particles. Ai cannot occur in present tense examples; raa is normally
the deictic alternative to ai in past time examples.
With oblique NP:
(224) Ko Mokoia te moutere i kau atu ai a
eq Mokoia the island T/A swim away part. pers
Hinemoa i Rotorua
Hinemoa from Rotorua
‘Mokoia is the island to which Hinemoa swam from Rotorua’

(Raa could be substituted for ai; one or other is obligatory.)

With locative genitive:
(225) Ka noho ia i teetahi raakau e noho nei
T/A sit IIIsg at a(sp) tree T/A sit proxI
he taangata i raro
a people at underneath
‘He sat in a tree under which some people were sitting’ (KM, 3)

(Ai cannot occur in place of nei here without altering the sense: it would then mean
‘where people were in the habit of sitting’; this use of ai appears to be a T/A marker.)
With notional IO:
(226) Kaaore te tamaiti i paatai ai te maahita
neg the child T/A ask part. the teacher
i te whakarongo
 Maori 56

T/A listen
‘The child that the teacher asked wasn’t listening’

(This is possible without ai or a deictic, according to my consultants, although they seem

to choose a particle when asked to translate such examples. Atu raa or atu nei can replace
ai.)
With DO possessive construction:
(227) I hoko mai ia i te whare a Hata i
T/A barter hither IIIsg DO the house gen Hata T/A
hanga ai
build part.
‘He bought the house that Hata built’

(Raa may be substituted for ai here.)

With DO relativization (principally younger speakers):
(228) I kite ia i te hooiho i tiaki ai te
T/A see IIIsg DO the horse T/A care part. the
tamaiti
child
‘He saw the horse that the child cared for’

(Raa may be substituted for ai.)

With subject:
(229) I whakawhaiti·tia ngaa moni i kohi·a
T/A collect·pass. the(pl) money T/A gather·pass.
(ai) e te komiti
part. by the committee
‘The money that was raised by the committee was collected up’

With DO of experience verb:

(230) I moe ia te wahine piirangi (ai)
T/A sleep IIIsg DO the woman T/A want part.
ia
IIIsg
‘He married the woman he wanted’

For most consultants, the ai in (229) and (230) is optional. Examples with ai in these
instances are relatively rare.

1.1.2.3.4.3 Deletion of the relativized element

This is the treatment of the relativized element if it functions as the subject or the DO of
an experience verb in the relative clause. Since the two commonest methods of
relativizing on the DO of a canonical transitive verb involve the promotion of that DO to
subject via either passive or the actor-emphatic, such elements are also deleted. The
examples illustrate these points:
 Syntax 57

With subject:
(231) I kite au i te tangata i hoko mai
T/A see Isg DO the man T/A barter hither
i too motokaa
DO sggenIIsg car
‘I saw the man who bought your car’

With DO of an experience verb:

(232) I tuutaki a ia ki te tamaiti i moohio
T/A meet pers IIIsg to the child T/A know
a Rewi
pers Rewi
‘He met the child that Rewi knew’

With DO promoted via passive:

(233) I hoko mai ia i te whare i
T/A barter hither IIIsg DO the house T/A
hanga·a e Hata
build·pass. by Hata
‘He bought the house that was built by Hata’

With DO promoted via actor-emphatic:

(234) I hoko mai ia i te whare naa Hata
T/A barter hither IIIsg DO the house actgen Hata
i hanga
T/A build
‘He bought the house that Hata built’

Deletion is also the normal treatment of non-locative genitives, although if they have
personal referents, they may also undergo pronoun replacement:
(235) I kaute ia i ngaa tamariki kua eke
T/A count IIIsg DO the(pl) children T/A reach
ngaa tau ki te tekau maa tahi
the(pl) year to the ten and one
‘He counted the children whose age had reached 11’
(236) He aha te ingoa o te waahi e waru
cls what the name gen the place num eight
maero te tawhiti atu i Te Araroa?
mile the distance away from Te Araroa
‘What is the name of the place whose distance from Te Araroa is eight miles?’

(Note, however, that not all genitives readily relativize, see Bauer, 1982, 328.)
Deletion is sometimes used for notional IOs (see the example in 1.1.2.3.4.2.3). In
addition, if ai and the deictics are subordinators, rather than pro-forms, all the examples
discussed in 1.1.2.3.4.2.3 would be instances of deletion of the relativized element, which
would then be the norm for Maori.
 Maori 58

1.2.2.3.5 Position of the pro-form for the relativized element

1.1.2.3.5.1 Same position

In all cases of replacement by a personal pronoun (see 1.1.2.3.4.2.1), the pronoun retains
the position in the clause normal for that type of NP. It must, however, be pointed out that
genitives follow their head when full NPs, but precede when pronominal. The pronominal
copy of a genitive thus appears on the opposite side of the head noun, cf.
(237) ngaa tau o ngaa tamariki
the(pl) year gen the(pl) children
‘the age of the children’
(238) oo raatou tau
plgenIIIpl year
‘their age’

1.1.2.3.5.2 Adjacent to the head noun

No pro-element in Maori occupies this position.

1.1.2.3.5.3 Moved to post-verbal position

In all cases where ai or a deictic particle appears in an adjective clause, it occupies the
position immediately following the verb. (This fact is used as an argument against the
pro-form analysis in Bauer, 1982).

1.1.2.3.6 Headless relative clauses

I have argued (Bauer, 1991) that emphatic ko structures (see 1.11.2.1.4) as opposed to
topicalizing ko structures (see 1.12) should be analyzed as involving a headless relative
clause as the subject in an equative predication. This analysis must still be regarded as
controversial. The reader is referred to 1.11.2.1.4 and the original argument, as it would
be out of place to repeat the arguments here.
Maori also uses possessive-relative clauses in contexts where the head consists of a
determiner which in the plural has no overt form, and which thus approach headless
status, although I do not believe this is the most appropriate analysis. This construction is
used in certain types of indirect statement (see 1.1.2.2.3) when a DO is involved, and is
also an alternative to the headless relative clause discussed in the previous paragraph in
emphatic contexts (see 1.11.2.1.4). These differ from the type of construction noted by
Clark (1976, 62–3) for languages like Hawaiian in that the Maori examples involve a
possessive, while the Hawaiian ones do not. The Hawaiian type does not occur in Maori.
Structures like ‘He said who could go’ which are headless relative clauses in English
are readily enough translated into Maori, but the Maori equivalents are interrogative, not
relative clause structures (see 1.1.2.2.3). Ahakoa ‘although’ is also used in some instances
where English has a headless relative, eg.
(239) Hoatu ahakoa maa wai e tae mai
give although intgen who T/A arrive hither
 Syntax 59

‘Give it to whoever comes’

(My consultant preferred the other order: Ahakoa maa wai e tae mai hoatu. Note also the
use of the actor-emphatic here with an intransitive verb, and no ai. I have no explanation
to offer for this.)

1.1.2.3.7 Elements that can be relativized

All noun phrases in the sentence can be relativized in at least some instances. Examples
of subjects, direct objects of experience verbs, direct objects of canonical transitives
(recall that these are most commonly relativized by promotion to subject, and note that in
some respects they are less readily relativized than oblique NPs), notional IOs, oblique
NPs, locative and non-locative genitives, and the object of comparison have been given
in 1.1.2.3.4 above.
There are some restrictions on certain subcategories, the main ones being as follows:
1. The e-phrase of the passive is generally not relativized, though the following
example is attested:
(240) I te poo ka puta mai te keehua i
at the night T/A appear hither the ghost T/A
kuku·a ai ia
haunt·pass. part. IIIsg
‘At night, the ghost he was haunted by appeared’

2. Many non-locative genitives cannot be relativized. It appears that such genitives can
only readily be relativized if they occur in stative, intransitive or non-verbal sentences.
Thus the following, in a transitive sentence, is unacceptable:
(241) *Ka maatakitaki a Maarama i te tamaiti i
T/A gaze pers Marama DO the child T/A
ngau (ai) te hooiho i a Rewi
bite part. the horse DO pers Rewi
‘Marama gazed at the child whose horse had bitten Rewi’

3. Only human objects of comparison can be relativized, a consequence of the

requirement that the pronoun-retaining strategy be used.
It is also possible to relativize on the subjects of embedded sentences, eg
(242) Ko teehea te hira i pooheehee a
top. which(sg) the shield T/A think mistakenly pers
Hata ka wini i Te Kaha?
Hata T/A be won cause Te Kaha
‘Which shield did Hata mistakenly think would be won by Te
Kaha?’ (TR2, 161)

However, it does not appear possible to relativize on other constituents of embedded

sentences.
 Maori 60

1.1.2.3.8 Movement of other items with the relativized element

The only potential “moved” element is ai and the deictics which alternate with it. No
other elements accompany them to post-verbal position.

1.1.2.3.9 Non-finite relative clauses

There is one construction which answers this description, namely an infinitival
construction which may be introduced by hei, although this is frequently elided. It is used
for clauses with future time reference, eg.
(243) Kei te piirangi·tia he tangata (hei) tiaki i te
T/A want·pass. a man for care DO the
whare
house
‘A person to look after the house is wanted’

The construction is characterized by the following:

– there is no T/A/M marker, although hei, when present, has a future reference
– the active/passive distinction is not made
– the subject is always omitted
– adverbials take their normal main-clause form
– main clause word order is maintained.
This same construction is sometimes used when there is no future implication, but in such
cases, it implies a habitual connection between the head and the description, and
corresponds to a compound, eg.
(244a) Ko te tangata horoi i te motokaa taku
eq the man clean DO the car sggenIsg
matua
parent
‘The car-cleaner is my father’

cf.
(244b) Ko te tangata horoi motokaa taku matua
eq the man clean car sggenIsg parent
‘The car cleaner is my father’

In this context, the second construction is fully productive, while the first is not; the first
may well be a recent innovation. In examples like these, hei is impossible.
No other non-finite constructions are possible in adjective clauses.
 Syntax 61

1.1.2.4 Adverb Clauses

1.1.2.4.1 General marking of adverb clauses

In the following sections, various different semantic relationships expressed in adverb
clauses are treated. There appear to be very few generalizations which hold true of all
types. However, there are only two positions for adverb clauses in relation to main
clauses: preceding and following. Most types of clause show a preference for one
position over the other, and in some cases, only one position is possible. In some
semantic areas, there is a conjunction or conjunctions specific to that type, but this is not
always the case.

1.1.2.4.2 Particular types of adverb clause

1.1.2.4.2.1 Time clauses

1.1.2.4.2.1.1 Past time clauses

One of the commonest constructions for past time adverb clauses is a nominalization, eg.
(245) I toona tae·nga mai, ka mahi·a e
at sggenIIIsg arrive·nom hither T/A make·pass. by
ahau he kapu tii
Isg a cup tea
‘When he arrived, I made a cup of tea’

(For the characteristics of such nominalizations, see 1.1.2.2.6.) The nominalization can
also follow the main clause. The i which introduces such nominalizations is frequently
elided, eg.
(246) Te tae·nga atu o ngaa tamariki, ka
the arrive·nom away of the(pl) children T/A
whakatika ia
stand IIIsg
‘When the children came, he got up at once’

It is also very common to find that an English ‘when’ clause is translated in Maori by a
coordinate main clause, indicating a succession of events eg.
(247) Ka mutu, ka tomo ki te whare
T/A finished T/A enter to the house
‘When [that] was over, [they] entered the house’ (TWh, 20)

This is particularly common in narrative.

There are several less common alternatives, some of which convey more specific time
information.
 Maori 62

A finite construction with ana following the verb can be used to indicate ‘as soon as’,
eg.
(248) Tae (kau) mai ana ia, ka mahi·a e
arrive as soon as hither T/A IIIsg T/A make·pass. by
ahau he kapu tii
Isg a cup tea
‘As soon as he arrived, I made a cup of tea’

Such clauses normally precede, although it is possible for them to follow the main clause.
A nominalization introduced by noo can be used to translate ‘when’ clauses, but it
brings with it a sense of cause, eg.
(249) Noo toona tae·nga mai, ka mahi·a e
actgen sggenIIIsg arrive·nom hither T/A make·pass. by
ahau he kapu tii
Isg a cup tea
‘When (and because) he arrived, I made a cup of tea’

Such noo nominalizations are apparently restricted to initial position. Their

characteristics are as for i nominalizations.
A further construction, which appears to be non-finite, is also found in narratives. It
consists of the verb stem with no T/A marker. However, this construction may be
accounted for by ellipsis of ka. It certainly appears true that ka can always be inserted,
and no other modifications to normal clause structure are made. Doubt, then, remains
about its finiteness or otherwise. The subject is frequently omitted, since it is often
predictable from context, but may be included in its usual position, following the verb,
eg.
(250) Tae noa ki te peka·nga, peka tonu hoki ia
arrive exactly to the turn·nom turn still also IIIsg
‘When [they] reached the turning, he turned also’ (TWh, 19)
(251) Tae moata atu, e moe tonu ana ngaa
arrive early away T/A sleep still T/A the(pl)
tamariki
children
‘When [we] arrived early, the children were still sleeping’

It is also possible to use a finite clause introduced by ka, with raa hoki ‘there also’
modifying the verb, eg.
(252) Ka tae mai raa hoki ia, ka mahi·a
T/A arrive hither dist also IIIsg T/A make·pass.
e ahau he kapu tii
by Isg a cup tea
‘When he arrived, I made a cup of tea’

This structure appears to be the least common, and always precedes the main clause.
 Syntax 63

All the examples above show punctual past time clauses. Other constructions are used
to express non-punctual events. Three constructions are possible for durative events,
which involve the T/A marker e…ana or kei te/i te, eg.
(253) I a au e kaukau ana, ka tangi mai
at pers Isg T/A swim T/A T/A sound hither
te pere o te tatau
the bell gen the door
‘While I was having a bath, the doorbell rang’

Alternatively, the time clause in (253) may be replaced by

(253a) E kaukau tonu ana au…
T/A swim still T/A Isg
‘While I was having a bath…’

The third construction of this kind involves merely the juxtaposition of two main clauses,
ie. coordination, eg.
(254) Kei te kaakahu ngaa waahine raa, kei te titiro
T/A clothe the(pl) women dist T/A look
whakatau mai ki a ia
intently hither to pers IIIsg
‘While the women were putting on their clothes, they gazed intently at him’ (TWh, 19)

In this case, it is possible to change the order of the clauses (with adjustment of the
pronominalization) without a change in meaning.
To express the simultaneity of past events, the commonest construction uses a clause
introduced by Kaatahi…ka ‘just when’. Kaatahi occurs in sentence initial position,
followed usually by the raised subject, and ka precedes the verb, eg.
(255) Kaatahi tonu taku ringa ka mau ki te
just when exactly sggenIsg hand T/A seize to the
waea, ka tangi hoki te pere o te tatau
phone T/A sound also the bell gen the door
‘As I picked up the phone, the doorbell rang’

The kaatahi clause must come first Alternatively, the construction with ana ‘as soon as’,
illustrated above, can serve this purpose.

1.1.2.4.2.1.2 Non-past time clauses

Three constructions are commonly used for non-past time clauses. However, in contrast
to past time clauses, the nominalization is not commoner than the other patterns. The first
construction uses the subjunctive kia, and in some instances, ai is required in the main
clause following the verb, eg.
(256) Kia tae mai koe, ka kai (ai) taatou
subj arrive hither IIsg T/A eat part. Iplincl
‘When you get here, well have a meal’
 Maori 64

It is not entirely clear what the function of ai is in this construction, but it is normally
used when the T/A marker of the main clause is ka. Thus (256) is judged preferable with
ai. It may serve to mark that the action of this clause is subsequent to (or dependent upon)
that of the kia clause. Ai is not required if the main clause is actor-emphatic, but this may
be because ai is a post-verbal particle, and the actor-emphatic construction is probably a
non-verbal sentence type (although Waite has recently argued otherwise, see 1.11.2.1.7):
(257) Kia tae mai te moni, maa·ku e hoko
subj arrive hither the money intgen·Isg T/A barter
mai (*ai) he koha maa·u
hither part. a gift intgen·IIsg
‘When the money arrives, I’ll buy you a present’

Kia time clauses always precede the main clause.

Alternatively, ina(a) (speakers vary as to the length of the final vowel) can be used in
place of kia. The appropriate analysis of ina is unclear. Biggs (1969, 35) suggests that it
is a T/A marker restricted to future time subordinate clauses, and suggests it has the force
‘if and when’. Williams’s Dictionary classifies it as a conjunction, which accords with
Biggs’s gloss. It is glossed as a conjunction here. See also 2.1.3.2.1. Ai is not required
with ina, but is preferred if the ina clause is in initial position. Like kia clauses, ina
clauses prefer first position, but ina clauses may occur in final position, eg.
(258) Maa·ku e hoko mai he koha maa·u
intgen·Isg T/A barter hither a gift intgen·IIsg
inaa tae mai te moni
when arrive hither the money
‘I’ll buy you a present when the money arrives’
(259) Inaa tae mai koe, ka kai ai taatou
when arrive hither IIsg T/A eat part. Iplincl
‘When you get here we’ll have a meal’

Future time nominalizations are introduced by hei. This hei appears to be prepositional.
These nominalizations follow the pattern for other prepositional types, rather than that of
the hei nominalizations discussed in 1.1.2.2.6. Ai is again preferred, though not
obligatory, in main clauses with ka (but not with the actoremphatic). Hei nominalizations
always occur in initial position, eg.
(260) Hei toou tae·nga mai, ka kai ai
at(fut) sggenIIsg arrive·nom hither T/A eat part.
taatou
Iplincl
‘When you get here, we’ll have a meal’

1.1.2.4.2.1.3 Habitual time clauses

It is necessary to treat these separately in Maori, since the possibilities do not seem to be
predictable from those for non-habitual clauses.
Kia may be used, and ai is normally required in the main clause if the T/A marker is
ka, eg.
 Syntax 65

(261) Kia maranga ake au, ka whakapuare ai au

subj arise away Isg T/A open part. Isg
i te matapihi o tooku ruuma moe
DO the window gen sggenIsg room sleep
‘When I get up, I always open my bedroom window’
(262) Kia hauhake·tia ngaa riiwai, ka wehe·a
subj harvest·pass. the(pl) potato T/A separate·pass.
ai ngaa purapura
part. the(pl) seed
‘When potatoes are harvested, the seeds are set aside’

Alternatively, ana may be used, implying ‘as soon as’. In this case, the main clause may
be introduced by ka, as above, or may also be marked with (kau) ana:
(263) Maranga ake ana au, whakapuare kau ana
arise away T/A Isg open as soon as T/A
au i te matapihi
Isg DO the window
‘When I get up, I open the window’

It is sometimes possible to use a nominalization introduced by i, eg.

(264) Wehe·a ai ngaa purapura, i te
separate·pass. habit. the(pl) seed at the
hauhake·tanga i ngaa riiwai
harvest·nom DO the(pl) potato
‘The seeds are set aside when potatoes are harvested’

In this construction, the main clause ai appears to function as a habitual marker, hence its
gloss. As illustrated, the nominalization may be final, though initial position is the norm.
If the sense is ‘whenever’, then ia ‘each’ is used to introduce the nominalization, eg.
(265) Ia haere·nga oo·ku ki Tuuranga, he tikanga
each move·nom gen·IIsg to Gisborne a custom
tonu naa·ku te mau puutete putiputi maa
still actgen·Isg the carry bunch flower intgen
too·ku whaea
sggen·Isg mother
‘Whenever I go to Gisborne, I (always) take a bunch of flowers for my mother’

These nominalizations with ia are the only adverb clauses encountered medially, but are
usually judged awkward in that position, eg.
(266) Whakapuare ai au, ia maranga·tanga ooku, i
open habit. Isg each arise·nom genIsg DO
te matapihi
the window
‘I open the window whenever I get up’
 Maori 66

Such notions are frequently expressed merely by the juxtaposition of two clauses, ie. by
coordination, eg.
(267) Ka mahi te iwi i te kai, ka kawe maa
T/A work the tribe DO the food T/A bring intgen
Poroumaataa i te hinu, i te ika, me eeraa
Poroumata DO the game DO the fish with those
atu kai katoa
away food all
‘When the tribe procured food, they brought for Poroumata game, fish, and all other kinds of
food’ (TWh, 17)

1.1.2.4.2.2 Manner clauses

The commonest construction for manner clauses uses the introductory phrase me te mea
(lit. ‘with (?if) the thing’) ‘as though’, which is sometimes strengthened by tonu ‘indeed’
and often followed by nei ‘here’. The clause which follows this subordinator does not
differ in its structure from main clauses. Manner clauses always follow the main clause:
(268) Ka haere raatou ki waho o te whare herehere
T/A move IIIpl to outside gen the house tie
me te mea nei, e haere ana ki te karakia
as though proxl T/A move T/A to the church
‘They left the prison as if they were going to church’
(269) E tohu ana ia i mua o ngaa
T/A sign T/A IIIsg at front gen the(pl)
matua raa, me te mea tonu ka whatiwhati
battalion dist as though indeed T/A break
te taiaha i roto i ngaa ringa
the taiaha at inside at the(pl) hand
‘He was gesticulating before the battalions as if the taiaha would break in pieces in his
hands’ (TWh, 22)

If the subject of the subordinate clause is identical to that of the main clause, the manner
clause may take the form of a non-finite ki te construction, as in the following example:
(270) I whakawhiti atu a ia me te mea nei ki
T/A cross away pers IIIsg as though proxI to
te poohi i tana reta
the post DO sggenIIIsg letter
‘He crossed over as if to post a letter’

It must be noted, however, that if it is possible to avoid a clause by the use of some
nominal expression, this is done. For instance
(271) Ka haere ia i te huarahi me te kaha ki
T/A move IIIsg at the path with the strong to
te whiowhio haere
the whistle move
‘He walked down the road whistling loudly’
 Syntax 67

1.1.2.4.2.3 Purpose clauses

Clauses expressing purpose are introduced by kia, the subjunctive marker. Ai is inserted
post-verbally in the vast majority of such clauses, although if the verb in the subordinate
clause is kite ‘see’, ai is usually omitted, and it is sometimes omitted with passives and
neuter verbs.
(272) I whakatangi ia i tana puutoorino kia
T/A play IIIsg DO sggenIIIsg flute subj
kite·a ai ia
discover·pass. part. IIIsg
‘He played his flute so that he would be found’
(273) I tere tana haere kia kore ai ia e
T/A fast sggenIIIsg move subj neg part. IIIsg T/A
tae toomuri
arrive late
‘She walked quickly so that she wouldn’t be late’

It is possible for purpose clauses to precede the main clause, but this requires the
insertion of a second ai following the main clause verb, except when the main clause T/A
marker is ka or e…ana, eg.
(274) Kia maarama ai tana kite mai i ahau,
subj clear part. sggenIIIsg see hither DO Isg
e mau ai au i te koti whero
T/A carry part. Isg DO the coat red
‘So that she will see me easily, I am wearing a red coat’

Purpose clauses are frequently non-finite, introduced by ki te if the subject of both

clauses is the same. The subject in the purpose clause is obligatorily deleted. (The
restrictions on verb types discussed in 1.1.2.2.2.4 appear to hold true here also.) Eg.
(275) E haere ana ahau ki te taaone ki te hoko
T/A move T/A Isg to the town to the barter
mai i teetahi koha
hither DO a(sp) gift
‘I am going to town to buy a present’

Ki te constructions occur only in final position.

A further possibility, dependent again upon like subjects, is the verb stem, non-
introduced, followed by ai. This is most commonly used for intransitive verbs, though not
exclusively, eg.
(276) Ka tau ia ki te tuarongo moe ai
T/A betake IIIsg to the inner end of house sleep part.
‘She betook herself to the inner end of the house to sleep’ (TWh, 20)
(277) Ka haere ia ki te kiitini tao ai i
T/A move IIIsg to the kitchen cook part. DO
ngaa kai
the(pl) food
‘She went to the kitchen to cook the food’
 Maori 68

1.1.2.4.2.4 Cause clauses

The chief construction for clauses expressing cause involves the use of i te mea (lit ‘from
the thing’), ‘because’ as the subordinating conjunction. The cause clause, which occurs in
final position, is otherwise structured as a main clause, eg.
(278) I hoki maatou ki te kaainga, i te mea e
T/A return Iplexcl to the home because T/A
ua ana
rain T/A
‘We returned home because it was raining’
(279) E puru ana au i taku koti mahana i te mea
T/A put T/A Isg DO sggenIsg coat warm because
kei te tuungaawiri au
T/A shiver Isg
‘I am putting on my warm coat because I am shivering’
(280) I toomuri maatou, i te mea i puni te
T/A late Iplexcl because T/A blocked the
huarahi i ngaa horonga
road from the(pl) slip
‘We were late because the road was blocked by slips’

It is also possible in the past tense to use a nominalization introduced by i ‘from’,

provided the verb has a subject. In this case, ai appears after the nominalization. This
construction is always in final position, eg.
(281) I toomuri maatou i te puni·nga ai o
T/A late Iplexcl from the block·nom part. gen
te huarahi i ngaa horonga
the road from the(pl) slip
‘We were late because of the blocking of the road by slips’

Again, constructions involving nominal expressions seem to be preferred wherever

possible to these clausal forms of expression. Thus (278) was elicited only after a specific
request for a verbal construction if possible. The first form provided was
(282) Naa te ua maatou i hoki ai ki te
actgen the rain Iplexcl T/A return part. to the
kaainga
home
‘We returned home because of the rain’

1.1.2.4.2.5 Condition clauses

Condition clauses are frequently introduced by mehemea ‘if’. (Note that, although
normally written as one word, this appears to be morphologically transparent: me he mea,
lit. ‘as though a thing’.) Mehemea clauses can be in initial or final position, but prefer
initial position. If the clause refers to a condition on a past event, the clause takes the
form mehemea+fronted subject+T/A marker+verb+… If the clause refers to a condition
on a non-past event (including eternal truths), the clause can take the form (mehemea+)ki
 Syntax 69

te+verb+ subject+… The fronting of the subject is not obligatory, though the norm. There
does not appear to be a realis/irrealis distinction to judge from the data I elicited from my
consultants. Williams (1862, 37–8) suggests that mehemea+clause is used for
counterfactuals, and ki te mea for uncertainty. He associates ki te with future time. Biggs
(1969, 122) associates ki te with future time and uncertainty. The following examples
will illustrate the problem, when compared with the examples in Biggs and Williams:
(283) Mehemea ahau i haere, kua mate au
if Isg T/A move T/A dead Isg
‘If I had gone, I would have been killed’
(284) Mehemea i tae mai ia, ka kite au
if T/A arrive hither IIIsg T/A see Isg
‘If he had come, I would have seen [him]’
(285) Mehemea naa·na i tahae ngaa moni,
if actgen·IIIsg T/A steal the(pl) money
tamaiti kino ia
child bad IIIsg
‘If he stole the money, he’s a naughty child’
(286) Mehemea ka rui·a ngaa kakano i te waa
if T/A plant·pass. the(pl) seed at the time
nei, ka tupu
proxI T/A grow
‘If you plant the seeds now, they will grow’
(287) Mehemea ki te haere atu koe maa teenaa
if to the move away IIsg intgen that
ara, ka tae atu koe ki te awa
path T/A arrive away IIsg to the river
‘If you walk along that track, you will reach the river’
(288) Mehemea koe ki te whakawera i te wai,
if IIsg to the heat DO the water
ka puta te korohuu
T/A appear the steam
‘If you heat water, steam will appear’

My consultant provided the same Maori translation for the following:

(289) Meina naa·na i mau te wati nei, ka
if actgen·IIIsg T/A take the watch proxI T/A
kite·a i tana kaainga
see·pass. at sggenIIIsg home
‘If he took the watch, it will be at his house’
‘If he had taken the watch, it would have been at his house’

The difference between the two English sentences appears not readily capturable in
Maori.
Other conjunctions are used in some dialects instead of mehemea, eg. me, ki he mea,
meina. In future examples, ina can also be used (cf. 1.1.2.4.2.1.2), eg.
 Maori 70

(290) Ina tahuri mai ia, kii·ngia atu kia tatari

if turn hither IIIsg say·pass. away subj wait
‘If he calls in, tell him to wait’

1.1.2.4.2.6 Result clauses

Clauses expressing result are introduced by naa reira, noo reira ‘by/from/belonging to
that aforementioned place’, which appear to be interchangeable variants in most
instances, or naa wai ‘by/from/belonging to whom’. Naa wai does not appear to have
personal associations, but has a time connotation not found in other uses of wai: it
implies a prolonged cause. The clauses under these conjunctions are most commonly
introduced by the T/A marker ka. If they are introduced by i, ai is inserted post-verbally.
The subjects of these clauses may be fronted to a position following the conjunction.
Such clauses occur in final position only.
(291) I moe ia i runga i te papa raakau
T/A sleep IIIsg at top at the floor tree
naa wai ka mamae tona tuaraa
actgen which T/A hurt sggenIIIsg back
‘He slept on the bare boards and so he got a sore back’
(292) He pai tonu noo teeraa tangata, aa, he
cls good indeed actgen that man and cls
whakaaro nui hoki noo·na ki ngaa mahi hei
thought big also actgen·IIIsg to the(pl) deed for
pai·nga moo te motu naa reira ia ka
good·nom intgen the island actgen there IIIsg T/A
whakaari·a hei mema paaremata
nominate·pass. for member parliament
‘Because he was such a good man, and had done so much for the people, they nominated
him for Parliament’
(293) I noho ia i runga i te maunga i te
T/A stay IIIsg at top at the mountain at the
poo, naa reira i hemo ai i te
night actgen there T/A die part. from the
makariri
cold
‘He spent the night on the mountain, and so he died of cold/exposure’

Alternatively, result clauses may be expressed by aa ‘and’, followed by a clause without

T/A marker, with tonu ‘indeed’ following the verb. Thus, instead of the naa reira clause
in the last example, it is possible to substitute:
(293a) …aa hemo tonu atu i te makariri
and die indeed away from the cold
‘And died of cold’

1.1.2.4.2.7 Clauses of degree

These are most frequently expressed in the form of relative clauses.
 Syntax 71

1.1.2.4.2.7.1 Comparative clauses

These take the form of relative clauses constructed using the possessive relative
construction (see 1.1.2.3.4.2.3). The head NP is introduced by the preposition i which
normally introduces comparative phrases, eg.
(294) He iti noa ake te whakaahua i taaku
cls small quite up the portrait compar sggenIsg
i whakaaro ai
T/A think part.
‘The portrait was smaller than I had imagined’
(295) He tere noa ake te motokaa i tana
cls fast quite up the car compar sggenIIIsg
i koorero mai ai ki ahau
T/A say hither part. to Isg
‘The car is much faster than he told me’

However, there is a tendency to avoid this construction if possible, and use a nominal
expression, eg.
(296) He pai kee au ki te raranga kete i te
cls good contr Isg to the plait kit compar the
whatu korowai
weave cloak
‘I make baskets better than I make cloaks’ (more lit. ‘I am better at basket-making than
cloak-weaving’)

1.1.2.4.2.7.2 Equative clauses

Equative clauses are introduced by kia, the subjunctive marker, if they refer to a not-yet-
realized event. The likeness is usually indicated by peeraa ‘like that’, although in some
circumstances rite ‘like’ may be used. The remainder of the clause may take the form of a
nominalization, but most frequently is a relative clause. The following examples
illustrate:
(297a) Tapa·hia ooku makawe kia pee·raa i
cut·pass. plgenIsg hair subj like·dist compar
too tapa·hanga i oo·na
sggenIIsg cut·nom DO plgen·IIIsg
‘Cut my hair just like you (will) cut hers’
(297b) Tapa·hia oo·ku makawe kia rite ki too
cut·pass. plgen·Isg hair subj like to sggenIIsg
tapa·hanga i oo·na
cut·nom DO plgen·IIIsg
‘Cut my hair just like you (will) cut hers’
(298) Maa·ku e hanga he tuuru kia pee·raa i
intgen·Isg T/A build a stool subj like·dist compar
taa Himi i hanga ai moo tana
sggen Jim T/A build part. intgen sggenIIIsg
tamaahine
 Maori 72

daughter
‘I’ll make a chair just like Jim made his daughter’

If the clause refers to a realized event, then kia is not used. Peeraa is non-introduced, but
rite is preceded by i in negative contexts, and may be preceded by he in positive contexts.
The function of i is not entirely clear, but it seems likely that it is the T/A marker. A
relative clause or a non-verbal construction follows the comparative particle. The head of
the relative clause is a noun stipulating the basis of comparison, eg.
(299) Kaaore aa raatou neera i rite te roa ki
neg plgenIIIpl nail T/A like the long to
aaku e whai ana
plgenIsg T/A chase T/A
‘They don’t have nails as long as I’m after’
(300) Kua kite anoo koe i te taraute pee·raa te
T/A see again IIsg DO the trout like·dist the
nui i taana e koorero raa?
big compar sggenIIIsg T/A speak dist
‘Have you ever seen a trout as big as he described?’
(301) He neera aa raatou (he) rite tonu ki aa taatou
cls nail plgenIIIpl cls like indeed to plgenIplincl
te roa
the long
‘They had nails as long as ours’

If it is possible to avoid a clause by using some nominal expression, this is done, eg.
(302) Pee·raa i toona matua te rite o
like·dist compar sggenIIIsg parent the like gen
tana heru i oona makawe
sggenIIIsg comb DO plgenIIIsg hair

‘He combed his hair as his father did’ (more lit. ‘His manner of combing his hair is like
his father’)

1.1.2.4.2.8 Concession clauses

Since these appear to differ from the other clause types discussed above, they are treated
separately here. They are introduced by ahakoa, ‘although’, which Biggs (1969, 70)
analyses as aha ‘what’+koa, a manner particle of intensity or emphasis. The remainder of
the clause has the same form as a main clause. They most commonly precede the main
clause, but may follow it.
(303) Ahakoa i tae toomuri mai koutou, hei
although T/A arrive late hither IIpl for
aha!
what
‘Although you arrived late, never mind!’
 Syntax 73

(304) Ahakoa kaatahi tonu a Hata ka moohio, tere

although just indeed pers Hata T/A know fast
tonu tana whakaae
indeed sggenIIIsg agree
‘Although Hata had only just found out, he quickly agreed’ (TR2, 72)
(305) He pai tonu te wawata ahakoa kaaore he
cls good still the day-dream although neg a
moni
money
‘It is quite pleasant day-dreaming, although there is no money’

1.1.2.4.2.9 Exploratory clauses

Maori has a further type of adverbial clause which I have tentatively called exploratory.
Semantically, such clauses have links with purpose clauses, but they are formally distinct
from the clauses in 1.1.2.4.2.3. They are translation equivalents of English ‘to see
whether…’, but do not involve indirect questions as the English does. They are
introduced by me kore ‘if neg’, and the T/A marker is e in all the examples I have found,
and i was rejected by my consultant even in the past-time example:
(306) Kia tere te oma, me kore koe e wini
subj fast the run if neg IIsg T/A win
‘Run fast, to see whether you will win’ (TR2, 104)

If the main clause is declarative, rather than imperative, ai is required post-verbally, eg.
(307) Haere mai ai raatou, me kore o raatou mate
move hither part. IIIpl if neg plgenIIIpl lack
e ora
T/A well
‘They come to see whether their ailments will be cured’

Note, however, that there is no ai in the actor-emphatic main clause in the following
example:
(308) Naa raatou i tahu te mea nei ki
actgen IIIpl T/A set on fire the thing proxI with
te maati, me kore e/*i toro
the match if neg T/A burn
‘They put a match to the substance to see whether it would burn’

1.1.2.4.3 Non-finite adverb clauses

(See the rider to the term non-finite in 1.1.2.2.6, and the discussion in 2.1.3.5.) Since non-
finite clauses are in many instances in complementary distribution with finite adverb
clauses, rather than alternatives to them, it has been necessary to include discussion of
them in 1.1.2.4.2. Various non-finite constructions occur in adverb clauses, but their
characteristics appear to be identical to those found when the non-finite constructions
 Maori 74

occur with other functions. These were described, with the exception of the ‘plain stem’
construction, in 1.1.2.2.6. For convenience, the distribution of non-finite types in adverb
clauses is summarized here:
Plain stem: past time clauses (1.1.2.4.2.1.1)
purpose clauses (1.1.2.4.2.3)
Nominalizations: past time clauses (1.1.2.4.2.1.1)
non-past time clauses (1.1.2.4.2.1.2)
habitual time clauses (1.1.2.4.2.1.3)
cause clauses (1.1.2.4.2.4)
equative clauses (1.1.2.4.2.7.2)
Ki te: manner clauses (1.1.2.4.2.2)
purpose clauses (1.1.2.4.2.3)
condition clauses (1.1.2.4.2.5)

Of these types, only the characteristics of the ‘plain stem’ construction require
specification here:
– all T/A/M categories are lost, and the active/passive distinction is not made in the verb
– the verb occurs in the stem form
– the subject may be omitted if it is coreferential with that in the main clause; the subject
may occur as an unmarked NP or as a possessive; whether these should count as the
same construction is unclear
– the other arguments that occur do so in main clause form and order
– in purpose clauses, ai is added post-verbally
– adverbials may be present, and occur in main clause form and position.
Some doubt remains as to whether this construction as it appears in past time clauses
involves ellipsis of ka. However, it appears that there is no motivation for ellipsis of a
T/A marker in purpose clauses.

1.1.2.5 Sequence of tenses

It is difficult to be sure whether Maori exhibits ‘sequence of tenses’, but there appear to
be a large number of restrictions on tenses in subordinate clauses, and an attempt is made
to specify these restrictions here. Some of them seem to depend on the type of verb in the
subordinate clause, and sometimes the tense appears to influence the possibilities for the
construction of clauses. It is extremely difficult to determine the facts in this area, and the
account below should not be taken as providing a complete picture. In many cases, the
main clause verb selects a particular tense reading of the subordinate T/A marker.

1.1.2.5.1 Sequence of tense in noun clauses

There are a number of restrictions which apply to noun clauses, and it is not always the
case that the same restrictions apply in all types of noun clause. In particular, the negative
and possibly the actor-emphatic, which may be analysed as containing noun clauses, have
their own set of restrictions, discussed in the detailed treatment of those constructions
 Syntax 75

(see 1.4 and 1.11.2.1.7). In noun clauses functioning as DO, or the subject of a passive,
eg. in an environment like

‘I know you will wash the dishes’, and other variants, a number of restrictions apply, see
Table 2.

1.1.2.5.2 Sequence of tense in relative clauses

Because of the complexity of relativization in Maori, and the rather unpredictable
interaction of tense with other factors, it is difficult to be certain that any account of this
area is accurate or comprehensive. However, a four-item test frame was used to check
patterns of acceptability. In all cases, the subject is relativized on, since it is clearly
easiest to relativize on subjects in Maori. Frame 1 has a passive verb in the relative
clause, Frame 2 has an active transitive verb, Frame 3 an active intransitive, and Frame 4
a stative intransitive. The sentences in (310) are the frames used:
Table 2: Sequence of tense in noun clauses
Subordinate Clause
Main kei te i te i e ka kua e…ana
Clause
↓
kei te √ √ pass. A-E Neg √ √ √
i te * √ √ A-E Neg √ pass. √
i √ √ A-E prob. A-E Neg √ pass. √
e (Neg only) √ √ √ * fut pass. fut
ka √ √ pass. A-E Neg fut pass. fut
kua √ √ pass. A-E Neg fut pass. √
e…ana √ √ A-E prob. A-E Neg fut pass. fut

Key to Table 2:
Pass.: with this combination, the verb in the subordinate clause must be in the passive form if it is
inherently transitive
A-E only possible if the subordinate clause is actor-emphatic in
Neg: form, or if the entire construction is negative
A-E if the verb in the subordinate clause is inherently transitive,
 Maori 76

prob.: the actor-emphatic construction is likely for the subordinate clause

Neg e is possible as the main clause marker only in certain
only.: negative constructions
fut: this combination is possible, but imposes a ‘future’ interpretation on the subordinate clause
marker
√: occurs without restriction
*: non-occurrent

In Frame 3, the two particles were sometimes required to complete the sense, sometimes
not required, but preferred, and sometimes not necessary at all, either singly or together.
Table 3 summarizes the findings. Such generalizations as can be made are noted here.
The two T/A markers most restricted in their use in relative clauses are kei te and i te, ie.
those Biggs calls (1969, 86) the “pseudo-verbal continuous” markers. Kei te was accepted
only with passive and active transitive relative clauses, and then only in some non-past
main clause contexts. I te was rejected for all but active transitives in some non-past main
clause contexts. I co-occurs with all main clause markers, but is not always acceptable
with active transitive relative clauses, and (somewhat surprisingly, I think) not always
with the intransitive frame. There appear to be no worthwhile generalizations about the
distribution of e. Ka has a tendency not to occur with active transitives, as does kua,
which also had a high rejection rate with intransitives. E…ana in most contexts is
possible with any type of subordinate clause verb except statives.
In addition, it must be noted that other positions for relativization may show more
restrictions than subjects, or different ones. In particular, it appears that it may be
possible to relativize directly on DOs in non-past contexts (for some discussion, see
Bauer, 1982, 317f).
 Syntax 77

1.1.2.5.3 Sequence of tense in adverbial clauses

In most instances, the tense marker of the adverbial clause, if there is one, is determined
by the construction. In several cases, the construction used for past events differs from
that for non-past. The restrictions found have been noted in the individual sections on
types of adverbial clause. It should also be borne in mind that possibilities for tense in
adverbial clauses appear to be subject to certain pragmatic limitations.
Table 3: Sequence of tense in relative clauses
Subordinate Clause
Main Clause kei te i te i e ka kua e…ana
kei te
1 * * √ * √f √ √
2 * * * √ * * √
3 * * * √ * * √
4 * * √ * √ √ *
i te
1 * * √ √ √f √ √
2 * * * * * * √
3 * * * √ * √ √
4 * * √ * √f √ *
i
1 * * √ AE √f √ √
2 * * √ AE * * √
3 * * * * √ * √
4 * * √ * √f √ *
e (Neg)
1 √ √ √ √f √ √
2 √ √ √ √ √f ? √
3 * * * √ * √ √
4 * * √ √ √f √ *
ka
1 √ ? √ √ √ √ √
2 √ √ √ * √ √ √
3 * * √ √ √ √ √
 Maori 78

4 * * √ √fc √ √ *?
kua
1 √ * √ √ √ √ √!
2 √ √ √ √ √ √ √
3 * * √ √ √ √ √
4 * * √ √fc √ √ √!
e…ana
1 √ * √ √f √f √ √
2 √ √ √ √ * * √
3 * * √ √ √im √ √
4 * * √ √ * √ *

Key
AE: this subordinate clause marker can only occur in an actoremphatic context; f:
imposes a ‘future’ reading on the marker; fc: imposes a reading of ‘future certainty’ on
the marker; im: imposes a reading ‘has just’; Neg: this main clause marker is only
possible in certain negative constructions; √: possible; *: impossible; ?: uncertainty about
acceptability; !: special restrictions, namely Frame 1: implies that the speaker has seen
the act; Frame 4 implies that the action of the subordinate clause is contemporaneous
with the time specified by kua.

1.2 STRUCTURAL QUESTIONS

1.2.1 Internal Structure of the Sentence

1.2.1.1 Copular Sentences

Copular sentences in Maori have the basic structure predicate-subject. There is never an
overt verbal form, and predicates are of four basic types: they are introduced by he, ko, a
preposition, or no special marker. Even ‘existence’ statements have no copula or
existence verb,e.g.
(311) Ko Ihoa
eq Jehovah
‘Jehovah exists’
(312) Kaahore he taniwha
neg a taniwha
‘There are no taniwhas’
(313) Aae, he taniwha
Yes a taniwha
 Syntax 79

‘Yes, there are taniwhas’

Note that the first and last of these are subjectless; this is less clearly the case for the
negative (see 1.4).

1.2.1.1.1 Copular sentences with nominal complement

1.2.1.1.1.1 Overt be-copula

Maori has no copulas.

1.2.1.1.1.2 Marking of the predicate noun

Depending on the semantic relation between the subject and predicate, the predicate noun
is marked with he or ko. Ko marks equative predicates, and he classifying or attributive
predicates. (Hei marks future classifications in some dialects.)
(314) He maahita ia
cls teacher IIIsg
‘He is a teacher’
(315) He tamaiti koretake ia
cls child lazy IIIsg
‘He is a lazy child’
(316) Ko te rooia teenei
eq the lawyer this
‘This is the lawyer’
(317) Ko Pani ahau
eq Pani Isg
‘I am Pani’

Possession sentences take he, eg.

(318) He hooiho toona
cls horse sggenIIIsg
‘He has a horse’

In those dialects which use hei as the marker for future classifications, there are examples
like
(319) Hei maahita ia
cls(fut) teacher IIIsg
‘He will become a teacher’

1.2.1.1.1.3 Order of Constituents

As illustrated above, the basic order is predicate-subject The subject is an unmarked NP.
It is also possible to front the subject, but this is a focussing or topicalizing construction,
 Maori 80

see 1.11, 1.12 and Bauer 1991. If the subject is definite, it is preceded by ko; if indefinite,
by nothing.
(320) Ko teenei he whare pai
top. this cls house good
‘This is a good house’

For many speakers, when the subject of an equative sentence is fronted, it appears that
the equative ko is omitted. Some of these speakers deny the grammaticality of examples
like
(320a) Ko teenei ko te rooia
top. this eq the lawyer
‘This is the lawyer’

and correct to
(320b) Ko teenei te rooia
eq? this the lawyer
‘This is the lawyer’

which appears to be an equative sentence with non-fronted subject. The problems with
the analysis of such sentences are discussed, though not resolved, in Bauer, 1991, 10–11.
See 1.13.2.5 for word order with complex predicates of this type.

1.2.1.1.2 Copular sentences with adjectival complement

1.2.1.1.2.1 Overt be-copula

Maori has no copulas.

1.2.1.1.2.2 Marking of the complement adjective

The predicate adjective is normally marked with he. He is omitted if tino ‘very’ modifies
the adjective. He is also replaced by emphatic te in examples of an exclamatory nature,
e.g.
(321) He whero teenei
cls red this
‘This is red’
(322) Tino pai te maahita
very good the teacher
‘The teacher is very kind’
(323) Teeraa tangata te moomona
that man the fat
‘That man is fat!’

(Note that (323) has marked constituent order.)

 Syntax 81

It must also be noted that other structures frequently correspond to sentences of

English with complement adjectives. Exclamatory noun phrases are common, eg.
(324) Te turituri o ngaa kaa!
the noise gen the(pl) car
‘Cars are noisy’

If tense, aspect, or modalities other than present are stressed, verbal structures are
sometimes resorted to. Compare
(325) Tino nui teeraa tangata i mua
very big that man at before
‘That man was very fat’

and
(326) Rere ana te mataku i te ruu o te
strong T/A the fright from the shake gen the
whenua
earth
‘The earthquake must have been very frightening’

1.2.1.1.2.3 Order of constituents

See 1.2.1.1.1.3. Examples of subject-fronting are
(327) (Ko) teenei he whero
top. this cls red
‘This is red’
(328) Ko teeraa tangata tino nui i mua
top. that man very big at before
‘That man was very fat’

See 1.13.2.5 for word order with complex predicates of this type.

1.2.1.1.3 Copular sentences with adverbial complement

1.2.1.1.3.1 Overt be-copula

Maori has no copulas.

1.2.1.1.3.2 Marking of the complement adverbial

The complement adverbial generally takes the form of a prepositionally marked
predicate. However, there is also a special class of deictic adverbs which occurs in such
sentences. The majority of the latter are time words, and many can be analysed as
containing a preposition, see 2.1.1.6.1. (332) illustrates this special category.
(329) Kei roto i te motokaa raa ia
at(pres) inside at the car dist IIIsg
 Maori 82

‘She is in that car’

(330) I te hui te Pirimia
at the meeting the PM
‘The Prime Minister was at the meeting’
(331) Noo Te Kaha a Tuu
actgen Te Kaha pers Tu
‘Tu belongs to Te Kaha’
(332) Aapoopoo te hui
tomorrow the meeting
‘The meeting is tomorrow’

Tense distinctions are encoded in the prepositions.

1.2.1.1.3.3 Order of constituents

See 1.2.1.1.1.2. Parallel examples are
(333) Ko te Pirimia i te hui
top. the PM at the meeting
‘The Prime Minister was at the meeting’
(334) Ko Tuu noo Te Kaha
top. Tu actgen Te Kaha
‘Tu belongs to Te Kaha’

Where the prepositional phrase is complex, involving a local noun as in (329), the subject
normally occurs following the local noun, especially if the subject is brief. Thus (335) is
a more natural order than (329):
(335) Kei roto ia i te motokaa raa
at(pres) inside IIIsg at the car dist
‘She is in that car’

See 1.13.2.5 for further details.

1.2.1.1.4 Copular sentences without overt be-copula

All copular sentences in Maori fall into this category.

1.2.1.1.4.1–2 Expression of verbal categories

In classifying sentences, the distinction realized vs. irrealis (or actual vs. anticipated: the
best characterization is arguable) is marked for some speakers by the distinction he/hei,
see the examples in 1.2.1.1.1.2. Other speakers use verbal alternatives for irrealis
sentences in this category, eg.
(336) Ka maahita pea ia
T/A teacher perhaps IIIsg
 Syntax 83

‘He might become a teacher’

In locative prepositional sentences, tense is marked by the preposition (cf. 2.1.1.5.1):

(The choice between ko/hei is dialectal.) This is also true for temporal location, although
in most contexts a replaces hei/ko as the future marker. However, present temporal
location requires verbal or adverbial specification. Compare (338), (339) with (340):
(338) I te Mane raa te hui
at(pt) the Monday dist the meeting
‘The meeting was on Monday’
(339) A te Mane te hui
at(fut) the Monday the meeting
‘The meeting will be on Monday’
(340) E hui ana
T/A meet T/A
‘The meeting is going on now’

In addition, the four possessive prepositions maa, naa, moo, noo encode an
actual/intended distinction (see further 1.10.5):
(341a) Noo Rewi te hooiho nei
actgen Rewi the horse proxI
‘This horse is Rewi’s’
(341b) Moo Rewi te hooiho nei
intgen Rewi the horse proxI
‘This horse is for Rewi’

Apart from this, verbal distinctions are not made in copular sentences. Inflections of the
voice and context determine the appropriate interpretation.

1.2.1.1.5 Omission of be-copula

Not applicable.
 Maori 84

1.2.1.1.6 Types of copula

There are no copulas, but there are different types of predicate, which correspond to
distinctions made in other languages by means of copulas. These have been exemplified
above.

1.2.1.1.6.1 Defining
It appears that this is the major function of he-predicates. However, they have additional
functions, and I have therefore called them ‘classifying’: they specify a class to which the
subject belongs.

1.2.1.1.6.2 Identity
ko-predicates have this function. (I have used the term ‘equative’ for this function.)

1.2.1.1.6.3 Role
He and hei predicates have this function.

1.2.1.1.6.4.1 Numerical
Predicates in such sentences have no specific marker, although if the number is 2–9, it is
preceded by e, or toko- if the things counted are people (for details see 2.1.6). (For some
speakers, toko- is obligatory for people, for others it is optional. Toko- forms may
optionally be preceded by e.)
(342) E whaa ngaa kurii
num four the(pl) dog
‘There are four dogs’ (More literally, The dogs are four [in number]’.)
(343a) E rua aa maaua tamariki
num two plgenIdlexcl children
‘We have two children’

or
(343b) (E) toko·rua aa maaua tamariki
num pnum·two plgenIdlexcl children
‘We have two children’
(344) Tekau maa tahi ngaa taangata purei hoka
ten and one the(pl) people play soccer
‘Eleven people play soccer’

1.2.1.1.6.4.2 Other
Sentences which are expressed in English with other copulas, such as ‘become’, ‘grow’
etc are expressed in Maori through verbal sentences, eg.
(345) Kei te koomaa rawa atu ia
 Syntax 85

T/A pale very away IIIsg

‘He is becoming paler’
(346) Kei te ngoikore ia
T/A weaken IIIsg
‘She is growing weaker’

1.2.1.2 Verbal sentences

1.2.1.2.1 Subjectless verbs and dummy subjects

There are no dummy subjects in Maori. There are a few constructions where the verb is
subjectless, at least on the surface, or can be, eg.
(347) E ua ana
T/A rain T/A
‘It is raining’

but te ua ‘the rain’ can occur as subject.

(348) E paki ana
T/A sunny T/A
‘It is sunny’

but compare
(349) Kei te pupuhi te hau
T/A blow the wind
‘It is windy’
(350) Ka roa, ka haere atu ia
T/A long T/A move away IIIsg
‘After a while, he left’

In addition, contextual deletion of the subject is extremely common, particularly in

narrative, eg.
(351) Ka haere taua taurekareka ki te tiki wai, ka
T/A move that servant to the fetch water T/A
utu i te tahaa ki te wai
dip DO the calabash to the water
‘The servant went to fetch water; [he] dipped the calabash into the water’ (H, 8)

A further interesting case of a superficially subjectless sentence type is the actor-

emphatic sentences with ‘cognate object’ verbs, see 1.11.2.1.7, examples (910) and (911).
There are also examples like the following, where the analysis is in doubt:
(352) E rua paringa o te tai i te raa
num two flowing gen the tide at the day
‘There are two tides a day’
 Maori 86

This may be a prepositional predicate (i te raa) with a fronted subject (the rest), but it is
possible that it is a numerical predicate without a subject. It is not clear what arguments
could resolve this.

1.2.1.2.2 Verbs without direct objects

There are several classes of verb in Maori which do not take direct objects.
Firstly, there are verbs which are intransitive, either active, eg. haere ‘move’, oma
‘run’, noho ‘stay’, or stative, eg. pai ‘good’, nui ‘big’, hee ‘wrong’. These verbs require
only one nominal argument,the subject, eg.
(353) Kei te oma raatou
T/A run IIIpl
‘They are running’
(354) Ka nui ngaa toenga miiti
T/A big the(pl) remnant meat
‘There’s plenty of left-over meat’

Such intransitives are distinguished from verbs which take DOs in a variety of ways
syntactically, eg. the associated imperative form, (see 1.1.1.3), the form of the possessive
in nominalizations (see 1.10.6).
Secondly, there is a class of verbs which have been variously referred to in the
literature as ‘participles’ (Williams, 1862, 48), ‘neuter verbs’ (Hooper, 1982), and ‘stative
verbs’ (Biggs, 1969). The middle term is adopted here, since I accept Hooper’s
arguments that the term ‘stative’ is semantically misleading. Like the intransitives, these
verbs require only one nominal argument, but that argument is normally the semantic
‘patient’ rather than ‘agent’. Examples of neuter verbs are mutu ‘finished’, mahue ‘be left
behind’ (and see 2.1.3.1.5 for further details). Their use is illustrated by sentences like
(355) Kua mutu te hui
T/A finished the meeting
‘The meeting is over’
(356) Kua mahue ia
T/A be left behind IIIsg
‘He was left behind’

Both intransitives and neuter verbs may be transitivized by the prefix whaka-, although
the result of this process has in some instances been lexicalized, and is therefore no
longer mere transitivization see 2.1.3.1.3.1.1. From haere ‘move’ comes whakahaere ‘to
cause to move’ =‘to run (eg. a meeting)’; from ora ‘well’ comes whakaora ‘to cure,
heal’; from mutu ‘finished’ comes whakamutu ‘leave off, cause to end’.
In addition, there are many verbs which may be used either transitively or
intransitively without a change in form, eg. kai ‘eat’, moe ‘sleep, sleep with’, whio
‘whistle’, waiata ‘sing’. Compare
(357) Kei te kai raaua
T/A eat IIIdl
‘They are eating’
 Syntax 87

and
(358) Kei te kai raaua i ngaa kuumara
T/A eat IIIdl DO the(pl) kumara
‘They are eating the kumaras’

Both transitive and intransitive constructions for the imperative occur with these verbs,
see 1.1.1.3.1.
There is a further construction in Maori which results in a sentence without a
grammatical DO, although the DO is semantically present, namely object incorporation,
see 2.1.3.6.12. If the DO in Maori is indefinite, it is normally incorporated into the verb,
to form a construction which is intransitive with respect to most syntactic criteria, but see
kai-agentives in 2.2.1. This formation of compound verbs is completely productive: any
transitive verb undergoes it, eg.
(359) E tuhituhi reta ana ia
T/A write letter T/A IIIsg
‘She is writing letters/a letter’ (lit ‘She is letter-writing’)

Note that the discontinuous T/A marker e…ana surrounds the verb and the semantic DO.
(There appears to be no associated imperative form for this construction.)
There is one further group of verbs in Maori which do not have canonical DOs,
although they have traditionally been regarded as transitive, and although they normally
have two nominal arguments. They are referred to here as ‘experience’ verbs, eg. moohio
‘know’, piirangi ‘want’, kite ‘see’, maumahara ‘remember’. Like verbs with canonical
DOs (called “canonical transitives” here), they take an unmarked subject NP, and a
second NP marked with either i (eg. kite ‘see’ (Exp), whaangai ‘feed’ (Canon.)) or ki
(moohio ‘know’ (Exp), karanga ‘call’ (Canon.)). However, the commonest marker for
experience verbs is ki, while the commonest marker for canonical transitives is i. These
second arguments differ in their grammatical patterning with respect to a number of
syntactic phenomena, eg. question-formation, relativization, imperativization,
reflexivization, and the two groups of verbs pattern differently with respect to the actor-
emphatic, nominalization, and complementation. (For a summary of these differences,
see Bauer, 1984). Whether these differences are ultimately best accounted for as a
transitivity cline phenomenon (see Hopper & Thompson, 1980), or as DO versus nonDO
is not an issue which can be tackled here. However, it is at least questionable whether
these verbs are rightfully regarded as having direct objects.

1.2.1.2.3 Indirect object

The category indirect object is not a distinct one in Maori. Notional indirect objects are
oblique NPs introduced by the preposition ki ‘to’, which is the normal goal marker in
Maori, eg.
(360) Homai ki a maatou toou waka
give to pers Iplexcl sggenIIsg canoe
‘Give us your canoe’
 Maori 88

(Homai, and hoatu are exceptional in not having a passive suffix, though in all other
respects this sentence is passive, ie. the normal form for a transitive imperative.)
Other oblique uses of ki appear in 1.2.1.2.4, and ki can also occur with canonical DOs,
and with the second nominal argument of experience verbs (see 1.2.1.2.2).
I know of no verbs in Maori with which a notional indirect object is obligatory.

1.2.1.2.4 Other arguments of verbs

There do not appear to be any clear criteria which distinguish arguments from adjuncts in
Maori (see eg. 2.1.3.1.1.1 for a discussion of what phrases can become subject in the
passive). Thus while DOs, for example, do not have the same syntactic properties as all
types of oblique NP, they share any given property with at least some other oblique
phrase type. However, the ability to undergo fronting may possibly serve as a criterion:
DOs and IOs cannot be fronted, while phrases expressing location, time, source, goal and
beneficiary, for example, can (see also 1.2.1.2.6). If this criterion is used, the passive
agent is an argument, and so is the bold phrase in the following, as it cannot be fronted
without change of sense:
Cause with neuter verbs
(361) Kua mahue ia i te pahi
T/A be left behind IIIsg cause the bus
‘He was left behind by the bus’

Two other types of phrases are not readily fronted, and thus must be regarded as possibly
arguments:
Instrument
(362) Kua poro·ngia te raakau e ia ki te toki
T/A fell·pass. the tree by IIIsg instr the axe
‘He has felled the tree with the axe’

Comitative
(363) Ka tae mai te taraka o Hata me ngaa
T/A arrive hither the truck gen Hata with the(pl)
teepu, me ngaa tuuru
table with the(pl) stool
‘Hata’s truck arrived with the tables and chairs’ (TR2, 86)

Certain types of goal are also marginal fronted, and may thus be arguments:
(364) Ka mau a Maama i a maatou ki te
T/A take pers Mummy DO pers Iplexcl to the
one
beach
‘Mummy is taking us to the beach’

While the vast majority of other phrases in Maori are prepositional, the neuter verbs
allow an unmarked phrase which is not the subject. Whether this is a nominal argument is
rather in doubt—the derivation of these sentences has been the cause of considerable
 Syntax 89

disagreement (see Hooper, 1982, 1984a; Waite, 1989). The construction in question is
illustrated in the following examples.
(365) Kua pau ngaa paraoa i ngaa kiore
T/A exhausted the(pl) bread cause the(pl) rat
te kai
the eat
‘The bread has been eaten up by the rats’
(366) Kua pau ngaa paraoa te kai e ngaa kiore
T/A exhausted the(pl) bread the eat by the(pl) rat
‘The bread has been eaten up by the rats’

1.2.1.2.5 Combinations of Su, DO, IO, and other arguments found

Essentially, all possible combinations can occur. However, in genuine text, it is unusual
to find more than three arguments/adjuncts present, although constructed examples which
break this limit are accepted readily enough, eg.
(367) I te Tuurei raa, ka tuhi au ki taku
at the Tuesday dist T/A write Isg to sggenIsg
teina me taku pene hou, i runga i te pepa
sister with sggenIsg pen new at top at the paper
kakariki, i roto i taku waahi mahi·nga
green at inside at sggenIsg place work·nom
‘On Tuesday, in my office, I wrote to my sister with my new pen, on a piece of green paper’
(368) I nuku·hia e au ngaa ahi wera mai i
T/A move·pass. by Isg the(pl) ash hot hither from
ngaa koohatu ki te pouaka ki toku hawera
the(pl) stone to the box instr sggenIsg shovel
‘I moved the hot ashes from the stones to the box with my shovel’

As arguments/adjuncts are frequently omitted when they can be understood from the
context, the number of arguments/adjuncts actually expressed is frequently very few.

1.2.1.2.6 Order of constituents

The basic word order for Maori is VSO, and obliques most normally follow the direct
object if there is one, although time adverbials seem to prefer sentence initial position.
However, a variety of other orders is also possible, some of these involving special
constructions.
Firstly, for topic or emphasis (see 1.11 and 1.12) definite subjects may be fronted
using the particle ko, and indefinites without special marking, eg.
(369) Ko Rewi kei te aawhina i a Pani
top. Rewi T/A help DO pers Pani
‘It is Rewi who is helping Pani’

Secondly, many oblique NPs may precede the verb, especially adverbials of time and
place (see 1.2.1.2.4). In some instances, this fronting is accompanied by the use of ai
 Maori 90

post-verbally, but this depends on the tense marker. The best generalization appears to be
that with the relative time markers, ka, kua, and e…ana, no ai appears, but with absolute
markers (i, e) ai is used, although there is considerable native speaker variation in the use
of this particle, eg.
(370) I taua poo, ka tuu te kanikani
at detaph night T/A stand the dance
‘That night, there was a dance’
(371) Noo te poo i tae mai ai
actgen the night T/A arrive hither part.
‘At night, [it] arrived’

(Time phrases with noo appear always to precede.)

With future time adverbials, ai is used where there is no tense marker. This quite
probably arises from the elision of e, which does not readily occur with main clause verbs
in modern Maori:
(372) A te wiki tuatahi o Tiihema, mutu ai
at(fut) the week first gen December finished part.
te kura
the school
‘School will finish in the first week of December’

An example with a place phrase fronted is

(373) I reira, ka tuu te tira ki te tina
at there T/A stand the travelling party to the dinner
‘There, the travelling party had lunch’ (TR2, 57)

Not infrequently, when a place phrase is fronted, the subject is also fronted, without
being topicalized:
(374) Kei raro te toa o Te Kaha e
at(pres) below the champion gen Te Kaha T/A
putu ana
spread out T/A
‘The Te Kaha champion is spread out on the ground’ (TR1, 152)
(cf. the ‘normal’ order: E putu ana te toa o Te Kaha kei raro.)

Other oblique NPs are also sporadically fronted, although it appears that the passive agent
and i causers with neuter verbs cannot be. The naa/maa/noo/moo set are also frequently
fronted, and often take the subject with them, eg.
(375) Naa te makariri raaua i hoki mai ai
actgen the cold IIIdl T/A return hither part.
‘They came back because of the cold’
(376) Moo te kuia teenaa ka tika
intgen the old woman that T/A right
‘That would be fine for an old lady’
 Syntax 91

It appears that it is not possible to front a direct object directly, ie.

(377) *I ngaa manu kei te whaangai ia
DO the(pl) bird T/A feed IIIsg
‘She is feeding the birds’

(The inclusion of the particle ai post-verbally does not improve the grammaticality of
this, either.) The DO can, of course, be fronted by first promoting it to subject via the
passive, eg.
(378) Ka whaangai·a ngaa manu (e ia)
T/A feed·pass. the(pl) bird by IIIsg
‘The birds were fed by her’
(379) Ko ngaa manu ka whaangai·a e ia
top. the(pl) bird T/A feed·pass. by IIIsg
‘It was the birds that were fed by her’

Given the relatively high frequency of passives in Maori texts, it is perhaps misleading in
one sense to imply that the DO is normally the third element in Maori sentence structure:
in the passive, the notional DO is second. This fact is also reinforced by the object-
incorporation of indefinite DOs, which also effectively puts them in second place.
It should also be mentioned that the actor-emphatic construction has the effect of
putting the actor in sentence initial position:
(380) Naa Rewi i whaangai ngaa manu
actgen Rewi T/A feed the(pl) bird
‘Rewi fed the birds’

This construction, together with ko-fronting of subjects ensures that, despite relatively
fixed word-order, subjects are not prevented from moving into other positions than the
post-verbal slot to which they are confined by VSO order.
The basic Su—O order can also be affected by ‘heavy constituent’ shift, see 1.13.2.4.
Heavy NP shift is also responsible in a few cases for DOs following other oblique NPs,
see 1.13.2.2.
It is not clear whether there are any hard and fast rules relating to the ordering of
oblique constituents amongst themselves, although the following constraints appear to be
followed in the vast majority of cases:
(i) In general, arguments semantically closer to the verb appear nearer the verb than more
peripheral arguments.
(ii) The passive agent phrase has great freedom, and can precede the grammatical subject
(iii) A source NP must precede a goal NP.
These points are illustrated in the following examples:
(381) I runga i te paruone ka mahi·a e ia he
at top at the dirt T/A make·pass. by IIIsg a
raina mai i te keeti ki te whare ki
line hither from the gate to the house instr
tana raakau
 Maori 92

sggenIIIsg stick
‘He drew a line in the dirt from the gate to the house with his stick’
(382) Kia kauwhau·tia atu iana, e ahau, ki a koe
subj recite·pass. away now by Isg to pers IIsg
‘Let me now recite [this] to you’ (H, 5)
(383) Otiraa, i moe·a anoo e Whakaue a
then T/A sleep·pass. again by Whakaue pers
Rangiuru
Rangiuru
‘Then Whakaue slept with Rangiuru again’ (H, 5)

1.2.1.3 Adverbials

1.2.1.3.1 Types of adverbials

1.2.1.3.1.1 Adverbs
Most of the single-word adverbs are a set of adverbial particles closely associated with
the verb and always in immediate post-verbal position. Their semantico-syntax is
extremely complex, and they often appear to function as aspectual particles. (For a
thorough exploration of one sub-set, with many insightful comments on others, see Mutu-
Grigg, 1982.)
These particles appear to fall into four groups:
manner directional deictic emphatic
eg. rawa, noa mai, atu nei, naa hoki, anoo
tonu, kee ake, iho raa anake ?ana
kau, pea

The manner particles passivize in agreement with passive verbs, e.g.

(384) I peehi·a rawa·tia ngaa waahine
T/A oppress·pass. intens·pass. the(pl) women
‘The women were severely oppressed’

The directional and deictic particles are limited to those listed, but the ana of the T/A
marker e…ana occupies the deictic slot, as does the post-verbal particle ai. The
characterization ‘emphatic’ is rather uncertain, as is the membership of this set. Some
examples of frequent combinations, only translatable in context are rawa atu, mai nei, nei
anoo, tonu hoki.
There are also a few adverbs which regularly precede the verb head, of which the
commonest are tino ‘very’, aata ‘carefully’, maatua ‘first’, aahua ‘somewhat’, eg.
(385) Ka aahua pukuriri a Tamahae ki a Rewi
T/A somewhat angry pers Tamahae to pers Rewi
 Syntax 93

‘Tamahae was somewhat angry with Rewi’ (TR2, 5)

There is also a group of time adverbs, including inaianei ‘now, immediately’, inanahi
‘yesterday’, aapoopoo ‘tomorrow’, aakuanei ‘later’. While these are always written as
one word, they are in many cases transparently prepositional phrases, eg.
(386) aa poo poo
at(fut) night night
‘tomorrow’

and in other cases, there is evidence that they have arisen from prepositional phrases.
This is also true of the interrogative adverbs aawhea, inaawhea, etc. (see
1.1.1.2.2.1.1.2.3).
In addition, other single word adverbs occur post-verbally and may be from any word
class (see 1.16), eg.
(387) Ka moe pai ia
T/A sleep good IIIsg
‘He slept soundly’
(388) E haere taki·tahi ana raatou
T/A move gp·one T/A IIIpl
‘They went singly’
(389) Ka noho manuhiri te tangata nei i Rurunui
T/A stay visitor the man proxI at Rurunui
‘This man remained as a guest at Rurunui’ (W, 197)

These are manner adverbials, and I have found no examples where they co-occur with the
manner particles described above. They appear to occupy the manner slot. I failed to
elicit examples with directional particles, but the ana of the e…ana T/A marker follows,
as do emphatic particles.

1.2.1.3.1.2 Prepositional phrases

Prepositional phrases (post-positions do not occur) are the commonest type of adverbial
in Maori (see also 2.1.1.5). In addition to the simple prepositional phrases which have
been extensively illustrated in 1.2.1.2.4, mention must also be made here of the complex
prepositional phrases which involve what Williams calls ‘local nouns’ (1862, 13) and
Biggs calls ‘locatives’ (1969, 53). With the exclusion of place names, local nouns are a
closed class, of which the commonest members are runga ‘top’, roto ‘inside’, raro
‘underneath’, waho ‘outside’, mua ‘front’, muri ‘behind’. (Williams gives a fuller list
(1862, 13), and it should be noted that he includes forms like nanahi which in more
recent times are normally written as one word with their preposition.) Unlike other nouns,
these locatives occur with no determiner, immediately following locative prepositions.
They are frequently followed by another prepositional phrase, which may repeat the first
preposition, but which may also be introduced by i or o. I is increasingly common,
although earlier writers suggest that there may be semantic distinctions of a subtle kind
between these prepositions (see eg. Harawira, 1950, 36–7), eg.
 Maori 94

(390) Kei te iri te whakaahua i runga i te pakitara

T/A hang the picture at top at the wall
‘The picture is hanging on the wall’
(391) I eke teeneki tangata ki runga i Maataatua
T/A land this man to top at Mataatua
i muri i tana tahuri·tanga
at behind at sggenIIIsg capsize·nom
‘This man boarded Mataatua after its capsizing’ (KW, 2–3)

(Note that this contains two relevant examples.)

Note also that, as applied to time, mua is time before, and muri, time after. They often
combine with the adverbial particles discussed in 1.2.1.3.1.1 to specify points in time, eg.
i mua tata tonu ake ‘only just before’, i muri tonu mai ‘a little time after’. The initial
preposition in such phrases is sometimes elided (see further 2.1.1.6).

1.2.1.3.1.3 Cases of noun phrases

Noun phrases in Maori are not marked morphologically for case.

1.2.1.3.1.4 Finite and non-finite adverbial clauses

Both finite and non-finite adverbial clauses occur. These have been discussed in 1.1.2.4.

1.2.1.3.2 Position of adverbials

The relative position of pre- and post-verbal particles is discussed in 1.2.1.3.1.1. There is
some discussion of the positions of prepositional phrases in 1.2.1.2.6. Perhaps it is worth
indicating here the range of orders deemed acceptable in one particular instance. The
unmarked order is
(392) I haere ia ki te motokaa me tana toko
T/A move IIIsg to the car with sggenIIIsg stick
‘He walked to the car with the aid of a stick’

Other possible orders of these constituents are

(392a) I haere ia me tana toki ki te motokaa
(392b) Me tana toko ka haere ia ki te motokaa
(392c) Ki te motokaa ka haere ia me tana toko
(392d) Ki te motokaa me tana toko ka haere ia

Each of these of course has special discourse preconditions. (Hohepa, 1967 contains
many examples of possible orders of constituents in Maori sentences.)
The position of finite and non-finite adverb clauses is discussed in 1.1.2.4.
 Syntax 95

1.2.1.3.3 Obligatory adverbials

Firstly, in non-verbal sentences, forms which are elsewhere adverbials function as
predicates, and these are, naturally, obligatory, eg.
(393) Kei runga i te teepu te pukapuka whero
at(pres) top at the table the book red
‘The red book is on the table’

However, it is probably misleading to consider this an obligatory adverbial.

Secondly, it appears that the post-verbal adverbial particles are obligatory on some
occasions, although it is not clear how those occasions are best specified. For instance,
with a verb such as haere, no direction of movement is implied. If the direction is
contextually clear, no adverbial is needed, but if the context does not make the direction
clear, an adverbial particle is required. Note that homai and hoatu ‘give’ include the
directional particles mai and atu. There are also many instances where the manner
particles are apparently obligatory, at least in the sense that native speakers reject
sentences from which such particles are removed. However, changes in context often lead
to changes in the particles used. A proper study of these particles would involve a major
study of discourse, and a panel of native speakers, and it has not been possible for me to
undertake this.

1.2.2 Adjective phrases

1.2.2.1 Operational definition for the adjective phrase

It is doubtful whether there is any operational definition of the adjective phrase, and
possibly doubtful whether this category is found in Maori—Biggs, for instance, does not
recognise it (Biggs, 1969). There is no affixation or marking which distinguishes
adjective forms from other forms in Maori. In what follows, I will outline those respects
in which adjectives resemble other word classes, and point out patterns which may serve
to distinguish them (see also 1.16.4).
In many cases, semantic adjectives function like verbs. Compare
(394) Ka pai ia ki te oma
T/A good IIIsg to the run
‘He is good at running’

and
(395) Ka tiimata ia ki te oma
T/A start IIIsg to the run
‘He starts to run’

However, pai, but not tiimata, can be pre-modified by tino ‘very’. This probably serves to
distinguish semantic adjectives from intransitive verbs, which they otherwise resemble,
but there are some verbs (eg. moohio ‘know’) which can be modified by tino.
 Maori 96

In non-verbal predicates, they are not readily distinguishable from nouns, compare
(396) He nui teeraa tangata
cls big that man
‘That man is big’

and
(397) He maahita teeraa tangata
cls teacher that man
‘That man is a teacher’

In this context, both forms can be modified by tino ‘very’, in the case of (397) producing
the sense ‘a really good teacher’.
In attributive position, nouns and adjectives are indistinguishable, at least at first sight.
They both occupy post-nominal position, compare
(398) Ko teenei teetahi whare nui
top. this a(sp) house big
‘This is a big house’

and
(399) Ko teenei teetahi whare miraka
top. this a(sp) house milk
‘This is a milking shed’

However, in this particular instance (and in many others), forms like that in (399) cannot
be modified by tino ‘very’, while forms like that in (398) can. Note, however, that tino
can modify head nouns, as in
(400) ngaa tino waahi uu·nga ki uta
the(pl) very place land·nom to shore
‘the main places for getting ashore’ (TR2, 119)

and I presume that it must therefore be possible for modified nouns to occur as modifiers.
Perhaps a more reliable distinguishing feature is that, while adjectives cannot be strung
together in Maori, noun modifiers can be strung with adjectives, or with each other
(although strings of noun modifiers in Maori, as in other languages, are almost always
analysable as binary groupings). Examples of multiple modification of head nouns appear
below in 1.2.5.2.
Finally, it may be possible to distinguish between nouns and adjectives by means of
the comparative and superlative structures; compare the two sets below:
(401a) Nui (noa) atu teenei whare
big quite away this house
‘This house is bigger’
(401b) Tino nui rawa atu teenei whare
very big intens away this house
‘This house is biggest’
 Syntax 97

(402a) *Maahita (noa) atu teeraa tangata

teacher quite away that man
‘That man is a better teacher’
(402b) *Tino maahita rawa atu teeraa tangata very teacher intens away that man
‘That man is the best teacher’

These criteria may in fact serve to provide sufficient evidence to define the category of
gradable adjectives, but they do so only in combination, and it is not clear that this is
equivalent to defining the class of adjectives, as opposed to the central members of that
class.

1.2.2.2 Adjectives that take arguments

No adjective in Maori to my knowledge takes an obligatory argument. However, there
are optional arguments of an oblique kind with many adjectives, especially arguments
which specify the domain of applicability of an adjective. These most commonly take the
form of a prepositional phrase introduced by ki. (Note that when ki is followed by an NP
introduced by te, the result may be indistinguishable from a non-finite verb phrase, and
thus the proper characterization of some of these adjectival arguments is in doubt—there
appears to be systematic ambiguity in these constructions.) Adjectives appear to be able
to take arguments regardless of the syntactic construction in which they occur, ie.
whether they occur in the predicate of verbal or non-verbal sentences, or attributively.
However, in some instances, the proper analysis seems to me to be in doubt. Examples
follow:
(403) Ka kawa ia ki te waiata
T/A bitter IIIsg to the sing
‘He is awful at singing’
(404) Tino pai ia ki te taakaro
very good IIIsg to the play
‘She is very good at sport’
(405) Ko teenei teetahi whare tino rahi kei teenei
top. this a(sp) house very large at(pres) this
ao
world
‘This is the biggest building in the world’

While these three seem to me to show adjectives with arguments in all three syntactic
environments, the following presents problems of analysis:
(406) He tamaiti pai ia i te kura
cls child good IIIsg at the school
‘She is a child who is good at school’

Note the alternative word-orders:

(406a) He tamaiti pai i te kura ia
(406b) I te kura he tamaiti pai ia
 Maori 98

It is not clear whether i te kura here is an argument of pai, or whether it is an adverbial.

While the intonation probably establishes that it is an adverbial in (406b), my consultant
denied that there was a semantic difference associated with the difference in word-order.
This kind of difficulty in determining the proper analysis of prepositional phrases in
sentences containing adjectives is not uncommon, and must be borne in mind throughout
this discussion of arguments of adjectives.

1.2.2.2.1 Adjectives in subjectless sentences

When an adjective functions as a verbal or non-verbal predicate, the subject may be
contextually omitted, eg.
(407) Ka pai
T/A good
‘[That’s] good’

When adjectives occur in any of the few apparently subjectless sentences of Maori, they
function as the predicate, eg.
(408) Ka roa
T/A long
‘Time passed’

1.2.2.2.2 Adjectives with DOs

There are no clear examples of adjectives with DOs (but see 1.2.2.2.4).

1.2.2.2.3 Adjectives with IOs

Since Maori does not clearly distinguish IOs from other oblique NPs, there are no clear
instances of adjectives with IOs (but see 1.2.2.2.4).

1.2.2.2.4 Other arguments with adjectives

Many adjectives may optionally take a prepositional phrase which specifies the domain
of application of the adjective. This is most commonly a phrase introduced by ki. (Note
that, since there are some DOs introduced by ki, some of these phrases may appear like
DOs; and since ki normally marks notional IOs, some of these phrases may be
indistinguishable from IOs. However, there does not seem to be any clear evidence which
would support such a classification.) The ki phrase may have a head noun, or it may have
a head verb, ie. the ki phrase may be nominal or infinitival. These possibilities, and some
other possible prepositions are illustrated below:
(409) Tino pai ia ki tana whaea
very good IIIsg to sggenIIIsg mother
‘She is very good to her grandmother’
(410) Ka hari ia ki te haere mai
T/A happy IIIsg to the move hither
 Syntax 99

‘He will be happy to come’

(411) Ko teenei raakau ka pai moo te whakairo
top. this tree T/A good intgen the carve
‘This tree will be good for carving’
(412) Tino mataku ia i te taniwha
very afraid IIIsg from the taniwha
‘He is very frightened of the taniwha’

1.2.2.2.5 Combinations of arguments

Adjectives with more than one argument are extremely rare, and thus generalizations are
not readily apparent One example will have to suffice:
(413) Ko teenei awa he pai moo te hii i
top. this river cls good intgen the catch at
Niu Tiireni
NZ
‘This river is the best for fishing in New Zealand’

The phrase i Niu Tiireni can be in initial position, and whether it is a second argument for
pai is an open question.

1.2.2.2.6 Order of constituents

Normally, the argument follows the adjective. However, examples have already been
adduced (see 1.2.2.2 and 1.2.2.2.5) which indicate that it is not clear whether other orders
are permitted. It should perhaps be noted here that it is not uncommon for the sentence
subject to intervene between an adjective and its argument if the adjective is part of the
predicate of the sentence. This is particularly common if the subject is brief (cf. 1.13.2.5).
Thus there are two possible constituent orders in each of the following:
(414a) Kei te mataku i te kurii nui ngaa tamariki
T/A afraid from the dog big the(pl) children
nei
proxI
‘These children are afraid of the big dog’

or
(414b) Kei te mataku ngaa tamariki nei i te kurii nui.
(415a) He pai teenei ki te kai
cls good this to the eat
‘This is good to eat’

or
(415b) He pai ki te kai teenei
 Maori 100

1.2.2.3 Adverbial modification of adjectives

1.2.2.3.1 Adverbs
A variety of adverbs can modify adjectives. While most of them follow the adjective,
there are a few which precede, eg. tino ‘very’, aahua, ‘somewhat’. In particular, the
comparative and superlative are constructed using adverbial modifiers for adjectives. A
few examples are given here, and there are many others scattered throughout this text.
(416) I kite·a he pukapuka tino pai
T/A see·pass. a book very good
‘[I] found a very good book’
(417) I koorero ia i teetahi pukapuka aahua
T/A read IIIsg DO a(sp) book somewhat
kino
bad
‘He read a rather bad book’
(418) Homai ki ahau te mea nui ake naa
give to Isg the thing big away proxII
‘Give me the bigger piece’
(419) E hiahia ana ahau ki te mea tino nui rawa
T/A want T/A Isg to the thing very big intens
‘I want the very biggest piece’

1.2.2.3.2 Prepositional phrases

It does not seem possible to distinguish these from arguments of adjectives, already dealt
with in 1.2.2.2.4.

1.2.2.3.3 Cases of noun phrases

These do not occur in Maori.

1.2.2.3.4 Adverbial clauses

I was unable to elicit any clear cases of adverbial clauses modifying adjectives.

1.2.2.3.5 Relative order of adverbial and adjective

This was dealt with in 1.2.2.3.1. One further example here should suffice:
(420) Tino kaha rawa atu koe
very strong intens away IIsg
‘You are very strong’
 Syntax 101

1.2.2.4 Order of constituents if arguments and adverbials both occur

with adjectives
It is rather rare for multiple modification of adjectives to occur. However, one word
adverbials which follow the adjective always precede any prepositional phrase, eg.
(421) Rite tonu koe ki te poaka ki te kai
like intens IIsg to the pig to the eat
‘You’re just like a pig at eating’

1.2.3 Adverbial phrases

It is perhaps important to begin any discussion of adverbial phrases in Maori by pointing
out that there are many instances where English has adverbial phrases, but the
corresponding Maori expression uses some other syntactic construction. A few typical
examples are given here, but it must be borne in mind throughout the discussion that
there are often alternatives to adverbial phrases, and that these are often judged more
natural by native speakers.
Firstly, compound verbs often correspond to verb+adverb structures in English, eg.
(422) Ka takahi haere atu i oo raaua tapuae
T/A tread move away at plgenIIIdl footprint
‘[He] walked on, treading in their footprints’ (TWh, 19)
(423) Ko ia… e titiro whakatau atu ana
top. IIIsg T/A look intently away T/A
‘He was gazing intently’ (TWh, 19–20)

Secondly, adjective+noun combinations in Maori often correspond to verb+adverb

combinations in English, eg.
(424) Kaatahi ka kaha te haere a Ruataupare
then T/A strong the move gen Ruataupare
maa
and others
‘Then Ruataupare and her party walked faster’ (TWh, 19)

Thirdly, sequences of verbs also correspond to verb+adverb combinations. The first verb
in such sequences is frequently an adjective used predicatively, eg.
(425) Ka roa e tuu ana
T/A long T/A stand T/A
‘[He] remained standing for a long time’ (TWh, 20)

1.2.3.1 Operational definition

Adverbial phrases cannot be distinguished by their position from other phrases: they
follow the forms they modify in most cases, as do other types of modifiers. Certain
adverbial phrases can be fronted in the sentence, but this seems only to be true of those
 Maori 102

that are, or arise from, prepositional phrases, such as inanahi ‘yesterday’. The one-word
adverbials which form part of the verbal complex (see 1.2.1.3.1.1) are fixed in position.
Adverbials are not distinct in form: there is no morphological marking to characterize
them.
They can be modified themselves by the same group of adverbials which can modify
adjectives and in some cases nouns. And since nouns and verbs are largely distinguished
by the constructions in which they are used (see 1.16.1, 1.16.3), adverbs cannot clearly be
distinguished from adjectives in terms of the forms which they modify.
Thus it appears that there is no possible operational definition for adverbial phrases.

1.2.3.2 Modification of adverbials by adverbials

1.2.3.2.1 Adverbs
Intensifying adverbs can modify other adverbs. It also seems likely that, in constructions
involving strings of one-word adverbs in post-verbal position, later adverbs modify those
preceding, although native-speaker intuitions on this matter are extremely difficult to tap,
as these forms do not have readily specifiable meanings.
(426) I haere tika tonu ki te whare kai
T/A move straight intens to the house eat
‘He went straight to the eating house’
(427) Ka tino maauiui rawa atu tana tinana
T/A very tired intens away sggenIIIsg body
‘His body was very, very weary’

1.2.3.2.2 Adverbial modification by prepositional phrases

Does not occur.

1.2.3.2.3 Adverbial modification by cases of NPs

Maori has no cases.

1.2.3.2.4 Adverbial modification by adverbial clauses

There are no clear instances of this. In examples like the following, it seems more likely
that the adverbial clause modifies the verb than the first adverbial:
(428) Aata haere tonu ia rite tonu ki te tamaiti
slow move intens IIIsg just intens to the child
e haere ana kia tango·hia ana niho
T/A move T/A subj pull·pass. plgenIIIsg tooth
‘He walked as slowly as a child going to have his teeth pulled
out’
 Syntax 103

1.2.3.3 Relative order of modified and modifying adverbials

Apart from a few like tino ‘very’, maatua ‘first’, aata ‘carefully, slowly’, the modifier
follows the form modified. This has been exemplified in the preceding sections.

1.2.3.4 Restrictions on the modification of adverbials by adverbials

Since the possible types of modifier are so restricted, there are no further restrictions to
be discussed here.

1.2.4 Prepositional phrases

All such phrases are prepositional and not postpositional in Maori.

1.2.4.1 Operational definition

It is possible to list the prepositions of Maori (see also 2.1.5):
i maa a ko
ki naa o whaka
kei moo me
hei noo e

However, some of these have functions other than being clearly prepositional. In
particular, i occurs as the marker of past tense in verbal phrases. (Although it is probably
true that, historically, absolute tense marking derives from prepositions, this position is
not readily supportable synchronically.) In general, prepositional phrases have the form
prep+NP; thus i te hui ‘at the meeting’ is formally distinct from i hui ‘met’ (a verb
phrase). However, particularly but not exclusively in Eastern dialects, the forms i te and
kei te function verbally. Thus the string
(429) Kei te mahi ia
at(pres) the work IIIsg

has two possible parsings, with two (closely associated) meanings: if kei is a preposition,
and te mahi a noun phrase, (429) means ‘He is at work’, whereas if kei te is a present
progressive marker, and mahi a verb, (429) means ‘He is working’. Because of the
systematic ambiguity of such phrases, Biggs (1969) calls this verbal use of kei/i te the
‘pseudoverbal continuous’. An additional complication in the use of this criterion is that
local nouns are not preceded by a determiner, and thus prepositional phrases with local
nouns are not formally distinct from verb phrases introduced by the same forms,
compare:
(430) I raro ia i te teepu
at underneath IIIsg at the table
‘He was under the table’
(431) I noho ia i te teepu
 Maori 104

T/A stay IIIsg at the table

‘He sat at the table’

However, since local nouns other than place names are a fairly small closed set, in
practice this second problem does not cause difficulties in the definition of prepositional
phrases.
It must also be noted that the combination of ki+te introduces infinitives, and so there
are occasions on which an infinitival complement is indistinct from a prepositional phrase
(although in most instances, accompanying arguments serve to identify the infinitive).
Consider
(432) Ka haere ia ki te kai
T/A move IIIsg to the food/eat

This could mean either ‘He went to eat’ or ‘He went to the food’, (but ngaa kai would be
more expected in the latter case).
In verbal sentences, prepositional phrases normally follow the verb and the subject (a
non-prepositional phrase), but may precede the verb. In non-verbal sentences where the
prepositional phrase functions as the predicate, it precedes the subject, thus appearing in
the same sentence position as the verb. Thus neither position nor construction nor form is
sufficient in itself to define unambiguously the prepositional phrase.

1.2.4.2 Prepositions and arguments

1.2.4.2.1 Prepositions without arguments

No prepositions occur without arguments in Maori.

1.2.4.2.2 Prepositions with more than one argument

No preposition in Maori has more than one argument

1.2.4.2.3 Prepositions with arguments other than noun phrases

It appears that prepositions can only introduce arguments with at least some nominal
characteristics (see also 1.1.2.2.2.2). Thus a preposition can introduce nominalized
sentential structures, as in the following example:
(433) Kei te mea hoki ia ki toona kore i paatai
T/A say intens IIIsg to sggenIIIsg neg T/A ask
ki ngaa tamariki raa ko teewhea a Ruataupare
to the(pl) children dist top. which(sg) pers Ruataupare
‘He was asking himself why he had not questioned those children as to which was
Ruataupare’ (TWh, 19)

(It must be noted, however, that examples of this kind are rare.) The commonest
examples of this type are the infinitives, which are introduced by ki (a preposition)+te
 Syntax 105

(the unmarked article). The occurrence of the article in this construction marks it as
having nominal characteristics. It should also be noted that when i/kei te is used as a tense
marker, the presence of te again marks the construction as having certain nominal
characteristics.
It is perhaps worth noting the existence of examples like the following, which may at
first sight appear to show that a preposition can have a prepositional phrase as an
argument:
(434) Ka kamo ngaa whatu o ngaa taangata o te
T/A wink the(pl) eye gen the(pl) men gen the
ihu ki o te taa me o te taa ki o
prow to plgen the stern with plgen the stern to plgen
te ihu
the prow
‘The men in the prow exchanged knowing glances with those in the stern, and those in the
stern with those in the prow’ (TWh, 17)

However, as the glosses show, not all the forms o are prepositions: those glossed ‘plgen’
would have the form to in the singular, where the t- is the singular “definitive” marker (to
use Biggs’s term), indicating that these forms are indeed nominal.

1.2.4.2.4 Stranding of prepositions if their NP is moved

No preposition can be stranded in Maori.

1.2.4.3 Modification of prepositions

Prepositions cannot be modified.

1.2.4.4 Prepositions that govern more than one case

Case is not marked in Maori except through prepositions, so prepositions do not govern
cases in Maori.

1.2.5 Noun Phrases

1.2.5.1 Operational definition for NPs

The vast majority of NPs in Maori consist of a determiner+ noun (+optional modifiers).
The determiners are te, ngaa, (t)eetahi, (t)aua, (t)eenei/naa/raa,.(t)eewhea, he, a, and the
paradigms of possessive determiners, eg. taaku, tooku, taku (for the full paradigm see
2.1.2.4.3). However, there are several groups of constructions which prevent this
structural criterion from providing a clear operational definition. Firstly, there are the
local nouns, raro ‘underneath’, runga ‘top’, etc. (including all place names) which do not
occur with determiners when they follow prepositions although they take the personal
article a when they function as subject (ie. in the only non-prepositional phrase in which
they can occur), eg.
 Maori 106

(435) Ka tae ki runga

T/A arrive to top
‘[He] reached the top’
(436) Ko teenei a runga
top. this pers top
‘This is the top’

Personal names only have a determiner following certain prepositions, namely i, ki, kei,
hei; elsewhere, they too have no determiner. NPs in lists also sometimes occur without a
determiner, see 1.16.1.
Secondly, as has been mentioned in other contexts (eg. 1.1.2.2.2.4, 1.2.2.2), te occurs
in infinitival phrases introduced by ki, the heads of which cannot be clearly classed as
nouns, since they take the arguments typical of verbs. The same is true of the “pseudo-
verbal” T/A markers, i/kei te: these constructions might as readily be considered
“pseudonominal”. One example of each will suffice:
(437) Ka haere ia ki te tiki wai moo ana
T/A move IIIsg to the fetch water intgen plgenIIIsg
tamariki
children
‘She went to fetch water for her children’ (R, 19)
(438) Kei te haere ia
T/A move IIIsg
‘She is going’

The bold phrases here are probably not to be regarded as NPs, as their heads have verbal
arguments.
There is a small group of nouns which name personal relations which inflect for plural
(see 2.1.1.8.6), but they are a tiny subset only of those forms which can function as the
heads of NPs, and so there is no morphological change which characterises nouns as a
whole.
Thus the only possible operational definition which will identify all and only the
nouns of Maori, is their occurrence with determiners. If the local nouns are excluded,
then the possibility of occurrence with any determiner other than te and he serves this
purpose. Since any local noun requires the personal article a when it occurs as a subject, a
conjunction of these two criteria will, it appears, provide a satisfactory operational
definition of a noun. However, there does not seem to be a satisfactory operational
definition of a noun phrase, since there is no mechanical way of distinguishing between
te and he as determiners and other uses of these forms.
 Syntax 107

1.2.5.2 Types of noun modifier

1.2.5.2.1 Adjective
A single adjectival modifier can occur, following the NP. The head noun must be
repeated (with or without the determiner), or a general noun such as mea ‘thing’ must be
used if a second modifier is added: strings of adjectival modifiers do not occur, eg.
(439) te tamaiti paku
the child small
‘the little child’
(440) teetahi tangata tino nui, tangata tino moomona
a(sp) man very big man very fat
‘a big, fat man’ (TR2, 151)
(441) he koowhatu nui, he mea taimaha
a stone big a thing heavy
‘a big, heavy stone’

1.2.5.2.2 Relative clause

Relative clauses are extremely common, and follow the head noun. Abundant
exemplification will be found in 1.1.2.3.

1.2.5.2.3 Possessive ‘adjective’

Maori has a complex system of possessive markers, which is discussed in detail in 1.10.
There are three sets of pre-nominal pronominal possessive ‘adjectives’ in Maori, an ‘a’
set, an ‘o’ set, and a partial set in which the a/o distinction is neutralized (see Biggs,
1955). The sets are listed and illustrated in 2.1.2.4.3. The choice between the three sets is
discussed in 1.10.1, 1.10.6, 2.1.2.4.2.5. Here, a few examples will suffice:
(442) too koorua whare
sggenIIdl house
‘your house’ (referring to two people)
(443) aa raatou pukapuka
plgenIIIpl book
‘their books’ (referring to three or more people)
(444) tana kurii
sggenIIIsg dog
‘his dog’

With NPs, there is a pre-nominal construction, which occurs readily with single proper
names, but is rather restricted with other types of noun phrases (see also 1.10.1):
(445) taa Turi kurii
sggen Turi dog
 Maori 108

‘Turi’s dog’

However, it is commoner to use a post-nominal genitive phrase:

(446) te kurii a Turi
the dog gen Turi
‘Turi’s dog’

This is the normal form for common nouns:

(447) te kurii a te tamaiti
the dog gen the child
‘the child’s dog’

The prenominal form can be used with some shorter common noun phrases, but it is
sometimes necessary to insert nei between the possessor and the head, as a sort of
boundary marker:
(448a) Ko teenei taa te tamaiti kurii
eq this sggen the child dog
‘This is the child’s dog’

but
(448b) Kua mate taa te tamaiti nei kurii
T/A dead sggen the child proxI dog
‘The child’s dog is dead’

Note the ungrammaticality of the following:

(449) *taa tamaiti kurii
sggen child dog

The t- prefix which occurs throughout the singular paradigms of possessive forms occurs
also in a number of other determiners, and is matched there, as here, by Ø in the plural.
There are good grounds for regarding it as the determiner te.

1.2.5.2.4 Article
There are a variety of articles in Maori, and the rules governing their use are not clear in
all cases. Johansen (1948, 8–18) provides interesting comments on a number of the more
controversial issues.
An article is required in a noun phrase in Maori except in the following instances:
(i) before local nouns, eg. raro ‘underneath’, runga ‘top’, including place names;
however, when these function as sentence subject, they take the personal article a.
(ii) pronouns and personal names do not require an article following prepositions other
than i, ki, kei, hei. (Note, however, that ahau, one form of the first person singular
pronoun, never takes an article, almost certainly because it includes the personal
article, ie. it is a·h·au, ‘personal article-linking consonant·Isg’.)
 Syntax 109

(iii) occasionally in lists, articles are omitted. It is not at all clear from the scanty data
under what circumstances this is permissible.
Some examples illustrating these points follow:
Local nouns:
(450) Kei roto i te kaapata te tioka
at(pres) inside at the cupboard the chalk
‘The chalk is in the cupboard’
(451) He whero a roto
cls red pers inside
‘The inside is red’

Pronouns:
(452) Ka kite au i a ia
T/A see Isg DO pers IIIsg
‘I saw him’
(453) I patu·a te tamaiti e ia
T/A beat·pass. the child by IIIsg
‘The child was hit by him’

Personal names:
(454) Ka tae mai a Hata
T/A arrive hither pers Hata
‘Hata arrived’
(455a) Ko Hata ka tae mai
top. Hata T/A arrive hither
‘It was Hata who arrived’

Compare this with

(455b) *Ko a Hata ka tae mai
(456) Kei te tatari ia ki a Hata
T/A wait IIIsg to pers Hata
‘She is waiting for Hata’

The articles are:

a ‘personal’ article, used before personal names, pronouns, and, under some
circumstances, local nouns;
te usually described as ‘definite singular’, but also used as the generic article (with plural
nouns in those few instances where the noun inflects for plural), and probably more
accurately described as the “elsewhere” article—te is used when the more semantically
specific articles are inappropriate (see further 2.1.1.10, and note that Johansen 1948, 8–
14 has some interesting data and comments in this connection);
ngaa plural definite in many contexts, although there also appear to be contexts where the
description “definite” seems inappropriate (see further 2.1.1.10);
 Maori 110

teetahi singular, indefinite, and usually specific, it is complex morphologically, consisting of te

‘the’+tahi ‘one’;
eetahi the plural corresponding to teetahi, showing the Ø plural formation characteristic of
many determiners in Maori, though some dialects have ngaatahi;
he indefinite, but with grammatical restrictions on its occurrence—it does not occur in the
subjects of active transitive sentences, nor following prepositions; when it contrasts with
teetahi, it appears to be non-specific (see further 2.1.1.10);
taua/aua singular/plural strongly anaphoric (see 1.5.1.5).

Whether hei is an article remains somewhat in doubt. Johansen suggests that it is (1948,
15–18), but the evidence is not overwhelming. Certainly, there are times when it
functions as a future-time form in contrast to he, but most instances are predicative.
Because NPs in general are not tense-marked, it does not contrast with other articles in
most sentence positions. Illustrations of the articles (see also 2.1.1.10) follow (the
personal article is illustrated above):
Te
Generic use:
(457) Kotahi hereni te utu moo te tamariki! E
one shilling the price intgen the children num
toru hereni te utu moo te pakeke!
three shilling the price intgen the adult
‘One shilling is the price for children! Three shillings is the price for adults!’ (TR1, 50)

Te where the description “definite singular” clearly is not appropriate:

(458) Kei te ruku raaua i te kooura
T/A dive IIIdl DO the crayfish
‘They are diving for crayfish’

Ngaa
Where the description “definite plural” is not clearly appropriate:
(459) E tiki ngaa pereti e whaa
imp. fetch the(pl) plate num four
‘Get four plates’

Teetahi
(460) Kua mau i a Tamahae teetahi ika paku
T/A caught cause pers Tamahae a(sp) fish little
‘A small fish has been caught by Tamahae’ (TR1, 39)
(461) Puuhi·a; kei mate teetahi tangata
shoot·pass. mon die a man
‘Shoot [it]; [it] might kill somebody’

Eetahi
(462) Kei te maatakitaki a Mere… i eetahi koti
 Syntax 111

T/A gaze pers Mere DO some(sp,pl) coat

‘Mere… is looking at some coats’ (TR1, 61)

He
(463) Ka hoko·na e ia he hikareti, he pepa,
T/A buy·pass. by IIIsg some cigarette some paper
he wai reka, he aha, he aha
some water sweet some what some what
‘He bought cigarettes, paper, soft drinks, and other things’

Taua
(464) Ka haere te mookai raa ki a Tutaanekai, ka
T/A move the slave dist to pers Tutanekai T/A
mea mai ia. “Kei whea hoki too wai?
say hither IIIsg at(pres) where again sggenIIsg water
Ka mea taua mookai, kua riro atu anoo.
T/A say detaph slave T/A taken away again
‘That slave went to Tutanekai, who said “Where is your water again?” That slave answered
that it had been taken away yet again’ (H, 9)

Aua (the example follows a report of a naming ceremony and explanations of the names)
(465) E mau nei anoo aua ingoa
T/A carry proxI still detaph(pl) name
‘Those names still remain’ (TWh, 17)

1.2.5.2.5 Demonstrative ‘adjective’

There are two possible constructions for demonstrative ‘adjectives’ in Maori, one of
which is no doubt more correctly analyzed as article+adverbial. However, since the two
are in free variation, and the choice depends on such factors as rhythm and euphony, both
are treated here.
(i) The deictic particles nei ‘near speaker’, naa ‘near hearer’, raa ‘distant from
speaker/hearer’ may be attached to the article te, forming teenei, teenaa, teeraa. These
forms are used before the noun:
(466) teenei pukapuka
this book
‘this book’

In the plural, the t-is dropped:

(467) eeraa kurii
those dog
‘those dogs’

(ii) The deictic particles may follow the head noun, while the articles te or ngaa precede:
(468) te pukapuka raa
 Maori 112

the book dist

‘that book’
(469) ngaa pukapuka naa
the(pl) book proxII
‘those books (near you)’

Williams (1862, 20) also classes ia ‘that’ with the demonstratives. It is most commonly
substantival, but when used for ‘each’ is clearly a determiner, since a noun phrase may
consist of ia+N, eg. ia raa ‘each day’, usually repeated.

1.2.5.2.6 Quantifiers
Quantifiers are not treated homogeneously in Maori (see also 2.1.6.6). The clearest
quantifier, katoa ‘all’, occurs in post-head position, eg.
(470) Kaaore anoo ia kia moohio ki ngaa tamariki
neg yet IIIsg subj know to the(pl) children
katoa
all
‘She does not yet know all the children’
(471) Kia ora koutou katoa
subj well IIpl all
‘Your health, all’, ‘Hello, everyone’

However, if katoa arises in subject NPs, it is normally floated to post-verbal position,

even across intervening constituents, see 2.1.6.6.
The form for ‘each’, ia, occurs in pre-nominal position, and functions as the
determiner. Williams (1862, 20) classifies it as a demonstrative, see 1.2.5.2.5, eg.
(472) Ko ia tamariki i whiwhi ki teetahi pukapuka
top. each child T/A receive to a(sp) book
‘Each child received a book’

Since, as this example shows, such constituents can be fronted with ko, they appear to
count as definite in Maori. Ia cannot be floated. Both ia and the noun it modifies can be
repeated to render ‘every’, although katoa is often used for that purpose:
(473) Ka whakatangi raaua i a raaua puu ia poo,
T/A play IIIdl DO plgenIIIdl flute each night
ia poo
each night
‘They played their flutes every evening’

‘Both’ is rendered by the numeral ‘two’, eg.

(474) I mate rawa ngaa tamaiti e rua
T/A die intens the(pl) child num two
‘Both boys were killed’
 Syntax 113

Note that numerals in NPs occur in the post-head position, like katoa.
‘Many’ is rendered either by maha ‘many, much’, or by the use of tini ‘very many’,
which is sometimes a numeral and sometimes a noun meaning ‘host, myriad’, or by the
use of mano ‘thousand’, which is a numeral. Sometimes tini mano is used for emphasis.
These are most commonly predicative, but an attributive example is:
(475) Kua ngaro ngaa ika maha o te moana nui
T/A missing the(pl) fish many gen the sea big
‘Many species of ocean fish have been depleted’

Tini occurs in post-nominal position, like maha in (475), and it can also occur as a head
noun with a nominal modifier as in
(476) I tae mai ngaa tini taangata noo
T/A arrive hither the(pl) many people actgen
Tuwharetoa
Tuwharetoa
‘The many people from Tuwharetoa arrived’

It appears that mano is normally predicative.

‘Much’ does not have a straightforward equivalent The distinction between countable
and uncountable is not made in Maori, and tini is often used to correspond to ‘much’. The
following withteenaa mea was proffered as another equivalent, eg.
(477) I runga i te papa e paatere ana teenaa at top at the floor T/A spill T/A that mea te honi
thing the honey
‘There was a lot of honey spilt on the floor’

Thus it appears that the only clearcut quantifiers in Maori are katoa and the numerals,
although maha and tini may be. Only katoa is subject to quantifier float.

1.2.5.2.7 Adverbials
There appear to be a few single-word adverbials which can modify nouns. The
commonest are undoubtedly the deictic particles discussed in 1.2.5.2.5. The following
examples illustrate some of the other forms found:
(478) ngaa rori pee·nei the(pl) road like-this
‘roads like this’
(479) He tangata kee koe, he tangata kee ahau a man contr IIsg a man contr Isg
‘We are different kinds of people’

Prepositional constructions are very common modifiers of nouns, especially, but by no

means exclusively, prepositional constructions indicating possession:
(480) te maauiui o te haere a Peetera the weary gen the move gen Peter
‘the weariness of Peter’s walk’ (TR2, 18)
(481) te huarahi ki te maatauranga the path to the knowledge
‘the path to knowledge’ (TR2, 166)
 Maori 114

(482) te roonihi i Mokoia the launch from Mokoia

‘the launch from Mokoia’ (TR2, 101)
(483) te whare kei te huringa the house at(pres) the corner
‘the house on the corner’

1.2.5.2.8 Emphatic words

It appears that tino is the only emphatic word which modifies nouns, eg.
(484) Ko teeraa te tino pukapuka i whiwhi ahau top. that the very book T/A get Isg
‘That’s the very book I received’

1.2.5.2.9 Comparative, superlative and equative structures

1.2.5.2.9.1 Comparatives
There are two comparative structures. The first uses atu or ake following the modifier.
Secondly, he+tino or rawa forms an ‘intense comparative’ (Williams, 1862, 25). Often
additional adverbial particles are required by the context. Comparative structures follow
the head noun. Examples are:
(485) Me mehua tana waewae nui ake
oblig measure sggenIIIsg foot big away
‘His bigger foot should be measured’
(486) I hoko mai ahau i teetahi pukapuka pai
T/A buy hither Isg DO a(sp) book good
kee noa atu
contr quite away
‘I bought a better book’
(487) he mea pai rawa i teenaa
a thing good intens compar that
‘a far better thing than that’

(See also 1.8 and 2.1.4.4.)

1.2.5.2.9.2 Superlatives
These use rawa or tino+rawa with te, ngaa, or another t-prefix determiner in either
singular or plural. Sometimes tino alone is used. Tino precedes the modifier, while rawa
follows. Sometimes atu or ake is added for pragmatic reasons, eg.
(488) I hoko mai ahau i teetahi pukapuka pai
T/A buy hither Isg DO a(sp) book good
rawa atu
intens. away
‘I bought the best book’
(489) Mau·ria mai ki ahau too heeki tino nui
 Syntax 115

bring·pass. hither to Isg sggenIIsg egg very big

rawa atu
intens away
‘Bring me your biggest egg’

1.2.5.2.9.3 Equative structures

It is not clear that Maori has equative structures other than relative clauses which act as
modifiers in noun phrases. Translations of English examples caused great difficulty.
Consider
(490) I hoko mai au i te pukapuka rite tonu
T/A buy hither Isg DO the book like still
te pai
the good
‘I bought as good a book’

Here the equative structure is a relative clause.

1.2.5.2.10 Other modifying structures

1.2.5.2.10.1 Nouns as modifiers

Maori makes extensive use of nouns modifying nouns. Whether such structures differ
from compounds is not clear. Because of the difficulty of ascertaining parts of speech, it
is not always clear that the modifiers are nouns. These problems are discussed further in
2.2 and 2.2.6.3. A few examples are given here:
(491) te whare kura
the house school
‘the school building’
(492) ngaa keene kiriimi
the(pl) can cream
‘the cream cans’ or ‘the cans of cream’
(493) ngaa toenga miiti
the(pl) remains meat
‘the remains of (the) meat’
(494) eetahi atu kai moana
some(pl,sp) away food sea
‘other sea-foods’

Perhaps it should be mentioned here that tuu, ‘kind of’, also takes a noun modifier:
(495) he tuu rimu
a kind rimu
‘a kind of rimu’
(496) eenei tuu whakaaro
 Maori 116

these kind thought

‘these kinds of thoughts’

Note that kura maahita ‘school master’, transliterated as a phrase and sometimes written
as one word, breaks the normal head—modifier order of Maori, as does the phrase
pukapuka toara ‘bookstore’, which one consultant said was standard in earlier years in
some areas.

1.2.5.2.10.2 Verbal constructions

It is very common for verbal constructions to modify nouns (cf. 1.1.2.3.9). These take a
variety of forms. Plain stem constructions occur (see 1.1.2.4.3 for details), eg.
(497) te taaima hoki ki te kaainga
the time return to the home
‘return-home time’
(498) te whare here-here
the house tie·dup
‘prison’

If notional subjects occur in these constructions, they appear as possessive phrases (as
with the corresponding nominalizations) eg.
(499) te taaima tae o te tira ki Te Kaha
the time arrive gen the team to Te Kaha
‘the arrival time of the travelling party at Te Kaha’ (TR2, 58)

If direct objects of transitive verbs occur, they normally use the ‘object incorporation’
construction (see 1.2.1.2.2 and 2.1.3.6.12), eg.
(500) teetahi poo mahi moni
a(sp) night make money
‘a fund-raising evening’ (TR2, 85)
(501) ngaa taangata toa tope raakau
the(pl) men champion cut tree
‘the champion wood-cutters’ (TR2, 156)

If appropriate, a non-incorporated DO occurs:

(502) Ka tae mai te taraka mau i ngaa
T/A arrive hither the truck carry DO the(pl)
taangata ki te ngahere
men to the bush
‘The truck to take the men to the bush arrived’ (TR2, 42)

Examples with transitive verbs accompanied by their subjects are relatively rare, but take
the possessive a:
(503) te taima kuti hipi a te toa
 Syntax 117

the time cut sheep gen the champion

‘the champion’s shearing time’

Sometimes looser semantic connections are also expressed in this way:

(504) te koaea whakataetae waiata Maaori o Te Arawa
the choir compete song Maori gen Te Arawa
‘the Te Arawa choir competing at Maori songs’ (TR2, 157)

The verbs in these constructions are always active in form, though Williams comments
(1862, 46) that they may be active or passive in sense, but his example with passive sense
appears to be a full relative clause, and I have not been able to find or elicit an example
that is clearly passive in sense.
Sometimes a nominalization of the verb is used, eg.
(505) ngaa tino waahi uu·nga ki uta
the(pl) main place land·nom to shore
‘the main places for getting ashore’ (TR2, 119)

1.2.5.2.10.3 Apposition
Apposition is particularly common with a proper name and a description, eg.
(506) I konei ka noho Te Rauparaha, te rangatira
at here T/A stay Te Rauparaha the chief
whakahirahira
great
‘Te Rauparaha, the great chief lived here’

The other order is also possible, eg.

(507) Kua oma atu tana wahine, a Hera, ki te
T/A run away sggenIIIsg woman pers Hera to the
awhi i a ia
embrace DO pers IIIsg
‘His wife, Hera, has run out to embrace him’ (TR1,152)

Apposition between two common noun phrases can occur, and as above, consists of
juxtaposition of the two NPs, eg.
(508) I ngaro ngaa whenua, te whanau tika o te
T/A lose the(pl) land the birth right gen the
iwi
people
‘They lost their land, the birth-right of the people’

1.2.5.2.10.4 -a-
It is perhaps appropriate to mention here a construction which may be a compound-
forming one, with a linking element. It is exemplified by
 Maori 118

(509) waiata aa ringa

song link hand
‘action song’
(510) hui aa tau
meet link year
‘annual meeting’

Stowell (1911, 13) gives the following examples:

(511) mahi aa ringa
work link hand
‘handicraft’
(512) haere aa waewae
move link foot
‘walk’

and comments that “aa speaks of the manner, means, process”. Unfortunately, this does
not shed any light on the grammar of the construction. However, the formation appears to
be still productive (see further 2.2.6.3).

1.2.5.3 Possibility of occurrence of more than one of each type of

modifier
It is not possible to have more than one adjective, comparative or superlative in the same
NP (see also 1.2.5.2.1). Either the head noun is repeated, or a prop is used, or the second
adjective is given the intonation of a separate tone group, often with the addition of hoki
‘also’, eg.
(513) Te tangata moomona, mangere ia
the man fat lazy IIIsg
‘He is a fat, lazy man’

If the adjectives belong to different sub-classes, a prop is always required, eg.

(514) Ko teenei he huruhuru whero, he aataahua hoki
top. this cls feather red cls beautiful also
‘This is a beautiful red feather’

Where both adjectives belong to the same subclass, I have found no evidence of ordering
preferences. Where different subclasses are involved, gradable adjectives are normally
put in the second phrase.
While it is possible to have two comparatives or two superlatives in juxtaposed
phrases, as with adjectives, such constructions are clearly marginal in Maori, and
speakers tend to avoidthem. Examples which should be taken in that light are:
(515) Ka hoko ia i te pahikara tino nui, tino
T/A buy IIIsg DO the bike very big very
turituri rawa atu
noisy intens away
 Syntax 119

‘He bought a bigger, noisier motorbike’

(516) Ko ia te kootiro tino kino te paruparu, tino
top. IIIsg the girl very bad the dirt very
kino te turituri i roto i te ruuma
bad the noise at inside at the room
‘She is the dirtiest, noisiest girl in the room’

It is possible to have two or even more relative clauses, eg.

(517) Engari, kei hea te tuna i piihuka·tia nei,
but at(pres) where the eel T/A hook·pass. proxI
i paoa·tia nei, i maka·ia nei ki takahi?
T/A bash·pass. proxI T/A throw·pass. proxI to bank
‘But where is the eel that was hooked, that was bashed, that was thrown to the bank?’

The only restrictions on ordering appear to be those imposed by external factors, such as
temporal ordering.
It is not possible to have two possessive adjectives, nor two articles, quantifiers, or
demonstratives.
Various combinations of single-word adverbials occur, especially where the second is
one of the three deictic adverbs. More than one prepositional phrase can occur, eg.
(518) I noho ia i raro i te puke i te whare
T/A stay IIIsg at under at the hill at the house
taha matau
side right
‘He lives in the house on the right under the hill’

1.2.5.4 Combinations of nominal modifiers

Basically, the only restrictions on combinations of modifiers apart from those discussed
in 1.2.5.3 are restrictions on the pre-head determiners, namely articles, possessives and
demonstratives, which cannot co-occur, ie. there is only one determiner slot, filled by
either an article or a demonstrative or a possessive adjective. A few examples of
combinations of modifiers are given here; any attempt at exhaustive illustration of
combinations would be too space-consuming. article+adj+quantifier
(519) ngaa manuhiri tuuaarangi toko·maha
the(pl) visitor distant pnum·many
‘the many visitors from afar’

article+adj+quantifier+prepositional phrase
(520) ngaa maatua nui katoa a teeraa
the(pl) battalion big all gen the other
‘all the main battalions of the enemy’

article+adjective+quantifier+relative clause
 Maori 120

(521) ngaa taangata rongonui katoa i kite·a e ia

the(pl) men famous all T/A see·pass. by IIIsg
‘all the famous people he saw’

However, it must be pointed out that heavy NPs in Maori are not the norm, except those
containing relative clauses. In a count of the first 200 NPs in a text, 51% had two
constituents; phrases with one and three constituents accounted for about 23% each, and
2% had four constituents; no larger phrases occurred. In a count of a comparable quantity
of an English text of a similar sort, 2-constituent phrases made up 45%, with 3-
constituent phrases 26%. Those with 1 constituent accounted for about 20%, but 14
phrases with 4 constituents, and 6 with 5 constituents occurred, ie. about 10% had 4 or
more, as opposed to 2% in the Maori.

1.2.5.5 Order of head and modifiers

Order has been discussed in 1.2.5.2. It may be summarized here in the following schema:
1 2 3 4 5 6
Det (article, demonstr., Noun Adj (incl. noun Quant Adv (single Phrasal and sentential
poss’ives) head mods) word) constr’ns

1.3 COORDINATION

1.3.1.1 Sentence Coordination

1.3.1.1.1 And-coordination
The commonest means of and-coordination in Maori is juxtaposition of the coordinated
elements with no formal marker. The final phrase of the first conjunct does not have
sentence-final falling intonation, eg.
(522) I mahi·a e ia te kaapura, ka
T/A make·pass. by IIIsg the fire T/A
whakamaoa·tia aa raatou kai
cook·pass. plgenIIIpl food
‘He made a fire and they cooked their food’
(523) I te whitu karaka, ka tangi te pere, ka
at the seven clock T/A sound the bell T/A
tiimata te mahi
start the work
‘At 7am, the bell rang and work started’

If there is a desire to emphasize the time lapse between the conjuncts, aa may be inserted
between them. The length of this vowel can be varied to indicate the relative time lapse
involved. This construction is most often found in narratives and can be best glossed as
‘and then’ or ‘after a while’, eg.
 Syntax 121

(524) Ka whakataa i toona manawa, aa

T/A give breath DO sggenIIIsg heart and then
ka mutu te ngenge o oona pokihiwi
T/A finished the weariness gen plgenIIIsg shoulder
‘[She] took breath, and at last the weariness of her shoulders disappeared’ (H, 8)

Aa also occurs in some contexts where no time lapse is involved, and in such instances, it
looks very like a coordinating conjunction, eg.
(525) Toko·rima oona tuaakana, aa, kotahi o
pnum·5 plgenIIIsg older siblings and one gen
raatou he wahine
IIIpl a woman
‘He had five older siblings, and one of them was a woman’ (TR2, 188)

Some previous scholars commenting on coordination in Maori have treated aa as a

coordinating conjunction, eg. Williams (1862, 58), Reedy (1979, 82ff). Biggs (1969, 97)
classes it under the heading ‘interjections and interjectory phrases’, and Hohepa (1967,
95) calls it a ‘conjunctive link’, which may or may not be a coordinating conjunction in
disguise. Reedy regards aa as the basic means of sentence coordination, and deletes it to
derive juxtaposition. However, I have found several textual examples which do not
clearly involve coordination, eg.
(526) Aa, ko te waahi i uu ai ia, he
now top. the place T/A land part. IIIsg a
waiariki, ko Waikimihia te ingoa
hot-spring eq Waikimihia the name
‘Now the place where she landed was a hot-spring called Waikimihia’ (H, 8)
(527) Aa ka ngaro te ngenge, ka kauhoe anoo
then T/A lost the weariness T/A swim again
a Hinemoa
pers Hinemoa
‘At last the weariness passed, and Hinemoa swam on again’ (H, 8)

I therefore believe that there is room for doubt as to the status of aa. Given the use of
juxtaposition for coordination at phrase level, and the statistical frequency of
juxtaposition at sentence level, I think aa may be better analysed as a coordination
marker inserted optionally in some conjoined constructions, rather than deleted from
most. It may, however, be the case that aa is on the increase as a sentence coordinator
under the influence of English. Speakers vary considerably in the extent to which they
use aa. Those who use it most appear to be male, especially those who frequently speak
in public. It may thus be a time-gaining expression. My female consultant very rarely
uses it.
In sentence coordination, hoki can be added to the second conjunct if the sense ‘and
also’ is to be emphasized. Hoki usually follows the verb, though Williams describes its
position as “after the first important word in the sentence” (Williams, 1862, 58). For
instance:
(528) Ka peita raaua i te whare, ka hoko·na hoki
 Maori 122

T/A paint IIIdl DO the house T/A buy·pass. also

te motokaa hou
the car new
‘They painted the house and also bought a new car’

It must also be noted that where two English conjuncts express a temporal sequence,
Maori speakers often choose non-conjoined structures to express them, preferring a time
adverbial clause, eg.
(529) Me whakamanuhiri ia i tana hoki·nga
oblig welcome IIIsg at sggenIIIsg return·nom
mai
hither
‘She came back and they welcomed her’ (More lit. ‘She was welcomed on her return’)

The converse can also be true:

(530) Ka kii ngaa puku, ka moe
T/A full the(pl) stomach T/A sleep
‘When [their] stomachs are full, [they] sleep’

This English translation seems more natural than the more literal ‘Their stomachs are full
and they sleep’, since the context implied habituality.
Questions, like declaratives, are coordinated by juxtaposition.
And-coordination of imperatives is rather more complex. If two or more true
imperatives (ie. not with me and not statives with kia) are coordinated, then a special
construction may be used for all but the first, namely a declarative with ai post-verbally,
thus:
(531) Mau·ria oo kete, ka whakakii ai ki
take·pass. plgenIIsg basket T/A fill part. with
te hua raakau
the fruit tree
‘Take your baskets and fill them with fruit’

This construction is not available if the second imperative is negative. In this case, both
conjuncts have their normal form, eg.
(532) Haere atu, kaaua anoo e hoki mai!
move away neg again T/A return hither
‘Go away and never come back!’

With the first imperative negative and the second positive, judgements of the
acceptability of the declarative with ai varied, but this construction was never preferred to
straight juxtaposition.
Even with two positive imperatives, the declarative plus ai is not compulsory. Thus
(533a) E noho ki raro, tuhi·a te reta naa
imp sit to underneath write·pass. the letter proxII
‘Sit down and write that letter’
 Syntax 123

was judged more probable than

(533b) E noho ki raro, ka tuhi ai te reta
imp sit to underneath T/A write part. the letter
naa
proxII
‘Sit down and write that letter’

In such cases, the juxtaposition involves two clearly separate intonation patterns, and the
second may better express the “and” of the English.
When three positive declaratives are involved, it is possible for the second and third to
take either form, but the third cannot take the imperative form if the second used the
declarative+ai, thus:
(534a) Tiki·na too koti, kaakahu·tia, ka uru
fetch·pass. sggenIIsg coat dress·pass. T/A enter
atu ai ki roto i te motokaa
away part. to inside at the car
‘Get your coat, put it on, and get into the car’
(534b) Tiki·na too koti, ka kaakahu ai, ka
fetch·pass. sggenIIsg coat T/A dress part. T/A
uru atu ai ki roto i te motokaa
enter away part. to inside at the car
‘Get your coat, put it on, and get into the car’
(534c) Tiki·na too koti, kaakahu·tia, uru atu ki
fetch·pass. sggenIIsg coat dress·pass. enter away to
roto i te motokaa
inside at the car
‘Get your coat, put it on, and get into the car’
(534d) *Tikina too koti, ka kaakahu ai, uru atu ki roto i te motokaa

1.3.1.1.2 but-coordination
Engari is used for but-coordination, and occurs between the two conjuncts. In the vast
majority of cases, consultants accept either order for the conjuncts (provided restrictions
on anaphora are not contravened), even in instances where the English conjuncts are not
readily reversible, as the second pair of examples below shows, eg.
(535a) Taimaha te tataa o te ua, engari kaahore i
heavy the beat gen the rain but neg T/A
roa ka mutu
long T/A finished
‘The rain was heavy but it didn’t last long’

or
(535b) Kaahore i roa ka mutu te ua, engari
neg T/A long T/A finished the rain but
taimaha te tataa
heavy the beat
 Maori 124

‘The rain didn’t last long, but it was heavy’

(536a) I karanga atu au ki a ia kia
T/A call away Isg to pers IIIsg subj
aawhina·tia ahau, engari kaahore i taea
help·pass. Isg but neg T/A able
‘I asked her to help me, but she wasn’t able to’

compare
(536b) Kaahore ia i taea, engari ka karanga atu
neg IIIsg T/A able but T/A call away
ahau ki a ia kia aawhina·tia ahau
Isg to pers IIIsg subj help·pass. Isg
‘I asked her to help me, but she wasn’t able to’

1.3.1.1.3 or-coordination
Raanei is used for or-coordination. It usually occurs in the final conjunct, usually
following the verb (or predicate in non-verbal sentences). It may also occur in any
preceding conjunct(s) in the corresponding sentence position, eg.
(537) Me haere atu (raanei) ia ki toou kaainga,
oblig move away or IIIsg to sggenIIsg home
me haere mai raanei koe ki konei
oblig move hither or IIsg to here
‘He could come to your house or you could come here’
(538) He hoa, he hoariri raanei ia?
cls friend cls foe or IIIsg
‘Is he a friend or foe?’

However, it is also possible to have or coordination by implication, with no formal

marker. It is thus formally indistinct from and coordination, see the example in 1.3.1.2
below.

1.3.1.1.4 Other types of coordination

There do not appear to be any further types of coordination in Maori.

1.3.1.2 Position and number of coordinators

There is some doubt over and-coordinators. Aa appears to function in this way on some
occasions, but not always. Hoki appears not to be a coordinator. Ø, with non-final
intonation at the end of the first conjunct is the commonest coordinator. There appear to
be no restrictions on the number of conjuncts which may be coordinated by juxtaposition.
In but-coordination, there is always one coordinator, engari, and my consultants rejected
coordination with engari involving more than two conjuncts (see also Reedy, 1979, 122).
In orcoordination, one coordinator is usual, though not obligatory, regardless of the
 Syntax 125

number of conjuncts, but raanei may occur in any combination of conjuncts. (Reedy
(1979, 116ff) regards the final occurrence of raanei as obligatory.) These points are
illustrated in the following examples, in addition to those in 1.3.1.1:
(539) Ka mahi·a e ia te kaapura, ka
T/A make·pass. by IIIsg the fire T/A
whakamaoa·tia ngaa kai e ngaa waahine, ka
cook·pass. the(pl) food by the(pl) women T/A
noho raatou, ka kai katoa
sit IIIpl T/A eat all
‘He built a fire and the women cooked the food and they sat down and all ate [their meal]’
(540) Me turituri (#1) koe, me haere (#2) koe ki
oblig quiet IIsg oblig move IIsg to
too ruuma, me haere atu (#3a) koe ki
sggenIIsg room oblig move away IIsg to
waho (#3b)
outside
‘Either be quiet or go to your room or go outside’

In (540), raanei can potentially occur in the positions marked (#). Note that there are two
possible positions in the third conjunct—only one, not both, can be filled on any one
occasion. My consultant readily accepted all the following: (i) no occurrence of raanei;
(ii) raanei in (#3) only; (iii) raanei in (#1) and (#2); (iv) raanei in (#1) and (#3); (v)
raanei in (#2) and (#3); (vi) raanei in (#1), (#2) and (#3). There was a little more doubt
over (vii) raanei in (#1) only and (viii) raanei in (#2) only—these were clearly less
‘normal’ than patterns involving raanei in (#3).

1.3.1.3 Coordination of major sentence categories

As the different types of coordination involve different processes, this material is
presented in three subsections parallel to those in 1.3.1.1.

1.3.1.3.1 And-coordination
This is a complex area in Maori, where the rules depend in part on the function of the
item, and in part on its form.
Predicates, both verbal and non-verbal are coordinated by juxtaposition with no
coordinator. Aa sometimes occurs between the conjuncts.
(541) I te poo roa, ka waiata, ka kanikani a
at the night long T/A sing T/A dance pers
raatou
IIIpl
‘They sang and danced all night long’
(542) Kei tooku whare, kei tooku ruma mahi
at(pres) sggenIsg house at(pres) sggenIsg room work
hoki aku pukapuka
also plgenIsg book
 Maori 126

‘My books are at home and at my office’

(543) Tino aataahua, tino pai ia
very beautiful very good IIIsg
‘She is very beautiful and very good’

It is not essential to have hoki in (542), although it clearly makes the sentence easier to
process.
Unless proper names are involved, NPs introduced by prepositions are coordinated by
juxtaposition, with the preposition repeated:
(544) Ka whakamihi raatou ki te nui, ki te pai hoki
T/A praise IIIpl to the big to the good also
o ngaa kai
gen the(pl) food
‘They praised the abundance and the quality of the food’
(545) Ka kite raatou i te ngaawhaa, i te whare
T/A see IIIpl DO the boiling pools DO the house
whakairo, i te aha, i te aha
carved DO the what DO the what
‘They saw the boiling pools, the carved houses and other things’ (TR2, 102)

Hoki in (544) makes for ease of comprehension, and is used in any case where structural
ambiguity could otherwise result
Subjects (ie. NPs without a preposition) are occasionally coordinated without any
marker, ie. simply by juxtaposition, but it is much more usual to conjoin subjects by me
‘with’ or by the appropriate pronoun+ko (see below) eg.
(546) Kei te oma te maahita, (#) ngaa tamariki i
T/A run the teacher the(pl) children at
runga i te one
top at the beach
‘The teacher and the children are running along the beach’

The position marked (#) in (546) can be filled by Ø, in which case the conjoined NPs are
in separate breath groups, or by me or raatou ko (see below).
In the following instance, which involves coordinated sentences, rather than merely
subjects, but where the verb has been deleted under identity in the second conjunct, the
second subject is marked with topic-switch ko (see 1.12.1):
(547) I reira, ka tangi teetehi ki te toorino, ko
at there T/A sound a(sp) to the long flute spec
teetehi ki te kooauau
a(sp) to the short flute
‘There, one played the long flute and the other the short flute’ (H, 7)

Regardless of function, if two proper names are conjoined, or if a proper name and
pronoun are conjoined, there is a construction which may be specified as
(name+) appropriate pronoun+[ko+name]n
eg.
 Syntax 127

(548) Ko au te paapaa o Tuu raaua ko Mihi

eq Isg the father gen Tu IIIdl spec Mihi
‘I am the father of Tu and Mihi’
(549) Kei te noho a Tuu raatou ko Hine, ko Pau
T/A sit pers Tu IIIpl spec Hine spec Pau
‘Tu, Hine and Pau are sitting’
(550) Kei te haere raatou ko Mere maa ki te
T/A move IIIpl spec Mere and others to the
taaone
town
‘He/she and Mere and some others are going to town’
(551) Kei te aha koorua ko Tame?
T/A what IIdl spec Tame
‘What are you and Tame doing?’
(552) Kei te tatari koe i a maaua ko Tame?
T/A wait IIsg DO pers Idlexcl spec Tame
‘Are you waiting for Tame and me?’

Thus raaua (IIIdl) is used for coordination of two third persons, raatou (IIIpl) for three or
more third persons, koorua (IIdl) for addressee and one other, koutou (IIpl) for addressee
and two or more others, maaua (Idlexcl) for self and one other, maatou (Iplexcl) for self
and two or more others, and taatou (Iplincl) for self, addressee and one or more others.
While it is clear that ‘personal pronoun+ko’ coordination is basically restricted to
human referents, Reedy notes its extension to personifications and certain metaphorical
contexts (Reedy, 1979, 99ff). However, there is also a comitative preposition me in
Maori, which can be used to coordinate non-human entities, eg.
(553) Ka nui te miiti, te heeki, me te tuna
T/A big the meat the egg with the eel
‘There is plenty of meat, eggs and eels’

It appears to be particularly common as a coordinator of subjects, but not confined

exclusively to that position. Thus:
(554) He tohunga a Kupe moo ngaa mahi hanga
cls expert pers Kupe intgen the(pl) work build
whare whakairo, hahau waka, me eeraa atu mahi
house carve fashion canoe with those away work
o te tuawhenua
gen the land
‘Kupe was an expert at the activities of building meeting houses, fashioning canoes and
other activities of the land’ (KH, 1)

Nowadays, me is being used increasingly to coordinate NPs with human referents,

although this is still decried as non-traditional by some older speakers, eg.
(555) Moohio tonu te taakuta me ngaa neehi, kaaore
know sure the doctor with the(pl) nurse neg
 Maori 128

ia e ora
he T/A well
‘The doctor and nurses knew for sure that he wouldn’t recover’ (TR2, 57)

These older speakers prefer te taakuta raatou ko ngaa neehi. Nevertheless, me in such
constructions is widespread. It is, however, much more generally frowned upon as the
coordinator for proper names and pronouns, though it is heard. This is generally
considered a clear case of interference from English.
Me is also increasingly common as the coordinator of oblique NPs, instead of the
juxtaposition with repeated preposition described above, eg.
(556) I haere atu a ia ki Ingarangi me
T/A move away pers IIIsg to England with
Amerika
America
‘He travelled to England and America’

1.3.1.3.2 But—coordination
But-coordination of sentence constituents is not common: sometimes two-sentence
versions are used, and on other occasions, the alternative strategies below. However, in
the instance I contrived to elicit, engari was used as the coordinator, between two oblique
constituents:
(557) Kaahore ia i haere ki te kura, engari ki te
neg IIIsg T/A move to the school but to the
tau·nga o ngaa wakarererangi
land·nom gen the(pl) aeroplane
‘He went not to school, but to the airport’

In instances where but-coordination is avoided, it is sometimes treated as and-

coordination, with the contrast implicit, as in
(558) I paatere te miraka i a Mere, eehara
T/A spill the milk cause pers Mere neg
i a Hone
cause pers John
‘Not John, but Mary spilt the milk’ (more lit ‘Mary spilt the milk, not John’)

In other cases, the contrastive adverb kee is used:

(559) Kaahore ia i koorero i te pakiwaitara,
neg IIIsg T/A speak DO the story
mea waiata kee
thing sing contr
‘He didn’t read, but sang the story’ [more literally, ‘He didn’t read the story, it was a sung
thing instead’]

This is not necessarily an exhaustive list of the ways of avoiding engari as a constituent
coordinator.
 Syntax 129

1.3.1.3.3 Or—coordination
Or-coordination of sentence constituents appears to follow the same pattern as or-
coordination of sentences: raanei is normally added to the final conjunct, following the
first major constituent. However, it can be omitted, and the choice merely implied. Other
possible positions for raanei parallel those outlined for (540) in 1.3.1.2, eg.
(560) Haere mai koe ki konei, ki too kaainga, ki
move hither IIsg to here to sggenIIsg home to
te marae raanei?
the marae or
‘Will you come here, or shall I come to your place, or shall we meet at the marae?’
(561) Naa Mere, naa Hone raanei i paakarukaru
actgen Mary actgen John or T/A damage
‘[It] must have been damaged either by Mary or by John’
(562) Ka aawhina a Mere (raanei), a Marama raanei
T/A help pers Mary or pers Marama or
i a Pani
DO pers Pani
‘Either Mary or Marama will help Pani’

1.3.1.4 Comparison of the expression of coordination and

accompaniment
As will be clear from what was said in 1.3.1.1.1 and 1.3.1.3.1, coordination and
accompaniment basically did not coincide in older Maori, but there was always some
common ground. In modern Maori, there is a strong tendency for the comitative me to
take over the role of coordinator for many sentence constituents, although not, as far as I
am aware, to be used as a sentence coordinator, nor as a predicate coordinator. Examples
of the use of me from the older text Tu-Whakairi-ora, given here with H.W.Williams’
translations will indicate that, even in older texts, me occurred in places where
coordination is not readily distinguished from comitative:
(563) Ka tuu ngaa tamaahine i te mataihi katau o
T/A stand the(pl) daughter at the front right gen
te marae, me too raaua kookaa, me Hine-maurea
the marae with sggenIIIdl mother with Hine-maurea
‘His daughters stood at the right of the front of the house, in the court, with their mother
Hinemaurea’ (TWh, 19)
(564) Kei te poohiri te iwi me ngaa tamaahine
T/A greet the tribe with the(pl) daughter
‘The tribe with the daughters were waving a welcome’ (TWh, 20)
(565) Kei te tuu te iwi me ngaa tamaahine, kei te
T/A stand the tribe with the(pl) daughter T/A
wehi i a Te Aotakii
fear DO pers Te Aotaki
‘The tribe and the daughters were still standing, being in awe of Te Aotaki’ (TWh, 20)
 Maori 130

(566) Kei te titiro te iwi ki ngaa tohu o te rangatira,

T/A look the tribe to the(pl) sign gen the chief
o te toa, ki te taa-kootuku, ki te
gen the champion to the plume-white heron to the
pare-kaarearea, aapiti·tia ai, pou·a
crest-sparrow-hawk put side by side·pass. part. stick·pass.
ai ki te upoko, me te kaakahu paepaeroa,
part. to the head with the garment ornamented cloak
uhi·a iho te maahiti, me te
cover·pass. down the dog-hair cape with the
taiaha-o-kura ki te ringa
taiaha-gen-feather to the hand
‘The tribe was gazing at the emblems of high birth and bravery, the plumes of white crane,
and crest of sparrow-hawk feathers, ranged close together, and stuck into his hair; with the
highly ornamented cloak, and dogskin cape worn over it, and the decorated taiaha in his
hand’ (TWh, 20)

It is interesting to note, in this last example, that Williams felt it necessary to indicate the
change in Maori from coordination to concomitance in the English translation by means
of a semi-colon. He clearly felt that the two constructions were not on a par. From
examples such as those adduced here, it can be seen that it is not entirely influence from
English which has caused the change, although English is widely blamed. Also, it would
appear that non-personal subjects have always been coordinated with me, and this has no
doubt contributed significantly to the change in the use of me nowadays.
When I pursued the distinction between coordination and concomitance, my
consultant felt that the following two examples were exactly equivalent in what they
implied:
(567) I haere atu a Hone raaua ko Piri ki te
T/A move away pers John IIIdl spec Bill to the
tangihanga
tangi
‘John and Bill went to the tangi’
(568) I haere atu a Hone me Piri ki te
T/A move away pers John with Bill to the
tangihanga
tangi
‘John and Bill went to the tangi’

It is apparently not the case that the distinction between ‘John and Bill went to the tangi’
and ‘John went to the tangi with Bill’ can be captured by the distinction between raaua
ko and me.

1.3.1.5 Structural parallelism in coordination

While constituents are most readily coordinated if they have parallel structures, this is not
always an absolute requirement.
 Syntax 131

1.3.1.5.1 Coordination of adjectives and participial constructions

As Maori has no participial constructions, the question does not arise. However, it should
perhaps be remembered that modifying adjectives cannot be strung together in Maori, but
require a separate head, see 1.2.5.2.1.

1.3.1.5.2 Coordination of nouns and nominalized constructions

These can be coordinated, either by the more traditional method of juxtaposition, or with
me, eg.
(569) I rongonui·ngia a Kupe moo tana
T/A famous·pass. pers Kupe intgen sggenIIIsg
toa, moo/me tana hopu·tanga i te
bravery intgen/with sggenIIIsg catch·nom DO the
wheke
octopus
‘Kupe is famous for his bravery and his catching of the octopus’
(570) Koorero atu ana raaua ki too raaua paapaa ki ngaa
talk away T/A IIIdl to sggenIIIdl father to the(pl)
tohu o te rangatira, me ngaa tohu o te
sign gen the chief with the(pl) sign gen the
toa, me te whai tonu mai ia i muri
champion with the follow still hither IIIsg at behind
i a raaua
at pers IIIdl
‘They described to their father the emblems of high birth and bravery, and how he had
persisted in following after them’

Note that there is minimal nominalization of the final constituent here, since the subject is
retained in its basic form.

1.3.1.5.3 Coordination of different types of adverbial

The closed sets of adverbial particles cannot be coordinated with each other, or with other
adverbs. However, the open set of one-word adverbs can coordinate with prepositional
phrases and nominalizations, eg.
(571) Ka riingi mai ia aapoopoo me a te
T/A ring hither IIIsg tomorrow with at(fut) the
Raahoroi
Saturday
‘He will phone me tomorrow and on Saturday’
(572) Ka kite koe i a ia aapoopoo me te
T/A see IIsg DO pers IIIsg tomorrow with the
hoki·nga mai i Sydney
return·nom hither from Sydney
‘You will see her tomorrow, and when she returns from Sydney’
 Maori 132

Finite adverbial clauses can also be coordinated with prepositional phrases, eg.
(573) Inaa utu·a mai ahau, me a te
when pay·pass. hither Isg with at(fut) the
Kirihimete, maa·ku e hoko mai eetahi
Christmas intgen·Isg T/A buy hither some(pl)
putiputi
flower
‘When I get paid, and at Christmas, I’ll buy you some flowers’

Whether these constructions can be coordinated without me, I have been unable to
determine.

1.3.1.5.4 Coordination of active and passive verbs

It is impossible in Maori to coordinate verb forms: only verb phrases can be coordinated,
ie. constructions involving minimally a T/A marker and a lexical verb. But given the
normality of identical subject deletion, it is often not possible to determine a difference
between the coordination of two V constituents and the coordination of two sentences
with identical subject deletion. It is certainly possible to find examples of active and
passive V constituents coordinated with identical subject deletion, as in:
(574) I haere mai ia, ka whakaora·ngia
T/A move hither IIIsg T/A heal·pass.
‘He came and was healed’

This could be followed in discourse by:

(575) Ka whakaora·ngia, ka hoki ki te kaainga
T/A heal·pass. T/A return to the home
‘[He] was healed and returned home’

where neither conjunct requires a subject. However, it is not possible to replicate in

Maori the sort of construction in the English ‘Bill shot and was shot by Peter’.

1.3.1.5.5 Verb categories that cannot be coordinated

If the coordination of verbal constituents outlined in 1.3.1.5.4 counts as coordination of
verbs, then it can be said that it is possible to coordinate any verbal categories that can
share a common subject. (However, it might also be claimed that it is impossible to
coordinate any verbal categories.) Some examples showing different verb categories in
conjoined structures:
(576) Ka mutu, ka mihimihi ki te iwi
T/A finished T/A greet to the tribe
‘That ended, he greeted the tribe’ (TWh, 22)

In (576), mutu is a neuter verb, and mihimihi is transitive. The example is interesting,
because the verbs do not share the same subject, but both are derivable from context. It is
 Syntax 133

perhaps also worth noting that, while this is structurally coordination, it is not
semantically.
(577) Ka patu·a, ka mate
T/A slay·pass. T/A die
‘He was slain and died’ (TWh, 17)

In (577), patua is passive, mate stative intransitive.

(578) I tiimata te pikitia, ka mutu hoki i te
T/A start the picture T/A finish also at the
waa pai ki ahau
time good to Isg
The movie started and finished at times that suited me’

In (578), tiimata is intransitive, and mutu is neuter.

However, there are certain verb categories which are unlikely to share a common
subject, most notably neuter verbs and active transitives. Coordination of these does not,
therefore, occur, except under special circumstances like those in (576).

1.3.2.1–2 Sentence elements omissible under identity in coordination

Subjects are readily omitted under identity, eg.
(579) Ka haere ia, ka tae iho ki te wai, ka
T/A move IIIsg T/A arrive down to the water T/A
mea
say
‘He went, and arrived at the water, and spoke’ (H, 9)

The controller of the deletion does not have to be a subject itself, eg.
(580) I hoko mai a Rona i te ika, ka
T/A buy hither pers Rona DO the fish T/A
whakamaoa·tia (e ia)
cook·pass. by IIIsg
‘Rona bought the fish and cooked it’

In this example, the passive agent can be deleted under identity, as well.
Direct objects can be omitted under identity, eg.
(581) Ko Rona kei te whakatikatika i ngaa kai, ko
eq Rona T/A prepare DO the(pl) food eq
Hone kei te whakamaoa
John T/A cook
‘Rona is preparing and John is cooking the food’

However, without the ko-fronting of the subjects, this example was not judged
acceptable. Notional DOs are readily omitted under identity in the actor-emphatic, eg.
(582) Naa Hone i hopu te ika maa·ku, naa
actgen John T/A catch the fish intgen·Isg actgen
 Maori 134

Rona i whakamaoa
Rona T/A cook
‘John caught the fish for me, and Rona cooked it for me’

This last example will also serve to illustrate the deletion under identity of an oblique NP
(and note also the passive agent in (580) above), together with the following:
(583) I hoki mai a Hone maa runga i te
T/A return hither pers John actgen top at the
wakarererangi i Tauranga, aa, i hoki maa
plane from Tauranga and T/A return intgen
raro mai a Mere
underneath hither pers Mary
‘John returned from Tauranga by plane, and Mary returned on foot’

The naa/maa NP in the actor-emphatic can also be deleted under identity:

(584) Naa Hone i keri te whenua, i rui anoo
actgen John T/A dig the land T/A plant again
hoki ngaa kakano
also the(pl) seed
‘The ground was dug by Hone and the seeds were planted by him’

However, the support of anoo hoki appears to be required here to ensure that the second
conjunct is interpreted as an actor-emphatic construction.
Verbs are also fairly readily deleted under identity in actor-emphatic constructions, eg.
(585) Naa Rona ngaa kai i mau mai, naa
actgen Rona the(pl) food T/A bring hither actgen
Hone hoki ngaa wai inu
John also the(pl) water drink
‘Rona brought the food and John the drinks’

The verb can be deleted under identity in canonical verbal sentences also, but some
marker of the deletion is required, eg.
(586) Mau mai ana a Rona i ngaa kai, kei
bring hither T/A pers Rona DO the(pl) food prep?
a Hone hoki ngaa wai inu
pers John also the(pl) water drink
‘Rona brought the food and John the drinks’

The appropriate gloss for kei here is uncertain. It appears unlikely to be the locative kei,
because that is usually prohibited from past-time contexts, and it cannot be the mood
marker which accompanies predicates. The following textual example uses me for
support:
(587) Ka mutu, ka mihimihi ki te iwi, me te
T/A finished T/A greet to the tribe with the
iwi ki a ia
tribe to pers IIIsg
 Syntax 135

‘That ended, he greeted the tribe, and the tribe him’ (TWh, 22)

Thus it appears that any constituent of a verbal sentence can be omitted under identity.
In non-verbal sentences, it would appear that all types of predicate can be omitted, but
the support of (anoo) hoki is needed. The coordinator under these circumstances is me,
eg.
(588) He whero ngaa paatuu, me te tuanui (anoo) hoki
cls red the(pl) wall with the ceiling again also
‘The walls are red and the ceiling is too’
(589) Kei a Hone eetahi paoro, me a Mere
at(pres) pers John some(pl) balls with pers Mary
hoki
also
‘John has some balls and Mary has some too’

The subject of a classifying or prepositional sentence can be omitted, although (anoo)

hoki is required. There is no overt coordinator, eg.
(590) He kaiako a Mere, he whaea anoo hoki
cls teacher pers Mary cls mother again also
‘Mary is a teacher and a mother’
(591) Kei roto ngaa kapu i te kapata, kei
at(pres) inside the(pl) cup at the cupboard at(pres)
runga anoo hoki i te teepu
top again also at the table
‘The cups are in the cupboard and on the table too’

In equative sentences, it is possible to use the same construction as in (590), but

consultants felt that pronoun substitution was more natural:
(592a) Ko Hone te kaapene, te whurupeeke anoo hoki
eq John the captain the fullback again also
‘John is the captain and the fullback’
(592b) Ko Hone te kaapene, ko ia anoo hoki te
eq John the captain eq IIIsg again also the
whurupeeke
fullback
‘John is the captain and the fullback’

1.3.3 Elements omissible under coordination identity in:

1.3.3.1 Noun phrase

The head noun is omissible, eg.
(593) Taku hiahia he paraka whero me te kikurangi
sggenIsg wish a block red with the blue
 Maori 136

(hoki)
also
‘I want a red block and a blue one’
(594) Ko ngaa whakapapa o taku paapaa i te
top. the(pl) genealogy gen sggenIIIsg father at the
tuatahi; aa ko oo taku maamaa i muri mai
first and top. gen sggenIIIsg mother at after hither
‘The genealogy of my father is first, and then that of my mother follows’ (HHM, 1)

Note that after the preposition me, the indefinite determiner he is impossible, and is
replaced by te or teetahi.
The quantifier katoa is also omissible, although it is omissible from the first, rather
than the second conjunct, eg.
(595) I pau ngaa koti me ngaa pootae katoa
T/A exhaust the(pl) coat with the(pl) hat all
i ahau te kohikohi
cause Isg the collect
‘I collected all the hats and coats’

(Katoa can also be floated into the verb phrase.)

Determiners, including demonstratives and possessives, must be repeated. An
adjective cannot be deleted from the second conjunct, but can from the first. Thus to
ensure the reading where both ‘hat’ and ‘coat’ are ‘red’, in the following, whero must be
in both conjuncts or in the second:
(596) Taku hiahia he koti (whero) me te pootae
sggenIsg wish a coat red with the hat
whero
red
‘I want a red coat and hat’

Without the second whero, the colour of the hat is unspecified, and it appears that
prosodic means cannot be used to ensure the reading ‘red’ for both.

1.3.3.2 Adjective phrase

The head, together with any one-word pre- or post-modifiers can be omitted. The
coordination can be by juxtaposition, or with aa, but requires hoki for support, eg.
(597) Tino pai ia ki te kaukau, ki te whutupaoro
intens good IIIsg to the swim to the football
hoki
also
‘He is very good at swimming and at football’
(598) He toa rawa ia ki te kaukau, aa, ki te
cls champion intens IIIsg to the swim and to the
whutupaoro hoki
 Syntax 137

football also
‘He’s a real champion at swimming and at football’

A prepositional phrase expressing the field of application of two adjectives can be

omitted from the first conjunct, eg.
(599) I kakama ia, i pai (hoki) ki te raranga
T/A fast IIIsg T/A good also to the weave
‘She was very fast and neat at weaving’

If a modifying adverb occurs with the first of such adjectives, there is ambiguity as to
whether it applies to both adjectives, eg.
(600) Tino kakama ia, pai hoki ki te raranga
very fast IIIsg good also to the weave
‘She is very fast and very neat at weaving’ or ‘She is very fast and (not necessarily very)
neat at weaving’

1.3.3.3 Adverb phrase

With complex prepositional phrases, the adnominal phrase can be omitted under identity,
eg.
(601) Ka aahei raatou ki te haere ki waho, ki
T/A able IIIpl to the move to outside to
roto hoki i ngaa whare
inside also at the(pl) house
‘They are allowed to go in and out of the buildings’

The head locative phrase can also be omitted, eg.

(602) Kuhu·a ki raro i ngaa pira me ngaa
hide·pass. to underneath at the(pl) cushion with the(pl)
whaariki
mat
‘Hide them under the cushions and the mats’

However, it is not possible to omit the complement of a preposition:

(603) *I naa raro ia ki me i te mahi
T/A actgen underneath IIIsg to with from the work
‘He walks to and from work’

Other types of adverb phrase are so rarely found conjoined that I have been unable to
elicit any relevant data.
 Maori 138

1.4 NEGATION

1.4.1 The expression of sentence negation

Sentence negation in Maori is a rather complex phenomenon, involving a variety of
negators and a superficially unpredictable set of relations between the T/A markers of
negatives and affirmatives. The negative forms encountered are:

kaaore ~ kaahore ~ kaare ~ kaao (the last only as a reply to a question): the
commonest negator
kore: a strong negative, often equivalent to ‘never’
kaua ~ kauaka: to negate imperatives
eehara: to negate certain classes of non-verbal sentences
kiihai: associated only with i ‘past tense’, and used instead of kaahore
in this context only in some dialects
tee: found in some older texts, and still used in at least the Te Aupouri
dialect in some rather ill-defined contexts.

The relationships between the T/A markers of negatives and affirmatives can be listed
thus:
Affirmative Negative
i i
ka i/e (depending on time reference)
i te i te
kei te i te
e…ana e…ana
kua kia

Hohepa (1969a) has argued persuasively that negatives with kore and eehara in Maori
must be analyzed as involving two predications, and Biggs points out (1969, 76) that the
analysis can be extended to negatives with kaahore. In essence, the argument is that hara,
kore and hore have independent existence in Maori, as stative verbs, whose semantics
indicate falseness. In negative sentences in Maori, these stative verbs are preceded by a
tense-marker which is normally written as a part of the negator, except with kore, which
is the only negator which occurs with more than one T/A marker. Thus eehara is e+ hara
and kaahore is ka+hore. Hohepa argues that the affirmative sentence is embedded as a
subject clause in these negative predications. The subject of the affirmative is normally
(although not obligatorily) raised to become the surface subject of the negative
predication. (Waite, 1987 attributes this step in the argument to Chung, 1970.) While
there are certain details of Hohepa’s analysis which are not above question, the essence
 Syntax 139

of his outline undoubtedly holds. (Further discussion of this analysis can be found in
Bauer, 1981a, 95ff, Chung, 1978, 132ff and Waite, 1987, 79–85.)
Some examples of negatives (a) and corresponding affirmatives (b) are given to
facilitate further discussion.
Negatives with kaahore:
(604a) Kaahore a Hera i te whakarongo
neg pers Hera T/A listen
‘Hera is/was not listening’

negates both
(604bi) I te whakarongo a Hera
T/A listen pers Hera
‘Hera was listening’
(604bii) Kei te whakarongo a Hera
T/A listen pers Hera
‘Hera is listening’.
(605a) Kaahore taatou e haere ana aapoopoo
neg Iplincl T/A move T/A tomorrow
‘We are not going tomorrow’

negates
(605b) E haere. ana taatou aapoopoo
T/A move T/A Iplincl tomorrow
‘We are going tomorrow’
(606a) Kaahore anoo he taangata kia tae mai
neg yet a people subj arrive hither
‘Nobody has arrived yet’

negates
(606b) Kua tae mai he taangata
T/A arrive hither a people
‘Some people have arrived’

(See also (1875) in 2.1.3.2.1.2.) As will be seen from the examples, the negator kaahore
takes sentence-initial position, the affirmative subject is in second position, and the
affirmative verb and other constituents follow in the order of the affirmative. Kaahore
also negates the prepositional class of non-verbal sentences, with the exception of the
possessive prepositions, naa, maa, noo, moo. In practice this means that kaahore negates
the locative prepositions, eg.
(607a) Kaahore ia i te kura
neg IIIsg at(past) the school
‘He was not at school’

negates
 Maori 140

(607b) I te kura ia
at(past) the school IIIsg
‘He was at school’

An example without subject raising is

(608) Kaaore e tipu te hua whenua ki reira
neg T/A grow the fruit land to there
‘Vegetables will not grow there’ (TR2, 40)

It appears likely, on the basis of a small sample, that the subject is not raised only if it is
semantically prominent
One of the advantages of Hohepa’s bi-clausal analysis of negatives is that the
relationship between the tense markers of corresponding affirmatives and negatives can
then be seen as a reflection of the wider distribution of these particles in Maori. Thus ka
and kua are excluded from a variety of subordinate clause types, and replaced by i/e and
kia respectively. I have argued elsewhere (Bauer, 1981a, 172ff) that kei te and i te are
neutralized to i te in the environment of other T/A markers. Thus the observed
distribution of T/A markers is in conformity with the bi-clausal analysis, while it is not
well explained without that analysis.
Negatives with kore
(609a) E kore e roa, ka haere raatou
T/A neg T/A long T/A move IIIpl
‘It won’t be long before they go’

negates
(609b) Ka roa, ka haere raatou
T/A long T/A move IIIpl
‘It will be a long time before they go’;
(610a) Ka kore anoo au e rongo ki te reo
T/A neg again Isg T/A hear to the voice
wahine e karanga ana
woman T/A call T/A
‘I will never again hear the voices of women calling in welcome’

negates
(610b) Ka rongo anoo ahau i te reo wahine e
T/A hear again Isg DO the voice woman T/A
karanga ana
call T/A
‘I will once again hear the voices of women calling in welcome’

The change here from ki to i as the preposition marking the complement of rongo is not
readily explained, and is merely one manifestation of variation between these
prepositions which does not seem to be predictable, nor constant from one speaker to
another (see further 2.1.1.2.4, 2.1.1.5.1).
 Syntax 141

(611a) Kore rawa a Pou i whakarongo

neg intens pers Pou T/A listen
‘Pou didn’t listen at all’

negates
(611b) I whakarongo a Pou
T/A listen pers Pou
‘Pou listened’

Again subject raising is not obligatory:

(612) Kore rawa i whakahoki·a atu e Maahia ngaa
neg intens T/A return·pass. away by Mahia the(pl)
kura raa
red feather dist
‘Mahia never returned those red feathers’ (P, 3)

Hohepa notes (1969a, 31) that kore can be preceded by T/A markers other than e: he cites
kua, i and the subordinator kia. However, in the vast majority of my textual examples, it
is preceded by no T/A marker at all. This may be the result of the general demise of e as a
verbal particle in main clauses: I suggest that such clauses are best analyzed with an
elided e. However, some speakers feel that in such examples, the T/A marker ka is
required, which fits with the general pattern of ka replacing e as a marker for main
clauses. Note that only the T/A markers i and e occur in the embedded sentence under
kore.
Negatives with kaua:
(613a) Kaua e koorero teka!
neg T/A talk false
‘Don’t tell lies!’

negates
(613b) Koorero teka!
speak false
‘Tell lies!’

The following example, where the subject of the embedded clause is not the addressee,
shows subject raising, as with other negatives; the subject in such cases can also be ko-
fronted, appearing before kaua.
(614a) Kaua te tamaiti raa e piki ake i te taiapa
neg the child dist T/A climb up DO the fence
raa!
dist
‘The child must not climb that fence!’

negates
(614b) Me piki ake te tamaiti raa i te taiapa raa
 Maori 142

oblig climb up the child dist DO the fence dist

‘The child should climb that fence’

Subject raising also occurs when the negative imperative is directed to the addressee if
the embedded sentence is formulated as a passive, as in the following Biblical example
(thanks to Tipene Chrisp for drawing this to my attention):
(615) Kaua ahau e panga·a a te waa o te
neg Isg T/A throw·pass. at(fut) the time gen the
koroheke·tanga
old man·nom
‘Cast me not off in the time of old age’ (PT, Nga Waiata, 71:9)

Note also the following example with kaua embedded, corresponding to the affirmative
with an embedded me clause:
(616a) Ka mea ia kaua e hoko·na e raatou te
T/A say IIIsg neg T/A buy·pass. by IIIpl the
hooiho raa
horse dist
‘He said they should not buy that horse’

which negates
(616b) Ka mea ia, me hoko e raatou te hooiho
T/A say IIIsg oblig buy by IIIpl the horse
raa
dist
‘He said they should buy that horse’

In the positive the verb following me cannot have the passive termination, although the
remainder of the sentence is in effect passive, but the corresponding negative with kaua
does not show the same restriction, and the verb is passive in form.
As would be predicted from the association with imperatives, kaua always co-occurs
with the T/A marker e. Despite the fact that an analysis of kaua as ka+ua looks
superficially likely, Chung (1970) argues that it is monomorphemic, like kore, because it
can be preceded by kia in embedded clauses, which indicates that ka cannot be in the T/A
slot, eg.
(617) Engari, ka whakaaro·tia kia kaua e
but T/A decide·pass. subj neg T/A
whakamate·a maa te patu
kill·pass. intgen the club
‘However, it was decided not to kill him with a club’ (KWh, 1)

Negatives with eehara:

Eehara negates equative sentences, classifying sentences and also some prepositional
sentences. Examples of negative equative sentences:
(618a) Eehara teeraa i a Tamahae
neg that at pers Tamahae
 Syntax 143

‘That’s not Tamahae’

negates
(618b) Ko Tamahae teeraa
eq Tamahae that
‘That is Tamahae’;
(619a) Eehara ngaa kuki i a Pani maa
neg the(pl) cook at pers Pani and others
‘The cooks are not Pani and company’

negates
(619b) Ko Pani maa ngaa kuki
eq Pani and others the(pl) cook
‘Pani and company are the cooks’

As will be seen from the examples, negation of equative sentences does not follow
precisely the same pattern as negation of the verbal predicates with kaahore, kore and
kaua, which would give the ungrammatical (618c) instead of (618a):
(618c) *Eehara teeraa ko Tamahae
neg that eq Tamahae
‘That is not Tamahae’

The predicate takes the form i+NP, and I presume the i to be the neutral locative
preposition, though there is no good evidence to support that assertion.
Eehara is also used to negate classifying or attributive sentences. As with the negation
of equative sentences, the predicate of the affirmative is introduced by i. In addition, the
NP of this i-phrase is always introduced by te, which I take in this construction to be
generic (see 1.2 5.2.4, 2.1.1.13). Consider the following examples:
(620a) Eehara te whare i te whero
neg the house at the red
‘The house is not red’

negates
(620b) He whero te whare
cls red the house
‘The house is red’

Note that the raising of the subject (te whare) is not obligatory, although it is with
pronominal subjects, as in the following example:
(621a) Eehara raatou i te tamariki
neg IIIpl at the children
‘They are not children’

negates
 Maori 144

(621b) He tamariki raatou

cls children IIIpl
‘They are children’

Note the ungrammaticality of the following, relating to (620) and (621):

(620c) *Eehara te whare he whero
*Eehara te whare te whero
(621c) *Eehara raatou i ngaa tamariki
*Eehara raatou he tamariki

When eehara negates sentences with prepositional predicates introduced by maa, naa,
moo, noo, however, eehara most commonly merely precedes the corresponding
affirmative, without further changes, eg.
(622ai) Eehara naa·na te pukapuka nei
neg actgen·IIIsg the book proxI
‘This is not his book’

negates
(622b) Naa·na te pukapuka nei
actgen·IIIsg the book proxI
‘This is his book’

Subject raising can occur if the subject is non-pronominal, eg.

(623) Eehara te pukapuka nei naa te maahita raa
neg the book proxI actgen the teacher dist
‘This book doesn’t belong to that teacher’

However, some speakers use the same type of construction as for equative and classifying
sentences to negate prepositional sentences, eg.
(622aii) Eehara teenei pukapuka i a ia
neg this book at pers IIIsg
‘This is not his book’

Head (1989,90) states that this is used only by older speakers. Negatives with kiihai:
This past-tense-only negator is the norm in that context in Northern dialects, but is not
used in other dialects. One example will suffice:
(624a) I te whaanau·tanga o Maaui, kiihai i
at the birth·nom gen Maui neg T/A
piirangi toona whaea ki a ia
want sggenIIIsg mother to pers IIIsg
‘When Maui was born, his mother did not want him’ (KM, 1)

negates
(624b) I te whaanau·tanga o Maaui, i piirangi
 Syntax 145

at the birth·nom gen Maui T/A want

toona whaea ki a ia
sggenIIIsg mother to pers IIIsg
‘When Maui was born, his mother wanted him’

The main interest of the form kiihai is that it does not so obviously lend itself to analysis
as T/A marker+stative verb, although Hohepa has suggested an analysis involving
segment re-ordering, but without argument (Hohepa 1967, 35), and Chung has analyzed
it as kii+hai (Chung, 1970, 72–3). She links kiihai to an archaic negative cited by
Williams, hai, and suggests that the kii is related to the ki of ki te, although this does not
satisfactorily account for its restriction to cooccurrence with i. However, this does not
seem to be sufficient reason to jettison the higher predicate analysis of negation with
kiihai.
Negatives with tee:
Such negatives occur sparsely in older texts and are still used by at least Te Aupouri in
some contexts. It has not been possible to gather sufficient data to be sure of the
restrictions on this construction, but it seems to occur only in sentences containing reason
constituents. That this has not always been the case can be seen from textual examples
like the following:
(625) Aa, kimi noa, kimi noa, tee kite·a
then search intens search intens neg see·pass.
‘Then [they] searched and searched, but [she] was not to be
seen’ (R, 20)

Modern examples (Shane Jones, personal communication) are:

(626) He aha koe tee wawahi ai he wahie maa
cls what IIsg neg cut part. a firewood intgen
taatou?
Iplincl
‘Why didn’t you cut some firewood for us?’
(627) Noo teenei raa i mate ai taku matua,
actgen this day T/A die part. sggenIsg parent
naa konei ahau tee aahei te haere atu
actgen this Isg neg able the move away
‘My father died today, and that’s why I am not able to come’

From the data available it appears that this negative construction does not follow the
patterns described for other negators. In particular, it does not appear to take an
embedded sentence as its subject. It appears to occupy the T/A slot associated with the
main verb. Clark notes (1976, 87) that this appears to be a reflex of a negator which he
reconstructs as PPN *ta’e, still common in the Samoic subgroup. Clark concludes (1976,
95) that this negator was a higher verb in PPN, and was originally preceded and followed
by tense markers. No trace of these tense markers appears in the data available for Maori
tee.
 Maori 146

1.4.2 Constituent negation

In contrastive contexts, constituent negation uses the sentence negators kaahore, kore or
kiihai. However, there seems to be a preference for negating the full proposition, and
using engari ‘but’ with a positive constituent for contrast. Thus, while the (a) forms
below are possible, the (b) forms are apparently more natural forms of expression:
(628a) Ko Rona kee i koorero ahau, kaahore i
top. Rona contr T/A speak Isg neg DO
a Hata
pers Hata
‘I spoke to Rona, not Hata’
(628b) Kaahore ahau i koorero ki a Hata, engari
neg Isg T/A speak to pers Hata but
ki a Rona kee
to pers Rona contr
‘I spoke to Rona, not Hata’

(Note that in (628a), there appears to be DO fronting with ko, and a change from ki for
the recipient of the conversation to i. I have no explanation for either of these departures
from the expected, but my consultant rejected my “regularization”.)
(629a) Ko Rona kei te haere mai, kaahore a Hone
top. Rona T/A move hither neg pers John
‘Not John, but Rona will be coming’

(The negative constituent could also take the form e kore a Hone.)
(629b) Kaahore ko Hone ki te haere mai, engari
neg top. John to the move hither but
ko Rona
top. Rona
‘Not John, but Rona will be coming’

There are no negative proforms in Maori corresponding to forms like nobody, nothing in
English. To render such sentences as ‘I saw nobody’, a sentence negation is required, eg.
(630) Kaahore au i kite tangata
neg Isg T/A see person
‘I didn’t see any people’

However, kore forms negative compounds such as koretake ‘good-for-nothing’ (lit. ‘no
reason’), kore kai ‘refuse food’, ‘hungry’. (Williams, in his Dictionary, discusses kore as
a suffix, but I can find no trace of this in present-day Maori, where it appears to be a
prefix.) Such forms can be nouns, adjectives, verbs or adverbials, although most appear
to be lexicalized in just certain of these functions, eg.
(631) He koretake ia
cls good-for-nothing IIIsg
‘He is a good-for-nothing’
(632) I kaukau kore-kaakahu maatou
 Syntax 147

T/A swim neg-clothes Iplexcl

‘We swam naked’
(633) E kore kai ngaa taangata
T/A neg food the(pl) people
‘The people won’t eat’

Kore prefixation is probably not productive in Maori. When I enquired if I could render ‘I
came without a knife’ as
(634) I haere kore·maaripi ahau
T/A move neg-knife Isg
‘I came without a knife’

(as an alternative to the normal rendering

(635) Kaahore ahau i haere mai me te maaripi
neg Isg T/A move hither with the knife
‘I didn’t come with a knife’),

my suggestion evoked laughter, but I was told I “might get away with it”. I suspect that
native speakers could get away with this jocularly.

1.4.3 More than one negation element

It is impossible to have more than one sentence negation in a simple predicate. However,
one of the kore compounds can occur in a negated sentence, and the result is then
positive, eg.
(636) Kiihai maatou i kaukau kore-kakahu
neg Iplexcl T/A swim neg-clothes
‘We didn’t swim naked’

It is obviously rare to find a negative embedded predicate in a negative higher sentence,

and my consultant had great difficulty in processing such examples and in deciding
whether the result was positive or negative, although the grammaticality of the
construction did not seem to be in doubt, eg.
(637) ?Kaahore ia i hiahia kia kore ia e
neg IIIsg T/A desire subj neg IIIsg T/A
tuutaki i toona iraamutu
meet DO sggenIIIsg cousin
?‘She doesn’t want not to meet her cousin’

Accordingly, it is not possible to comment with any certainty on this topic.

 Maori 148

1.5 ANAPHORA

1.5.1 Means of expressing anaphora

In this area, as in many others, the options available for anaphoric reference to humans
differ from those available for non-humans to some extent. In particular, pronouns are
generally not available for non-humans. However, there are also other differences not so
readily accounted for.

1.5.1.1 Deletion
Deletion is the commonest method of making anaphoric reference in Maori, especially in
older narratives concerning humans once the identity of the participants is well-
established. In the vast majority of instances, deletion occurs from subject position, but
deletion of the passive agent is also frequent Probably the most extended example in the
literature is the one quoted by Hohepa (1970) and (1981) from the story of Wairangi,
concerning the flight of Parewhete. (The last part of this can be found in 1.12.3.) Another
example is:
(638) Te hoki·anga mai o Tama-te-kapua me te
the return·nom hither gen Tama-te-kapua with the
teina i te kaukau, karanga noa i
younger sibling from the swim call in vain DO
taa raaua mookai, kore·kore ana. Ui atu ki
sggenIIIdl pet neg·dup T/A ask away to
too raaua paapaa, kaaore hoki teeraa i moohio ki
sggenIIIdl father neg also that T/A know to
hea. Kaatahi ka kii atu ki a Tama-te-kapua
where then T/A say away to pers Tama-te-kapua
kia haere ki te paa o Uenuku raaua ko Toi,
subj move to the pa gen Uenuku IIIdl spec Toi
teeraa pea i whai i a raaua.
that perhaps T/A follow DO pers IIIdl
Ka tae ki te paa o Uenuku raaua ko Toi
T/A arrive to the pa gen Uenuku IIIdl spec Toi
ka tiimata te karanga haere i taa raaua mookai,
T/A start the call move DO sggenIIIdl pet
“E Poo, e Poo.” Ka tae ki waho o te
voc Po voc Po T/A arrive to outside gen the
whare o Toi ka rongo i te auee. Kaatahi
house gen Toi T/A hear DO the cry then
ka tomo atu ki roto me te karanga haere
T/A enter away to inside with the call move
atu, “E Poo, e Poo.”
away voc Po voc Po
‘When Tama-te-kapua and his younger brother returned from swimming, [they] called and
called their pet, but [it] was nowhere to be found. [They] asked their father, but he didn’t
know where [it] was either. Then [he] said to Tama-te-kapua that [he] should go to Uenuku
 Syntax 149

and Toi’s pa; perhaps it had followed them.

When [they] reached Uenuku and Toi’s pa, [they] started calling their pet as [they] went,
“Po, Po”. When [they] arrived
outside Toi’s house, [they] heard crying. Then [they] went in, calling as [they] went, “Po, Po.”’
(KP, 40–41)

Hohepa suggests that such sections of text marked by deletions constitute the natural
units of Maori discourse above the sentence. The progression from name (to pronoun) to
deletion is characteristic. In the text Wairangi, deletion accounted for about 60% of the
anaphoric references. Texts involving non-human participants do not have as high a
percentage of deletions, but deletion is typically the largest category.
While deletion is overwhelmingly used for human referents, it can also be used for
non-human entities, as the following passage shows:
(639) Ka whiu te kai a te tangata whenua, ka
T/A gather the food gen the people land T/A
haere te iwi raa ki te kai. He kotahi te move the tribe dist to the eat a one the
kuumara i roto i te rourou maa ngaa
kumara at inside at the basket intgen the(pl)
taangata toko·rua. Ka pau, ka noho i roto
people pnum·2 T/A exhausted T/A stay at inside
i too raatou whare.
at sggenIIIpl house
‘The home people prepared food and the visitors went to partake thereof. There was one
small round basket containing one kumara to two men. When [the food] was eaten they
rested in their house’ (W, 198)

The fact that deletion is not restricted to subjects and passive agents is shown by the
following excerpt discussing a pet shark:
(640) Ko te mahi hoki a toona ariki e haere
top. the work also gen sggenIIIsg master T/A move
tonu i ngaa raa katoa ki te titiro, ki te
intens at the(pl) day all to the look to the
whaangai, ki te karakia i ngaa karakia taniwha
feed to the chant DO the(pl) chant taniwha
‘Its master used to visit it every day to see how [it] was getting on and to feed [it], as well as
to recite his karakias over [it] such as were used for taniwhas’ (TP, 89)

1.5.1.2 Deletion with element marked on the verb

Not applicable.

1.5.1.3 Ordinary personal pronoun

This method of anaphoric reference is principally available for persons in Maori, (though
personification sometimes allows the category to be extended). Personal pronouns are
frequently found between the full name and deletions. They also tend to be used where
 Maori 150

there is more than one protagonist, or where another potential referent intervenes. In the
following texts, the relevant pronouns are in bold type:
(641) Kaatahi ka rere a Wairangi ki te tute i
then T/A leap pers Wairangi to the shove at
te whare, kore rawa i ngaoko. Kaatahi ia
the house neg intens T/A stir then IIIsg
ka whai kupu, “He whare koohuru teenei”
T/A utter word cls house murder this
‘Then Wairangi threw his weight against [the side of] the house [to shake it], but it never
yielded in the slightest. Then he spoke, “This is a house for murder”’ (W, 198)
(642) Kua moohio tonu hoki a Ao-kehu i haere
T/A know indeed still pers Ao-kehu T/A move
atu a Tamaahua-rererangi ki te tiki atu i
away pers Tamaahua-rererangi to the fetch away DO
a ia hei patu i taua taniwha
pers IIIsg for slay DO det aph taniwha
‘Indeed Ao-kehu had easily divined the object of Tamaahua-rererangi’s visit—that he came
to fetch him to slay the taniwha’ (TP, 91)

This particular text contains occasional uses of ia for the taniwha, eg.:
(643) Kaatahi ka tino moohio te taniwha nei, kua
then T/A intens know the taniwha proxI T/A
patu·a toona ariki e eetahi iwi, kua tae
kill·pass. sggenIIIsg chief by some(pl) tribe T/A arrive
hoki te tohu ki a ia
indeed the sign to pers IIIsg
‘He then felt quite sure that his master had been killed by strange people, because a sign had
come to him’ (TP, 89)

However, in both Wairangi and Tutae-poroporo, 66% of the total occurrences of

pronouns are possessive pronouns, where the option of deletion does not occur, and this
despite the fact that the pronouns form 16% of the anaphoric occurrences in Wairangi,
but 26% of the total in Tutae-poroporo. Thus possessive contexts provide the major locus
for pronominal anaphoric reference.
In addition, the form koia, which appears to contain ia ‘IIIsg’ is used to refer to
preceding text. (Note, however, that Williams, 1862, 20 classes ia as a demonstrative, as
well as a personal pronoun):
(644) Ko·ia anoo i roa ai te noho i te
eq·IIIsg emph T/A long part. the stay at the
raakau
tree
‘So that’s why [it] stayed so long in the tree’ (KM, 3)

Pronouns are, of course, a normal means of exophoric reference, ie. of referring to

entities not found in the text, but in the discourse situation.
 Syntax 151

1.5.1.4 Reflexive pronoun

There are no reflexive pronouns in Maori, although anoo ‘again’ (a reinforcing particle)
can be added to personal pronouns to serve this purpose, if required (see 1.6). It is typical,
however, that neither of the texts analyzed for anaphoric usage contained any occurrences
of reflexive anoo with a pronoun. It can also be added to non-pronominal NPs, see
1.5.1.5. Anake ‘alone’, ‘only’ is in some environments an alternative to anoo.

1.5.1.5 Special anaphoric pronoun

Maori has a special anaphoric determiner, taua (sg)/aua (pl), and there is also a special
anaphoric pronoun, reira, for reference to time and place (see 2.1.2.8.1). The extent to
which taua/aua are used seems to be a matter of stylistic preference: some people use
them very sparingly, while others use them quite widely. The two texts Wairangi and
Tutae-poroporo are representative of the two styles: there is one occurrence only in
Wairangi (out of a total of 175 anaphoric references), but taua/aua make up
approximately 14% of the anaphoric references in Tutae-poroporo. This may be in part a
reflection of the fact that Tutae-poroporo is frequently concerned with natural features
(eg. rivers), to which the possibility of reference by deletion and personal pronouns is
very limited, whereas Wairangi is principally concerned with human protagonists.
Taua/aua can be reinforced by anoo ‘again’, to give an expression with the force ‘that
very same’, or by the deictic particles nei, naa, raa, to give the equivalent of ‘that there’.
The following examples illustrate:
(645) Kaatahi ka mea atu a Tamaahua-rererangi ki
then T/A say away pers Tamaahua-rererangi to
a raatou, “kotahi te tangata i rongo i
pers IIIpl one the man T/A hear cause
a au; he toa taua tangata ki te patu
pers Isg a champion det aph man to the kill
taniwha…”
taniwha
‘Tamaahua said to them, “There is a man that I’ve heard of; he is skilled at killing
taniwhas”’(TP, 91)
(646) …kaatahi ka haere i roto i te awa o
then T/A move at inside at the river gen
Whanga-nui, ka ahu ki runga o taua
Whanganui T/A move directly to top gen det aph
awa, ka tae ki Te Paparoa—he taaheke kei
river T/A arrive to Te Paparoa a rapids at
runga o te awa o Whanganui—ka noho ki
top gen the river gen Whanganui T/A stay to
reira taua taitaahae nei
there det aph monster proxI
‘Then he went into the river, and went up that river, and reached Te Paparoa—a rapid in the
upper reaches of the Whanganui River—and there that monster stayed’ (TP, 90)

1.5.1.6 Demonstratives
 Maori 152

These are also used to provide anaphoric reference, either as determiners (eg. teenei
whare ‘this house’), as modifying adverbs, (eg. te whare nei ‘the house here’), or as
heads of NPs (teenei ‘this’). In the texts I examined, demonstratives accounted for 14–
16% of the anaphoric references. They are sometimes accompanied by the name of the
referent in apposition, if there could be any doubt as to the intended referent. They are
typically, but by no means exclusively, used when other referents intervene, eg.
(647) I a Wairangi e ngaro ana i Kaawhia, ka
at pers Wairangi T/A absent T/A at Kawhia T/A
tae mai teetehi tangata rangatira o Ngaati-Maru,
arrive hither a(sp) man chief gen Ngati-Maru
ko Tupeteka te ingoa, he whanaunga ki a
top. Tupeteka the name a relative to pers
Pare-whete. Ka noho manuhiri te tangata nei i
Pare-whete T/A stay visitor the man proxI at
Rurunui. E rua ngaa poo e noho ana i
Rurunui num 2 the(pl) night T/A stay T/A at
te kaainga, i te ata ka titiro atu a
the home at the morning T/A look away pers
Puuroku e piri ana te kookoowai i te
Puroku T/A stick T/A the red ochre at the
paapaaringa o Tupeteka, noo Pare-whete. Kua
cheek gen Tupeteka actgen Pare-whete T/A
moohio a Puuroku kua tae·a te wahine
know pers Puroku T/A take·pass. the woman
raa e Tupeteka
dist by Tupeteka
‘Whilst Wairangi was absent at Kawhia, there arrived a certain
man of rank of the Ngati-Maru tribe, Tupeteka by name, and he was related to Parewhete.
This man remained as a guest at Rurunui. When he had stayed two nights at the village, in
the morning, Puroku saw that the red ochre sticking to Tupeteka’s cheek belonged to
Parewhete. Puroku realized that that woman had yielded to Tupeteka’ (W, 197)
(648) “E tatari ana kia tae ake te ope, kei te
T/A wait T/A subj arrive down the group T/A
hoe ake i roto o Waihou. Ka tae
paddle down at inside gen Waihou T/A arrive
ake ka patu·a a Ngaati-Raukawa. Maa
down T/A kill·pass. pers Ngati-Raukawa intgen
teeraa kee te kai e mahi·a nei, maa
that contr the food T/A make·pass. proxI intgen
Ngaati-Maru.” “Aa, hei aawhea raa te tae
Ngati-Maru ah at(fut) when(fut) dist the arrive
mai ai kia hohoro ai te patu iho i
hither part. subj quick part. the kill down DO
eenei, i a Ngaati-Raukawa?”
these DO pers Ngati-Raukawa
“‘We are waiting for the party which is paddling up the Waihou River. When they arrive the
Ngati-Raukawa will be killed. The feast we are preparing is for them, the Ngati-Maru.”
“Ah, and when will they arrive so that we may speedily destroy these Ngati-Raukawa?”’(W,
199)
 Syntax 153

Demonstratives can be used exophorically, and they are the sole means of cataphoric
reference (ie. reference forwards in text) in the texts examined. Examples of cataphoric
reference are:
(649) Ko te whakapapa teenei o Ao-kehu…
top. the genealogy this gen Ao-kehu
‘This is Ao-kehu’s genealogy…’ (TP, 93)

where the following text outlines the genealogy, and

(650) Ko teenei tangata ko Wairangi noo Ngaati
top. this man top. Wairangi actgen Ngati-
Raukawa
Raukawa
‘This man, Wairangi, belonged to Ngati-Raukawa’ (W, 197)

1.5.1.7 Definite noun phrase

Definite noun phrases can refer anaphorically, especially if the head noun refers to an
inanimate, eg.
(651) Ko Wairangi te tangata whakamutunga ki te
top. Wairangi the man last to the
tomo i te whare. Ka titiro a Wairangi, ko
enter at the house T/A look pers Wairangi top.
ngaa poupou o te whare he koohurihuri
the(pl) upright gen the house cls sapling
kahikatea. Kaatahi ka rere a Wairangi ki te
white pine then T/A leap pers Wairangi to the
tute i te whare, kore rawa i ngaoko
shove DO the house neg intens T/A stir
‘Wairangi was the last man to enter the house. Wairangi noticed that the side-posts of the house
(supporting the rafters) were composed of white-pine saplings. Then Wairangi threw his weight
against the side of the house, but it didn’t yield in the slightest’ (W, 198)

1.5.1.8 Other forms

Maori also has a series of forms like too mua, too muri, which consist of the singular
determiner t-, the possessive oo, and a local noun. Mua is ‘the front’, muri ‘the rear’.
These forms can be used like ‘the former’ and ‘the latter’. They have plurals without the
t- prefix. Too runga, too raro can be used similarly for ‘the top one’, ‘the lower one’.
These can be used of both human and non-human referents, eg.
(652) E rua oo raaua whare, kei Pooneke teetahi,
num 2 plgenIIIdl house at(pres) Wellington one(sp)
aa kei Heretaunga teetahi. Ko too mua
and at(pres) Napier one(sp) top. sggen front
e riihi·tia ana, aa too muri e
T/A lease·pass. T/A and sggen behind T/A
noho·ia ana e raaua ake
 Maori 154

live·pass. T/A by IIIdl intens

‘They have two houses, one in Wellington and one in Napier. The former is rented out,
while they live in the latter’

1.5.2 Use of the various means of anaphora

1.5.2.1
Within the clause
The only ways to express anaphora within a clause are pronouns, or pronouns+the
reflexive particles. Contrary to the norms for Maori, personal pronouns appear to be used
to refer to non-humans in this context, eg.
(653) I tunu paraoa a Hata maa·na (anoo/anake)
T/A bake bread pers Hata intgen·IIIsg again/alone
(hoki)
also
‘Hata baked bread for himself’

(The sentence is grammatical with neither, either or both of the bracketed items.)
(654) E waiata ana te manu ki a ia
T/A sing T/A the bird to pers IIIsg
(anoo/anake)
again/alone
‘The bird sang to itself’

It appears that when the anaphoric form is a direct object, anake is not possible as an
alternative to anoo, and functions less semantically central to the verb are more likely to
occur with one or more of these support forms, and non-human referents also increase the
likelihood of a support form. The ki-phrases with experience verbs require anoo support,
although canonical DOs do not, and notional IOs do not. The antecedent normally
precedes the anaphor within the clause, see further 1.6. Compare
(655) I tapahi a Maramai i a iai (anoo)
T/A cut pers Marama DO pers IIIsg again
‘Marama cut herself’
(656) I tapahi a iai i a Maramak
T/A cut pers IIIsg DO pers Marama
‘Shei cut Maramak’

It is not possible for the ia in (656) to refer forwards to Marama.

Cataphoric reference within the clause is comparatively rare, but always uses the
demonstrative:
(657) Ko teenei tangata, ko Wairangi noo
top. this man top. Wairangi actgen
Ngaati-Raukawa
 Syntax 155

Ngati-Raukawa
‘This man, Wairangi, belonged to Ngati-Raukawa’ (W, 197)

1.5.2.2 Between coordinate structures

Some of the information relevant to this section has already been presented in 1.5.1, in
the discussion and examples of the individual means of anaphoric reference. Where there
is only one protagonist, and that protagonist is human, there is an overwhelming tendency
for deletion to occur, unless a possessive structure is involved. Where two possible
referents are present, deletion is not always possible because of resultant ambiguity,
although it occurs in (638), nor is it possible to pronominalize both NPs if they require
the same pronoun. Consider the following examples:
(658) Naa Honei a Pirik i meke, ka meke
actgen John pers Bill T/A punch T/A punch
anoo a Pirik i a Honei
again pers Bill DO pers John
‘John punched Bill and Bill punched John back’

The second conjunct cannot read

(658a) …*ka meke anoo iak i a iai
T/A punch again IIIsg DO pers IIIsg
‘…and he punched him back’

Nor can Hone be pronominalized:

(658b) …*ka meke anoo a Pirik i a iai
T/A punch again pers Bill DO pers IIIsg
‘…and Bill punched him back’

(although (658b) is possible meaning ‘Bill punched himself’.) Piri can be

pronominalized, but such examples are described as possibly confusing:
(658c) …ka meke anoo iak i a Honei
T/A punch again IIIsg DO pers John
‘…and he punched John back’

Pronominalization of Hone under these circumstances is only possible if the second

conjunct is passive:
(658d) …ka meke·a anoo ia e Piri
T/A punch·pass. again IIIsg by Bill
‘…and he was punched back by Bill’

Only the subject in this conjunct can be pronominalized.

If the second conjunct is also actor-emphatic in form, ie.
(658e) …naa Piri a Hone i meke anoo
actgen Bill pers John T/A punch again
‘…and Bill punched John back’
 Maori 156

Hone can readily be pronominalized:

(658f) …naa Piri ia i meke anoo
actgen Bill IIIsg T/A punch again
‘…Bill punched him back’

However, it is also possible (though perhaps somewhat marginal) to pronominalize the

actor-emphatic NP, Piri:
(658g) …naa·na a Hone i meke anoo
actgen·IIIsg pers John T/A punch again
‘…and he punched John back’

If both conjuncts are straightforwardly transitive, as in the following:

(659) E meke·meke ana a Hone i a Piri,
‘T/A punch·dup T/A pers John DO pers Bill
huri mai ana a Piri ki te meke·meke i
turn hither T/A pers Bill to the punch·dup DO
a Hone
pers John
‘John is hitting Bill, and Bill is hitting him back’

the most natural rendering is to pronominalize the final Hone: i a ia. It is possible to
pronominalize Piri in the second conjunct if Hone in the second conjunct is not
pronominalized. If both are pronominalized, the result is judged not ungrammatical but
confusing. However, if the NPs are pronominalized with different pronouns, all three
patterns of anaphoric pronominalization are possible, with pronominalization of both
judged best, eg. in
(660) E meke·meke ana a Hone i ngaa mahanga,
T/A punch·dup T/A pers John DO the(pl) twins
huri mai ana ngaa mahanga ki te meke·meke
turn hither T/A the(pl) twins to the punch·dup
i a Hone
DO pers John
‘John is punching the twins and they are punching him back’

the following pairs of possibilities occur at the two bold type positions:
ngaa mahanga Hone
raaua ia
raaua Hone
ngaa mahanga ia

In contexts where deletion is not possible, because the information would be

unrecoverable (eg. unpredictable adverbial constituents), or not desirable because of
possible confusion, and where pronominalization is not possible because the referents are
nonhuman, then a demonstrative or the anaphoric determiner (t)aua can be used, eg.
 Syntax 157

(661) …ka tae ki Te Paparoa—he taaheke kei

T/A arrive to Te Paparoa cls rapid at(pres)
runga o te awa o Whanga-nui—ka noho ki
top gen the river gen Wanganui T/A stay to
reira taua taitaahae nei… Heoi, ka noho ka
there det aph monster proxI well T/A stay T/A
roa; kaatahi ka hoe mai ngaa waka o
long then T/A paddle hither the(pl) canoe gen
runga o taua awa ki waho, ka tae ki
top gen det aph river to outside T/A arrive to
te waahi i noho ai taua nanakia nei,
the place T/A stay part. det aph monster proxI
kaatahi ka whawhati·a mai;…
then T/A snap·pass. hither
‘…and reached Te Paparoa, a rapid in the upper part of the Wanganui River, and there this
monster stayed…Well, he remained there for a long time; then the canoes from up the
river came paddling down, and reached the place where that monster was, then they were
snapped up ’ (TP, 90)

(Note that taniwha, taitahae and nanakia are used interchangeably throughout this story
to refer to the taniwha.) Sometimes demonstratives, pronouns or the anaphoric determiner
are used where deletion is possible, and no confusion would result, eg. taua taitaahae nei
above, so no absolute rules for determining the use of these forms can be postulated.

1.5.2.3 Between superordinate and subordinate clauses

In noun clauses, the choice between the complementizers kia and ki te depends on
whether the subjects in the main and subordinate clauses are identical or not (see
1.1.2.2.2.3, 1.1.2.2.2.4). If they are identical, ki te is used, and the subject of the
subordinate clause is obligatorily deleted. If the complementizer is kia, and the main
clause subject is coreferential with an NP in the kia-clause, the subordinate clause NP is
pronominalized, eg.
(662) I hiahia a Hone kia aawhina a Pita i
T/A desire pers John subj help pers Peter DO
a ia
pers IIIsg
‘Hone wanted Peter to help him’

(Pronominalization is also used as an alternative to deletion in those unexplained

instances where kia introduces a clause with a subject identical to that in the main clause,
see 1.1.2.2.2.3, (175), (177).
Where the noun clause is a full sentence, pronominalization of coreferential NPs
occurs in the subordinate clause, regardless of the syntactic function of these NPs, eg.
(663) I moohio a Hone, ka aawhina·tia ia e
T/A know pers John T/A help·pass. IIIsg by
Pita
Peter
 Maori 158

‘John knew that Peter would help him’

(664) I kite·a e Hone i mauaahara a Pita ki
T/A see·pass. by John T/A bear grudge pers Peter to
a ia
pers IIIsg
‘John discovered that Peter bore a grudge against him’

In ki te clauses, the rules that apply to simple sentences appear to hold, eg.
(665) I whakaaro a Hone ki te pupuhi i a
T/A decide pers John to the shoot DO pers
ia (anoo)
IIIsg again
‘John decided to shoot himself’

If a nominalization is embedded in a sentence with two possible referents, ambiguities

result from pronominalization, eg.
(666) Tino kino a Hone ki a Pita moo tana
very bad pers John to pers Peter intgen sggenIIIsg
tinihanga i a ia
cheat DO pers IIIsg
‘John hated Peter for his tricking of him/himself’

This is ambiguous as to whether Peter tricked himself or John. Tana, however, can
apparently only be understood as referring to Peter. With embedded nominalizations, the
antecedent must be in the main clause and the anaphor in the subordinate clause. In
(667) I whakatuu ia i te patu·nga e Hone o
T/A boast IIIsg DO the kill·nom by John gen
te mango
the shark
‘He boasted about John’s killing of the shark’

ia cannot be understood as Hone.

In all the instances discussed so far in 1.5.2.3, there are no alternative orders of the
clauses possible.

1.5.2.3.1 In the order superordinate clause—subordinate clause

All the previous material in 1.5.2.3 is relevant here. Consider also the following adverbial
examples, which contrast with 1.5.2.3.2, where examples have parallel numbers. (668) is
a nominalization:
(668a) I haere mai a Pitai i te karanga a
T/A move hither pers Peter at the call gen
Hone (i a iai)
John DO pers IIIsg
‘Peter came when John asked him’
 Syntax 159

(669a) Ka aawhina a Honei meina ka karanga

T/A help pers John if T/A call
mai a Pita (i a iai)
hither pers Peter DO pers IIIsg
‘John will help if Peter asks him’
(669b) Ka aawhina a iai meina ka karanga mai
T/A help pers IIIsg if T/A call hither
a Pita i a Honei
pers Peter DO pers John
‘John will help if Peter asks him’

Note that the bracketed constituents in these examples can be omitted, ie. deletion is an
alternative to pronominalization. Note also that (669a) and (669b) show that, in these
clauses, pronominalization can work in either direction. The same is true of the
following, in a different clause type, and with actor-emphatic structure:
(670a) Maa Honei e whakatika te tuuru kia
intgen John T/A mend the chair subj
tae mai iai
arrive hither IIIsg
‘John will mend the chair when he arrives’
(670b) Maa·nai e whakatika te tuuru, kia tae
intgen·IIIsg T/A mend the chair subj arrive
mai a Honei
hither pers John
‘John will mend the chair when he arrives’

Deletion is not an alternative to pronominalization in (670a), possibly because the

antecedent there is not the subject of the main clause.

1.5.2.3.2 In the order subordinate clause—superordinate clause

Compare the examples in 1.5.2.3.1 with the corresponding examples below, which have
been numbered to facilitate the comparison:
(668b) I te karanga a Hone ki a Pitai, ka haere
at the call gen John to pers Peter T/A move
mai iai
hither IIIsg
‘When John asked Peter, he came’
(668c) I te karanga a Hone (i a iai), ka haere
at the call gen John DO pers IIIsg T/A move
mai a Pitai
hither pers Peter
‘When John asked him, Peter came’
(669c) Meina ka karanga mai a Pitai i a
if T/A call hither pers Peter DO pers
Hone, ka aawhina iai
 Maori 160

John T/A help IIIsg

‘If Peteri calls John, hei will help’
(669d) Meina ka karanga mai a Pita (i a iai),
if T/A call hither pers Peter DO pers IIIsg
ka aawhina a Honei
T/A help pers John
‘If Peter calls him, John will help’
(669e) Meina ka karanga mai iai i a Hone,
if T/A call hither IIIsg DO pers John
ka aawhina iai
T/A help IIIsg
‘If he calls John, he will help’
(670c) Kia tae mai a Honei maa·nai e
subj arrive hither pers John intgen·IIIsg T/A
whaka·tika te tuuru
cause-right the chair
‘When John arrives, he will mend the chair’
(670d) Kia tae mai iai, maa Honei e
subj arrive hither IIIsg intgen John T/A
whakatika te tuuru
mend the chair
‘When he arrives, John will mend the chair’

In all these cases, pronominalization can proceed in either direction. Deletion is not
possible in the actor-emphatic example (670c) because of the constraint against the
stranding of prepositions. The preposition in this construction cannot be omitted, as it can
in other types of construction. Deletion is not possible in (668b) and (669c), although it is
in (668c) and (669d). In (669e), the second ia cannot be deleted; if the first ia is deleted,
then the second ia is understood as referring to Hone, and not as coreferential with the
deleted subject

1.5.2.4 Between different subordinate clauses

If there is only one possible referent, pronominalization (or deletion from subject
position) occurs, eg.
(671) I whakaaro a Himii, ka hiahia iai ki te
T/A decide pers Jim T/A want IIIsg to the
tuutaki i te tangata i tinihanga i a iai
meet DO the man T/A cheat DO pers IIIsg
‘Jim decided he wanted to meet the man who had cheated him’

Neither ia can be deleted.

If there are two possible referents in the main clause, as in
(672) Naa Himii i karanga ki te tangatak kia riingi
actgen Jim T/A call to the man subj ring
 Syntax 161

mai iak muri tanak tae·nga ki te

hither IIIsg after sggenIIIsg arrrive·nom to the
kaainga
home
‘Jim asked the man to phone him after he got home’

the Maori equivalent is not ambiguous, as the English is. The only possible understanding
is that indicated by the identity subscripts. This is also true if the final clause is finite,
rather than a nominalization:
(672a) …inaa tae iak ki te kaainga
when arrive IIIsg to the home
‘…when he got home’

This is true even if Himi is introduced into the second clause, ie. if it reads
(672b) …kia riingi iak ki a Himi
subj ring IIIsg to pers Jim
‘…that he should phone Jim’

Compare these with the following:

(673) I karanga atu a Himii ki te tangata ka
T/A call away pers Jim to the man T/A
riingi atu muri tanai tae·nga ki te kaainga
ring away after sggenIIIsg arrive·nom to the home
‘Jim told the man he would ring after he got home’

The deleted subject of the second clause can only be understood as ‘Jim’.
Consider, in contrast, the following set (NB tangata functions as head of the
predicate):
(674a) I hiahia a Himi hei taakuta a Pirii
T/A desire pers Jim cls(fut) doctor pers Bill
inaa tangata iai
when adult IIIsg
‘Jim wants Bill to be a doctor when he grows up’
(674b) I hiahia a Himi, inaa tangata a Pirii,
T/A desire pers Jim when man pers Bill
hei taakuta iai
cls(fut) doctor IIIsg
‘Jim wants Bill to be a doctor when he grows up’
(674c) I hiahia a Himi, inaa tangata iai, hei
T/A desire pers Jim when man IIIsg cls(fut)
taakuta a Pirii
doctor pers Bill
‘Jim wants Bill to be a doctor when he grows up’

and the version with raising (not possible for all speakers):
 Maori 162

(674d) I hiahia a Himi i a Pirii hei

T/A desire pers Jim DO pers Bill cls(fut)
taakuta (iai) inaa tangata iai
doctor IIIsg when adult IIIsg
‘Jim wants Bill to be a doctor when he grows up’

(Deletion of the raised-to-DO subject is evidently not compulsory—my consultant felt

that, if the speaker was being emphatic, or doubly cautious to ensure the message was
understood, the pronominal copy could occur.) The other logically possible combination
of antecedent and anaphor does not mean the same, as the translation below indicates:
(674e) I hiahia a Himii hei taakuta iai inaa
T/A desire pers Jim cls(fut) doctor IIIsg when
tangata a Piri
adult pers Bill
‘Jim wants to be a doctor when Bill grows up’

1.5.2.5 Between different sentences

Much data relevant to this section has already been presented in the textual examples in
1.5.1. The following examples indicate the situation when there are two possible
referents:

Because Piri is not the subject of the second sentence in the Maori (oti is a neuter verb
(see 1.2.1.2.2), and the second sentence is more literally translated The harvesting of the
vegetables had already been completed by Bill for her’), a simple pronoun is understood
as coreferential with the subject of the first sentence, and anoo is required to secure
reference to the nearest antecedent Compare (675) with the following:
 Syntax 163

Here both subjects are ko-fronted, but this does not affect the anaphoric processes.
Neither a simple pronoun nor pronoun+anoo can refer back to the first subject across the
intervening same-clause subject.

1.5.3 Elements next to complementizers

In Maori, with VSO word order, it is normally the verb which occurs next to
complementizers. However, if ko-fronting or the actor-emphatic puts an NP into this
position, normal pronominalization can occur. Deletion cannot, as these special
constructions are for emphasis of the NP:
(677) Me riingi mai a Hone meina ko ia te
oblig ring hither pers John if top. IIIsg the
hiahia ki te motokaa
wish to the car
or
…meina noo·na te hiahia ki te motokaa
if actgen·IIIsg the wish to the car
‘John will have to phone me if he wants the car’

(Note that neither if-clause in the Maori is a finite verbal construction.)

(678) Ko Hone te tangata pirau meina naa·na i
top. John the man rotten if actgen·IIIsg T/A
patu te kurii
kill the dog
‘John’s a rotter if he killed the dog’

1.6 REFLEXIVES

1.6.1 The expression of reflexivity

Reflexivity is expressed in Maori either by ordinary non-reflexive personal pronouns, or
by these pronouns+anoo ‘again’ or (an)ake ‘only’. (Note that the form ake is
homophonous with ake ‘up’, but since consultants claim that, in reflexive contexts, it is
“the same” as anake ‘only’, I will presume that it is a clipped form of anake, and gloss it
 Maori 164

accordingly.) Possessive pronouns cannot be followed directly by these support forms,

although (an)ake can sometimes occur following a head noun preceded by a possessive
and have the force of ‘own’.

1.6.2 The scope of reflexivity

Since anaphora and reflexivity are not clearly distinguished, the scope of reflexivity is
unclear. It appears that pronoun+anoo cannot refer to an NP in a previous sentence.
However, in instances of embedding with raising, identical-subject deletion or anaphoric
deletion, the antecedent will not necessarily appear in the same clause in surface
structure, eg.
(679) Ko taku hiahia ki a Mere kia titiro ki a
top. sggenIsg wish to pers Mary subj look to pers
ia anoo i roto i te karaehe whakaata
IIIsg again at inside at the glass reflection
‘I want Mary to look at herself in the mirror’
(680) I mea atu a Marama ki a Merei, kia
T/A say away pers Marama to pers Mary subj
tiaki i a iai anoo
care DO pers IIIsg again
‘Marama told Mary to look after herself

1.6.3 Reflexive element is a verbal affix

Not applicable.

1.6.4 Position of the reflexive element

The pronoun (+anoo/anake) which functions as a reflexive element in Maori occurs in
the same sentence position as non-reflexive items with that function.

1.6.5 Relations between antecedent and reflexive

Not all of the relations specified in the Questionnaire are relevant for the description of
Maori, and for this reason, the subsection numbers of the Questionnaire are not used
below. However, as far as possible, the Questionnaire’s order of relations is preserved.
All IOs are prepositional, so the category ‘IO (zero- or case-marking)’ does not exist
While IOs are not formally distinct from prepositional adverbials, these two categories
have nevertheless been treated distinctly here because of the possibility of differences in
semantic salience being allied to differences in accessibility to reflexives. Next, the
description ‘copular complement’ is inappropriate in relation to Maori, since it implies
the existence of a copula. However, since relevant structures exist, they are covered under
the category ‘non-verbal predicate’. The relations of ‘subject complement’ and ‘object
complement’ can be expressed in Maori by an embedded non-verbal predicate with hei.
These structures do not contain referential NPs, and so cannot take part in reflexive
 Syntax 165

relations. In addition, it is necessary to treat the NPs in the actor-emphatic construction. It

was advisable to find some neutral terminology for the NPs in this construction for
labelling purposes, since there is no agreement in scholarly circles on the grammatical
relations involved (see 1.11.2.1.7): the semantic agent is labelled NPA(gent), and the
notional DO is labelled NPP(atient). Because many of the relations do not co-occur, the
data is presented in three separate grids: canonical verbal sentences, actor-emphatic
sentences and non-verbal sentences. Numbers in the grids refer to the examples which

1.6.5A Reflexives in canonical verbal sentences

Table 4: Reflexives in canonical verbal sentences
Reflexives
Antecedents Su Mod DO Mod IO Mod Ag Mod Adv Mod
↓ Su DO IO Ag Adv
Su 681 682 683 684 * 685 686 687

(+) (+) + (+) (+) + (+)

Mod 688 689 690 691 692 693 694 695
Su
DO * 696 697 698 699 700

(+) + (+) + (+)

Mod * * * * 701 702
DO
IO 703 * 704 * 705 706 707

+ +
Mod * * * * * * * *
IO
Ag 708 * 709 710 711

(+) (+)
Mod * * * * 712 713
Ag
Adv * 714 * * * 715 * 716

Mod * * * * 717 718 * *

Ag
Key: Adv=prepositional adverbial phrase; Ag=passive agent; DO= direct object; IO=indirect
object; Mod=modifier of head of; Su= subject;+= requires a support form; (+)=optionally takes a
support form; *=reflexive relation cannot occur; =logically impossible
 Maori 166

follow. It seems important to indicate those places where Maori uses a plain personal
pronoun as opposed to a pronoun+a support form. This has been shown below by adding
a ‘+’ to those example numbers which require the support forms, and ‘(+)’ to those which
may optionally take them. Reflexives are listed horizontally, and antecedents vertically.

While the judgements given about coreference do not permit the formulation of absolute
rules, there are a number of clear tendencies. Most importantly, the antecedent normally
precedes the reflexive. (In a few instances specified below in the discussion of the
examples, a reflexive relation was obtainable only if the NPs concerned were ordered to
conform to this principle.) The notable exception to this is that in many cases, it is
possible for a pronoun occurring in the modifier of a higher-ranked NP (especially the
subject modifier) to be understood as coreferential with a noun functioning as the head of
a lower-ranked NP, although no other reflexive relations are possible in that direction
between the constituents of those NPs. In some cases, a reflexive relation cannot exist
between particular constituent types if the pronoun has a potential referent in a more
accessible position, but can exist if there is no competing NP. I have indicated this in the
discussion of those instances where I had data to support this. It is likely that reflexive
relations are possible under these circumstances in other instances as well, but I lack the
data to demonstrate this. There is rather less regularity in the lower orders. This is
partially a reflection of the fact that IO is an artificial category for Maori. In many
instances, the only possible adverbial constituents were beneficiaries, which are not
clearly distinct from IOs, and may even be more salient: with hoatu ‘give’, for instance,
with the usual IO preposition ki ‘to’, the recipient may have temporary possession only,
whereas with maa ‘for’, the beneficiary is stipulated to have permanent ownership.
Examples for canonical verbal sentences
ANTECEDENT IS SUBJECT
Reflexive is DO:
(681) Kei te horoi a Mere i a ia (anoo/anake)
T/A wash pers Mary DO pers IIIsg again/only
‘Mary washed herself’

It appears that, without the support form, the sentence is ambiguous between a reflexive
and a non-reflexive reading.
Reflexive is Mod DO:
(682) Kei te horoi a Mere i oona kaakahu (anake)
T/A wash pers Mary DO plgenIIIsg clothes only
‘Mary is washing her (own) clothes’

Anoo ‘again’ cannot replace anake here, or in any modifier.

Reflexive is IO:
(683) I hoatu a Mere i te kete maa·na ake
T/A give pers Mary DO the kit intgen·IIIsg only
‘Mary gave herself the kit’

Anoo cannot replace ake here.

Reflexive is Mod IO:
 Syntax 167

(684) I hoatu a Mere i te keke ki tana

T/A give pers Mary DO the cake to sggenIIIsg
tama (anake)
son only
‘Mary gave her (own) son a cake’

Anoo here would function as an adverbial modifier to hoatu, and indicate repetition of the
action. If maa ‘for’ replaces ki, the sentence then implies that the cake was for the son to
eat. As it stands, it implies that it was possibly given for some other purpose.
Reflexive is Mod Ag:
(685) I puuhi·a a Hone e tana tuakana
T/A shoot·pass. pers John by sggenIIIsg older brother
(ake)
only
‘John was shot by his (own) brother’

Anoo is not possible as a replacement for ake.

Reflexive is Adv:
(686a) Ka hoko a Hone i te motokaa moo·na
T/A buy pers John DO the car intgen·IIIsg
anake
only
‘John will buy a car for himself’

Anake is obligatory here, and anoo is not possible as a substitute. However, in the
following example where the reflexive is in the cause-phrase with a stative verb, anoo is
compulsory and anake an impossible substitute:
(686b) I whara a Hone i a ia anoo
T/A injure pers John DO pers IIIsg again
‘John was injured by himself’

Reflexive is Mod Adv:

(687) Ka hoko a Hone i te irirangi moo
T/A buy pers John DO the radio intgen
tana motokaa (anake)
sggenIIIsg car only
‘John will buy the radio for his (own) car’

ANTECEDENT IS MODIFIER OF SUBJECT

Reflexive is DO:
(688) E patu ana te hoa o Hone i a ia
T/A beat T/A the friend gen John DO pers IIIsg
‘John’s friend is hitting him’

If anake is added, then ia refers to John’s friend. Anoo would indicate that the action took
place repeatedly.
 Maori 168

Reflexive is Mod DO:

This is not possible if the head of the subject phrase agrees in number with the DO
modifier. Thus in (689i), tana cannot refer to John.
(689i) E horoi ana te hoa o Hone i tana
T/A clean T/A the friend gen John DO sggenIIIsg
motokaa (anake)
car only
‘John’s friend is cleaning (only) his own car’

However, if the head of the subject phrase is dual or plural, the situation is different:
(689ii) E horoi ana ngaa hoa o Hone i
T/A clean T/A the(pl) friend gen John DO
tana motokaa
sggenIIIsg car
‘John’s friends are cleaning his car’

Reflexive is IO:
(690) I hoatu te tama a Hone i te kai
T/A give the son gen John DO the food
maa·na
intgen·IIIsg
‘John’s son gave him the food’

If anake is added, -na refers to te tama.

Reflexive is Mod IO:
The situation is parallel to (689):
(6911) I hoatu te tama a Hone i teetahi kai
T/A give the son gen John DO a(sp) food
maa toona hoa
intgen sggenIIIsg friend
‘John’s son gave his own friend food’

ie. toona can only refer back to te tama, compare

(691ii) I hoatu ngaa tama a Hone i teetahi kai
T/A give the(pl) son gen John DO a(sp) food
maa toona hoa
intgen sggenIIIsg friend
‘John’s sons gave his friend some food’

Reflexive is Ag:
(692) I puuhi·a te hoa o Hone e ia
T/A shoot·pass. the friend gen John by IIIsg
‘John’s friend was shot by him’

The addition of anoo makes ia refer to te hoa.

Reflexive is Mod Ag:
 Syntax 169

See also (689) and (691):

(693i) I puuhi·a te hoa o Hone e tana
T/A shoot·pass. the friend gen John by sggenIIIsg
paapaa
father
‘John’s friend was shot by his own father’

In (693i), tana can refer only to te hoa, but in (693ii), tana refers to Hone:
(693ii) I puuhi·a ngaa hoa o Hone e tana
T/A shoot·pass. the(pl) friend gen John by sggenIIIsg
paapaa
father
‘John’s friends were shot by his father’

Reflexive is Adv:
See also (689), (691), (693). In
(6941) I koorero teka noa te hoa o Hone ki
T/A talk false indeed the friend gen John to
a ia (anake)
pers IIIsg only
‘John’s friend lied to himself’

it is not possible for ia to refer to Hone, but compare

(694ii) I koorero teka noa ngaa hoa o Hone ki
T/A talk false indeed the(pl) friend gen John to
a ia (anoo)
pers IIIsg again
‘John’s friends lied to him’

Here ia refers to Hone, with or without the support of anoo. Anake was rejected here.
Reflexive is Mod Adv:
(695) Ka haere te hoa o Hone ki toona whare
T/A move the friend gen John to sggenIIIsg house
‘John’s friend came to his house’

(695) was deemed ambiguous, as the English is. However, the addition of anake
establishes te hoa as the antecedent of toona.
ANTECEDENT IS DO
Reflexive is Mod Su:
(696a) E meke·meke ana tana tama (anake) i
T/A punch·dup T/A sggenIIIsg son only DO
a Hone
pers John
‘His (own) son is hitting John’
 Maori 170

However, my consultant was much more reluctant to accept parallel examples with dual
or plural reflexives, eg. in
(696b) I aawhina aa raaua tamariki i a Hone
T/A help plgenIIIdl children DO pers John
raaua ko Mere
IIIdl spec Mary
‘Their children helped John and Mary’

a non-reflexive interpretation was deemed far more likely than a reflexive one. I have no
explanation to offer for this inconsistency.
Reflexive is IO:
(697) E whakamoohio ana ahau i a Hone ki a
T/A introduce T/A Isg DO pers John to pers
ia (anoo)
IIIsg again
‘I am introducing John to himself’

Although anoo was not judged compulsory here, it was strongly preferred.
Reflexive is Mod IO:
(698) E whakaari atu ana ahau i a Hone ki
T/A show away T/A Isg DO pers John to
toona paapaa
sggenIIIsg father
‘I showed John to his father’

The addition of anake would give the sense ‘only to his father’.
Reflexive is Adv:
(699) I koorero ahau ki a Hone moo·na
T/A talk Isg to pers John intgen·IIIsg
anoo/anake
again/only
‘I told John about himself’

Either anoo or anake is required here.

Reflexive is Mod Adv:
(700) I koorero ahau ki a Hone moo tana
T/A talk Isg to pers John intgen sggenIIIsg
iriiri·tanga
christen·nom
‘I told John about his christening’

It is possible to add anake in final position, but that requires the possessive form toona.
ANTECEDENT IS MOD DO
Reflexive is Adv:
(701) I koorero ahau ki te iraamutu o Hone
T/A talk Isg to the nephew gen John
 Syntax 171

intgen·IIIsg
moo·na
‘I told John’s nephew about him’

This is ambiguous as to whether -na refers to Hone or to some other previously

mentioned referent.
Reflexive is Mod Adv:
(702) I koorero ahau ki te iraamutu o Hone moo
T/A speak Isg to the nephew gen John intgen
toona iriiri·tanga
sggenIIIsg christen·nom
‘I told John’s nephew about his christening’

This is ambiguous: toona could refer either to Hone or to the iraamutu.

ANTECEDENT IS IO
Reflexive is Mod Su:
(703) I hoatu tana tama i te paihana ki a
T/A give sggenIIIsg son DO the poison to pers
Hone
John
‘His son gave John poison’

This is ambiguous: tana could refer to Hone or to some other previously mentioned
antecedent.
Reflexive is Mod DO:
(704) I hoatu ahau ki a Hone i toona maaripi
T/A give Isg to pers John DO sggenIIIsg knife
ake
only
‘I gave John his knife’

The reflexive reading is only possible if the IO constituent precedes the DO, and only
with the support form ake.
Reflexive is Ag:
(705) I hoatu te maaripi ki Hone e tana
T/A give the knife to pers John by sggenIIIsg
tuakana
brother
‘The knife was given to John by his brother’

Reflexive is Adv:
(706) I hoatu ahau i te maaripi ki a Hone
T/A give Isg DO the knife to pers John
maa·na anake
intgen·IIIsg only
 Maori 172

‘I gave John the knife for himself’

Reflexive is Mod Adv:

(707) I hoatu ahau i te maaripi ki a Hone
T/A give Isg DO the knife to pers John
maa tana tuakana
intgen sggenIIIsg brother
‘I gave John a knife for his brother’

ANTECEDENT IS MOD IO
There are no possible reflexive relations with the modifier of the IO as antecedent
ANTECEDENT IS AG:
Reflexive is Mod Su:
(708) I runga i te hee, ka puuhi·a e Hone
at top at the error T/A shoot·pass. by John
tana hoa
sggenIIIsg friend
‘By mistake, John’s friend was shot by him’

The reflexive reading is only possible if the agent phrase precedes the subject.
Reflexive is Mod IO:
(709a) I hoatu te maaripi ki toona hoa e Hone
T/A give the knife to sggenIIIsg friend by John
‘The knife was given to his friend by John’

This is ambiguous: -na could refer to Hone or to some other previously-mentioned

antecedent. However, with the phrases in the other order, and an appropriate support
form, a purely reflexive reading can be obtained:
(709b) I hoatu te maaripi e Hone ki toona hoa
T/A give the knife by John to sggenIIIsg friend
ake
only
‘The knife was given by John to his own friend’

Reflexive is Adv:
(710) I puuhi·a te kereruu e Hone maa·na (ake)
T/A shoot·pass. the pigeon by John intgen·IIIsg only
‘The pigeon was shot by John for himself’

The support form is required to ensure a reflexive reading; without it, the sentence is
ambiguous between a reflexive and a non-reflexive reading.
Reflexive is Mod Adv:
(711) I puuhi·a te kereruu e Hone maa tana
T/A shoot·pass. the pigeon by John intgen sggenIIIsg
hoa
 Syntax 173

friend
‘The pigeon was shot by John for his friend’

ANTECEDENT IS MOD AG:

Reflexive is Adv:
(712) I puuhi·a te kereruu e te hoa o Hone
T/A shoot·pass. the pigeon by the friend gen John
maa·na
intgen·IIIsg
‘The pigeon was shot by John’s friend for him’

The addition of the support form anoo to (700) would mean that -na referred to the hoa,
ie. would render ‘The pigeon was shot by John’s friend for himself’.
Reflexive is Mod Adv:
(713) I puuhi·a te kereruu e te hoa o Hone
T/A shoot·pass. the pigeon by the friend gen John
maa tana tuakana
intgen sggenIIIsg brother
‘The pigeon was shot by John’s friend for his brother’

This is ambiguous: tana could refer to Hone or to hoa.

ANTECEDENT IS ADV:
Reflexive is Mod Su:
(714) I koorero·rero tana tuakana moo Hone
T/A talk·dup sggenIIIsg brother intgen John
‘His brother talked about John’

This is ambiguous between a reflexive reading and a non-reflexive one.

Reflexive is Mod IO:
(715) I hoatu ahau i te maaripi ki tana hoa
T/A give Isg DO the knife to sggenIIIsg friend
maa Hone
intgen John
‘I gave the knife to his friend for John’

This is ambiguous between a reflexive reading and a non-reflexive one.

Reflexive is Mod Ag:
(716) I waiho·tia te kereruu ki a Hone e
T/A leave·pass. the pigeon to pers John by
tana hoa
sggenIIIsg friend
‘The pigeon was left with John by his friend’

ANTECEDENT IS MOD ADV:

Reflexive is IO:
 Maori 174

(717) E whakaari atu ana ahau i te maaripi ki

T/A show away T/A Isg DO the knife to
a ia i roto i te whare o Hone
pers IIIsg at inside at the house gen John
‘I was showing the knife to him at John’s house’

This is ambiguous between a reflexive reading and a non-reflexive one.

Reflexive is Mod IO:
(718) E whakaari atu ana ahau i te maaripi ki
T/A show away T/A Isg DO the knife to
tana hoa i roto i te whare o Hone
sggenIIIsg friend at inside at the house gen John
‘I was showing the knife to his friend at John’s house’

This is ambiguous between a reflexive reading and a non-reflexive one.

1.6.5B Reflexives in actor-emphatic sentences

Table 5: Reflexives in actor-emphatic sentences
Reflexives
Antecedents↓ NPA Mod NPP Mod IO Mod Adv Mod
NPA NPP IO Adv
NPA 719 720 721 722 723 724

+ + +
Mod NPA 725 726 727 * 728 *

NPP * 729 * 730 * 731

Mod NPP * 732 733 * * 734

IO * 735 * * 736 737

(+)
Mod IO * * * *

Adv * * * *

Mod Adv * * * *

Key: Adv=adverbial prepositional phrase; IO=indirect object; Mod= modifier of the head of;
NPA=the NP preceded by maa/naa; NPP= the unmarked NP; Su=subject;+= require support forms;
(+)= optionally take support forms; *=reflexive relation cannot occur; = pairing is logically
impossible. Empty boxes show the same distribution as in canonical transitive sentences.
 Syntax 175

In the actor-emphatic construction, the head of NPA can be the antecedent in a reflexive
relation with any other slot. Reflexives as heads require anoo support, while those in
modifiers do not. The modifier of NPA is accessible as the antecedent to other heads, but
not to modifiers lower than the subject NPP (which on some analyses is in an embedded
clause in deep structure, though it may be raised in surface structure) can be the
antecedent only if the reflexive is in a lower modifier. I have no explanation to offer for
the fact that the pattern for NPP modifiers, which appears to be the same, breaks down
according to my consultant’s judgement when the reflexive is in the IO slot. When the
antecedent is lower-ranked than NPP, reflexivization is generally not possible. The IO
and adverbial slots here behave as they do in canonical transitive sentences. Two
examples are given for comparison.
Examples for actor-emphatic sentences
ANTECEDENT IS NPA
Reflexive is NPP:
(719) Naa Hone i pupuhi ia anoo
actgen John T/A shoot IIIsg again
‘John shot himself’

Anoo is obligatory here to obtain a reflexive reading.

Reflexive is Mod NPP:
(720) Naa Hone i pupuhi tana tama
actgen John T/A shoot sggenIIIsg son
‘John shot his son’

No support forms are possible here: anake would give ‘his only son’.
Reflexive is IO:
(721) Naa Hone i hoatu he takoha maa·na anoo
actgen John T/A give a present intgen·IIIsg again
‘John gave himself a present’

Anoo is obligatory here to obtain a reflexive reading.

Reflexive is Mod IO:
(722) Naa Hone i hoatu he takoha ki tana
actgen John T/A give a present to sggenIIIsg
tama
son
‘John gave his son a present’

Reflexive is Adv:
(723) Naa Hone i pupuhi te manu maa·na
actgen John T/A shoot the bird intgen·IIIsg
anoo/anake
again/only
‘John shot the bird for himself’

One or other of the support forms is required for a reflexive reading.

 Maori 176

Reflexive is Mod Adv:

(724) Naa Hone i pupuhi te manu i roto i
actgen John T/A shoot the bird at inside at
tana maara
sggenIIIsg garden
‘John shot the bird in his garden’

ANTECEDENT IS MOD NPA

Reflexive is NPP:
(725) Naa te tuakana o Hone i pupuhi ia
actgen the brother gen John T/A shoot IIIsg
‘John’s brother shot him’

This is ambiguous between a reading where ia refers to Hone and where ia refers to a
third person. (Note also that tuakana is ‘older brother’, although for the sake of brevity,
this is not indicated in the gloss here or below.)
Reflexive is Mod NPP:
(726) Naa te tuakana o Hone i pupuhi tana
actgen the brother gen John T/A shoot sggenIIIsg
tama
son
‘John’s brother shot his son’

While the most likely reading is that where tana refers to tuakana, it could also be read as
referring to Hone, or to some third person.
Reflexive is IO:
(727) Naa te tuakana o Hone i hoatu he takoha
actgen the brother gen John T/A give a present
maa·na
intgen·IIIsg
‘John’s brother gave him a present’

This is ambiguous as to whether -na refers to Hone or some third person. The same is
true if the IO is of the form ki a ia ‘to him’, implying temporary possession only.
Reflexive is Adv:
(728) Naa te tuakana o Hone i pupuhi te manu
actgen the brother gen John T/A shoot the bird
maa·na
intgen·IIIsg
‘John’s brother shot the bird for him’

Here, -na could refer to the tuakana, to Hone or to a third person. Anoo and anake would
be read as sentence modifiers.
ANTECEDENT IS NPP
Reflexive is Mod NPA:
(729) Naa tana paapaa a Hone i pupuhi
 Syntax 177

actgen sggenIIIsg father pers John T/A shoot

‘His father shot John’

Reflexive is Mod IO:

(730) Naa Mere i hoatu a Hone ki tana
actgen Mary T/A give pers John to sggenIIIsg
paapaa
father
‘Mary gave John to his/her father’

This is ambiguous, as the English translations indicate.

Reflexive is Mod Adv:
(731) Naa Mere i pupuhi a Hone i roto i
actgen Mary T/A shoot pers John at inside at
tana maara
sggenIIIsg garden
‘Mary shot John in his/her garden’

Like (730), this is ambiguous.

ANTECEDENT IS MOD NPP
Reflexive is Mod NPA:
(732) Naa tana tuakana i pupuhi te manu a
actgen sggenIIIsg brother T/A shoot the bird gen
Hone
John
‘His brother shot John’s bird’

This is ambiguous. Tana could refer to Hone or to some third person previously
mentioned.
Reflexive is IO:
(733) Naa Piri i hoatu te manu a Hone ki a
actgen Bill T/A give the bird pers John to pers
ia
IIIsg
‘Bill gave John’s bird to him’

Reflexive is Mod Adv:

(734) Naa Piri i pupuhi te manu a Hone i
actgen Bill T/A shoot the bird pers John at
roto i tana maara
inside at sggenIIIsg garden
‘Bill shot John’s bird in his garden’

This is ambiguous: tana could refer either to Piri or to Hone.

ANTECEDENT IS IO
Reflexive is Mod NPA:
 Maori 178

(735) Naa tana tuakana i hoatu he takoha ki

actgen sggenIIIsg brother T/A give a present to
a Hone
pers John
‘His brother gave a present to John’

Reflexive is Adv:
(736) Naa Mere i hoatu te maaripi ki a Hone
actgen Mary T/A give the knife to pers John
maa·na (anoo)
intgen·IIIsg again
‘Mary gave John a knife for himself’

Anoo, while not essential, is preferred.

Reflexive is Mod Adv:
(737) Naa Mere i hoatu te maaripi ki a Hone
actgen Mary T/A give the knife to pers John
maa tana tuakana
intgen sggenIIIsg brother
‘Mary gave John a knife for his brother’

1.6.5C Reflexives in non-verbal sentences

Table 6: Reflexives in non-verbal sentences
Reflexives
Antecedents↓ N-V Mod Su Mod Adv Mod
Pred N-V Su Adv
Pred
N-V Pred 738 739 740 741

+ +
Mod N-V Pred 742 * 743 744

Su 745 746 747 748

+?
Mod Su * 749 * *

Adv * 750 751 *

?
Mod Adv * * * *

Key: N-V=non-verbal. Symbols as before. ?=some uncertainty exists over this pairing, see
examples. Other abbreviations as before (see 1.6.5A, 1.6.5B).
 Syntax 179

This is ambiguous: tana could refer to either Mere or Hone.

In non-verbal sentences, the predicate NP is more accessible to reflexivization than the

subject, at least as far as modifier position is concerned, and the modifier of the predicate
NP is the only position which can have a reflexive when the antecedent is in an adverbial
phrase.
Examples for non-verbal sentences
ANTECEDENT IS NON-VERBAL PREDICATE NP
Reflexive is Subject:
(738) I roto i te pikitia, ko Waka (anoo) ia (anoo)
at inside at the picture eq Waka again IIIsg again
‘In the film, Waka was himself’

This was judged easier to process if both anoo’s are present, though either (but not both)
can be omitted.
Reflexive is Mod Subject:
(739) Kei a Hone tana tuakana
at(pres) pers John sggenIIIsg brother
‘His brother is with John’

Reflexive is Adv:
(740) Kei a Hone he koti moo·na anoo
at(pres) pers John a coat intgen·IIIsg again
‘John has a coat for himself’

Reflexive is Mod Adv:

(741) Kei a Hone he koti moo tana tuakana
at(pres) pers John a coat intgen sggenIIIsg brother
‘John has a coat for his brother’

ANTECEDENT IS MOD NON-VERBAL PREDICATE NP

Reflexive is Subject:
(742) Kei te tuakana o Hone a ia
at(pres) the brother gen John pers IIIsg
‘He, [John], is with John’s brother’

This is ambiguous: ia could refer to Hone or to a third person previously mentioned.

Reflexive is Adv:
(743) Kei te tuakana o Hone he koti moo·na
at(pres) the brother gen John a coat intgen·IIIsg
‘John’s brother has a coat for him’

This is ambiguous: -na could refer to the tuakana (the most likely interpretation), or to
Hone, or to someone else previously mentioned.
Reflexive is Mod Adv:
 Maori 180

(744) Kei te tuakana o Hone he koti moo

at(pres) the brother gen John a coat intgen
tana hoa
sggenIIIsg friend
‘John’s brother has a coat for his friend’

Again, this is ambiguous: tana could refer to the tuakana, to Hone or to some third
person.
ANTECEDENT IS SUBJECT
Reflexive is non-verbal predicate NP:
(745) Ko ia a Hone
eq IIIsg pers John
‘John is him’

Reflexive is Mod non-verbal predicate NP:

(746) Kei tana kaainga a Hone
at(pres) sggenIIIsg home pers John
‘John is at his (own) house’

Reflexive is Adv:
(747) He maahita a Hone, i taana anoo
cls teacher pers John from sggenIIIsg again
‘John is a teacher, according to himself’

Even if i taana anoo is in initial position, the reflexive relation obtains. Note, however,
that this is a sentence modifier, rather than a predicate modifier, which proved impossible
to construct.
Reflexive is Mod Adv:
(748) He kai·mahi pai a Hone i tana kaainga
cls ag·work good pers John at sggenIIIsg home
‘John is a good worker at his home’

ANTECEDENT IS MOD SUBJECT

Reflexive is Mod non-verbal predicate NP:
(749) Kei tana tuakana te koti o Hone
at(pres) sggenIIIsg brother the coat gen John
‘His brother has John’s coat’

This is ambiguous: tana could refer to Hone or to a third person.

ANTECEDENT IS ADV:
Reflexive is Mod non-verbal predicate:
(750) Kei tana hoa he koti moo Hone
at(pres) sggenIIIsg friend a coat intgen John
‘His friend has a coat for John’
 Syntax 181

This is ambiguous between a reflexive and a non-reflexive reading.

Reflexive is Subject:
The only data involves a sentence modifier:
(751) He maahita ia i taa Hone
cls teacher IIIsg from sggen John
‘John is a teacher, according to himself’

1.6.6 Reflexive relations within nominalized clauses

Reflexive relations can exist within nominalized clauses, but the possibilities appear to be
far more restricted than in simple sentences. It appears that the only possible antecedent
is the underlying subject, and anoo support is often considered necessary in the
nominalization
where in the corresponding simple sentence it would be optional. Consider the
following examples:
(752) I mea mai a Tuu ki a maatou i te
T/A say hither pers Tu to pers Iplexcl DO the
koohuru·tanga o Pou i a ia anoo
kill·nom gen Pou DO pers IIIsg again
‘Tu told us abut Pou’s killing of himself’

In the corresponding simple sentence anoo would be optional.

(753) I mea mai a Tuu ki a maatou moo
T/A say hither pers Tu to pers Iplexcl intgen
te hopu·tanga o Pou i te hooiho moo·na
the catch·nom gen Pou DO the horse intgen·IIIsg
anoo
again
‘Tu told us about Pou’s catching a horse for himself’

Without anoo here, the sentence would be ambiguous as to whether the horse was for Tu
or Pou.
(754) I mea mai a Tuu ki a maatou moo
T/A say hither pers Tu to pers Iplexcl intgen
te mau·nga o Pou i a ia anoo i roto
the catch·nom gen Pou DO pers IIIsg again at inside
i tana rore
at sggenIIIsg trap
‘Tu told us about Pou catching himself in his own trap’

There is some doubt here as to whether tana could also refer to Tu.
Note the reflexivity relations in the following:
(755) I mea mai a Tuu ki a maatou i te
T/A say hither pers Tu to pers Iplexcl DO the
koohuru·tanga o Pou i tana tuakana
kill·nom gen Pou DO sggenIIIsg brother
 Maori 182

‘Tu told us about Pou’s killing of his brother’

In (755), tana can refer only to Tu, and not to Pou. Note also the impossibility of the
subject modifier as antecedent:
(756) I mea mai a Tuu ki a maatou moo
T/A say hither pers Tu to pers Iplexcl intgen
te hopu·tanga o te teina o Pou i te
the catch·nom gen the brother gen Pou DO the
hooiho moo·na (anoo)
horse intgen·IIIsg again
‘Tu told us about Pou’s younger brother catching a horse for himself’

With or without anoo, naa cannot refer to Pou, but refers to the teina.

1.6.7 Reflexive relations within ordinary NPs

It is not possible to have two possessive forms:
(757) *tana hoa o Hone
sggenIIIsg friend gen John
‘his, John’s, friend’

Nor is it possible to have a reflexive relation between a possessive determiner and an NP

in a relative clause. Thus in
(758) I kite ahau i tana koti i waiho·tia ki
T/A see Isg DO sggenIIIsg coat T/A leave·pass. to
konei e Hone
here by John
‘I saw his coat that was left here by John’

tana cannot be understood as John’s; it must be some other person’s.

However, when pronoun substitution occurs as a relativization strategy, reflexive
relations obtain with the antecedent (see also 1.1.2.3.4.2), eg.
(759) I kite ahau i te koti o Hone naa·na i
T/A see Isg DO the coat gen John actgen·IIIsg T/A
waiho ki konei
leave to here
‘I saw John’s coat which he left here’

Support forms like anoo are never used under these circumstances.
It is also possible for a prepositional phrase modifying a noun to be coreferential with
a genitive modifier, eg.
(760) Tino nui te aroha o te kootiro ki a ia
very big the love gen the girl to pers IIIsg
anoo
again
‘The girl’s love of herself is very great’
 Syntax 183

Without anoo, ia in (760) would not be understood reflexively.

(761) Tino nui te aroha o te kootiro ki tana
very big the love gen the girl to sggenIIIsg
tungaane
brother
‘The love of the girl for her brother is very great’
(762) He maaripi te takoha a Hone maa·na (anoo)
cls knife the gift gen John intgen·IIIsg again
‘John’s gift to himself was a knife’

In this last example, anoo was not judged obligatory, but its addition was clearly
preferred.

1.6.8 Reflexive structures without overt antecedent

First and second person reflexive structures occur without overt antecedents, but it is not
clear that third person structures can. The following examples illustrate:
(763) Tino pai te pakiwaitara moo·ku ake anoo
very good the story intgen·Isg only again
‘The story about myself was really good’

Either ake or anoo is possible here as an alternative to ake anoo.

(764) He poorangi ki te patu i a koe anoo
cls stupid to the kill DO pers IIsg again
‘It’s stupid to kill yourself/oneself

1.6.9 Other uses of reflexive forms

It will be clear from the comments elsewhere in 1.6 that the reflexive support forms all
have other uses.

1.6.9.1. Reflexive pronoun as emphatic pronoun

This is possible; the examples in 1.6.8 are illustrations.

1.6.9.2 Reflexive verb form as detransitivizer

Not applicable.

1.6.9.3 Other functions

Anoo indicates reprise or emphasis in many contexts other than reflexive ones (see
1.2.1.3.1.1, and also examples (383), (464), (465), (470), (644), (799)), eg.
(765) I karanga anoo ia
T/A call again IIIsg
 Maori 184

‘She called again’

(766) Kua kaa anoo ngaa raiti
T/A burn again the(pl) light
‘The lights have come on again’

It is standard in negations relating to kua, to mean ‘yet’:

(767) Kaahore anoo te raa kia whiti
neg again the sun subj rise
‘The sun is not yet up’
Anake is used for ‘only’ in non-reflexive contexts, eg.
(768) Ko Hata anake kua haere atu
top. Hata only T/A move away
‘Only Hata has left’

1.7 RECIPROCALS

1.7.1 The expression of reciprocity

Reciprocity is expressed in Maori by the use of the third person pronouns raaua (IIIdl)
and raatou (IIIpl)+anoo ‘again’, or by the related possessive pronouns oo/aa raaua,
oo/aa raatou either without a support form, or with anoo ‘again’ or occasionally (an)ake
‘only’ following the head noun. Reciprocals are thus not formally distinct from dual or
plural reflexives. Whether the sense of such constructions is understood as reciprocal or
reflexive, or as non-specific between the two depends on what makes best sense in real
world terms. Thus, presented with:
(769) I tunu keke a Hone raaua ko Mere maa
T/A bake cake pers John IIIdl spec Mary intgen
raaua anoo
IIIdl again
my consultant could understand this to mean
‘John and Mary (together) baked cakes for themselves’
or ‘John and Mary baked cakes for each other’
or ‘John and Mary each baked cakes for themselves’

In contrast, presented with (770), there was no doubt that the sense was reciprocal:
(770) Kei te tuu ngaa whaea o Hone raaua ko Mere
T/A stand the(pl) mother gen John IIIdl spec Mary
i te taha i a raaua anoo
at the side at pers IIIdl again
‘Mary and John’s mothers are standing beside each other’

However, in (771), the sense was taken as exclusively reflexive:

 Syntax 185

(771) Kei a Hone raaua ko Mere oo raaua koti

at(pres) pers John IIIdl spec Mary plgenIIIdl coat
‘John and Mary have their coats’

Rendering ‘John and Mary have each other’s coats’ requires spelling out, eg. ‘John has
Mary’s coat and Mary has John’s coat’. Even examples where the reflexive interpretation
is less likely pragmatically than the reciprocal one are deemed ambiguous, eg.
(772) Naa Hone raaua ko Mere i patu a raaua
actgen John IIIdl spec Mary T/A beat pers IIIdl
anoo
again
‘John and Mary hit themselves/each other’

Thus it seems fair to say that reflexive interpretations are normally preferred to reciprocal
ones, and that reciprocity is treated as a special case of reflexivity.
The idea of reciprocity can also be expressed in other ways. The following extract
indicates the kinds of possibility that exist, as well as providing textual exampes of the
pronominal type discussed above:
(773) …kua kite·kite noa ake hoki raaua i a
T/A see·dup freely away emph IIIdl DO pers
raaua, i ngaa waa e hui·hui ai ngaa
IIIdl at the(pl) time T/A meet·dup part. the(pl)
taangata o Rotorua. I aua huihui·nga, ka
people gen Rotorua at det aph(pl) gather·nom T/A
kite a Hinemoa i a Tuutaanekai, ka titiro
see pers Hinemoa DO pers Tutanekai T/A look
atu, ka titiro mai, aa, aahua pai ki too
away T/A look hither and appearance good to sggen
teetehi ngaakau, ki too teetehi ngaakau; e wawata
a(sp) heart to sggen a(sp) heart T/A yearn
puku ana teetehi me teetahi, ki a raaua ake…
stomach T/A a(sp) with a(sp) to pers IIIdl only
i titiro whakatau ai raaua teetehi ki teetehi
T/A look intently part. IIIdl a(sp) to a(sp)
‘…they frequently happened to see each other at the times when the people of Rotorua
foregathered. At these gatherings, Hinemoa saw Tutanekai, they looked at each other, and
each liked the other’s appearance; they deeply desired each other. ‘…they looked intently at
each other’ (H, 6)

1.7.2 Scope of reciprocity

The comments in 1.6.2 apply. A corresponding example with a reciprocal-only
interpretation is
(774) I hiahia a Hone raaua ko Mere ki te
T/A desire pers John IIIdl spec Mary to the
tuutaki i a raaua anoo
meet DO pers IIIdl again
 Maori 186

‘John and Mary wanted to meet each other’

1.7.3 Reciprocal element a verbal affix

Not applicable.

1.7.4 Position of reciprocal element

See 1.6.4.

1.7.5 Relations between antecedent and reciprocal

Basically, the pattern is as described in 1.6.5. However, my consultant was much less
ready to accept cases of reciprocity as possible when the reciprocal was in modifier
position than with reflexives. In addition, the greater reluctance to accept reflexive
relations with dual or plural pronouns preceding the antecedent noted in 1.6.5 in relation
to (696) has obvious implications here: no cases of reciprocal interpretations were
obtained when the reciprocal preceded the antecedent

1.7.6 Reciprocal relations within nominalizations

See 1.6.6. As noted in 1.7.1, reflexive interpretations, if feasible, over-ride possible
reciprocal interpretations. A reciprocal example is:
(775) I mea mai a Tuu ki a maatou moo
T/A say hither pers Tu to pers Iplexcl intgen
te opu·tanga o Pou raaua ko Pani i ngaa
the throw·nom gen Pou IIIdl spec Pani DO the(pl)
koowhatu ki a raaua anoo
stone to pers IIIdl again
‘Tu told us about Pou and Pani throwing stones at each other’

1.7.7 Reciprocal relations within ordinary NPs

See 1.6.7. Again, reflexivity over-rides reciprocity. The best example I obtained was
deemed ambiguous:
(776) Rite·rite tonu ngaa takoha a Mere raaua ko
like·dup indeed the(pl) gift pers Mary IIIdl spec
Hone maa raaua anoo
John intgen IIIdl again
‘John’s and Mary’s gifts to themselves/each other were the same’

1.7.8 Reciprocals without overt antecedents

See 1.6.8. Again, the clearest example is ambiguous:
(777) Kaua e whaka·mamae i a koorua anoo
 Syntax 187

neg T/A cause·hurt DO pers IIIdl again

‘Don’t hurt yourselves/each other!’

1.7.9 Other uses of reciprocal forms

See 1.6.9.

1.8 COMPARISON

See also 1.1.2.4.2.7 and 2.1.4.4.

1.8.1 Means of expression of comparison

Maori uses a comparative element associated with the parameter of comparison, and a
preposition associated with the noun forming the standard of comparison. (However, the
distinction between preposition and particle made in the Questionnaire is not entirely
clear for Maori.) The comparative element is one of the following forms: atu ‘away’, ake
‘up’, iho ‘down’, kee ‘different’, ‘other than the expected’. The choice between them is
determined on semantic grounds. The clearest case is iho which is used with adjectives
with negative or inferior senses, such as kino ‘bad’, and in some contexts only, iti ‘small’.
Ake tends to be associated with desirable qualities, although this is not an absolute rule.
Kee is used when difference is emphasised, and is the least common. Atu appears to be
the unmarked choice, but there are examples where it is not acceptable, such as (780)
below. Any of these elements may be reinforced by noa preceding them, and kee on
occasion is reinforced by ake, in which case the order is kee ake. The preposition
associated with the standard of comparison is i. The examples illustrate:
(778) Nui noa atu tana whare i tooku
big indeed away sggenIIIsg house compar sggenIsg
‘His house is a lot bigger than mine’

(Ake is also possible here.)

(779) Horo atu ia ki te oma i ahau
fast away IIIsg to the run compar Isg
‘He is faster at running than me’

(The phrase ki te oma can also come in final position; ake is an alternative to atu.)
(780) Iti iho te utu o te motokaa o Mere
small down the price gen the car gen Mary
i too Hoo
compar sggen Joe
‘Mary’s car is/was cheaper than Joe’s’

(Ake is possible, but atu is marginal here.)

 Maori 188

On occasions, my consultant used ki for the comparative preposition. In most cases, ki

was rejected as an alternative to i, but if the two NPs being compared are both given in
full, ki was the first preference, although i was always accepted as an alternative, eg.
(781) He kuuiti noa atu ngaa ara ki/i ngaa
cls small intens away the(pl) path compar the(pl)
ara e tau·nga ana ia
path T/A accustom·pass. T/A IIIsg
‘The roads were narrower than he was accustomed to’

It should perhaps be pointed out here that in Maori a comparative is often used when no
explicit standard of comparison is given, including instances where English would have a
simple adjective, eg.
(782) Ko koe roa ake
top. IIsg tall away
‘You are the taller’
(783) Pai rawa atu te rangi nei
good very away the day proxI
‘It’s a lovely day’

In covert comparisons like (783), rawa usually accompanies the comparative.

1.8.2–4 Omission of elements under identity

To describe this aspect of comparatives, it is necessary to consider separately the three
possible types of construction which occur as comparative elements, namely a plain NP,
a nominalization, and a possessive relative clause.
Within a plain NP, all repeated material except the determiner can be (and normally is)
omitted. Consider the “full” version in (784a):
(784a) Nui ake too raatou whare whakairo i
big away sggenIIIsg house carve compar
too taatou whare whakairo
sggenIplincl house carve
‘Their meeting house is bigger than our meeting house’

Whare whakairo would normally be omitted from the i-phrase. Thus (784b) is the
preferred form. However, it is possible to retain this phrase as in (784a).
(784b) Nui ake too raatou whare whakairo i
big away sggenIIIpl house carve compar
too taatou
sggenIplincl
‘Their meeting house is bigger than ours’

It is rather less clear that whare alone (without whakairo) can be retained without change
in meaning, unless the context makes the intention entirely clear; the addition of nei ‘this’
 Syntax 189

either following too taatou or following whare seemed to improve the probability of
retaining the sense of (784a):
(784c) Nui ake too raatou whare whakairo i
big away sggenIIIpl house carve compar
too taatou whare nei
sggenIplincl house proxI
‘Their meeting house is bigger than our house here’

With a nominalization, the same sort of pattern holds. Thus the “full” version in (785a) is
the least likely:
(785a) Tere noa ake tana taa·nga i te raakau
fast indeed away sggenIIIsg fell·nom DO the tree
i tauu taa·nga i te raakau
compar sggenIIsg fell·nom DO the tree
‘He’s faster at felling trees than you are at felling trees’

The most normal rendering omits all the repeated material:

(785b) Tere noa ake tana taa·nga i te raakau
fast indeed away sggenIIIsg fell·nom DO the tree
i tauu
compar sggenIIsg
‘He’s faster at felling trees than you are’

It is also possible to repeat just the nominalized verb:

(785c) Tere noa ake tana taa·nga i te raakau
fast indeed away sggenIIIsg fell·nom DO the tree
i tauu taa·nga
compar sggenIIsg fell·nom
‘He’s faster at felling trees than you are at felling’

This set of possibilities is valid not only for -Canga nominalizations, but also for plain
stem nominalizations, see (789) below.
Within relative clauses, the repeated material must be omitted:
(786) Moohio noa ake te tamaiti raa i taa
know indeed away the child dist compar sggen
te kaiako i whakaaro ai
the teacher T/A think part.
‘That child is cleverer than his teacher thinks’

It is not possible to include mention of te tamaiti or of moohio in the possessive relative

clause.
Note that, in all cases, the determiner (and of course the contrastive element) in these
comparative NPs must be retained.
 Maori 190

1.8.5 Types of comparative structure

The two types of structure, comparative particle+reduced comparative clause and
preposition+standard of comparison differ only in the form of the construction following
i. Thus the following are all equivalent:
(787) He nui noa atu te utu moo teeraa pootae
cls big indeed away the price intgen that hat

Compare these with a plain NP:

(788) He nui noa atu te utu moo teeraa pootae
cls big indeed away the price intgen that hat
i taku pootae
compar sggenIsg hat
‘That hat is more expensive than my hat’

Note, however, that the plain stem alternative (like the final alternative in (787) above)
can be accompanied by sentential arguments, though the bracketed elements are normally
omitted, eg.
(789) Tere noa ake tana taa i te raakau i
fast indeed away sggenIIIsg fell DO the tree compar
taau (taa (i te raakau))
sggenIIsg fell DO the tree
‘He’s faster at felling trees than you are’

1.8.6 Correlative comparison

The expression of correlative comparison involves verbal structures introduced by ka,
rather than the non-verbal structures which are the norm elsewhere for comparatives. The
first item of comparison is followed by atu, and the second by rawa atu (occasionally
merely by atu), thus:
(790) Ka tini atu, ka pai rawa atu
T/A many away T/A good intens away
‘The more, the better’
 Syntax 191

(791) Ka nohinohi atu te tamaiti, ka kaha rawa

T/A small away the child T/A strong intens
atu ki te pararee
away to the bawl
‘The smaller the child, the louder the yelling’
(792) Ka nui atu tana kai, ka moomona (rawa)
T/A big away sggenIIIsg food T/A fat intens
atu ia
away IIIsg
‘The more he eats, the fatter he gets’

1.9 EQUATIVES

See also 1.1.2.4.2.7.2 and 2.1.4.4.1.

1.9.1 The expression of equation

The most basic types of equative sentence involve the use of one of two forms. The first
is the verb rite ‘be like, resemble’, often reinforced by tonu ‘indeed’, and the preposition
ki to introduce the parameter of equation. The second involves the forms
peenei/peenaa/peeraa, which consist of the deictic particles nei/naa/raa suffixed to the
bound root pee- ‘like’. These too function as verbs in these sentences. With pee- forms,
the parameter of equation is introduced by i. The structure is thus parallel to English ‘X
resembles Y in respect of Z’. The dimension of equality may be expressed as a quality
possessed by the objects equated, or it may appear as an (unmarked) NP. Some equative
sentences permit a considerable variety of word orders. The T/A marker is very
commonly omitted, especially with pee-. Consider the following examples:
(793a) (I) rite tonu oona wae ki ooku te
T/A like indeed plgenIIIsg foot eql plgenIsg the
nunui
big
‘His feet are just as big as mine’

The following alternatives are also possible: (793)(b) has ko-fronting of the subject,
(793)(c) has the equative phrase in final position, and (793)(d-f) show the quality in a
possessive phrase:
(793) (b) Ko oona wae (i) rite tonu ki ooku te nunui
(c) (I) rite tonu oona wae te nunui ki ooku
(d) (I) rite tonu te nunui o oona wae ki ooku
(e) (I) rite tonu ki ooku wae te nunui o oona
(f) (I) rite tonu ki ooku te nunui o oona wae
 Maori 192

However, the normal order is to place the equative phrase after the standard. Compare the
parallel sentence with pee-:
(794) (I) pee·raa tonu oona wae i ooku te
T/A like·dist indeed plgenIIIsg foot eql plgenIsg the
nunui
big
‘His feet are just as big as mine’

A parallel set of alternative word orders is also possible with peeraa…i.

(795) I rite tou whakahiihii ki tou tungaane
T/A like sggenIIsg cheek eql sggenIIsg brother
‘You are as cheeky as your brother’

If two actions are equated, so that verbal structures, rather than plain NPs are involved,
there is rather more variety of structure found. The forms peenei/peenaa/peeraa seem to
be commoner than rite, but both are possible. They are preceded by kia ‘subjunctive’ for
not-yet-accomplished clauses, and Ø for accomplished clauses. The particle i (not ki as
with simple NPs) introduces a possessive relative clause even if rite is used. Consider the
following examples:
(796) Me horoi te hakihaki pee·raa i taa te
oblig wash the wound like·dist eql sggen the
rata i mea mai ai
doctor T/A say hither part.
‘The wound should be bathed as the doctor said’
(797) Maa·ku e horoi taku ngeru kia pee·naa i
intgen·Isg T/A wash sggenIsg cat subj like·proxII eql
taau e horoi naa i taau peepi
sggenIIsg T/A wash proxII DO sggenIIsg baby
‘I’ll wash my cat just the way you wash your baby’

Other variants with rite have been exemplified in 1.1.2.4.2.7.2.

It is apparently also possible to mark both halves of the equation with an equative
construction, although there is insufficient data to be sure under what circumstances this
is possible:
(798) Kia rite te tapahi i te koorari kia pee·raa i
subj like the cut DO the flax subj like·dist eql
taa Pita
sggen Peter
‘Cut the flax as Peter is cutting it’

When actions are equated according to a standard, the equative form is usually pee-,
which may be preceded by he, although this is often omitted, as indeed the pee- form may
be. The equative particle is me in such constructions, eg.
(799) He pai a ia ki te koorero tara (he) pee·raa
cls good pers IIIsg to the talk story cls like·dist
anoo me tana pai ki te poi
 Syntax 193

again with sggenIIIsg good to the poi-dance

‘She is as good at telling stories as she is at performing poi dances’
(800) Tino kino a Pou ki te tunu kai me Pani ki
very bad pers Pou to the cook food with Pani to
te tapahi kakaaihe i waho
the cut grass at outside
‘Pou is as bad at cooking the dinner as Pani is at cutting the lawn’

Sometimes the equative particle in such constructions is i:

(801) He nui tana maatau·ranga ki te aru poaka
cls big sggenIIIsg know·nom to the hunt pig
puihi i tana maatau·ranga ki te hii ika
wild eql sggenIIIsg know·nom to the catch fish
‘He is as skilled at pig-hunting as he is at catching fish’

However, nominalized versions of sentences like these are often preferred, eg (802)
below is regarded as preferable Maori to (799) above:
(802) I rite tonu tana pai ki te koorero tara
T/A like indeed sggenIIIsg good to the talk story
ki tana pai ki te poi
eql sggenIIIsg good to the poi-dance’
‘She is as good at telling stories as she is at performing poi dances’

1.9.2–4 Omission of elements under identity

As with comparatives, it is necessary to treat this material separately for the different
types of construction.
With plain NPs and both stem and -Canga nominalizations, the facts are as for
comparatives, see 1.8.2–4. The following are parallel examples with equatives.
Plain NPs:
(803) Rite tonu te nui o too raatou whare whakairo ki
like indeed the big gen sggenIIIsg house carved eql
too taatou (whare whakairo)
sggenIplincl house carved
‘Their meeting house is just as big as ours’

The bracketed material is optional. As with comparatives, it is not clear that whare
without whakairo in the ki-phrase retains synonymy. Stem nominalizations:
(804) Rite tonu taku taa i te raakau ki taau
like indeed sggenIsg fell DO the tree eql sggenIIsg
(taa (i te raakau))
fell DO the tree
‘I felled the tree just like you’

As with comparatives, the norm is the minimal form; the maximum form is rather
unusual.
 Maori 194

-Canga nominalizations:
(805) Rite tonu taku taa·nga i te raakau ki
like indeed sggenIsg fell·nom DO the tree eql
taau (taa·nga (i te raakau))
sggenIIsg fell·nom DO the tree
‘My felling of the tree was just like yours’

Again, the minimal form is the norm, and the maximal form highly redundant.
With possessive relative structures, only the minimal structure is possible:
(806) Rite tonu te moohio o te tamaiti ki taa
like indeed the know gen the child eql sggen
te kaiako i whakaaro ai
the teacher T/A think part
‘The child is just as clever as the teacher thought’

However, with ‘double’ equatives, like (798) above, it is possible to repeat material,
although the more material that is repeated, the more marginal the construction becomes:
(807) Kia rite te tapahi i te koorari kia pee·raa
subj like the cut DO the flax subj like·dist
i taa Pita (i tapahi raa (i te koorari))
eql sggen Peter T/A cut dist DO the flax
‘Cut the flax like Peter’

Again, as with comparatives, the determiner and the contrastive element in equatives
must be retained in the equative section of the sentence.

1.9.5 Types of equative structure

The two types of equative structure, equative particle+reduced equative clause and
preposition+standard of equation, differ chiefly in the construction following the
preposition/particle. Note that there is some variation in the particle itself with equatives,
unlike comparatives: while rite takes the particle ki with plain NPs and nominalizations
(see examples in 1.9.1 and 1.9.2–4), it takes i with full clauses, and while pee- takes i in
most cases, it sometimes occurs with me.

1.9.6 Correlative equatives

My consultant did not know of any such construction, but proffered the following as
possible translations:
(808) Ka moomona, ka harikoa
T/A fat T/A happy
?‘As fat, as happy’
(809) Ka ngenge, ka whakangari
T/A tired T/A quarrelsome
?‘As tired, as quarrelsome’
 Syntax 195

I have questioned the appropriateness of the translations, as it appears that these Maori
sentences (which do not contain overt equatives) are more equivalent to ‘When you are
fat, you are happy’, and ‘When you are tired, you are quarrelsome’, which do not convey
the link in degree which the English correlatives convey. It is thus probably wise to
conclude that Maori does not have a correlative equative construction.

1.10 POSSESSION

1.10.1 The construction of sentences expressing possession

All sentences attributing possession in Maori are non-verbal. Maori has no verbs
equivalent to own,have or possess in English. Beyond that, there are few generalizations
that can be made. Possession sentences in Maori involve a set of intersecting contrasts:

location vs. ownership (see 1.10.3)

actual vs. intended (see 1.10.5)
a vs. o (see 1.10.6)
specific vs. non-specific (see 1.10.7)
singular (vs. dual) vs. plural (see below)

The intersection of these categories is complex. There is a primary distinction between

location and ownership. (“Temporary possession” in the Questionnaire’s terms is
expressed as location in Maori, but since the implications of the term “temporary
possession” seem to me inappropriate for these locative sentences, the term is avoided in
the discussion.) None of the other distinctions is relevant for location, although location
sentences have a set of tense distinctions which do not apply to any categories of
ownership. The a vs. o distinction intersects with all categories of ownership. The actual
vs. intended distinction applies to ownership of specifics. (There is some doubt as to
whether it applies to non-specifics, see below.) The singular vs. plural distinction is
marked in the possessive phrase only in non-specifics, although this distinction is also
marked when possessive phrases are attributive. Fig 1 may help to make the relationships
clearer. In Fig 1, the square brackets indicate that one choice must be made from each
system. Any three-term outcome from the large braces is possible. The singular/plural
distinction here relates to the number of the possessee. (Number is also marked on
possessors.)
The following examples illustrate one of each type of possessive sentence; the
individual contrasts are discussed in detail in the sections that follow. Only one sentence
is provided here out of the two possibilities arising from the A/O choice, as other
illustrations are given later.
Location:
(810) Kei a Hone te koti
at(pres) pers John the coat
‘John has the coat’
 Maori 196

Fig 1: Overview of possession

Intended Specific Ownership:

(811) M·oo Hone eenei huu
int·Ogen John these shoe
‘These shoes are for John’

Actual Specific Ownership:

(812) N·aa Hone te pukapuka naa
act·Agen John the book proxII
‘That book belongs to John’

Singular Non-specific Ownership:

(813) He hooiho t·oo Tohe
cls horse sg·Ogen Tohe
‘Tohe has a horse’

The gloss for the form too indicates that it consists of the determiner t(e) (sg)+o, the O-
class genitive marker. The e of the determiner is always elided.
Plural Non-specific Ownership:
(814) He pukapuka aa Pou
cls book [pl]·Agen Pou
‘Pou has some books’

The gloss here has ‘pl’ in square brackets to indicate that there is no morph in the Maori
corresponding to it: the deletion of the t- of the singular is what marks this form as plural.
 Syntax 197

Biggs (1955, 347) gives an example with an n- form as a non-specific possessive

sentence. However, to my consultant, Biggs’s example is not a complete sentence. I have
not been able to ascertain whether this is a matter of dialectal variation or whether there
are particular contextual requirements for its use. The following two textual examples
make it clear that noo occurs with non-specifics under some circumstances, but I have
found no parallel instances with moo/maa.
(815) …he wiri hoki n·oo·na i te maaeke…
cls tremble also Act·Ogen·IIIsg from the cold
‘…she was trembling with cold…’ (H, 8)
(816) …he puuhaehae n·oo raatou
cls jealousy Act·Ogen;IIIpl
‘…they were jealous’ (H, 11)

It may thus be the case that while n- forms occur in place of definite forms with non-
specifics in some instances, the actual/intended distinction is not made for non-specifics.
Not all of the contrasts above are marked both predicatively and attributively.
Location cannot be expressed attributively, and nor can the intended/actual distinction, ie.
if there is a time dimension involved, the construction will be predicative. Such phrases
can be post-modifiers of heads through relativization, though. Note that in that position
they are not formally distinct from other prepositional phrases, although I do not think
there is any doubt that they remain underlying predicatives.
Some discussion is also called for here of the structure of possessive phrases in
sentences expressing, but not attributing possession. The possessive pronouns of Maori
are treated in 2.1.2.4.3. However, it may be noted here that there are two distinct groups,
the main set, which is marked for the a/o distinction, and a minor set neutralizing this
distinction. The main set marks the following distinctions:
– person (I, II, III), number (sg, dl, pl) of possessor
– number of possessee (sg, pl)
– the relation between them (a, o).
The neutral set (glossed N for ‘neutral’) is only available for singular possessors, and
marks the following distinctions:
– person (I, II, III) of possessor
– number of possessee (sg, pl).
There are special clitic forms of the singular pronouns which are required in combination
with the markers of the A/O distinction. In previous sections, the glosses of possessive
forms have not indicated the A/O distinction, since this is always evident from the Maori
form. Here, however, fuller glosses are provided,using the following conventions:
(817a) t·oo raatou whare
sg·Ogen;IIIpl house
‘their house’

where the information relates to

possessee·relation;possessor.
 Maori 198

The semi-colon is used in the gloss where there is a separate orthographic word in the
standard orthography of Maori. Here, as elsewhere in the determiner system of Maori, the
distinction between singular and plural (here, of the possessee) is marked by t- vs. Ø;
compare (817a) with (817b):
(817b) oo raatou whare
[pl]·Ogen;IIIpl house
‘their houses’

If the possessive phrase is attributive, and the possessor is a full NP rather than a
pronoun, there are two possible constructions. The first is a post-head prepositional
phrase:

Alternatively, the possessive phrase can be cliticized to the determiner, so that the
possessive preposition fuses with the determiner, te, which loses its -e. The plural
corresponding to the t- possessive form is formed, as other plurals of t- determiners are,
by the absence of t-, so that all overt trace of the determiner is lost. Thus corresponding to
the above, we have
 Syntax 199

There are, however, complexity/length restrictions on this second construction. Thus the
following example is rejected:
(822a) *Kua kite·a t·aa te tangata i haere mai
T/A find·pass. sg·Agen the man T/A move hither
(ai) pukapuka
part. book
‘The man who came’s book has been found’

This example was rejected as unacceptable (although like multiple relative clause
embeddings in English, it is apparently well-formed according to grammatical rules). The
post-head construction must be used instead:
(822b) Kua kite·a te pukapuka a te tangata i
T/A find·pass. the book Agen the man T/A
haere mai (ai)
move hither part.
‘The book of the man who came has been found’

As might be expected, there is no absolutely clear cut-off point for this construction.
Consider the following judgements:
(823) *Kua kite·a taa te tamaiti n·oo Tonga
T/A find·pass. sg·Agen the child act·Ogen Tonga
pukapuka
book
 Maori 200

‘The Tongan child’s book has been found’

(824) *Kua kite·a t·aa te tamaiti maangere pukapuka
T/A find·pass. sg·Agen the child lazy book
‘The lazy child’s book has been found’
(825) ?Kua kite·a t·oo Pou raaua ko Pani hooiho
T/A find·pass. sg·Ogen Pou IIIdl spec Pani horse
‘Pou and Pani’s horse has been found’

All of these are improved by the addition of nei ‘proxI’ immediately preceding the head
noun. Nei appears to function as a sort of bracket, helping the listener to keep track of the
grammatical structure. With nei added, (825) is probably acceptable, although very much
on the limit of acceptability. Even with nei, (824) was rejected, although with some
doubt, but (823) was rejected with confidence. These sentences thus show a cline from
the clearly acceptable (820) and (821) to the clearly unacceptable (822a).
Possessors in the form of pronouns, on the other hand, can occur only in pre-head
position, cliticized to the determiner, and in the singular, requiring special clitic forms of
the pronouns:
(826a) *te whare o ia
the house Ogen IIIsg
*te whare oo·na
the house Ogen·IIIsg
*te o ia whare
the Ogen IIIsg house
*t·oo ia whare
sg·Ogen;IIIsg house
t·oo·na whare
sg·Ogen·IIIsg house
‘his house’
(826b) *te whare o raaua
the house Ogen IIIdl
*te o raaua whare
the Ogen IIIdl house
t·oo raaua whare
sg·Ogen;IIIdl house
‘their house’
(827a) *ngaa pukapuka a ia
the(pl) book Agen IIIsg
*ngaa a ia pukapuka
the(pl) Agen IIIsg book
*ngaa pukapuka aa·na
the(pl) book [pl]·Agen·IIIsg
aa·na pukapuka
[pl]·Agen·IIIsg book
‘his books’
(827b) *ngaa pukapuka a raaua
the(pl) book Agen IIIdl
 Syntax 201

*ngaa a raaua pukapuka

the(pl) Agen IIIdl book
aa raaua pukapuka
[pl]·Agen;IIIdl book
‘their books’

1.10.2 Alienable and inalienable possession

At first sight, the A/O distinction in Maori may appear to mark the distinction between
alienable and inalienable possession, since the relation between possessor and most
obviously alienable possessees is expressed by the A category. However, there are
examples which appear to go counter to this in both directions, and I have therefore
chosen to treat the A/O distinction as a different type of relation, rather than as a sub-type
of this category. However, the evidence for doing so is not overwhelming, and so the
issue is discussed a little further here. The major problem stems from the fact that
linguistic categories and natural world categories need not coincide exactly: there are
many well-known cases like German Mädchen ‘girl’ which is neuter in gender, not
feminine. Thus the finding of an individual item which does not match the real world
category assignment does not necessarily bear much weight. In the examples below, it is
necessary to bear in mind that it is not only this particular example which behaves as
illustrated, but many other items of the same category. Example (828) is notionally
alienable with the A category, (829) is notionally inalienable with O, (830) is notionally
inalienable with A, and (831) is notionally alienable with O:
(828) Mea taahae aa raaua pukapuka
thing stolen [pl]·Agen;IIIdl book
‘Their books were stolen’
(829) I horoi raaua i oo raaua makawe
T/A wash IIIdl DO [pl]·Ogen;IIIdl hair
‘They washed their hair’
(830) Mea taahae aa raaua peepi
thing stolen [pl]·Agen;IIIdl baby
‘Their baby was stolen’
(831) Mea taahae oo raaua kaakahu
thing stolen [pl]·Ogen;IIIdl clothes
‘Their clothes were stolen’

(Further examples which show that there is no oddity in O-category possessions (ie. the
most likely inalienables) changing hands are given in 1.10.5, examples (850) and (851).)
It could perhaps be argued that examples like these are a matter of perception, or
cultural custom—that certain things which take the O-category, like land and houses, still
belong to the original possessor, even if stolen. However, it is a little more difficult to
argue away differences like the following:
(832) aa raaua tama
[pl]·Agen;IIIdl son
‘their sons’
 Maori 202

(833) oo raaua maatua

[pl]·Ogen;IIIdl parents
‘their parents’

If the distinction between A and O was a distinction between alienable and inalienable,
then children would be alienable from their parents, even though the parents are not
alienable from the children. Perhaps more inportantly, this is not what is found in those
languages which clearly have the alienable/inalienable distinction. To judge by the data
in eg. Lynch (1973), body parts and kinship terms seem to be the canonical members of
the inalienable category in Melanesian languages, and the Maori classification of kinship
relations into two groups one in the A-category and one in the O-category does not seem
like the situation found in the Melanesian languages. Furthermore, the
alienable/inalienable distinction in Melanesian languages is usually associated with two
rather different syntactic constructions.
For all these reasons, I have decided to treat the A/O distinction as a different type of
phenomenon, see 1.10.6. Nevertheless, much of what Lynch has to say about the
relational nature of the alienable/inalienable distinction in Melanesian, the overlaps, and
the relevance of the notion of control seems to apply to Maori A/O as well.

1.10.3 Temporary and permanent possession

As was indicated in 1.10.1, temporary possession is not treated as possession at all by the
grammatical system of Maori. It is expressed as location. Consider
(834) Kei a Pou te koti
at(pres) pers Pou the coat
‘Pou has the coat’, ‘The coat is with Pou’

No information is given in this sentence about the owner of the coat in question. This
locative structure is readily compatible with subjects containing possessives, eg.
(835a) Kei a Pou te koti o Hone
at(pres) pers Pou the coat Ogen John
‘Pou has John’s coat’

(alternatively,
(835b) Kei a Pou t·oo Hone koti
at(pres) pers Pou sg·Ogen John coat
‘Pou has John’s coat’)
(836) Kei a Pou t·oo·na koti
at(pres) pers Pou sg·Ogen·IIIsg coat
‘Pou has his (own) coat’

Location is independent of the A/O distinction. Thus while the possessive phrases
(837a) te koti o Pou
the coat Ogen Pou
‘Pou’s coat’
 Syntax 203

(837b) te ngeru a Pou

the cat Agen Pou
‘Pou’s cat’

show the A/O distinction, if te koti and te ngeru are located with Pou, rather than
possessed by him, no distinction is made:
(838a) Kei a Pou te koti
at(pres) pers Pou the coat
‘Pou has the coat’
(838b) Kei a Pou te ngeru
at(pres) pers Pou the cat
‘Pou has the cat’

Permanent possession, on the other hand, (or ownership, as it was labelled in 1.10.1),
involves the use of the A/O categories. Compare the first example above, (834), with the
following:
(839) N·oo Pou te koti
act·Ogen Pou the coat
‘The coat is Pou’s’, ‘The coat belongs to Pou’

A sentence like this gives no information about the current location of the coat.
The temporal distinctions relevant to these sentence types are discussed in 1.10.5.

1.10.4 Possession relative to persons, animals and things

Fundamentally, there are no differences in the expression of possession relative to
persons, animals and things. However, since things relatively seldom have control over
possessions, they seldom occur in A-category constructions. Nevertheless, they are not
precluded from doing so a priori. When an O-category relation is involved, persons,
animals and things are treated identically.
Examples to illustrate in a variety of types of possessive sentence follow:
Non-specific possession:
(840a) He huu o te tangata
cls shoe Ogen the man
‘The man has shoes’
(840b) He huu o te hooiho
cls shoe Ogen the horse
‘The horse has shoes’
(840c) He wiira o te tuuru raa
cls wheel Ogen the chair dist
‘That chair has castors’

Actual Specific Possession:

(841a) N·oo te tangata raa te koti nei
 Maori 204

act·Ogen the man dist the coat proxI

‘This coat belongs to that man’
(841b) N·oo te kurii raa te koti nei
act·Ogen the dog dist the coat proxI
‘This coat belongs to that dog’
(841c) N·oo te teepu raa te aarai nei
act·Ogen the table dist the cover proxI
‘This cover belongs to that table’

Attributive construction:
(842) He parauri ngaa waewae o te
cls brown the(pl) leg Ogen the
tangata/kurii/teepu raa
man/dog/table dist
‘The man’s/dog’s/table’s legs are brown’

The following examples show that a car can be the possessor in an A relationship just as
a person or animal can:
(843a) He ika te kai a te tangata/ngeru raa
cls fish the food Agen the man/cat dist
‘That man’s/cat’s food is fish’
(843b) He hinu te kai a te motokaa
cls oil the food Agen the car
‘The car’s food is petrol’

However, the following situation is more typical:

(844a) Kua pakaru te riihi a te tangata/kurii
T/A broken the dish Agen the man/dog
‘The man’s/dog’s dish is broken’
(844b) Kua pakaru te riihi o te mea hurihuri
T/A broken the dish Ogen the mixer
‘The man’s/dog’s dish is broken’

The mixer does not control the bowl; the bowl is considered a part of it, and hence a
cannot be used. On the other hand, since parts of people and animals are possessed in O
relations, the “thing” is really not behaving differently. Compare:
(845a) He paihou t·oo te tangata raa
cls beard sg·Ogen the man dist
‘That man has a beard’
(845b) He paihou t·oo te kurii raa
cls beard sg·Ogen the dog dist
‘That dog has whiskers’
(845c) He waewae whakairo o te teepu raa
cls leg carved [pl]·Ogen the table dist
 Syntax 205

‘That table has carved legs’

On the other side of the coin, all three categories behave in the same way when they are
possessions. Thus there are members of each category which are normally found in A-
category relations, and other members of each category which are normally found in O-
category relations. The following sets of examples illustrate:
A-category:
(846a) te tama a Hone
the son Agen John
‘John’s son’
(846b) te kurii a Hone
the dog Agen John
‘John’s dog’
(846c) te kooauau a Hone
the flute Agen John
‘John’s flute’

O-category:
(847a) te matua o Hone
the parent Ogen John
‘John’s parent’
(847b) te hooiho o Hone
the horse Ogen John
‘John’s horse’
(847c) te heru o Hone
the comb Ogen John
‘John’s comb’

Thus the categories person, animal, thing are not relevant to the description of the
possessive system of Maori.

1.10.5 The expression of present and past possession

Time distinctions are marked in two types of possessive constructions in Maori, firstly in
the locatives used for what the Questionnaire calls “temporary possession” (see 1.10.3)
and secondly in the Actual/Intended distinction which is marked in sentences predicating
specific possession.
The time distinction in the locatives has been mentioned elsewhere (notably
1.2.1.1.4.1–2, and see also 2.1.1.5.1). It is illustrated here in the possessive context:
(848a) Kei a Hone te kii
at(pres) pers John the key
‘John has the key’
(848b) I a Hone te kii
at(pt) pers John the key
 Maori 206

‘John had the key’

(848c) Hei a Hone te kii
at(fut) pers John the key
‘John will have the key’

(The preposition in (848c) is subject to considerable dialectal variation, kei, ko, and ki, at
least being used instead of hei in other dialects. Note that in dialects which have kei, there
is no distinction between present and future, which are not distinguished in most
grammatical contexts in Maori.)
The Actual/Intended distinction is marked by the contrast n-/m-, where N- forms mark
either past or present (or both) ownership, and, M- forms mark future ownership. Since
this goes counter to the prevailing tendency in Maori to distinguish past from non-past in
verbal markers, other terms have been used here, which better reflect the use of these
forms. (The terminology used for this distinction is not standard: other terms found in the
literature include past/future (Hohepa, 1967, 22), accomplished/not yet realized (Biggs,
1969, 57) and actual or realized/virtual or prospective (Waite, 1990, 404).) The n- and m-
morphs never occur in isolation. They are always bound to one or other of the possessive
prepositions a or o, and the resultant forms are functionally prepositions. Thus there are
four terms in the set: maa, naa, moo, noo. (And since those possessive prepositions
require special clitic forms of the singular pronouns, so do the complex N- and M-
forms.) These forms are most commonly predicative:
(849a) M·aa Hone te kii nei
int·Agen John the key proxI
‘This key is for John’
(849b) M·oo Hone te hooiho raa
int·Ogen John the horse dist
‘This key is for John’
(849c) N·aa Hone te kii nei
act·Agen John the key proxI
‘This key is John’s’, ‘This key belongs to John’
(849d) N·oo Te Kao a Kare
act·Ogen Te Kao pers Kare
‘Kare belongs to Te Kao’, ‘Kare comes from Te Kao’

The following two contrastive examples illustrate the temporal distinction:

(850) N·oo Hone te hooiho i mua, n·oo·ku
act·Ogen John the horse at before act·Ogen·Isg
inaianei
now
‘The horse used to be John’s, but it’s mine now’
(851) N·oo·ku eenei huu inaianei; inaa kore e oo
act·Ogen·Isg these shoe now when neg T/A fit
ki ahau, m·oo·u
to Isg int·Ogen·IIsg
‘These shoes are mine now, but you can have them when they don’t fit me any more’
 Syntax 207

However, by relativization, the M-forms can be post-head modifiers:

(852) Kia tino nui te hooiho m·oo Pou
subj very big the horse int·Ogen Pou
‘The horse for Pou should be very big’

However, the N-forms were rejected in this construction, eg.

(853a) *Tino nui te hooiho n·oo Pou
very big the horse act·Ogen Pou
‘The horse that belongs to Pou is very big’

This is no doubt because such a message is usually encoded using a plain possessive, eg.
(853b) Tino nui te hooiho o Pou
very big the horse Ogen Pou
‘Pou’s horse is very big’

However, there is not an absolute embargo on such constructions with N-forms, so that
the following was accepted:
(854) He tungaane n·oo·ku teeraa
cls older brother act·Ogen·Isg that
‘That’s an older brother of mine’

It should also be noted here that the n- and m- prepositions have other uses in Maori
which are perhaps predictable from their basic ‘Actual/Intended Ownership’ senses. The
m- forms are used for beneficiaries; the -a forms are used in actor-emphatic sentences;
and the -a forms also appear in certain other types of causal contexts. Noo is also used in
some temporal constructions.

1.10.6 A and O Possession

These categories have probably had more pages of text devoted to them than any other
single topic in the Maori language. The most substantial contribution is Biggs (1955).
These forms certainly cause great difficulty for the learner and for the semi-speaker. The
account here differs in its perspective from previous accounts, rather than providing a
new account
These two categories express relations between possessor and possessee, not a
property of either the possessor or the possessee. The terms which have been used to
label the distinction look at the relation from one end or the other of the relation, and this
causes difficulties for the learner, at least. Consider the commonest labels , ‘dominant’
and ‘subordinate’ (see eg. Biggs, (1969, 43ff) and Hohepa, (1967, 24), and note that the
relationship is characterized differently in those two works, although they use the same
labels). If the possessor dominates the possessee, then the possessee can be seen as
subordinate to the possessor, and vice versa, so that the same relationship can be
characterized as either dominant or subordinate depending on point of view. These labels
are normally used from the point of view of the possessor. However, most treatments of
the subject proceed to list groups of items which are possessed dominantly or
 Maori 208

subordinately. Thus all the items listed under the head ‘dominant’ possession are
subordinate to the possessor, and vice versa. To avoid such confusion, the form labels A
and O are used here to identify the categories.
However, this is not to deny the relevance of the semantic distinction encapsulated in
the usual labels. The ‘A’ relationship does indeed seem to be characterized by the
dominance or control of the possessor over the possessee. However, I think that the ‘O’
relationship is not well-characterized by labels like ‘subordinate’. Rather, it is a
relationship where the possessor is not dominant, or does not control the possessee. If the
distinction is thus characterized as between dominant and non-dominant, or controlled/
not controlled, (where each of those represents a different point of view), it reflects much
better the fact that the A and O categories are not equal in the system. O is clearly the
unmarked relationship. This point has been made before (see Clark, 1976, 42–4), but has
not found its way into subsequent (pedagogical) grammars (eg. Foster, 1987; Head,
1989). The difference in markedness shows itself in particular in terms of distribution:
there are far more tokens of O-category possessives than A-category possessives in texts,
(in a brief story, Te Aohuruhuru, 75% of the possessive tokens are O-category), and far
more types of object have O-relations with possessors. (A measure of this is perhaps the
fact that Head’s thorough treatment characterizes 15 groups of O-category possessions
against 9 groups of A-category; he devotes twice as many pages to the illustration of O as
to A.) Yet another reflection of this difference in markedness, one difficult to
accommodate within previous proposals, is the fact that o is used for relations between
equals, where neither dominates, nor is dominated by the other.
The information necessary to master the system is thus the items which belong to the
marked A-category. Various characterizations can be given: Biggs says the A-category
applies where “the possessor is dominant, active or superior” (1969, 43); Hohepa says
(1967, 24) it is used for “those possessions to which the possessor (…) is dominant (eg.
small personal portable property, food) or which the possessor acquired in his lifetime
(eg. wife, children, husband, uninherited objects)”. Any different grammar of Maori will
give a slightly different characterization, but these are the two most general. Perhaps the
most useful addition to these general characteristics is a list of some of the items which
do not fall into the A-category though they might be expected to, at least by those
unfamiliar with traditional Maori culture. These include clothing and grooming
implements, houses (shelters in general), large pieces of furniture, means of transport
(including horses), water (for some speakers) and medicine. Perhaps a little more
predictably, parts of the body (including hands) and emotions (as distinct from actions)
fall outside the A-category, and thus take O.
These last remarks raise two further points about these categories. The first concerns
the culture reflected in these categories. It seems to me that much of the problem which
the A/O distinction presents is due to the sorts of possessions which arrived with
European settlers. For instance, prior to this, the canoe was the only means of transport,
and in general belonged to the tribe, not to an individual. It was thus clearly not
something which an individual controlled. However, with the arrival of the horse and
later the car, problems of classification arose: in one sense, these items might be in an
individual’s control, thus calling for A, and in another, they were functionally like
canoes, thus calling for O. In most instances like this, items have by convention become
associated with the unmarked category. Thus what may once have been relatively
 Syntax 209

transparent categories have become blurred. This, together with the virtual suppression of
the Maori language for most of the twentieth century, has left even the native speakers
with doubts about these categories. I suspect that the categories are applied nowadays in a
manner which resembles gender categories far more than they were prior to the general
demise of the language.
The second issue which must be aired here is the matter of variation. Relating back to
the issue of the classification of new items, variation can be found in the treatment of
some of the most borderline items. For example, a tractor can be seen as a means of
transport, or it can be seen as an implement controlled by the possessor. Both A and O
forms are attested for tractors. Thus Head (1989, 115) gives the example
(855) He whero te tarakihana a Hata
cls red the tractor Agen Hata
‘Hata’s tractor is red’

My consultant, however, produced

(856) Kua hemo t·oo raatou tarakihana
T/A be exhausted sg·Ogen;IIIpl tractor
‘Their tractor has broken down’

and denied that A was appropriate. Thus there is variation found between speakers.
Similarly, although wai is generally claimed to take O-class possession, so that Head
gives (1989, 111)
(857) Homai he wai m·oo·ku
bring a water int·Ogen·Isg
‘Bring some water for me’

my consultant produced
(858) N·aa·ku teeraa wai
act·Agen·Isg that water
‘That’s my (glass of) water’

and rejected nooku when asked. There is some tribal variation, which may account for
some of the divergence noted. Biggs also notes (1955, 341) a tendency for the A-category
to be preferred with first person possessors. Biggs’ discussion in that article suggests that
there was variation even before the demise of Maori. I suspect that the rigidity and
uncertainty that prevails today is the type of language death phenomenon discussed in
Campbell & Muntzel 1989. The “correct” use of the A/O distinction is regarded as a
shibboleth by many Maori speakers today.
Aside from individual or tribal variation, certain objects may occur with either
category as appropriate. Although grammars aimed at learners usually list objects as if
they always, immutably, occur in one category or the other, mention is normally made of
such pairs of examples as
(859a) M·aa·ku te waiata
int·Agen·Isg the song
‘The song is (sung/composed) by me’
 Maori 210

(859b) M·oo·ku te waiata

int·Ogen·Isg the song
‘The song is for/about me’
(860a) te rongoa o Pou
the medicine Ogen Pou
‘Pou’s medicine (for him to take)’
(860b) te rongoa a Pou
the medicine Agen Pou
‘Pou’s medicine (which he made/prescribed)’

I believe that this sort of variation was much more widespread than is usually recognized.
The following textual example shows the same noun in two different possessive relations,
according to circumstances. Pou asks Rua-kapanga, half man, half bird, for some feathers
from his body:
(861a) Ka paatai atu a Pou m·oo eetahi o
T/A ask away pers Pou int·Ogen some(pl) Ogen
oo·na kura
[pl]·O·IIIsg feather
‘Pou asked for some of his feathers’ (P, 2)

Note that oona refers to Rua-kapanga, and the O relationship holds between Rua-kapanga
and the feathers which are his natural covering. However, later Pou loses the feathers
from Rua-kapanga, and they are found by Mahia, and a dispute arises over their
ownership, now marked with A:
(861b) Kite mai ana a Pou i ngaa kura nei,
see hither T/A pers Pou DO the(pl) feather proxI
ka mea mai, N·aa·ku ngaa kura naa. Ka
T/A say hither act·Agen·Isg the(pl) feather proxII T/A
whakahoki atu a Maahia, “E, n·aa·ku tonu
reply away pers Mahia hey act·Agen·Isg indeed
aa·ku!”
[pl]·Agen·Isg
‘As soon as Pou saw these feathers, he said, “Those feathers belong to me”. Mahia replied,
“Come on now, what I have belongs to me’” (P, 3)

Rameka Cope (personal communication) provided the following examples of the A/O
distinction used to make a semantic distinction:
(862a) Mau·ria mai aa·ku hooiho
bring·pass. hither [pl]·Agen·Isg horse
‘Bring me my horses’
(862b) Mau·ria mai t·oo·ku hooiho
bring·pass. hither sg·Ogen·Isg horse
‘Bring me my horse’
 Syntax 211

where (862a) was understood as a request to bring the Clydesdales, the work horses, and
(862b) a request to bring the riding horse. Thus the same form, hooiho shifts from one
category to the other according to the nature of the relationship.
These are not isolated cases. Native speakers are reported to have arguments on
occasion about the appropriate form of the possessive for eg. a horse which has become
too old to be ridden or to pull a cart: is it appropriate to say taa raaua hooiho ‘their
horse’, (parallel to other pets), or too raaua hooiho ‘their horse’, (once a means of
transport, always a means of transport)? This suggests that these categories can (or at
least could) be manipulated by speakers of Maori to make a point. Thus to regard them as
absolute or fixed seems to me not to reflect their earlier use in Maori.
Before leaving the subject of the A/O distinction, it is necessary to consider their
distribution in nominalizations. The idea of control is relevant to their use in -Canga and
stem nominalizations. With -Canga, a is used for the nominalized subjects of active,
canonical transitive verbs, and o for other subjects, thus:
Active Canonical Transitive:
(863) te taa·nga a Pou i te raakau
the fell·nom Agen Pou DO the tree
‘Pou’s felling the tree’

Passive:
(864) te taa·nga o te raakau e Pou
the fell·nom Ogen the tree by Pou
‘the felling of the tree by Pou’

Experience:
(865) te moohio·tanga o Pou ki te mahi
the know·nom Ogen Pou to the work
‘Pou’s knowledge of the job’

Intransitive:
(866) te oma·nga o Pou
the run·nom Ogen Pou
‘Pou’s running’

Neuter verb:
(867) te riro·tanga o Rona
the seize·nom Ogen Rona
‘Rona’s seizure’

(rironga is also used instead of rirotanga.) In (864–7), no variation in the possessive

marker is possible. However, my consultant also accepted o .with active canonical
transitives, ie. as well as (863), the following is possible:
(868) te taa·nga o Pou i te raakau
the fell·nom Ogen Pou DO the tree
‘Pou’s felling of the tree’
 Maori 212

(863) and (868) are not, however, in free variation. Perhaps surprisingly, my consultant
felt that (863) was appropriate to a discussion of the event after it had taken place, and
that (868) was appropriate while it was going on. It is not clear whether this
generalization extends to other speakers, or even to other examples. However, it is clear
that for my consultant, o is the unmarked preposition in such nominalizations, contrary to
the usual textbook descriptions (eg. Biggs, 1969, 81, although the examples do not bear
out the description fully). The following textual examples also throw doubt on the normal
characterization of the choice as dependent on verb type: the first shows a transitive verb
with O, and the second an intransitive verb with A:
(869) N·oo te tatau·nga o Taranga i a·na
act·Ogen the count·nom Ogen Taranga DO [pl]·Ngen·IIIsg
tamariki…
children
‘When Taranga counted her children…’ (KM, 2)
(870) Ko oo·na pakikau kua pahuhu kee ki raro
top. [pl]·Ogen·IIIsg garment T/A slip off contr to below
i te kowhana·nga a ngaa ringaringa…
cause the toss·nom Agen the(pl) hand
‘Her garments had slipped down as a result of the tossing of her arms…’ (TA, 6)

(In this last example, it is important to the story that responsibility for her nakedness is
not her own.) Again, these are not isolated instances, but it seems to me that they
establish that control is crucial for the use of A, while O is the unmarked category, and
that verb category is not the crucial factor in the distribution of these prepositions.
With stem nominalizations, the pattern of possessives is different. It appears that
canonical active verbs, whether transitive or intransitive, take a for their subjects, but o is
found occasionally, see (1137). Experience verbs take o. Stem nominalizations cannot be
derived from passives or neuter verbs. Thus:
Active Canonical Transitive:
(871) te taa a Pou i te raakau
the fell Agen Pou DO the tree
‘Pou’s felling of the tree’

Passive:
(872) *te taa o/a te raakau e Pou
the fell O/Agen the tree by Pou
‘the felling of the tree by Pou’

Experience:
(873) te moohio o Pou ki te mahi
the know Ogen Pou to the work
‘Pou’s knowledge of the job’

Intransitive:
(874) te oma a/*o Pou
 Syntax 213

the run A/Ogen Pou

‘Pou’s run’

Neuter verb:
(875) *te riro o/a Rona
the seize O/Agen Rona
‘Rona’s seizure’

(This perhaps suggests that stem nominalizations may be more action-oriented than -
Canga nominalizations.)

1.10.7 Specific vs. non-specific

This distinction applies only to sentences predicating possession: the construction that is
commonly used for non-specific possessees (see (799) -(800)) is not the same as that used
for specific possessees (see (798)). This distinction has not generally been recognized in
descriptions of Maori possession as a fundamental one, but I believe that the fact that the
associated sentence types diverge so radically suggests that it is fairly basic. (It could
equally well be characterized in other ways, eg. as a type/token distinction, perhaps.
Nothing hangs on this.)
The chief linguistic interest of this distinction lies in the structure of the subject
constituent. Consider first the singular examples:
(876) He wai t·aa·ku
cls water sg·Agen·Isg
‘I have a (glass of) water’
(877) He wai t·aa Pou
cls water sg·Agen Pou
‘Pou has a (glass of) water’

The subject constituent in such examples consists of a determiner with a cliticized

preposition and either an NP or a cliticized pronoun. Such phrases thus lack a lexical
head. Plural examples, however, have no overt determiner, eg.
(878) He kai aa·ku
cls food [pl]·Agen·Isg
‘I have some food’
(879) He kai a Pou
cls food Agen Pou
‘Pou has some food’

They thus appear on the surface to consist of a prepositional phrase, a construction which
cannot normally function as a subject in Maori.
 Maori 214

1.11 EMPHASIS

1.11.1 The expression of emphasis

1.11.1.1 Non-contradictory emphasis

Non-contradictory sentence emphasis in Maori is expressed through supra-segmentals. If
the sentence begins with a particle, that particle is stressed rather more than in non-
emphatic utterances. The main stress occurs on the first lexical item of the predicate. This
is accompanied by high onset pitch (a jump up if a particle precedes), prolonged first
segment, and drawl right through the item, accompanied by falling intonation. The
remainder of the utterance remains low. The following examples illustrate (indicates
drawl):

Stop consonants are “prolonged” by strong aspiration or affrication.

In instances where a sentence is repeated emphatically because it was not heard the
first time, however, rather different patterns are observed. The sentence is normally
repeated with all major lexical items stressed by increased loudness:
(882) A: Naa Ruka i hiki te tapu
actgen Ruka T/A lift the tapu
B: ↑E?
Eh
A: Naa ‘Ruka i ‘hiki te ‘tapu
actgen Ruka T/A lift the tapu
‘A: Ruka’s lifting the tapu.
B: What?
A: Ruka’s lifting the tapu!’

The repetition may have more pitch movement in its final fall, but this is not necessary.
 Syntax 215

There are also various ‘sentence introducers’ used in discourse which draw attention
to the following sentence, eg. E hui! or Huhui! which express the speaker’s excitement
about or great interest in the events.

1.11.1.2 Contradictory emphasis

Contradictory emphasis is chiefly expressed by extra strong stress on the verb or
predicative element, accompanied by a fall from high pitch, and sometimes by drawl.
There may be an explicit denial with kaahore ‘not’. If there is a second verb in the
sentence, it is also likely to receive strong stress, but without the pitch movement which
accompanies the first verb. The following examples illustrate:
(883) A: E peita ana ahau i taku whare
T/A paint T/A Isg DO sggenIsg house
B: E peita ana koe i toou taiapa?
T/A paint T/A IIsg DO sggenIIsg fence

1.11.2.1 Constituent emphasis

Constituent emphasis is far from simple in Maori, since the method that can be used
depends on the grammatical function of the item to be emphasised.
 Maori 216

1.11.2.1.1 Stress/accent
Stress/accent cannot shift from constituent to constituent for emphasis in the Maori of
older speakers, although it is not uncommon to hear younger speakers using it as in
English. However, if the verb or predicate NP is to be emphasised, the pattern for
sentence emphasis can be used, ie. the two are not distinct, eg.
(886) Kaatahi ka “reti·a e Maaui te raa!
then T/A snare·pass. by Maui the sun
‘Then Maui snared the sun!’
(887) He “hooiho taa Hone i hoko mai ai
cls horse sggen John T/A barter hither part.
‘John bought a horse’ (more lit. ‘What John bought was a horse’)

However, sometimes in explicit denials where a correction is made, the contrasted lexical
items are stressed, eg.
(888) A: Naa Maaui i pupuhi te raa?
actgen Maui T/A shoot the sun
B: Kaahore. Naa Maaui i “reti te raa
neg actgen Maui T/A snare the sun
‘A: Did Maui shoot the sun?
B: No. Maui snared the sun.’
(889) Kaahore. I kite·a e ahau he “pene, e’hara
neg T/A see·pass. by Isg a pen neg
i te “mukumuku
prep the eraser
‘No. I found a pen, not an eraser’

The existence of patterns like this no doubt facilitates the introduction of stress on
constituents other than the verb/predicate phrase in other contexts.

1.11.2.1.2 Particle
Maori does not have any particles whose sole function is to mark emphasised
constituents. However, sometimes both pre- and post-posed deictic particles occur in
noun phrases, with emphatic effect, eg.
(890) I teeraa wiiki raa, ka oati ia ki te haere
at that week dist T/A promise IIIsg to the move
mai
hither
‘Last week, he promised to come!’

However, the particle alone is insufficient—front position is also required to effect

emphasis.
 Syntax 217

1.11.2.1.3 Movement without dislocation

Pure movement, unaccompanied by other syntactic changes does not seem to be a method
of emphasis in Maori, except possibly for adverbials of time, and to a lesser extent place.
It certainly seems to be the case that emphasised time adverbials in Maori always
occur in initial position, eg.
(891) I“nanahi raa, ka mea atu au ki a koe
yesterday dist T/A say away Isg to pers IIsg
‘I told you yesterday!’

Note the reinforcing deictic particle in (891). Place adverbials are sometimes emphasised
by movement to initial position:
(892) I “konei ka kite·a e ahau
at here T/A see·pass. by Isg
‘It was here that I saw it’

(This can be either contradictory or non-contradictory.) However, not all adverbials can
move to initial position, eg. in
(893) Kei te “mau a Maama i a maatou ki te
T/A take pers Mummy DO pers Iplexcl to the
one
beach
‘Mummy’s taking us to the beach!’

it is not possible to front ki te one. The possibility of movement seems to depend on

whether the place adverbial is an argument logically required by the verb (as in (893)), or
whether it provides “framing” (as in (892)): only those which provide “framing” can
move.

1.11.2.1.4 Clefting
The structure which appears to be the closest match for clefting in other languages uses
ko. I have argued (Bauer, 1991) that there are two functions for ko which have previously
not been clearly distinguished, a topicalizing function which involves a fronting process,
and a clefting function, which involves an equative ko predication, where the subject is a
headless relative clause. Thus an unemphatic structure like
(894a) I kite a Hone i te taahae
T/A see pers John DO the thief
‘John saw the thief’

can undergo clefting for emphasis to produce

(894b) Ko “Hone i kite i te taahae
eq John T/A see DO the thief
‘It was John who saw the thief’

which has the structure

 Maori 218

(894c) [Ko Hone]Pred [i kite i te taahae]su

This process applies only to definite sentential subjects. The fronted constituent
receives strong stress. This structure is possible in both contradictory and non-
contradictory contexts. Direct objects can be made available for ko-clefting by promotion
to subject via either the passive or the actor-emphatic.
Equally well, the possessive-relative construction can be used as subject in an equative
sentence introduced by ko, eg.
(895) Ko te “kooauau taa Hone i tohu ai
eq the flute sggen John T/A save part.
‘It was the flute that John saved’

This is also possible in both contradictory and non-contradictory environments. There is

clearly a close link between these two types of ko structure, but it is not clear that
structures like (895) constitute clefting, rather than pseudo-clefting.
Not all English cleft-sentences are translated by my consultants with ko-structures in
Maori. Certain constituents, notably the full range of adverbials, cannot be ko-clefted.
Nor is it the case that ko-clefting is always used as the equivalent for English clefted
subjects. In non-contradictory examples, the actor-emphatic is quite frequently used for
those sentences to which it can apply.

1.11.2.1.5 Pseudo-clefting
Structures with general nouns like mea ‘thing’, wahi ‘place’ are the closest equivalent to
pseudo-clefting, eg.
(896) Ko “Hone te mea moohio ki te whakautu
eq. John the thing know to the answer
‘John’s the one who will know the answers’

This is a non-verbal structure: an equative sentence with ko. This structure can be used in
both contradictory and non-contradictory contexts. It can also occur without ko-clefting,
as in
(897a) Mea “ngaungau naa·na ngaa taura ki
thing gnaw actgen·IIIsg the(pl) rope with
ana niho
plgenIIIsg teeth
‘What he did was gnaw the ropes with his teeth’

But the subject can be clefted, as in

(897b) Ko ngaa taura mea ngaungau naa·na ki
top. the(pl) rope thing gnaw actgen·IIIsg with
ana niho
plgenIIIsg teeth
‘The ropes were things gnawed by him with his teeth’

It is not restricted to subjects, as is clefting, eg.

 Syntax 219

(898) Ko te kooauau te mea i tohu ai a Hone

eq the flute the thing T/A save part. gen John
‘The flute was what John saved’

(Note the unexpected relative clause here, see 1.1.2.3.4.2.3.) However, not all
constituents can be pseudo-clefted. Thus adverbials that can be fronted cannot be pseudo-
clefted (cf. 1.11.2.1.3):
(899a) *Ko te one te waahi i kite ai a
eq the beach the place T/A see part. pers
Hone i te motokaa taahae
John DO the car steal
‘The beach was where John saw the stolen car’

This can only be emphasised in the following manner:

(899b) I te one, ka kite a Hone i te
at the beach T/A see pers John DO the
motokaa taahae
car steal
‘John saw the stolen car at the beach’

However, not all pseudo-clefts in English can be translated with these sorts of pseudo-
clefts in Maori. Verbal structures can be pseudoclefted using this pattern but non-verbal
structures cannot. Thus English sentences like ‘What she wants to be is a doctor’ cannot
be paralleled in Maori. A straightforward non-verbal structure is the only possibility:
(900) Ko tana “hiahia hei rata ia
eq sggenIIIsg wish cls(fut) doctor IIIsg
‘Her wish is to become a doctor’

1.11.2.1.6 Dislocation
It is rather difficult to determine whether dislocation is possible in Maori, because any
phrase boundary is a position for potential pause. The difficulty is compounded by the
fact that subjects which have undergone (topicalizing) ko-fronting often drop the ko in
informal dialogue, although if asked to repeat, speakers often insert the ko. Thus subjects
can appear to be left dislocated:
(901) A: “Tino pai a Hone
very good pers John
B: Hone, he “tino pai rawa atu
John cls very good indeed away
‘A: John’s very good
B: John’s really excellent’

In (901), B made a pause after Hone, but it is not clear that this constitutes dislocation.
Even if it is dislocation, the stress suggests that it is not dislocation for emphasis.
 Maori 220

However, there are some other instances where dislocation does seem to be the
appropriate analysis.
Consider the following:
(902) Maaui, naa·na i hii ake Te-Ika-aa-Maaui
Maui actgen·IIIsg T/A fish up the-fish-gen-Maui
‘Maui, he fished up the North Island of N.Z.’

Notice that a pronominal copy remains attached to the preposition. While this does
appear to be dislocation, it is not clear that it is dislocation for emphasis, rather than for
topic. The following exchange makes it fairly clear that dislocation is used for topics in
Maori:
(903) A: Kei te aha a Hone?
T/A what pers John
B: (A) Hone, kei te ruku·ruku kooura ia
pers John T/A dive·dup crayfish IIIsg
‘A: What is John doing?
B: John, he’s diving for crayfish’

It thus appears unlikely that left dislocation is a means of emphasis in Maori. However,
given the impossibility of deciding whether fronted adverbials are moved or dislocated,
this conclusion must remain qualified.
Right dislocation occurs in Maori, but only as a clarifying afterthought:
(904) Ka tino pai ia, a Hone
T/A very good IIIsg pers John
‘He’s great, John (is)’
(905) Ka pai teenaa tikanga, ki te whakapaipai
T/A good that custom to the tidy up
‘That’s a good idea, tidying up’

This is not a means of emphasis.

1.11.2.1.7 Actor-emphatic construction

This construction (A-E for short), as its name suggests, emphasises the actor. It is
primarily available for the actors of canonical transitives, but under certain ill-defined
circumstances, occurs with intransitives (see further below).
The most appropriate analysis of this construction is a matter of controversy. The
surface characteristics are described first, and then some of the analyses are outlined.
Two examples are given here to facilitate the discussion.
(906) Naa Maaui i here-here te raa
actgen Maui T/A tie·dup the sun
‘Maui snared the sun’
(907) Maa·ku koe e whaangai
intgen·Isg IIsg T/A feed
 Syntax 221

‘I’ll feed you’

In the actor-emphatic construction, the verb is active in form. The semantic actor or agent
appears in pre-verbal position, in a phrase introduced by one of the prepositions maa or
naa. (These are basically possessive prepositions. Both contain the possessive a, used for
relations where the possessor is in control, see 1.10.6. The n-/m-distinction marks actual
vs. intended possession, see 1.10.5). As is usual for these prepositions, if the agent is a
singular pronoun, a special pronominal form is cliticized to the preposition (see
2.1.2.1.10 and 2.1.2.4.3). The semantic patient normally follows the verb, and occurs
without a preposition (ie. in the form normally reserved for subjects in Maori). The
patient may also appear between the emphatic actor constituent and the verbal constituent
(compare subject raising with negatives, 1.4); this is normal if the patient is pronominal,
but by no means restricted to pronouns. The T/A marker is always in agreement with the
preposition of the emphatic constituent: with n-, i is obligatory, with m-, e is obligatory.
Thus tense appears to be marked twice in this construction.
Various analyses of this construction have been suggested. I restrict myself here to
published material, and will not consider in detail the arguments for the various positions.
(Waite 1990 contains a summary of previous positions and arguments.) The analyses can
be grouped by what their proponents regard as the predicate in this construction. There
are some scholars who treat the ‘T/A-marker+V constituent as the predicate, eg. Hohepa
(1967, 19–23), Waite (1990). In this view, the naa/maa -NP is non-subject, but agentive,
and merely the form of agentive phrase that occurs pre-posed to the verb. The
construction is thus regarded as akin to the passive (but note the ergative-type marking
implied by this analysis). The lack of case-marking of the patient is thus accounted for,
and the form of the verb has a parallel in sentences with me (which Clark (1976, 115)
suggests may contain the same m- as the A-E prepositions), eg.
(908) Me whaangai koe e ahau
oblig feed IIsg by Isg
‘I should feed you’

This account does not provide an explanation for the possible fronting of the Ø-marked
NP to the position immediately preceding the verb, but can presumably handle it, for
instance, as ko-fronting with obligatory deletion of ko in non-initial position. This
analysis provides a natural description of the possible appearance of the Ø-marked NP
ko-fronted to initial position, as in
(909) Ko te raa naa Maaui i here-here
top. the sun actgen Maui T/A tie·dup
‘It was the sun that Maui snared’

However, it does not offer any explanation of certain other phenomena, such as the
failure of ai to appear following the verb in such constructions when there is a fronted
reason adverbial.
Alternative analyses propose that the n-/m- NP is predicative. Chung (1978,175ff)
proposes a structure like that in Fig 2 for (906). The subject clause contains the agent as
subject, but obligatorily deleted. Amongst the arguments for this is the fact that negative
 Maori 222

A-E sentences have eehara as negator, and the fact that the only possible T/A markers
with the verb are i and e, which are the most common T/A markers in embedded
sentences. The patient NP is analysed as an underlying DO promoted to subject after the
deletion of the underlying subject, thus accounting for the form of the underlying DO.
The alternative order with underlying DO preceding the V is then treated as subject
raising.
Fig 2: A-E base structure 1

A third analysis is suggested by Clark (1976, 119ff) and Harlow (1986). Clark proposes
an original structure for actor-emphatics of the kind in Fig 3:
Fig 3: A-E base structure 2

Within the predicate constituent, the naa/maa NP is described as a genitive, and the
second constituent as “an additional clause” (1976, 119). The second NP is the sentence
subject, hence its marking. The alternative word order is accounted for by extraposition
of the second element of a complex predicate over a short subject (see 1.13.2.5).
However, Clark also suggests that reanalysis has taken place in many Eastern Polynesian
languages, to a structure like that in Fig 4 (1976, 122) (my example). Note that this is
rather similar to Chung’s proposal, although the two accounts differ considerably in the
proposed generation of this structure. Clark’s suggestion of this reanalysis is supported
by certain languages where the DO marker appears, and also by the extension of the
 Syntax 223

construction to intransitive verbs. Both of these occur to some extent in Maori. A

consultant from the Taranaki area uses i for some of the underlying objects in this
construction, although there appears to be variation in this individual speaker, and I do
not know how widespread this is in the community. It is also the case that, under some
circumstances, intransitives appear in this construction. Such extensions are discussed
next
Fig 4: A-E Reanalysis

Firstly, underlying transitives with what might best be described as cognate patient
deletion are readily accepted in the actor-emphatic construction:
(910) Maa Pita e karakia
intgen Peter T/A chant
‘Peter will perform the incantation’
(911) Maa Pani e karanga
intgen Pani T/A call
‘Pani will call in welcome’

(Note that these sentences lack a surface subject)

Secondly, in instances where an intransitive verb is accompanied by a locative phrase,
the A-E can be used, eg.
(912) Naa Pita i oma te one
actgen Peter T/A run the beach
‘Peter ran along the beach’
(913) E tika ana, maa Pita e haere ngaa marae
T/A right T/A intgen Peter T/A move the(pl) marae
o te motu
gen the island
‘It’s right that Peter is accepted on any marae in the country’

However, not all such instances are accepted by consultants. The following, for example,
were rejected:
(914) *Naa te tamaiti i tuu te teepu
 Maori 224

actgen the child T/A stand the table

‘The child stood on the table’
(915) *Naa te kurii i peke a runga i ngaa
actgen the dog T/A jump pers top at the(pl)
hipi
sheep
‘The dog jumped onto the sheep’s backs’

These show that whether the verb involves movement or not is not a sufficient criterion
to determine the possibility of using this construction, although almost all the cases
accepted involve movement.
Time phrases are also at times accepted in the unmarked NP slot, eg.
(916) Maa Pita e hoki mai a taihoa
intgen Peter T/A return hither pers while
‘Peter will return by and by’

(I am unsure whether the a before taihoa is the personal article, or the future time
preposition.)
A few further points about this construction must be made. There seems to be
considerable disagreement on whether neuter verbs are acceptable in the actor-emphatic:
Hooper (1982) reports that they are, although in my experience, many consultants reject
them. This may be a matter related to the age of the consultants. However, the anomalous
case-marking is absent in such constructions, since the neuter verbs already have case-
marking more like the A-E. In addition, ai is added in Hooper’s examples (1982, 55)
(glosses altered to conform to those in this volume):
(917) Naa Rewi i mahue ai te wahine
actgen Rewi T/A abandon part. the woman
‘Rewi abandoned the woman’

It must also be noted that Biggs records a further construction which in certain respects
resembles the actor-emphatic. The construction is introduced by noo, and the T/A marker
is e…ana. However, unlike the naa/maa A-Es, this construction always retains the DO
case marker i, and is confined to subordinate clauses. The following is one of Biggs’s
examples (1969, 74):
(918) Noo·na e whakakii ana i tana kete
actgen·IIIsg T/A fill T/A DO sggenIIIsg kit
ka kite·a e te tangata
T/A see·pass. by the man
‘While he was filling his kit, he was discovered by the man’

There are examples with intransitive verbs which resemble this, eg.
(919) …Noo·ku pea e takatakahi ana i roto i
actgen·Isg perhaps T/A stamp T/A at inside at
a ia, ka kii iho nei…
pers IIIsg T/A say down proxI
 Syntax 225

‘…it was perhaps because I was moving violently within her that she said…’ (TWh, 19)

It is not at all clear whether these are to be regarded as actor-emphatic constructions.

After this somewhat lengthy descriptive digression, it is now possible to return to the
relation between this construction and the expression of emphasis. The actor-emphatic
construction is not uncommonly used to translate English clefts, eg.
(920) Naa Maaui i herehere te raa. Naa Maaui
actgen Maui T/A tie the sun actgen Maui
i hii ake Te Ika-aa-Maaui
T/A fish up the fish-gen-Maui
‘It was Maui who tied up the sun. It was Maui who fished up the North Island’

This is particularly true in past-time contexts. I suspect that this interacts with the fact
that, for canonical transitives, the least marked form in the past tense is the passive, so
much so that speakers often reject the active. Because ko-clefting applies only to
subjects, the actor cannot be emphasised through ko-clefting in past-time contexts. Thus
in the present, the active is unmarked:
(921a) Kei te patu a Pou i te kurii
T/A beat pers Pou DO the sheep
‘Pou is beating the dog’

and ko-fronting can emphasise the subject:

(921b) Ko “Pou kei te patu i te kurii
eq Pou T/A beat DO the dog
‘It is Pou who is beating the dog’

However, in the past, the active structure is seldom used:

(922a) *?I patu a Pou i te kurii
T/A beat pers Pou DO the dog
‘Pou beat the dog’

It is usually replaced by the passive:

(922b) I patu·a te kurii e Pou
T/A beat·pass. the dog by Pou
‘The dog was beaten by Pou’

(see 2.1.3.1.1.2 for a discussion of the conditions for this rule). The passive agent in
Maori is not an emphatic constituent—consultants reject structures of this kind as suitable
structures in contexts where the agent is emphasised. Ko-clefting cannot apply to the e-
phrase of the passive, but only to the unmarked NP (te kurii in (922b)). Thus (922b) is
not possible in past contexts where the actor requires emphasis. The past actor-emphatic
fills the hole:
(923) Naa Pou i patu te kurii
actgen Pou T/A beat the dog
‘It was Pou who beat the dog’
 Maori 226

In a study of topic and focus in 19th century Maori texts, Harlow (1982) found that
almost all A-E constructions were focal, and the exceptions were non-past A-E examples.
Biggs (1969, 73) suggests that the A-E is an alternative to ko-clefting.
One might guess that this could also explain the commonness of the A-E with neuter
verbs reported by Hooper: a “basic” neuter verb construction like
(924) I mahue te kurii i a Rewi
T/A leave the dog cause pers Rewi
‘Rewi lost the dog’

allows for ko-clefting only of te kurii. Fronting of the cause phrase is ungrammatical,
with or without the insertion of ai:
(925) *I a Rewi i mahue (ai) te kurii
cause pers Rewi T/A leave part. the dog
‘Rewi lost the dog’

However, the A-E provides a means of emphasising the agent:

(926a) Naa Rewi i mahue ai te kurii
actgen Rewi T/A leave part. the dog
‘Rewi lost the dog’
(926b) Naa Rewi te kurii i mahue ai
actgen Rewi the dog T/A leave part.
‘Rewi lost the dog’

It could also explain the development of the noo…e..ana structure described by Biggs,
since that structure allows for the emphasising of the actor in certain types of subordinate
clause.

1.11.2.1.8 Combinations of means of emphasis

I have noted the interaction of stress with the other means available in the individual
sections above. Essentially, each means of emphasis has its accompanying stress pattern.
Stress thus acts to reinforce the emphasis provided by the structure.
The double deictic mentioned in 1.11.2.1.2 can in principle be used with any definite
NP other than proper names, but is not particularly common in my experience.

1.11.2.2 The means for emphasising element types

The methods of emphasis are determined by phrase function, rather than phrase type in
Maori.

1.11.2.2.1.1 Noun phrase

Noun phrases are not treated homogeneously in Maori with respect to emphasis. The
details have been discussed in 1.11.2.1. Here one further set of examples illustrating the
non-homogeneous treatment of the category NP is given. The NP te kaiako ‘the teacher’
 Syntax 227

is emphasised by different means in each of the following contexts: as subject of an

active canonical transitive (927), as object of a canonical transitive (928), and as subject
of a passive (929):
(927) Naa te “kaiako i whakatikatika te ruuma
actgen the teacher T/A tidy the room
‘The teacher tidied the room’

This uses the actor-emphatic, with stress on the agent NP, which is normal when that NP
is non-pronominal.
(928) Ko te “kaiako naa·na i meke
eq the teacher actgen·IIIsg T/A hit
‘He hit the teacher’

This uses the A-E to make the DO a subject, and then clefting with ko applies. The ko-
clefted NP receives strong stress, as normal.
(929) Ko te ‘kaiako te mea me whakawhiu
eq the teacher the thing T/A punish
‘The teacher was punished’

This uses pseudo-clefting to emphasise the subject; this was deemed more emphatic than
the clefted version here.

1.11.2.2.1.2 Adjective
Adjectives which are predicative are emphasised by the normal predicative means, ie.
strong stress on the predicate, accompanied by drawl and emphatic pronunciation of
initial segments, eg.

(Note that this example shows topicalization of the subject.) Adjectives which are
attributive are normally emphasised by fronting the NP in which they occur, and
assigning the strongest phrase stress to the adjective, eg.
(931) Ko te .wuuru ‘whero taku “hiahia!
top. the wool red sggenIsg wish
‘I want red wool’
(932) Oo ringa ‘paru me tino “horoi ki te
plgenIIsg hand dirty T/A very wash with the
hopi
soap
‘Wash your dirty hands very well with soap’
 Maori 228

However, the sentence stress still falls on the predicate, and this stress is often emphatic.

1.11.2.2.1.3 Verb
Verbs are emphasised by emphatic stress, usually accompanied by drawl and emphatic
initial segment This is true even when the verb is not in sentence initial position by virtue
of the construction, eg.
(933) He “reka ki a au te paramu!
cls sweet to pers Isg the plum
‘I love plums!’
(934) Kua “koohuru·ngia a Pani!
T/A murder·pass. pers Pani
‘Pani was murdered!’
(935) Naa Hone a Tuu i “taarona
actgen John pers Tu T/A strangle
‘John strangled Tu!’
(936) Kaatahi a Tuu ka “taarona·ngia e Hone
then pers Tu T/A strangle·pass. by John
‘Then Tu was strangled by John’

Verbs can sometimes be emphasised by pseudo-clefting. Thus an alternative to (934) is

(937) Mea “koohuru a Pani
thing murder pers Pani
‘Pani was murdered’

1.11.2.2.1.4 Adverbial
Like NPs, these are not treated homogeneously. Some information has been given in
1.11.2.1 concerning major types of adverbial. Those adverbials which can be fronted are
emphasised in that way, eg.
(938) Noo “muri i te hui ka ‘hoki a ia
actgen behind at the meeting T/A return pers IIIsg
‘He left after the meeting’

It appears possible to use emphatic stress with the heads of such fronted adverbials.
Adverbials which cannot be fronted cannot readily be emphasised. The normal
treatment would be to use a construction which marks the entire content as noteworthy,
and leave the hearer to deduce the reason for this construction, eg.
(939) I moohio koe, mea “tapahi te keke ki te
T/A know IIsg thing cut the cake with the
‘toki!
axe
‘And do you know, he cut the cake with his axe!’
 Syntax 229

Modifying adverbials are emphasised like attributive adjectives, ie. they receive the
strongest stress within the predicate phrase, eg.
(940) Ka ,inu ‘poaka mai!
T/A drink pig hither
‘[They] drank like pigs!’
(941) ‘Haere ‘rua mai!
move 2 hither
‘Walk in pairs!’

But note the alternative:

(942) Haere ‘e rua mai!
move num 2 hither
‘Walk in pairs!’

where the particle of the modifying phrase receives the major stress.

1.11.2.2.2.1
Constituents of the main clause
This information has been covered above in various sections in 1.11.2.1 and 1.11.2.2.
Most main clause constituents can be emphasised. However, certain adverbial types
cannot, and the passive agent cannot be emphasised.

1.11.2.2.2.2 Constituents of subordinate clauses

Many constituents of subordinate clauses cannot readily be emphasised. Predicative
constituents can be emphasised by prosodic means (strong stress, drawl, etc), eg.
(943) E mahara ana ahau i tino “kino ia ki
T/A remember T/A Isg T/A very bad IIIsg to
te mahi nama
the work number
‘I remember that he hated maths’

The only subject constituents that can be emphasised appear to be those that can be actor-
emphatic agents, which on some analyses makes them predicative, eg.
(944) I mahue ai ahau noo te mea “naa·u i
T/A leave part. Isg actgen the thing actgen·IIsg T/A
‘mau te mapi!
take the map
‘I got lost because you took the map!’

Left dislocation is possible for DOs in fairly simple constructions, eg.

(945) Too ‘tamaiti, ka “tuumanako au ki te tuutaki
sggenIIsg child T/A hope Isg to the meet
‘I hope to meet your son’
 Maori 230

However, it is not at all clear that this is emphasis rather than topicalization.
It does not seem possible to emphasise other constituents of any subordinate
constructions, finite or non-finite.

1.11.2.2.2.3 Constituents of NP
It appears that the phrase stress can be moved from the lexical head to an adjectival
modifier or to a determiner, to emphasise those constituents, eg.

Note however that the section given emphatic pronunciation includes the head.
(947) I paatai ahau ki “aua taangata
T/A ask Isg to det aph(pl) people
‘I asked those people’

However, it does not appear possible to emphasise by prosodic or other means more
complex post-head modifiers, such as prepositional phrases.

1.11.2.2.2.4 Constituents of coordinate constructions

Only the first element of a coordinate construction can be emphasised, and then only if
the coordinate construction as a whole can be emphasised, eg.
(948) Ko tana “whaea me tana tuahine i
eq sggenIIIsg mother with sggenIIIsg sister T/A
“tae mai
arrive hither
‘His mother and sister came’

1.11.2.2.2.5 Emphasising more than one constituent simultaneously

Whether it is possible to emphasise more than one constituent simultaneously depends on
the construction used. It appears impossible in non-verbal constructions or the actor-
emphatic. Thus in the following dialogue, only the predicate can be stressed (and hence
emphasised):
(949) A: Ko ‘Hata te roia
eq Hata the lawyer
B: Kaahore. Ko ‘Hemi te roia. Ko ‘Hata te
neg eq Jim the lawyer eq Hata the
kaiako
teacher
‘A: Hata’s a lawyer
 Syntax 231

B: No. Jim’s a lawyer. Hata’s a teacher’

Similarly in the actor-emphatic:

(950) A: Naa Hone i “pupuhi te tuuii
actgen John T/A shoot the tui
B: Kaahore. Naa Hone i “hopu te kereruu
neg actgen John T/A catch the pigeon
‘A: John shot a tui.
B: No. John snared a pigeon’

However, in canonical verbal constructions, it appears possible to emphasise subject and

DO simultaneously or predicate and DO simultaneously. To accomplish the first, the
subject must be ko-clefted, but if two constituents are emphasised, neither receives full
emphatic stress, eg.
(951) A: E aroha ana a Piri ki a Mere
T/A love T/A pers Bill to pers Mary
B: Kaahore. Ko ‘Piri e aroha ana ki a
neg eq Bill T/A love T/A to pers
‘Mihi. Ko ‘Hone e aroha ana ki a ‘Mere
Mihi eq John T/A love T/A to pers Mary
‘A: Bill loves Mary.
B: No. Bill loves Mihi. John loves Mary’
(952) A: I kake a Piri ki te teiteitanga
T/A climb pers Bill to the summit
B: Kaahore. I “kake a Piri ki te ‘taumata.
neg T/A climb pers Bill to the lookout
I whaka“aahua ia i te teiteitanga
T/A photograph IIIsg DO the summit
‘A: Bill climbed to the summit.
B: No. Bill climbed to the lookout. He photographed the summit’

It does not seem possible to emphasise two elements simultaneously in non-contradictory

contexts.

1.11.2.2.3 Material left behind with movement processes

It is necessary to answer separately for each movement process. Ko-clefting, which
applies only to subjects, leaves nothing behind. With adverbial movements, the particle ai
is sometimes left behind, but is post-verbal rather than in the position occupied by the
moved constituent. However, this does not occur unless the T/A marker of the verb is one
of the absolute tense markers, i and e (or in modern Maori, no T/A marker, which
 Maori 232

probably arises from the elision of e). With dislocation, a pronominal copy remains
behind in the slot occupied by the dislocated constituent.

1.11.3 Focus of yes-no questions

The focus of yes-no questions is indicated by the same method as the emphasis of that
constituent. This means that certain items cannot readily be the focus of yes-no questions,
and that the method varies from one constituent type to another, and from one
construction to another. A few sets of contrasting examples are given below.
(953a) Naa ‘Hone a Piri i patu?
actgen John pers Bill T/A beat
‘Did John hit Bill?’ (neutral)
(953b) Naa “Hone a Piri i patu?
actgen John pers Bill T/A beat
‘Did John hit Bill?’
(953c) I “tino ‘patu a Hone i a Piri?
T/A very beat pers John DO pers Bill
‘Did John (really) hit Bill?’
(953d) Ko “Piri naa Hone i patu?
eq Bill actgen John T/A beat
‘Did John hit Bill?’
(954a) Kei ‘konei a Piri?
at(pres) here pers Bill
‘Is Bill here?’ (neutral)
(954b) Kei “konei a Piri?
at(pres) here pers Bill
‘Is Bill here?’
(954c) Ko ‘Piri te mea kei ‘konei
eq Bill the thing at(pres) here
‘Is Bill here?’

(Ko-clefting of the basic structure here apparently produces an echo-question.)

(955a) E’aruaru ana a Hone i a Piri?
T/A chase T/A pers John DO pers Bill
‘Is John chasing Bill?’ (neutral)
(955b) Ko “Hone e aruaru ana i a Piri?
eq John T/A chase T/A DO pers Bill
‘Is John chasing Bill?’
(955c) Ko “Piri e aru·mia ana e Hone?
eq Bill T/A chase·pass. T/A by John
‘Is John chasing Bill?’ (lit. ‘Is Bill being chased by John?’)
(955d) E”aruaru ana a Hone i a Piri?
T/A chase T/A pers John DO pers Bill
‘Is John chasing Bill?’
 Syntax 233

(956a) I’konei a Piri inanahi?

at here pers Bill yesterday
‘Was Bill here yesterday?’ (neutral)
(956b) I’nanahi i konei a Piri?
yesterday at here pers Bill
‘Was Bill here yesterday?’

1.12 TOPIC

1.12.1 Means of indicating sentence topic

Maori is not a topic prominent language, and in general, topics are not marked:
(957) A: E haere ana a Hone ki te pikitia.
T/A move T/A pers John to the picture
B: Ka whaka·peehea a Piri?
T/A cause-how pers Bill
A: Ka noho ia ki te kaainga
T/A stay IIIsg to the home
‘A: John is going to the movies.
B: What about Bill?
A: He’s staying at home’

However, it appears that on occasion, left dislocation can be used for topics. Two
possible examples were given in 1.11.2.1.6, (902) and (903). However, this is by no
means a regular phenomenon, and many parallel examples are rejected by consultants.
Topic switch, however, is quite commonly marked in Maori by fronting with ko, a
process which applies principally but not exclusively to subjects. Such ko-phrases do not
receive the major sentence stress, which remains on the predicate. Examples illustrating
topic switch are:
(958) Mea unahi ngaa ika, ka kai·nga. Engari ko thing scale the(pl) fish T/A eat·pass. but top. ngaa
pero me ngaa whiore, naa raatou i the(pl) head with the(pl) tail actgen IIIpl T/A waiho noa
ki runga i te one leave indeed to top at the beach
‘They scaled the fish and ate them. As for the heads and tails, they just left them on the
beach’
(959) Mea taahae te TV me ngaa hiriwa, engari
thing stolen the TV with the(pl) silver but
ko ngaa rama mea inu katoa
top. the(pl) booze thing drink all
‘They stole the TV and the silver, but as for the booze, it was all drunk’

Texts frequently begin with several scene-setting statements, which frequently all show
topic-switch with ko-fronting, eg.
 Maori 234

(960) Ko teenei tangata ko Wairangi noo Ngaati

top. this man top. Wairangi actgen Ngati-
Raukawa. Ko toona kaainga ko Rurunui i te
Raukawa top. sggenIIIsg home eq Rurunui at the
takiwaa o Whare-puuhunga. Ngaa waahine a
district gen Whare-puhunga the(pl) women pers
Wairangi toko·rua, ko Pare-whete, ko Puuroku.
Wairangi pnum·2 eq Pare-whete eq Puroku
‘This man, Wairangi, belonged to Ngati-Raukawa. His home was at Rurunui, in the district of
Whare-puhunga. Wairangi had two wives, Pare-whete and Puroku’ (W, 197)

(Note that the final sentence here has fronting without ko; this is fairly common in
informal speech.)
There are also examples which appear to mark topic-switch with ko but without
fronting, eg.
(961) Ka mutu te haka a Tama-te-hura, ka puta
T/A finished the haka gen Tama-te-hura T/A appear
ko Upoko-iti, e tataki ana i taana haka
top. Upoko-iti T/A lead T/A DO sggenIIIsg haka
‘When Tama-te-hura had finished his haka, Upoko-iti appeared and led his haka’ (W, 200)

(Such examples are not confined to coordinate clauses, although they often seem to occur
when there is a degree of parallelism in the events narrated.)
There is another means of introducing a new protagonist into a scene, namely
introduction by teeraa. This has the force of English ‘Now, there was this…’. It is not
usually used at the beginnings of texts. Some examples are given below, with information
about the story immediately preceding the teeraa.
In Tutae-poroporo, the introduction discusses the acquisition of Tutae-poroporo by his
master, and his development into a taniwha. The second section begins:
(962) Aakuanei, teeraa teetahi ope-taua naa Whanganui
presently there a(sp) group-war actgen Whanganui
kei te haere mai
T/A move hither
‘Presently a war-party from Whanganui came along’ (TP, 89)

The second example is from Tu-whakairi-ora. Until this point in the narrative, the death
of Poroumata and the birth and growth to maturity of Tu-whakairi-ora have been the
focus of the story. A new element is introduced thus:
(963) Teeraa ngaa rongo aataahua o ngaa tamaahine
there the(pl) fame beauty gen the(pl) daughter
a Te Aotakii, o Ruataupare, raaua ko
gen Te Aotaki gen Ruataupare IIIdl spec
Auahi-kooata, kua hau noa atu ki Oopootiki
Auahi-koata T/A report indeed away to Opotiki
‘Now the tidings of the beauty of the daughters of Te Aotaki,
Ruataupare and Auahi-koata had spread even to Opotiki’
 Syntax 235

(TWh, 19)

In the story of Te Aohuruhuru, the heroine has just decided to commit suicide, and the
story continues:
(964) Naa, teeraa teetahi toka teitei e tuu ana i te
now there a(sp) rock tall T/A stand T/A at the
tahatika…
shore
‘Now, there is this lofty crag that stands near the shore…’(TA, 6–7)

This form of introduction, which apparently also fronts subjects only, is used to introduce
important new ingredients into the story. The following appears to be an occurrence in a
subordinate clause:
(965) Kua moohio te iwi o Te Aea teeraa a Ngaati
T/A know the tribe gen Te Aea there pers Ngati
Raukawa e haere atu
Raukawa T/A move away
‘The Te Aea people knew that Ngati-Raukawa would come’ (W, 197)

1.12.2 Elements topicalized by different means

Only noun phrases can be topicalized with ko, but on occasion other types of constituent
can be given nominal form in order to be topicalized. However, such cases are rare, and
constructed attempts to reproduce them were rejected, so they must be considered rather
marginal.
Only subject NPs can be freely topicalized, and even then subject NPs from
subordinate clauses cannot readily be topicalized. However, certain constituents of
subject NPs can be topicalized, but it is not clear precisely under what circumstances, nor
whether there are restrictions on which constituents can be topicalized.
Nothing is left behind to mark the topicalization of subjects or their constituents.
Some examples illustrating these points follow. The first shows two topicalized
subjects:
(966) He koroheke te tangata nei, ko tana
cls old man the man proxI top. sggenIIIsg
wahine he tuutuuaa, he mea tango mai e ia
woman cls low-born cls thing take hither by IIIsg
i te tangata i aroha·tia nui·tia e teenei
from the man T/A love·pass. big·pass. by this
wahine. Ko te take i tango·hia ai e ia
woman top. the reason T/A take·pass. part. by IIIsg
te wahine he pai, he aataahua, he wahine the woman cls good cls beautiful cls woman momoho
ki te mahi.
capable to the work
‘This man was old. His wife was of lowly birth, and had been taken by him from one she loved
dearly. The reason the woman was taken by him was that she was good, beautiful and
accomplished’ (TA, 6)
 Maori 236

Non-subject examples are few and far between. However, the following is the possessive
phrase of a subject:
(967) Ko te koroheke nei kua paawera noa ake te
top. the old man proxI T/A stir indeed up the
ngaakau ki te purotuta·nga o tana wahine
heart to the handsome·nom gen sggenIIIsg woman
taitamariki
young person
‘The old man’s heart glowed at his wife’s youthful beauty’ (TA, 7)

The following is an example of the topicalization of an adnominal phrase which is

embedded in a subject:
(968) Ko ngaa waka i mua raa, mate katoa ngaa
top. the(pl) canoe at front dist die all the(pl)
taangata o runga
people gen top
‘The canoes in front, all the people on them died’ (TP, 90)

There are occasional examples of topicalized embedded subjects, eg.

(969) Ko te ingoa nei naa Te Aotakii i tapa
top. the name proxI actgen Te Aotaki T/A name
‘This name was given by Te Aotaki’ (TWh, 23)

Note also the following, where the topicalization of the subject takes place within the
subordinate clause:
(970) Ka titiro a Wairangi, ko ngaa poupou o te
T/A look pers Wairangi top. the(pl) side post gen the
whare he koohurihuri kahikatea
house cls sapling kahikatea
‘Wairangi noticed that the side posts of the house were kahikatea saplings’ (W, 198)

Examples where non-subject-like concepts have been encoded as subjects and then
topicalized include:
(971) Otiraa ko toona noho e noho pononga ana
but top. sggenIIIsg stay T/A stay slave T/A
ki teenei koroheke, otiraa ko toona aroha
to this old man but top. sggenIIIsg love
pea e mau tonu ana ki te tangata i
perhaps T/A hold indeed T/A to the man T/A
aroha·tia nui·tia e ia
love·pass. big·pass. by IIIsg
‘But she lived the life of a slave with the old man, her heart still yearning for the loved one
from whom she had been torn’ (TA, 6)
(972) Kua moohio te iwi ko te mate o toona
T/A know the tribe top. the dead gen sggenIIIsg
tipuna, o Poroumaataa, ka haere·a e
 Syntax 237

ancestor gen Poroumata T/A move·pass. by

Tuu-whakairi-ora
Tu-whakairi-ora
‘The tribe knew that the death of his grandfather Poroumata was the reason Tu-whakairi-ora
was going’ (TWh, 19)
(973) Ka oho ngaa iwi, ko te mate o
T/A wake the(pl) tribe top. the dead gen
Poroumaataa te haramai a Tuu-whakairi-ora
Poroumata the arrive gen Tu-whakairi-ora
‘The tribes jumped to his meaning; avenging the death of Poroumata was the occasion of
Tu-whakairi-ora’s coming’ (TWh, 21)

1.12.3 Preference for topicalizing

Topicalization is not obligatory. In narrative texts, not all topicswitches are marked by ko
(or fronting without ko), eg. in the story of Wairangi, there is a long description of the
flight of Pare-whete, which ends thus:
(974) Ka whiti i Waihou, ka tae ki Te Aea. Ka
T/A cross DO Waihou T/A arrive to Te Aea T/A
moe i taana taane i a Tupeteka. Ka
sleep DO sggenIIIsg man DO pers Tupeteka T/A
kimi te iwi nei a Ngaati-Raukawa i a
search the tribe proxI pers Ngati-Raukawa DO pers
Pare-whete
Pare-whete
‘[She] crossed the river Waihou and reached Te Aea. There [she] married her husband,
Tupeteka.
This tribe, Ngati-Raukawa, searched for Pare-whete’ (W, 197)

The switch of topic from Pare-whete to the tribe is not marked by topicswitch ko,
although the translator (Te Rangihiroa) saw fit to begin a new paragraph at that point in
the translation. The amount of topicswitch marking varies from author to author. It
appears more likely to be marked if there is a quick succession of topic shifts than with
switches which occur after a sustained topic. Unpredictable shifts are also more likely to
be marked than those which the story leads the hearer to expect However, these are only
tendencies, and there are counter-examples to them all. (If one is eliciting data by getting
translations of sentences in isolation, topicalized versions are frequently proffered.)

1.13 HEAVY SHIFT

1.13.1 Movement processes for heavy constituents

It is impossible to determine what counts as heavy shift in Maori independent of one’s
stand on basic constituent order. Below I will provide data which may be accounted for
as heavy shift, and discuss the implications of the data in relation to three possible
hypotheses about base order.
 Maori 238

All potential movements which might be heavy shift in Maori move material
rightwards, toward the end of the sentence. While there are certain movement processes
which do genuinely seem to be heavy shift whatever the base order hypothesis, many of
the instances of movement of constituents appear to be better described as the movement
of light constituents leftward, rather than heavy constituents rightward. Movement
processes of the light kind are discussed in 1.14.

1.13.2 Structures subject to heavy shift

1.13.2.1 Adjective phrases

Attributive adjective phrases do not appear to be subject to heavy shift: they must remain
with the head noun. Predicative adjective phrases behave like other predicative phrases,
see 1.13.2.5.

1.13.2.2 Object NPs

These can appear in final position, following any oblique NPs, if they are heavy. Direct
and indirect speech is almost always treated thus if the addressee is specified, eg.
(975) Kaaore i roa e mahi haere ana kua paatai
neg T/A long T/A work go T/A T/A ask
ia ki te hoa, “Moo aawhea raa te whiu
IIIsg to the friend intgen when(fut) dist the make
kai nei maa Ngaati-Raukawa?”
food proxI intgen Ngati-Raukawa
‘They hadn’t been going along working for long when he asked his companion, “When is
the food to be ready for NgatiRaukawa?”’ (W, 198)
(976) Ka koorero a Toto ki ana kootiro,
T/A speak pers Toto to plgenIIIsg girl
mehemea e pai raaua ki te moe taane
if T/A good IIIdl to the sleep man
‘Toto asked his daughters if they wanted to get married’ (KH, 2)

This word order can be used even if actual words are not quoted, eg.
(977) Noo te ata ka koorero ki a Toto i
actgen the morning T/A tell to pers Toto DO
ngaa mea katoa
the(pl) thing all
‘In the morning [he] told Toto everything’ (KH, 4)

However, heavy shift of objects is not always possible. Thus the following example was
rejected. The phrase ki runga i te teepu was required to be in final position, following
tamariki:
(978) *E whaka·tuutuu ana ia ki runga i te teepu
T/A cause-stand T/A IIIsg to top at the table
i ngaa pukapuka e mau·ria mai ana ki
 Syntax 239

DO the(pl) book T/A bring·pass. hither T/A to

a ia e ngaa tamariki
pers IIIsg by the(pl) children
‘She is standing on the table the books which the children are bringing her’

The unacceptability of heavy shift with such an obviously heavy DO NP, and the
occurrence of DO-final order in (977), when the DO is not particularly heavy, calls into
question whether this process is really heavy shift. It might be better accounted for in
terms of leftward movement of NPs higher up the animacy hierarchy.

1.13.2.3 Adverb phrases

Since these are normally final, heavy shift does not apply.

1.13.2.4 Subject NPs

Subjects, if heavy, can appear in final position following a light DO or adverbial phrase.
However, whether this is heavy shift or not depends on the position taken as to the basic
structure of Maori (see further 1.13.3.1). Examples are:
(979) Kua whakahoki·a ki a ia te tamaiti i
T/A return·pass. to pers IIIsg the child T/A
mau·ria e ngaa pirihimana i te marama o
take·pass. by the(pl) policeman at the month gen
Oketopa
October
‘The child who was taken by the police in October has been returned to her’
(980) Kei te aawhina i a ia ngaa tamariki katoa
T/A help DO pers IIIsg the(pl) children all
o te taaone o Te Kao
gen the town gen Te Kao
‘All the children of Te Kao are helping her’

Subjects may also be shifted towards the right, without necessarily taking final position—
lengthy adverbs may still follow, eg.
(981) Ka puta katoa ki waho ngaa taangata o te
T/A appear all to outside the(pl) men gen the
paa ki te tahu kai
pa to the cook food
‘All the people of the pa came outside to cook food’ (H, 10)

A number of points are raised by this example. Firstly, note that quantifier float from the
subject is not hindered by the intervening adverbial constituent. Second, this subject
constituent is not especially heavy, which raises the issue of what counts as heavy in
Maori. It appears to be the case that this is a relative matter—a constituent is heavy if it is
heavier than other phrases in the environment, in this case te waho. Thirdly, this is the
kind of example which might just as well be accounted for as an example of light shift
 Maori 240

Long subjects very frequently appear after passive agents, either finally, or preceding
other adverbials, eg.
(982) Muri iho ka whakaara·hia e ia oona
behind down T/A wake·pass. by IIIsg plgenIIIsg
hoa koroheke o roto i te whare ki te
friend old gen inside at the house to the
maatakitaki ki te aataahuatanga o tana wahine
gaze to the beauty gen sggenIIIsg woman
‘Afterwards he roused his old cronies in the house to gaze at the beauty of his wife’ (TA, 6)

However, length/weight is not always the critical factor in such shifts, as the following
example indicates:
(983) Kaatahi ka whakanoho·hia e te tangata nei
then T/A place·pass. by the man proxI
toona mookai ki roto ki teetahi puna
sggenIIIsg pet to inside to a(sp) spring
‘Then this man placed his pet in a spring of water’ (TP, 89)

This also applies to subjects in actor-emphatic sentences, eg.

(984) Naa toona tupuna i whakaatu ki a ia
actgen sggenIIIsg ancestor T/A show to pers IIIsg
ko wai oona tuaakana
eq who plgenIIIsg older brothers
‘His grandfather pointed his brothers out to him’ (KM, 1)

Many of these examples could be equally well accounted for as leftward movement of
NPs high on the animacy hierarchy.

1.13.2.5 Predicates
In any non-verbal sentence, if the predicate is complex, then any constituents after the
first can be moved across the subject, to appear in rightmost position. Clark (1976, 119)
calls it extraposition and notes that it applies only over short subjects. It might be more
accurate to say that it cannot apply over long subjects, and is normal with short subjects
in the light of examples like (986) below. Such extraposition is the unmarked word-order
in these sentences, and if the predicate consists of more than two phrases, such
extraposition appears to be compulsory. The following is a range of types of example:
Complex prepositional phrase:
(985) Kei waenganui te taaone i te ngahere
at(pres) middle the town at the bush
‘The town is in the middle of the bush’ (TR2, 40)
(986) I runga ngaa momo kai katoa i te teepu
at top the(pl) kind food all at the table
‘All kinds of food were on the table’

Equative NP with two prepositional phrases:

 Syntax 241

(987) Ko te kooha teenei a Wairangi ki tana

eq the gift this gen Wairangi to sggenIIIsg
wahine
woman
‘This was Wairangi’s gift to his wife’ (W, 199)

Equative NP with relative clause:

(988) Ko te tohu teeraa i tika atu ai ia ki
eq the sign that T/A straight away part. IIIsg to
Waikimihia
Waikimihia
‘This was the sign that led her straight to Waikimihia’ (H, 8)

Classifying NP with prepositional phrase:

(989) He pahi teenei noo te kura
cls bus this actgen the school
‘This is a bus belonging to the school’

Classifying predicate with verbal-type adjuncts:

(990) He whakaako tana mahi i ngaa tamariki o
cls teach sggenIIIsg work DO the(pl) children gen
te kura o Te Kaha ki te tao kai
the school gen Te Kaha to the cook food
‘Her job was to teach cooking to the children of the Te Kaha school’ (TR2, 71)

Adjective with qualifying prep phrase:

(991) Tino tere te poti ki te haere
very fast the boat to the move
‘The boat is very fast’

1.13.3 Position of moved elements

The appropriate formulation of the position to which constituents are moved depends on
what the appropriate underlying structure for Maori is taken to be. If the structure is taken
to be as in Fig 5,
Fig . 5: Base structure 1

for all types of sentence, both verbal and non-verbal, then heavy shift does not apply at
all to subjects in Maori. Heavy subjects in verbal sentences would simply not undergo the
 Maori 242

rule which orders lighter subjects to the left of all constituents of the predicate other than
the phrase containing the lexical verb. Under this interpretation, all constituents of the
predicate other than the lexical verb phrase could be seen as undergoing heavy shift, and
this one rule would also account for the movement of non-head constituents of non-
verbal predicates to the right of subjects as well, although it should be remembered that in
the case of non-verbal predicates, the movement is optional for twophrase predicates.
Under this interpretation, the clearest case of heavy shift is Objects of certain types,
which do appear to be moved to the end of the constituent they are immediate
constituents of. However, it seems to me a little odd to propose as a basic order of
elements VOS, when that is marked both in terms of likely typological patterns, and in
terms of Maori, where it is a rare order of surface constituents. Its advantage is that it
allows for both verbal and non-verbal predicates to be treated alike.
If, on the other hand, the basic structure is taken as in Fig 6:
Fig 6: Base structure 2

then the data in 1.13.2.4 would be accounted for as heavy shift, and so would the data in
1.13.2.2. Both of these constituent types would then be moved to the end of the
constituent they are immediate constituents of. The movement of items from non-verbal
predicates would also appear to be heavy shift, but in this case, such constituents would
not be moved to the end of the constituent of which they were immediate constituents,
but to a position best specified as immediately post-subject. While this sort of hypothesis
is more intuitively satisfactory, in that it takes unmarked orders as basic and derives
marked structures by movement rules, it requires the postulation of different rules for
different sentence types.
If the basic structure was that in Fig 7:
Fig 7: Base structure 3

(and there seems least motivation for this internally from Maori), then there would have
to be a rule for “normal” subject placement, which could be following the head phrase of
the predicate. The data in 1.13.2.2 and 1.13.2.4 would thus be accounted for by heavy
shift, but the “extraposed” predicate structures of 1.13.2.5 would be “normal”, and the
optional structures with the predicate uninterrupted would have to be derived by some
 Syntax 243

additional rule: perhaps that subject placement could optionally follow the entire
predicate if the predicate was not heavy.
This is not the appropriate place to argue for one or other of these solutions. I wish
merely to draw attention to the fact that the data here is not necessarily all to be
accounted for as heavy shift. Indeed, if the order IO—DO in “speech” sentences is
accounted for in terms of animacy, it is possible that Maori shows no evidence of heavy
shift.

1.13.4 More than one heavy phrase

It is not possible to move more than one heavy phrase in Maori. If both subject and object
are heavy, they retain their basic positions, eg.
(992) Ka kite mai a Rangi-uru-hinga, he taniwha
T/A see hither pers Rangi-uru-hinga cls taniwha
teenei kei Moana-ariki e noho ana, i te
this at(pres) Moana-ariki T/A stay T/A DO the
kino o te mahi a Kupe ki a Hotu
bad gen the work gen Kupe to pers Hotu
‘Rangi-uru-hinga, who was a taniwha living in Moana-ariki saw the wickedness of Kupe’s
treatment of Hotu’ (KH, 3)

1.13.5 Heavy shift with elements next to complementizers

Heavy shift is relatively rare as a whole, and even rarer next to complementizers. Since
verbs immediately follow complementizers, the only potential heavy shift from this
position is predicate shift in relation to adjectival predicates. But since it is not the head
phrase which is moved, the situation simply does not arise.
In the position preceding complementizers, the only potential candidates for heavy
shift are subjects. It appears that they can be shifted over a short complementizer phrase,
whether the complementizer is kia or ki te, eg.
(993) E piirangi ana ki te haere mai he tangata
T/A want T/A to the come hither a man
noo Ingarangi tino moohio ki te tuhi·tuhi ture
actgen England very know to the write·dup law
‘A man from England who is very expert at legal drafting wants to come’
(994) E piirangi ana kia haere atu maatou he
T/A want T/A subj move away Iplexcl a
tangata noo Ingarangi tino moohio ki te
man actgen England very know to the
tuhi·tuhi ture
write·dup law
‘A man from England who is very expert at legal drafting wants us to come’

However, the versions deemed most natural had these heavy subject phrases fronted.
 Maori 244

1.14 OTHER MOVEMENT PROCESSES

1.14.1 Movements other than heavy shift

As foreshadowed in 1.13, there seem to be some movements in Maori better accounted
for as the movement of light material or phrases containing animate NPs leftward, rather
than of heavy material rightwards. Light shifts are discussed in 1.14.1.1. There are also a
number of instances where subjects move leftwards, but which are not accounted for by
the “packaging” strategies discussed in 1.11 and 1.12. Some of these are fairly clearly
raising, and are thus not merely movement processes. However, in other superficially
similar cases, there do not seem to be such strong grounds for treating the movements as
raising. These processes are treated in 1.14.1.2.

1.14.1.1 Light constituents

There are a number of cases in Maori where the various arguments which follow the verb
do not occur in the canonical order, namely subject—DO—obliques. In some such
instances, none of the NPs involved is sufficiently heavy to justify heavy shift as the
explanation. Intuitively, in such cases, what seems to happen is that informationally less
salient NPs are placed to the left of informationally more salient NPs, much in the way
that the theories of “communicative dynamism” proposed by Firbas and others (eg.
Firbas, 1959) would predict As a result of these movements, the order verb—agent—
patient often occurs with constructions in which agent is not mapped onto subject, and
patient is not mapped onto DO.
One of the commonest constructions to show non-canonical ordering is the passive.
The unmarked order is Vpass.—subject—agent, eg.
(995) I pooki·a te rangi e ngaa kapua
T/A cover·pass. the sky by the(pl) cloud
‘The sky was covered over by clouds’

However, examples showing the agent phrase preceding the subject are not at all
uncommon. A selection is given below:
(996) Kaatahi ka whiu·a e ia taana
then T/A throw·pass. by IIIsg sggenIIIsg
mookai ki te moana
youngest child to the sea
‘Then she threw her baby into the sea’ (TR2, 188)
(997) Ka tuku·na e Wairangi tana taurekareka ki
T/A send·pass. by Wairangi sggenIIIsg slave to
a Pare-whete
pers Pare-whete
‘His servant was sent by Wairangi to Pare-whete’ (W, 199)
(998) Kaatahi ka whakanoho·hia e te tangata nei
then T/A place·pass. by the man proxI
toona mookai ki roto ki teetahi puna
sggenIIIsg pet to inside to a(sp) spring
 Syntax 245

‘Then this man placed his pet in a spring of water’ (TP, 89)

(Notice that this example was adduced above as a possible example of subject heavy
shift.)
(999) Ka whakapaa·ngia atu e Rewi teetahi raakau ki te
T/A touch·pass. away by Rewi a(sp) stick to the
waewae o Tamahae
leg gen Tamahae
‘Rewi touched Tamahae’s leg with a stick’ (TR2, 6)

In all these cases, and indeed in very many such cases, the verb is followed not only by a
subject and passive agent, but by some other phrase as well. If the ‘normal’ position of
the passive agent is the last of the verbal arguments, then it appears that it is seldom in
that position (if ever) if two arguments intervene between it and the verb, even if those
arguments are short (as in (996) above). It must be taken into account that in the past
tense in Maori, the passive is the unmarked form of sentence for many transitive verbs
(see 2.1.3.1.1.2). This means that the active, which would put the agent immediately
following the verb is not in competition with the passive in these environments (ie. the
passive is not chiefly used, as it is in English, for “packaging” purposes). Thus, although
the passive form of the verb is required, the order verb—agent—affected object is
frequently preserved. Notice that examples like (998) can as well be accounted for by this
process as by heavy shift. It may also be relevant to note that in both (997) and (998), the
use of the marked order facilitates pronominalization, which generally proceeds
backwards in Maori (see 1.6).
Sometimes the passive agent occurs between the subject and some other adverbial:
(1000) Kua kiki·a te paaoro e Tawhiti ki a Piripi
T/A kick·pass. the ball by Tawhiti to pers Philip
‘The ball was kicked by Tawhiti to Philip’

Not all the examples of the marked ordering have more than two arguments, however, eg.
(1001) Kua kauhoe·tia e ia te awa
T/A swim·pass. by IIIsg the river
‘She has swum the river’
(1002) E hoko·na ana e au kia whaa
T/A buy·pass. T/A by Isg subj 4
‘I’m buying four’
(1003) Kaatahi ka whakakaa·ngia e Tane te raiti
then T/A switch on·pass. by Tane the light
‘Then Tane switched on the light’

In none of these cases does it seem reasonable to postulate heavy shift, but in all cases,
the Ø-marked NP is more salient to the discourse than the passive agent, and the passive
agent is higher in animacy. Either of these factors could plausibly account for the order.
The same phenomenon also occurs with neuter verbs, and the use of the non-canonical
order here also results in the verb—agent—patient ordering, eg.
 Maori 246

(1004) I mate i a Kupe te wheke nei ki

T/A kill cause pers Kupe the octopus proxI to
Raukawa
Cook Strait
‘Kupe killed this octopus in Cook Strait’ (TR2, 81)
(1005) I hinga katoa i a ia ngaa toa
T/A defeat all cause pers IIIsg the(pl) champion
o teeraa whenua
gen that land
‘All the champions of that country were defeated by him’
(TR2, 151)

While examples like (1005) can plausibly be accounted for as heavy shift, examples like
(1004) seem to be ordered by some other principle. Cause phrases with intransitives can
also be treated thus:
(1006) Haere atu, kei tangi i a koe te tamaiti
move away mon cry cause pers IIsg the child
nei!
proxI
‘Go away, you might make the child cry!’

The NPs in question in (1006) do not differ in animacy.

Notional IOs also sometimes precede subjects, eg.
(1007) Homai ki a maatou toou waka
give to pers Iplexcl sggenIIsg canoe
‘Give us your canoe’

Again, this does not really seem explicable as heavy shift, but is more plausibly analysed
as light or animate material moving leftwards.

1.14.1.2 Raisings and similar phenomena

There are several cases in Maori where arguments have been adduced to support subject
raising, most notably negation (see 1.4). An example is given here for ease of comparison
with other data presented in this section. (1008) is negative with the subject raised:
(1008) Kaahore te hau i te pupuhi
neg the wind T/A blow
‘The wind is not blowing’

Certain other instances of non-canonical order appear to be raising like these, notably
those with kaatahi. Three examples are given here, the first without subject raising, the
last two with subjects raised:
(1009) Kaatahi anoo ka mutu te pikitia
then again T/A finish the picture
‘The picture has only just finished’ (TR1, 52)
 Syntax 247

(1010) Kaatahi anoo ia ka kite he raakau kee

then again IIIsg T/A see cls stick contr
‘It was then that he saw it was a stick’ (TR2, 6)
(1011) Kaatahi a Tamahae maa ka rongo atu
then pers Tamahae and others T/A hear away
i a Hata e kanga ana
DO pers Hata T/A curse T/A
‘Then Tamahae and co. heard Hata cursing’ (TR2, 59)

The similarity of these to negatives in structure is fairly plain. It appears possible that
kaatahi is a higher predicate (including clitic T/A marker, ka), and that the lower
predication is the underlying sentential subject of the predication. The subject of the
lower predication can optionally be raised into the higher predication, and as with
negatives, the subjects can be ko-fronted. However, notice that the T/A marker of the
predication under kaatahi is ka, which is not associated primarily with subordinate
clauses, unlike those in negatives. This removes one of the main arguments for this
analysis. The case is thus less clear-cut than for negatives. The picture is complicated by
the fact that in many cases when adverbials are fronted, the subject can appear between
the fronted adverbial and the verb. It is not clear precisely what types of adverbial are
involved, but the following categories are commonly found showing subject attraction:
adverbials with naa, maa, particularly if they express causation; noo and moo in various
senses; time adverbials; reason question adverbials. These categories are illustrated
below:
(1012) Naa te makariri raaua i hoki mai ai
actgen the cold IIIdl T/A return hither part.
‘Because of the cold they came back’
(1013) Maa te huruhuru te manu ka rere
intgen the feather the bird T/A fly
‘With feathers, birds can fly’
(1014) Noo te waru karaka raatou i tae ai
actgen the eight clock IIIpl T/A arrive part.
‘At 8 o’clock they arrived’
(1015) Moo te kuia teenaa ka tika
intgen the old lady that T/A right
‘That would be fine for an old lady’
(1016) Inawhea koe i haere ai?
when(pt) IIsg T/A move part.
‘When did you go?’
(1017) Aapoopoo te pukapuka nei i runga i te teepu
tomorrow the book proxI at top at the table
‘Tomorrow, this book will be on the table’
(1018) He aha koe i hoki mai ai?
cls what IIIsg T/A return hither part.
‘Why did you come back?’
 Maori 248

(1019) He aha ngaa taangata o Hawaiki i hiahia

cls what the(pl) people gen Hawaiki T/A want
ai ki te haere mai?
part. to the move hither
‘Why did the people of Hawaiki want to come here?’ (TR2, 82)

Thus it will be seen that this is a widespread phenomenon in Maori. There is an

additional category for which the analysis seems to me very unclear, and which makes a
suitable starting place for a discussion of the problem surrounding these examples. Place
adverbials appear in initial position in examples like:
(1020) Kei mua a Tuu e haere ana
at(pres) before pers Tu T/A move T/A
‘Tu is walking in front’
(1021) I reira ia e kura ana moo te whaa
at(pt) there IIIsg T/A school T/A intgen the 4
tau
year
‘He was at school there for four years’

Superficially, a construction like this appears just like those above: there appears to be a
fronted adverbial, followed by the subject, followed by the verb. However, there are a
number of points to note. Firstly, since the T/A marker is e…ana, there is no ai to
indicate the fronted adverbial, and all the examples of this construction I have found bar
one (see (1023)) have e…ana. Thus the following was rejected:
(1022) *Inanahi nei i muri a Tuu i haere ai
yesterday proxI at behind pers Tu T/A move part.
‘Yesterday, Tu walked behind’

E…ana was required instead. However, the following textual example has no T/A
marker:
(1023) Ki konei taaua hii ai
at here Idlincl catch part.
‘We will fish here’ (KH, 2)

Secondly, in examples like (1020), kei mua is not the form the adverbial would take if it
was not fronted:
(1024) E haere ana a Tuu i/*kei mua
T/A move T/A pers Tu at(pt)/at(pres) front
‘Tu is walking in front’

Kei mua is, however, the expected form of predicative adverbial:

(1025) Kei mua a Tuu
at(pres) before pers Tu
‘Tu is in front’
 Syntax 249

This raises the possibility that such constructions are to be analysed as non-verbal
sentences with a verbal structure as some kind of adverbial or relative clause, ie. that kei
mua is predicate, a Tuu is subject, and e haere ana is an adverbial or relative clause.
Alternatively, one might suggest that kei mua is predicate, and e haere ana a Tamahae is
a subject clause, comparable to the structure proposed for negatives and by some for
actor-emphatics, so that (1020) involves subject raising. The possibility of adopting an
analysis like the first of these (which involves neither raising nor subject movement)
suggests that alternative analyses of a similar kind might be possible for (1012)– (1019)
as well. Or they too might be analysed as predicates with sentential subjects whose
subjects are then raised. An argument against this might be the presence of ai which
Chapin has argued served to mark fronted obliques in Proto-Polynesian (Chapin, 1974),
but since I have suggested elsewhere (Bauer, 1983) that this may be used as a marker of
subordination in modern Maori, no hard and fast analysis can be determined. This is not
the place to argue the pros and cons of the various analyses proposed. However, whether
there are subject movements in these sentences in Maori depends crucially on which of
these structural analyses is correct.

1.15 MINOR SENTENCE TYPES

Exclamatory sentence types have already been discussed in 1.1.1.4. Some of these are
minor sentence types. One further type must be mentioned here, and that is
nominalizations. -Canga nominalizations are very frequent as subordinate clauses, but
they can also be used as the main clause, often accompanied by a nominalized
subordinate clause, eg.
(1026) Ko te riri·nga teenei o Pou, ko toona
eq the angry·nom this gen Pou eq? sggenIIIsg
heke·nga hoki ki Maraenui
descend·nom also to Maraenui
‘Pou was very angry about this, and so he moved down the coast to Maraenui’ (P, 3)

While the first half of this is a non-verbal equative, the second consists of
ko+nominalization, which is used to assert the proposition it expresses, apparently just as
a verbal sentence would. The example that follows consists simply of two juxtaposed
nominalizations, the second introduced by ko:
(1027) Te mutu·nga o te haka a Pipito, ko te
the finish·nom gen the haka gen Pipito eq? the
puta·nga o Wairangi
appear·nom gen Wairangi
‘When Pipito’s haka had ended, then Wairangi appeared’ (W, 200)

Johansen (1948, 51ff) and Clark (1981, 73) also comment on this phenomenon, which is
shared with other Polynesian languages. Sometimes the main clause nominalizations
occur without ko (as subordinate clause nominalizations frequently do), eg.
(1028) Ka paatai ia ki ngaa tamariki raa; te kii·anga
T/A ask IIIsg to the(pl) children dist the say·nom
 Maori 250

mai ko Ruataupare raaua spec Auahi-kooata

hither eq Ruataupare IIIdl ko Auahi-koata
‘He questioned the children, and they told him it was
Ruataupare and Auahi-koata’ (TWh, 19)

There are also examples of stem nominalizations used as main clause assertions, eg.
(1029) Te puta ana mahara i te kore waka
the pass on plgenIIIsg thought at the neg canoe
‘Her thoughts turned back to her lack of canoe’ (H, 7)
(1030) …aapoopoo taaua te rongo ai i te koorero
tomorrow Idlincl the hear part. DO the talk
‘Tomorrow we shall hear the tidings’ (TWh, 20)

1.16 OPERATIONAL DEFINITIONS FOR WORD CLASSES

Maori shows a very high degree of overlap between word classes, or a very high degree
of zero conversion, ie. the same stem can normally operate as noun, verb, adjective, etc.
Because of a paucity of inflectional markers of word classes, and because of quite a high
degree of homophony amongst the functional particles, in many cases word forms cannot
be assigned unambiguously to word classes. This is true to such an extent that Biggs was
led to deny the relevance of many of the traditional parts of speech for the grammar of
Maori, and to establish his own set of categories for Maori (1969, 50ff). However, it
appears that in a number of cases, syntactic processes in Maori do refer to some of the
traditional classes which Biggs discards.

1.16.1 Noun
The majority of nouns in Maori are unchangeable in form, although a handful show
vowel reduplication in the plural, see 2.1.1.8.6.
Biggs’s class of nouns is defined in terms of cooccurrence: as forms which do not co-
occur with T/A particles, which do co-occur with definite articles (te, ngaa), which do
not occur with passive suffix, nor the personal article, and which cannot immediately
follow ki (following Clark, 1983). Biggs gives as examples ika ‘fish’, ngaru ‘wave’, and
raakau ‘tree’. If this narrow definition is maintained, then the description fails to capture
the fact that many but not all of Biggs’s class universals will behave like these nouns on
particular occasions, while some universals never do. It also fails to capture the fact that
there are meaning differences of a relevant kind associated with differences in syntactic
usage for universals.
Nouns in Maori are almost without exception accompanied by a determiner. However,
there are occasional exemptions which prevent this criterion from being used as even a
minimal criterion, eg.
(1031) …pau katoa i a ia te kai—kaakahu
exhausted all cause pers IIIsg the food clothes
atu, meremere atu; parawai atu; aha atu;
away meremere away fine mats away what away
 Syntax 251

katoa ngaa mea a te Maaori, haere katoa

all the(pl) thing gen the Maori move all
atu ki roto ki te koopuu o taua nanakia
away to inside to the stomach gen det aph monster
nei
proxI
‘He finished all the food—clothes, meremeres, fine mats, etc, all the things which a Maori
possesses, all went into the stomach of that monster’ (TP, 90)

Occurrence with a determiner is also far too inclusive a criterion for noun-hood in Maori,
since the determiner te appears in a number of constructions which are principally verbal,
eg. the kei te and i te continuous verb forms, and ki te infinitival constructions, where the
head is followed by verbal arguments rather than by typically nominal adjuncts.
However, it appears that co-occurrence with the determiners teetahi, eetahi (and probably
(t)aua and (t)eewhea) provides a rather more useful criterion for establishing nouns.
While it is true that most nouns in Maori have the potential to occur as the head of a
phrase under a preposition, this is not exclusively a feature of nouns either, as the heads
of infinitives introduced by ki te also fulfil this criterion.
One of the syntactic patterns which selects from Biggs’s class universals only those
which can also co-occur with (t)eetahi is the following construction:
(1032) Kaaore au i te moohio he aha he _______
neg Isg T/A know cls what cls
maa·ku
intgen·Isg
‘I don’t know what to _______’

Many stems are rejected in this slot, eg. horoi ‘clean’, hanga ‘build’, aawhina ‘help’,
while others are accepted, eg. waiata ‘sing’, kai ‘eat’, whakaaro ‘think’, paatai ‘ask’. All
those accepted can co-occur with (t)eetahi, while those that were rejected in (1032) were
also rejected with (t)eetahi. In other words, only those stems which are both nouns and
verbs can occur in this construction. Without these classes to refer to, it is not possible to
specify which forms are possible in this construction.
It must be noted that, under this definition of the noun, (ie. of co-occurrence with
(t)eetahi), proper names and local nouns do not belong to the class noun. They differ
from nouns in a number of ways, but whether they constitute separate word classes or
whether they are sub-classes of the class noun is not a matter easily resolved. They are
discussed in 1.16.7 and 1.16.8.

1.16.2 Pronoun
There are several types of pronoun in Maori. They can all be defined operationally by
listing. None of the groups has any clear-cut morphological defining features. Co-
occurrence restrictions and structural behaviour differ for each group.
 Maori 252

1.16.2.1 Personal Pronouns

There are eleven of these, classified into I, II, III persons; singular, dual and plural
number; and in the first person, into inclusive and exclusive. The forms are listed in
2.1.2.1.10. While there are certain tendencies towards recurrent marking: t- for inclusive;
m-for exclusive; ko- for second person, ra- for third person; -ua for dual, -tou for plural,
which indicate that these forms are morphologically complex, the singular forms are
irregular, as are certain other partial forms. Thus form cannot be used to define personal
pronouns.
In certain syntactic environments, personal pronouns take the personal article a,
namely when following a preposition ending with -i, and, in formal Maori only and even
then not for all speakers, when they function as subject Proper names take a under
basically the same conditions, although they require a in subject position regardless of
formality. Co-occurrence with the personal article therefore does not distinguish personal
pronouns from proper names, although patterns for coordination do: pronouns cannot be
coordinated using pronoun + ko, as proper names are, though pronouns coordinated with
me are accepted by native speakers.
Listing is the only practical operational definition for personal pronouns.

1.16.2.2 Possessive Pronouns

These are listed in 2.1.2.4.3, and the distinctions involved in their use have been
discussed in 1.10. Because of the special clitic forms of singular pronouns, they cannot be
defined as a group by reference to the forms of personal pronouns, although the dual and
plural forms can all be defined as having the form ‘t+aa/oo+ personal pronoun’ in the
singular, and that same form without the initial t in the plural. The singular forms share
the same clitic forms as the neutral set.
In both attributive position and independently, they are indistinguishable in use from
forms like (t)eetahi. Thus again, listing is the only possible operational definition.

1.16.2.3 Demonstratives etc

Demonstratives can function as pronouns. These take the form ‘(t)e+deictic’, where the
deictic is one of the forms nei, naa, raa, see 2.1.2.5. Other determiners can also be used
as pronouns, namely (t)eetahi and (t)eewhea. There are no morphological defining
features, since these share formal characteristics with possessive pronouns. These forms
are proforms for non-local nouns. Note that they are the forms used in the definition of
nouns. Since personal pronouns can on occasion refer to common nouns (see eg. 1.5.2.1,
(654)), these pronouns cannot be distinguished by the type of noun they can replace.
These forms can directly follow prepositions, including those ending in -i. However, this
does not distinguish them from local nouns and proforms. Listing is again the only
operational definition.
 Syntax 253

1.16.2.4 Local pronouns

Maori has an anaphoric local proform, reira, and a nonanaphoric locative set which
consist of the base ko+deictic, see 2.1.2.5. The local pronouns are pro-forms for local
nouns. In terms of their distribution, they are indistinguishable from the local nouns. In
particular, they follow prepositions directly, but take the personal article a if they
function as sentence subject Thus only listing can distinguish these from local nouns.

1.16.2.5 Interrogative pronouns

These cannot be defined operationally as a group. They lack homogeneity formally and
syntactically. Each probably belongs to the word class(es) it is a pro-form for.

1.16.3 Verb
Biggs’s classification of bases does not have a category ‘verb’, although he has a class of
statives. Those forms which are traditionally called verbs are part of his class
‘universals’. However, Hooper has argued (Hooper, 1982) that Biggs’s statives do not
constitute a homogeneous syntactic class, and it was pointed out in 1.16.1 that universals
do not.
Morphological criteria are considered first. The capacity of a base to appear with a
passive termination forms the crux of Biggs’s distinction between universals and statives
on the one hand and between universals and nouns on the other. Statives and nouns do
not take the passive termination; universals do. However, the usefulness of the passive
termination as an operational criterion is diminished by its occurrence on adverbials in
agreement with a passive verb. Thus, although ora ‘well’ is classified by Biggs as a
stative, it can appear with a passive termination, as shown by the following example:
(1033) E titiro tonu taku whatu, ka whakairi·a
T/A see still sggenIsg eye T/A hang·pass.
ora·tia
well·pass.
‘My eyes were still open and yet you suspended me alive’ (TWh, 18)

A second morphological modification possible to all verbal bases is the nominalization of

the form -Canga. This does not discriminate between neuter and other verbs, which is
perhaps why it was ignored by Biggs, although as it is a morphological feature, he might
have been expected to use it. However, the -Canga suffix, like the passive suffix, appears
as an agreement phenomenon on accompanying adverbials, which limits its usefulness as
an operational criterion, eg
(1034) E maumahara ana au ki te puta·nga
T/A remember T/A Isg to the appear·nom
ohorere·tanga mai o taua tangata i te kuuaha
sudden·nom hither gen detaph man at the door
‘I remember that man appearing suddenly in the doorway’
 Maori 254

Prefixation with whaka- must also be considered. Whaka- can be used with a variety of
non-transitive verbal bases to provide transitive verbs: from ora ‘well’ comes whakaora
‘cure’; from haere ‘move’ comes whakahaere ‘run (a function)’; from mutu ‘finished’
comes whakamutu ‘finish’. However, whaka- cannot be added to all verb bases: in
particular, it cannot be added to bases which already contain it. In addition, this prefix
(but see the discussion of prepositions in 1.16.5) can be added to local nouns, eg. from
muri ‘behind’ comes whakamuri ‘backwards’; and it can be added to phrases eg. in the
following example, whakatangata whenua has to be parsed as [whaka [tangata
whenua]]:
(1035) Ka ruuruu ana te manuhiri ki te tangata
T/A shake T/A the visitor to the man
whenua, kua whaka·tangata whenua anoo raatau
land T/A cause·man land again IIIpl
i a taatau
DO pers Iplincl
‘When the visitors shake hands with the hosts, they become hosts also’ (RR, 12)

Thus whaka- prefixation does not serve as a criterion for the class verb in Maori.
Co-occurrence with the T/A/M particles (which can be listed) can be used as an
operational criterion for verbs, although a certain degree of caution is necessary in the
application of this criterion. Notional statives do not occur with e…ana, and the ‘pseudo-
verbal’ continuous markers kei te and i te also occur with non-verbs. I is multi·functional
and can precede local nouns, and e is the form of the vocative marker, and precedes
certain numerals. Thus the most appropriate markers to use are ka and kua. Co-
occurrence with ka and kua does not discriminate between neuter and non-neuter verbs.
There do not appear to be clear-cut structural properties which could help to provide
an operational definition for the class verb, because of the functional overlap with non-
verbal predicates.
Before leaving the discussion of this class, it is perhaps necessary to add a word or two
in justification of the decision not to regard neuter verbs as a separate word class.
Mention has already been made of the fact that Hooper identifies clear syntactic
differences within Biggs’s class of statives between what she calls ‘neuter verbs’ and
other groups (Hooper, 1982). Biggs put all these in a separate group from other verbs on
the basis of one morphological criterion: failure to occur with the passive suffix. He
ignored the fact that they share with many of his universals the capacity to occur with
T/A markers, and to take the -Canga ending. Biggs’s division does not create
homogeneous groups in terms of syntactic behaviour: Hooper has shown diversity within
the stative class; Bauer (1983) can be read as demonstrating diversity within the
universals. It seems to me that the facts are more usefully accounted for by regarding
neuter verbs, experience verbs, and any other classes that might be distinguished by
future research as sub-classes of the category verb, rather than as distinct parts of speech,
since only then can the fact that they share the property identified above as criterial be
accounted for in a non-arbitrary fashion.
 Syntax 255

1.16.4 Adjective
See also the discussion of adjective phrases in 1.2.2.1. The only morphological
modification possible for adjectives (other than possible agreement with a passive or
nominalized verb) is reduplication, but that applies to other word classes too. Many co-
occurrence criteria also fail to identify any group of forms as adjectives. The ability to
follow a head noun is shared with nouns and quantifiers; the ability to occur following
T/A markers is shared with verbs; the ability to co-occur with intensifiers such as tino is
shared with nouns and possibly verbs. However, Hohepa (1969, 8–17) provides “an
almost complete listing” (1969, 8) of the potential members of the class, and an outline of
properties they share with neuter verbs and properties on which they differ. I select here
those properties which seem to me most useful. Firstly, Hohepa’s stative adjectives
undergo both partial and complete reduplication (1969, 15), whereas few other bases
undergo both, and stative adjectives show a different range of possibilities of meaning
under reduplication from other word bases. As predicates, stative adjectives do not allow
subjects with indefinite he (1969, 15). As predicates, stative adjectives take peeraa (etc)
as pro-forms, whereas verbs cannot (1969, 16). Adjectives can function as modifiers to
other adjectives and neuter verbs (1969,14).
Perhaps simpler as an operational definition is the capacity to fill the slot in a
comparative sentence like
(1036) Tino ______ ake teenei i teeraa
intens away this compar that
‘This is more ______ than that’

However, this does not define as adjectives precisely the same forms as Hohepa lists. The
majority of his stative adjectives are included, but the following from Hohepa’s list were
doubtful or rejected by my consultant: mate ‘unwell, sickly, dead’; motu ‘sever, break or
snap a cord’; rewa ‘melt’; tereii ‘float’; totohu ‘sink’; whati ‘snap, break a stick’. The
glosses Hohepa gives for these make it unclear that they are adjectives rather than
intransitive verbs. However, some (but by no means all) of Hohepa’s stative verbs
(=neuter verbs) were accepted in this frame: mahue leave behind’; u ‘fix, steady, land,
establish’. In addition forms not listed by Hohepa such as moohio ‘know, knowing,
clever’ and nanakia ‘scoundrel’, tata ‘near’ were accepted in this frame.
Thus there are several potential operational definitions for adjective in Maori,
although each seems to identify a slightly different group of forms as adjectives.
Nonetheless, there is a significant core selected by all the tests. and thus good grounds for
recognizing these as a group. However, it should be noticed that these forms also share
the criterial property for the class verb—ie. they occur with T/A markers such as ka and
kua. In 1.16.3 I argued that cooccurrence with ka and kua was sufficient to warrant
calling neuter verbs a sub-class of verbs. The same argument must therefore apply to
adjectives—ie. they are a sub-class of verbs in Maori.
 Maori 256

1.16.5 Preposition
The prepositions of Maori can be listed: they form a closed set, although the lists in
various sources do not totally agree; Williams, 1862, 16–17 contains four which are not
listed in Biggs, 1969, 54. They agree in including aa, oo, naa, noo, maa, moo, i, kei, hei,
ki, me and ko. It is perhaps surprising that Biggs omits e (the passive agent marker) as it
fits all his criteria; I see no reason to doubt that it is a preposition. The other three given
by Williams are raa ‘by way of, through’; too ‘up to’; and whaka ‘towards’. Williams’s
Dictionary gives just one citation for too as a preposition, and since it is preceded by ka
the T/A marker, it appears to be a verb and its inclusion as a preposition should probably
be taken as an error. Raa and whaka are interesting. They both occur most commonly
with local nouns, at least in more recent times, and are usually written as prefixes to these
forms. It appears highly likely that they were once prepositions which have become
prefixes. Whether this whaka is related to the transitivizing prefix is a moot point.
While this list of prepositions contains two groups with formal resemblances, namely
the possessive group constructed from aa, oo, m-and n-, and the locative group which
share -i, the remaining three me, e and ko do not obviously fit either group. Thus there
appears to be no morphological operational definition of a preposition.
They all occur, of course, in initial position in prepositional phrases. Since the phrase
can be defined in Maori in phonological terms, this definition is not as circular as it might
appear. They are all followed by NPs, which normally begin with a determiner. This can
potentially be used to distinguish verb phrases from preposition phrases, but entails the
claim that kei te… and i te… which form what Biggs calls the ‘pseudo-verbal’ continuous
are still prepositional phrases. However, this criterion also fails operationally because
local nouns follow prepositions directly. Thus the presence or absence of a determiner
following a particle cannot be used to determine whether that particle is a T/A marker or
a preposition.
Sentence position cannot be used to determine this either, since in non-verbal
sentences, prepositions can introduce the predicate, compare
(1037) I waho ia i te whare
at outside IIIsg at the house
‘She was outside the house’
(1038) I kite ia i te whare
T/A see IIIsg DO the house
‘She saw the house’

and note the dual interpretation possible for

(1039) Kei te mahi ia
at(pres) the work IIIsg
‘She is at work’
or
T/A work IIIsg
‘She is working’
 Syntax 257

Thus listing appears to be the only operational definition possible for prepositions.

1.16.6 Numeral/quantifier
Numerals and quantifiers (katoa, ‘all’ is the only uncontroversial member), both occur in
post-head position in NPs, see 1.2.5.2.6. There the resemblance ends. There is no shared
morphological material. The numerals cannot be defined morphologically either,
although patterns for numeral formation can be given, see 2.1.6 and for example, Head,
1989, 117ff. The numerals 2–9 can be prefixed by toko- for counting persons, and are
preceded by e in other contexts, but tahi ‘one’ and tekau ‘10’ fall outside this rule. Tahi is
prefixed by ko, and tekau by Ø. This is true in higher numbers as well. When the
numerals are in NPs, rather than predicative, they still take these same prefixes:
(1040) eenei waka e rua
these canoe num 2
‘these two canoes’

The numerals cannot be floated, unlike katoa, and hence never take the passive suffix in
agreement with a passive verb, although katoa does. Maha ‘many’ can also be prefixed
by toko, although it is not a numeral, nor clearly.a quantifier. The numerals can be
prefixed by tua-to form ordinals, but katoa cannot. They can also be prefixed by taki- to
indicate ‘in pairs’, ‘in threes’, etc., whereas katoa cannot.
There is thus no operational definition for this class, although the various affixation
processes for numerals probably serve as an operational definition for numerals, see
further 2.1.6.4–5.

1.16.7 Personal nouns

These can be defined as a group of bases which require the personal article as subject and
when they follow those prepositions which end in -i. The personal pronouns can be
included if the second criterion is used alone (see Biggs, 1969, 53). The interrogative wai
is also a personal noun by this definition. While Biggs classes these as a separate form
class, the question must be asked whether they are a subclass of noun, parallel to the
various verb sub-classes suggested in 1.16.3. Since they clearly have the same general
syntactic distribution as common nouns, there would seem to be a good argument for this
treatment: they can be heads of phrases functioning as subject; they can be the heads of
phrases which are complements to prepositions. They differ chiefly in the article they
select. They are probably therefore best treated as a subclass of nouns.

1.16.8 Local nouns

These can be defined as a group of bases which can follow prepositions directly (ie.
without an article). They include place names as well as the heads of complex
prepositional phrases. Apart from the place names, local nouns can be listed. Since
prepositions can be listed, these can clearly be defined operationally. The problem they
raise is simply whether they are another sub-class of noun (as their name implies) or
whether they constitute a separate part of speech. Besides following prepositions directly,
 Maori 258

the members of this class also take a, the personal article, when they function as subject
(ie. when there is no preposition). This is one of the positions in which personal nouns
take a, but the one in which personal pronouns usually occur without it. The same
arguments apply here as to personal nouns: they can occur in the same phrase-types as
canonical nouns, and therefore whatever decision is reached with respect to personal
nouns must also apply here.

1.16.9 Verbal particles

These can be listed. There is a partial overlap in the forms with other classes of particles
(i and e can both be prepositions, and ai at times appears to be anaphoric, for instance).
These particles are not totally consistent in position: while the majority precede the
lexical head, ai and ana do not. The overlap with prepositions was outlined in 1.16.5. (It
should perhaps be noted that it is probably not coincidental that there is overlap between
tense markers and temporal location markers.) However, these can probably be defined
operationally as forms which can be substituted for ka and kua, two verbal particles
which are not multi-functional.

1.16.10 Definitives
Biggs frequently talks in terms of the class ‘definitives’ (eg. 1969, 48), which can be
listed. These are forms which contain (t)e, with the exception of the irregular ngaa. The
te- forms are singular, and the e-forms are plural. Biggs thus defines the class in formal
terms. However, in functional terms it is not clear that they can be distinguished from
other articles, he ‘indefinite’, a ‘personal’ and the demonstrative ia ‘each’ (see 1.2.5.2.5).
One of these is required before non-local nouns under almost all circumstances in Maori,
but similar claims cannot be made for definitives: they are not a functional class. Thus
there is a conflict here between a class defined formally (which would exclude ngaa), and
a class defined functionally. All these forms seem to me probably best classified as one
word-class, determiners.
 2
Morphology

2.1 INFLECTION

2.1.1 Noun inflection

2.1.1.1 The general expression of the syntactic and semantic function of

NPs
In general, Maori marks noun function by accompanying particles, but it should be noted
that absence of such marking particles is often the chief indicator of function.

2.1.1.2.1 Bound affixes

Maori nouns do not appear with bound affixes to indicate function.

2.1.1.1.2 Morphophonemic alternation

Morphophonemic alternations are extremely rare in Maori, and never indicate noun
function.

2.1.1.1.3 Clitic particles

Maori nouns do not have clitic particles, although the question might at least be asked as
to whether the determiners could be analyzed in this way (see 2.1.8). However, since the
intensifier tino for instance, can be inserted between the determiner and noun, this
analysis is not readily supported.

2.1.1.1.4 Prepositions
Prepositions are the characteristic marker of noun function in Maori, although it must be
repeated that absence of a preposition is equally important. Prepositions mark almost all
non-subject functions. However, many Maori prepositions are multi-functional, so that
there is no one-to-one correspondence between preposition and syntactic/semantic
function, although Johansen (1948, 9–51) provides an interesting attempt at
rationalization.
 Maori 260

2.1.1.1.5 Word order

Word order alone does not indicate the major argument functions of nouns, although
there are characteristic, but not immutable, word orders associated with particular
functions. However, word order in Maori does mark the function ‘modifier’: modifiers
follow heads with a small number of listable exceptions.

2.1.1.1.6 Derivational processes

Derivational processes do not indicate the syntactic-semantic functions of nouns in
Maori.

2.1.1.1.7 Other
Because of the multiplicity of uses of prepositions in Maori, notably of i and ki, the actual
semantic function has sometimes to be deduced by pragmatic means—a knowledge of the
likely arguments of a verb, for example. Thus compare:
(1041) Kua horoi ia i te pahi
T/A clean IIIsg prep the bus
‘She has cleaned the bus’
(1042) Kua mahue ia i te pahi
T/A leave behind IIIsg prep the bus
‘She was left behind by the bus’
(1043) Kua hoki mai ia i te pahi
T/A return hither IIIsg prep the bus
‘She has come back from the bus’

As the translations indicate, each of the phrases i te pahi has a different role, while being
formally and positionally identical. Examples of this kind are extremely common, and
even on occasion involve ‘the same’ verb, eg. mau can be either a canonical transitive, as
in (1044) or a neuter verb, as in (1045):
(1044) Kei te mau a Pani i te ika
T/A carry pers Pani DO the fish
‘Pani is carrying the fish’
(1045) Kua mau te ika i a Pani
T/A catch the fish prep pers Pani
‘Pani has caught the fish’

In these last two examples, it is the absence of preposition with one NP versus the
presence of i with the other, together with pragmatic information, which allows the
determination of the functions involved.
 Morphology 261

2.1.1.1.8 Combinations
Word order and prepositions combine to indicate function, and pragmatic considerations
also play a part on most occasions, as illustrated above.
The function of modifier is indicated not only by word-order, but also by the absence
of determiners, which in itself does not specify a function. Consider
(1046) Kua pau te wai te inu e te kurii
T/A exhausted the water the drink by the dog
‘The dog has finished drinking the water’
(1047) Kua pau te wai inu i te kurii
T/A exhausted the water drink cause the dog
‘The dog has finished the drinking water’

The presence of te before inu in (1046) marks the start of a new phrase. Its absence in
(1047) is thus part of the marking of inu as a modifier.

2.1.1.2 The expression of specific syntactic functions

2.1.1.2.1 Subject of intransitive verb

The subject of all types of intransitive verb is a nonprepositional NP, ie. it is Ø-marked.
This is true whether the verb is an action intransitive as in (1048), an adjectival
intransitive as in (1049), a passive as in (1050), a neuter verb as in (1051), a transitive
with cognate object deletion as in (1052), or a transitive with object incorporation as in
(1053):
(1048) Kua tae mai ngaa manuhiri
T/A arrive hither the(pl) visitor
‘The visitors have arrived’
(1049) Kei te marino tonu te moana
T/A calm still the sea
‘The sea is still calm’
(1050) Kua karanga·tia ngaa manuhiri
T/A call·pass. the(pl) visitor
‘The visitors have been welcomed’
(1051) Kua mutu te hui
T/A finished the meeting
‘The meeting has finished’
(1052) Ka kai ngaa kereruu
T/A eat the(pl) pigeon
‘The pigeons are feeding’
(1053) Kei te hii ika tonu a Tane
T/A catch fish still pers Tane
‘Tane is still fishing’
 Maori 262

This is true whether or not the subject has control over the situation: compare (1050) and
(1053).

2.1.1.2.2 Subject of transitive verb

The subjects of all types of transitive verb are Ø-marked, just as those of intransitive
verbs are. Examples (1054) and (1055) illustrate canonical transitives, while (1056) is an
experience verb:
(1054) E kai ana a Mere i ngaa kooura
T/A eat T/A pers Mary DO the(pl) crayfish
‘Mary is eating the crayfish’
(1055) Ka kii mai raatou, “Aua hoki”
T/A say hither IIIpl neg also
‘They replied, “[We] don’t know’”
(1056) E piirangi ana ahau ki te motokaa nei
T/A want T/A Isg to the car proxI
‘I want this car’

It should perhaps be noted, however, that there are a number of constructions in Maori
where the agent is not coded as the subject (see 2.1.1.2.13).

2.1.1.2.3 Subject of copular construction

These, too, are Ø-marked in all types of non-verbal sentences (which are the translation
equivalent of copular constructions in other languages). A variety of types is illustrated:
(1057) Ko Kare taku ingoa
eq Kare sggenIsg name
‘My name is Kare’
(1058) He tino kino teenei pahi
cls intens bad this bus
‘This bus is shocking’
(1059) Kei too puku kee oo whakaaro
at(pres) sggenIIsg stomach contr plgenIIsg thought
‘Your thoughts are on your stomach instead’
(1060) Toko·rima oona tuaakana
pnum·5 plgenIIIsg older brothers
‘He has five older brothers’ (TR2, 188)

2.1.1.2.4 Direct object

The expression of direct object does not depend on the form of the subject, nor does it
depend on whether the subject is expressed or not
It appears that the direct object is marked sometimes by i and sometimes by ki, and
that while there are tendencies which allow for partially successful prediction of the
appropriate preposition, there are attested cases which go against these tendencies, and
 Morphology 263

some nativespeaker variation, even within an individual speaker. There is some

interesting and perceptive discussion in Johansen (1948, 28–51), and two unpublished
papers on this topic, Mark (1970) and Bauer (1983).
The most canonical DOs are marked in Maori with the preposition i:
(1061) Kei te mekemeke te kootiro i tana
T/A punch the girl DO sggenIIIsg
teina
younger sibling
‘The girl is punching her sister’
(1062) Kei te tapahi a Hata i te miiti
T/A cut pers Hata DO the meat
‘Hata is cutting the meat’

The occurrence of the preposition i, however, does not uniquely identify DOs, see
2.1.1.1.7. Quite a number of verbs show variation between i and ki. Sometimes this
appears to correlate with a distinction between a patient and a goal, as in the following
instance (from Mark, 1970, 3):
(1063) Ka kapo au i te puu
T/A snatch Isg DO the gun
‘I snatched the gun’
(1064) Ka kapo au ki te puu
T/A snatch Isg to the gun
‘I snatched at the gun’

However, in many of the parallel instances adduced by Mark, my consultants did not
agree with her distinctions. There are large numbers of examples with rongo ‘hear’, for
instance, where the distinction is unclear. Johansen (1948) discusses the Biblical uses of
rongo, and finds only partial support for a patient—goal distinction. He cites the parallel
accounts in Matthew and Luke of the young man told to give all his possessions to the
poor:
(1065) Naa, ka rongo taua taitamariki i taua
then T/A hear det aph young person prep det aph
kupu
word
‘Now, the young man heard these words’ (PT, Matiu 19, 22)

compare
(1066) Naa, ka rongo ia ki eenei mea
then T/A hear IIIsg prep these thing
‘Now, he heard these things’ (PT, Ruka 18, 23)

Similarly, both the following have been attested with tuutaki ‘meet’:
(1067) I hiahia a Hone raaua ko Mere ki te
T/A desire pers John IIIdl spec Mary to the
tuutaki i a raaua anoo
 Maori 264

meet prep pers IIIdl again

‘John and Mary wanted to meet each other’
(1068) Ka tuutaki ia ki taku tama
T/A meet IIIsg prep sggenIsg son
‘He met my son’

If there really was a patient—goal distinction to be made here, it seems to me that the
prepositions in this last pair could be expected to be reversed. One further example from
Hinemoa is worth quoting in this respect; it should be remembered that apposition (and
coordination) is effected in Maori by juxtaposition of the phrases with the preposition
repeated:
(1069) Naa ka mau ki te tahaa, i te tua·rua,
now T/A take prep the calabash prep the ord·2
ka utu i te tahaa
T/A dip prep the calabash
‘So [he] took the calabash, the second one, and dipped the calabash’ (H, 9)

Here the first phrase has ki, the parallel explanatory phrase has i. Functional
differentiation does not seem possible in the face of examples like this.
Verbs of ‘experience’ generally have their second argument marked with ki, eg.
piirangi ‘want’, moohio ‘know’, maumahara ‘remember’, but kite ‘see’ takes i (possibly
for reasons of euphony) although its syntactic behaviour is like the other experience
verbs. In addition, there are verbs like karanga ‘call’, mihi ‘greet’, aroha ‘to feel for’
which always take ki for their second arguments, eg.
(1070) Kaaore anoo ia kia moohio ki ngaa
neg again IIIsg subj know prep the(pl)
tamariki katoa
children all
‘She doesn’t yet know all the children’
(1071) Kua kite ahau i te auahi
T/A see Isg prep the smoke
‘I had seen the smoke’
(1072) E mihi ana ia ki ngaa kootiro
T/A greet T/A IIIsg prep the(pl) girl
‘She is greeting the girls’

While a goal reading seems intuitively possible for these, there are syntactic arguments
against any simple association of i with DO and ki with (oblique) goal. These arguments
are briefly treated below.
As has been illustrated, both i and ki NPs occupy the same sentence position, and NPs
with either marker can be promoted to subject via passive, eg.
(1073a) Ka pupuhi ia i te manu
T/A shoot IIIsg prep the bird
‘He shot the bird’
 Morphology 265

is more normally rendered by the passive

(1073b) Ka puuhi·a e ia te manu
T/A shoot·pass. by IIIsg the bird
‘The bird was shot by him’

Experience verbs also regularly occur in the passive, compare:

(1074) Ka piirangi ia ki ngaa mea katoa
T/A want IIIsg prep the(pl) thing all
‘He wants all the things’
(1075) Ka piirangi·tia e ia ngaa mea katoa
T/A want·pass. by IIIsg the(pl) thing all
‘All the things are wanted by him’

Thus the passive does not discriminate, but this is perhaps not surprising given its
widespread distribution in Maori.
With questioning, the strategies used are not determined by the preposition. Thus
aawhina ‘help’ takes a second argument in i, and karanga ‘call’ takes a second argument
in ki, but they use the same question strategy:
(1076) Ko wai taa Hata i aawhina ai?
eq who sggen Hata T/A help part.
‘Who did Hata help?’
(1077) Ko wai taa Hata i karanga ai?
eq who sggen Hata T/A call part.
‘Who did Hata call?’

Similarly, within the class of experience verbs, both piirangi (ki) ‘want’ and kite (i) ‘see’
behave identically with respect to questioning, although there are different preferences
from canonical transitives in some instances (see 1.1.1.2.2.1.1.2.2).
Similarly, relativization strategies treat, for instance, whaangai (i) ‘feed’ and karanga
(ki) ‘call’ identically, and moohio (ki) ‘know’ and kite (i) ‘see’ identically, although
relativization strategies differ for these two pairs (see 1.1.2.3.4 and Bauer, 1982).
Object incorporation applies irrespective of marking to canonical transitives, thus
tuhituhi ‘write’ takes i, but karanga ‘call’ takes ki:
(1078) E tuhituhi reta ana ia
T/A write letter T/A IIIsg
‘She is writing letters’
(1079) E karanga manuhiri ana ia
T/A call visitor T/A IIIsg
‘She is welcoming visitors’

This process applies much more sporadically to experience verbs, irrespective of their
marking.
With respect to reflexivization, only canonical transitives with i are normally read as
reflexive without anoo support. However, this is still not a distinction sensitive to an i/ki
 Maori 266

difference, since kite (i) ‘see’ and mihi (ki) ‘greet’ are alike in requiring anoo support for
a reflexive reading, eg. (where bracketed options belong together):
(1080) I mihi a Rewi ki a ia (anoo)
T/A greet pers Rewi prep pers IIIsg again
‘Rewi greeted him/(himself)’
(1081) I kite a Rewi i a ia (anoo)
T/A see pers Rewi prep pers IIIsg again
‘Rewi saw him/(himself)’

(See also 1.6).

The actor-emphatic also occurs with both verbs with i and verbs with ki. Thus pupuhi
(i) ‘shoot’ and karanga (ki) ‘call’ both readily occur in actor-emphatic sentences:
(1082) Naa Rewi i pupuhi te manu
actgen Rewi T/A shoot the bird
‘Rewi shot the bird’
(1083) Maa Pani e karanga ngaa manuhiri
intgen Pani T/A call the(pl) visitor
‘Pani will welcome the visitors’

Experience verbs, regardless of marking, do not appear in the actoremphatic.

Thus while canonical transitives and experience verbs frequently differ syntactically
(see Bauer, 1984), there is no such difference in syntactic behaviour between those verbs
that take ki and those verbs that take i. Thus the patient—goal distinction is not borne out
by the syntax, and both i and ki must be taken as marking the direct object in Maori if
useful generalizations are to be made about syntactic behaviour.
It must also be noted that direct and indirect speech does not take a prepositional
marker when it functions as DO. For examples, see 1.1.1.1.

2.1.1.2.5 Indirect object

Notional indirect objects in Maori can be marked with either ki or moo/maa, where the
distinction appears to be between goal and beneficiary. However, these phrases are
neither formally nor syntactically distinct from other oblique phrases marked with the
same prepositions. The unmarked sentence position is following any DO, but length or
animacy considerations often override this, eg.
(1084) Kei te paatai ia ki a Rewi, “He aha
T/A ask IIIsg prep pers Rewi cls what
teenei?”
this
‘He is asking Rewi, “What is this?”
(1085) Kei te whakamaarama ia i ngaa koorero
T/A explain IIIsg prep the(pl) talk
pakitara ki ngaa tamariki
fiction prep the(pl) children
‘She is explaining the stories to the children’
 Morphology 267

Examples with moo/maa are:

(1086) Kua mahi·a e Pani he kapu tii maa raatou
T/A make·pass. by Pani a cup tea prep IIIpl
‘Pani has made them a cup of tea’
(1087) Ka karakia ia i ngaa karakia moo
T/A incant IIIsg prep the(pl) incantation prep
ngaa taniwha moana
the(pl) taniwha sea
‘He chanted the incantations to/for the taniwhas of the sea’ (KWh, 2)

It is possible to have a ki goal and a moo/maa benefactive together, eg.

(1088) I hoatu ahau i te maaripi ki tana
T/A give Isg DO the knife prep sggenIIIsg
hoa maa Hone
friend prep John
‘I gave the knife to John’s friend for John’

The maa/moo forms imply possession, while the ki form implies location. Thus
(1089) I hoatu a Mere i te keke ki tana
T/A give pers Mary DO the cake prep sggenIIIsg
tama
son
‘Mary gave the cake to her son’

implies that she gave it to him to carry, or look after, but not necessarily to eat;
sometimes there is simply no expectation about what the recipient is to do with it
Compare this with
(1090) I hoatu a Mere i te keke maa
T/A give pers Mary DO the cake intgen
tana tama
sggenIIIsg son
‘Mary gave the cake to her son’

(1090) implies that he was given it for himself to keep, with the expectation that he will
eat it; it is his to treat as he sees fit.
Examples of the use of these prepositions in cases where they do not seem to encode
notional IOs, but which are syntactically indistinguishable from the IO instances are
readily found, eg.
(1091) Kua tae te waea ki a Hata
T/A arrive the telegram to pers Hata
‘The telegram has reached Hata’

The absence of a DO might be used as a criterion to separate IOs from such examples,
but that cannot distinguish IOs and the following:
(1092) Kei te here ia i a Paki ki a Pou
 Maori 268

T/A tie IIIsg DO pers Paki prep pers Pou

‘He is tying Paki to Pou’

Similarly,
(1093) Kei te waiata a Mere moo Marama
T/A sing pers Mary prep Marama
‘Mere is singing about Marama’

has no DO, but compare (1088) above with (1094) and recall that ki can mark DOs:
(1094) Ka paatai atu ia ki a Rehua moo
T/A ask away IIIsg prep pers Rehua prep
teetahi o aana tamariki
a(sp) gen plgenIIIsg children
‘He asked Rehua for one of his children’ (P, 2)

Thus indirect objects are not a category useful for the description of Maori.

2.1.1.2.6 Object of Comparison

Objects of comparison are marked by i, eg.
(1095) Roa ake koe i ahau
tall up IIsg compar Isg
‘You are taller than me’

See also 1.8.

2.1.1.2.7 Objects of equation

Objects of equation are marked with either i or ki depending on the equative verb, eg.
(1096) I pee·nei anoo te tiitere o te moana
T/A like-this again the pitch gen the sea
inanahi nei i/ki teenei raa
yesterday proxI eql this day
‘The sea yesterday was as rough as it is today’

(My consultant varied here on different occasions.)

(1097) I rite te tiitere o te moana inanahi ki
T/A like the pitch gen the sea yesterday eql
teenei raa
this day
‘The sea yesterday was as rough as it is today’

See 1.9 for further examples.

 Morphology 269

2.1.1.2.8 Other objects

There are no other formally distinct objects of verbs. However, the objects of experience
verbs are syntactically different from canonical DOs. The main differences are
catalogued in Bauer, 1984, and see also 1.1.1.2.2.1.1.2.2, and 1.1.2.3.4.

2.1.1.2.9 Complement of copular construction

These are all predicates in non-verbal sentences in Maori.

2.1.1.2.9.1 Defining
These all have the form he+head, eg.
(1098) He tohunga ia
cls tohunga IIIsg
‘He is a tohunga’

Not all the forms which occur as heads in this construction can be unambiguously
identified as nouns, and it is not entirely clear that this he is ‘the same’ as the indefinite
article, although it occurs in the same position in relation to the head. See also 1.2.1.1 and
Polinsky (1991).

2.1.1.2.9.2 Identity
These all have the form ko+NP. This ko is distinct from the ko which topicalizes subjects
(see 1.12), eg.
(1099) Ko Roimata ahau
eq Roimata Isg
‘I am Roimata’

Proper names are not preceded by the personal article when preceded by ko, but common
nouns require a determiner, which precludes regarding ko as a determiner (cf. the
problem with he raised in 2.1.1.2.9.1):
(1100) Ko te poo tika tonu teenei
eq the night right indeed this
‘This is certainly the right night’

See also 1.2.1.1.

2.1.1.2.9.3 Role
These have the form he+head for past or present roles, and hei for future roles, eg.
(1101) He kaiako ia
cls teacher IIIsg
 Maori 270

‘She is a teacher’
(1102) Hei kaiako ia
cls(fut) teacher IIIsg
‘She is going to be a teacher’

As for defining sentences (2.1.1.2.9.1) there is doubt as to whether he and hei are articles
here. The argument is complicated by the fact that hei (unlike he) is quite clearly a
preposition under some circumstances, and has a future locative sense in the clearest
prepositional instances. However, in those instances, the NP following it takes an article,
which is not the case in role sentences. The glosses here adopt the view that the two are
not to be equated, though the problem remains. (For some discussion, see Johansen,
1948, 15–18.) However, these NPs are non-referential.

2.1.1.2.9.4 Other copula verbs

Other types of non-verbal predicates are prepositional. They consist of a preposition+NP.
They have been illustrated in 1.2.1.1 (especially 1.2.1.1.3) and also in many other
sections.

2.1.1.2.10 Subject complement

Subject complements are rare, but take the form hei+head. They are probably to be
analyzed as embedded role sentences, eg.
(1103) I meinga a Hone hei rangatira
T/A cause·pass. pers John cls(fut) chief
‘John was made leader’

These subject complements are non-referential.

2.1.1.2.11 Object complement

Like subject complements, these are rare, and take the form hei +head. They are also
probably to be analyzed as embedded role sentences, eg.
(1104) Naa Hone a Paaora i whakatuu hei
actgen John pers Paul T/A erect cls(fut)
tauira maa·na
model intgen·IIIsg
‘John made Paul his model’

2.1.1.2.12 Objects governed by adjectives

Since adjectives function as verbs in Maori, objects of adjectives take the same form as
objects of verbs, ie. they are prepositional phrases introduced by i or ki, eg.
(1105) E kii ana te kapu i te wai
T/A full T/A the cup prep the water
 Morphology 271

‘The cup is full of water’

(1106) E nui taku whakamihi ki taku tamaiti
T/A big sggenIsg commend to sggenIsg child
‘I am proud of my child’

This is true even if the T/A marker is omitted, as is usual if tino ‘very’ precedes the
adjective, or if tonu ‘still’ follows it:
(1107) Tino pai ia ki tana whaea
very good IIIsg to sggenIIIsg mother
‘She is very good to her mother’

(See also 1.2.2.2.) However, it is not beyond question that such phrases are objects, see
2.1.1.2.13.

2.1.1.2.13 Agent in passive, neuter, stative and actor-emphatic

constructions
The agent in the passive is marked by the preposition e, eg.
(1108) Ka karakia·tia te kai e te minita
T/A incant·pass. the food by the minister
‘The food was blessed by the minister’

This is also the preposition used when the T/A marker is me, and the verb does not take a
passive ending:
(1109) Me whakakaahore e koe eenei tuu whakaaro
oblig refuse by IIsg these kind thought
‘You should resist these kinds of thoughts’ (TR2, 166)

With neuter verbs, the agent (or cause) is expressed as a preposition phrase introduced by
i, eg.
(1110) …kua mahue ia i toona iwi
T/A leave behind IIIsg cause sggenIIIsg tribe
‘He had been abandoned by his tribe’ (TR2, 121)

Agents or causes with adjectives also take the form of i+NP, eg.
(1111) Ka motu·motu noa iho ngaa hipi i
T/A cut·dup indeed down the(pl) sheep cause
a ia
pers IIIsg
‘The sheep were quite badly cut about by him’(TR2, 18)

Notice that there is therefore no necessary distinction in marking between agents and
patients with adjectival predicates, and in some cases, either reading is possible:
(1112) E mataku ana ia i te taniwha
 Maori 272

T/A frightened T/A IIIsg prep the taniwha

‘He is frightened of/by the taniwha’

Here, the phrase i te taniwha could have a causal reading, or it could be the object of
mataku. It should perhaps also be pointed out that i is the marker for sources, and there is
a fairly obvious link between source and cause. Thus (1105) could be glossed The cup is
full on account of the water’.
In actor-emphatic sentences, the agent is marked by maa/naa eg.
(1113) Naa·na aua aaporo i kohi, i kai
actgen·IIIsg det aph(pl) apple T/A gather T/A eat
‘He gathered and ate those apples’

This is not restricted to A-E sentences. Agents in non-verbal sentences can also be
marked with these prepositions, eg.
(1114) Naa raatou te waiata-aa-ringa i te marae
actgen IIIpl the song-lk-hand at the marae
‘The action-song at the marae was by them’

It is not clear, of course, that ‘agent’ is the most appropriate label for such phrases, but
the parallel with the actor-emphatic is clear enough. Naa and maa can also be used for
fronted causes and reasons in non-A-E verbal sentences, eg.
(1115) Naa te makariri raaua i hoki mai ai
actgen the cold IIIdl T/A return hither part.
‘They returned because of the cold’

This is particularly common with adjectival predicates:

(1116) Naa te kakama o Kapua i ora ai ia
actgen the alert gen Kapua T/A well part. IIIsg
‘She was saved by Kapua’s alertness’ (TR2, 121)

For a more detailed discussion of the expression of agency, see Bauer, 1981a, 152ff.

2.1.1.2.14 Topic
Topic is frequently not marked in Maori. However, one function of ko is to mark topics
under certain circumstances. For details see 1.12.

2.1.1.2.15 Emphasised element

Again, this is not associated with one particular marker, but under some circumstances,
ko marks this function, as do the maa/naa of the actor-emphatic. See further 1.11.
 Morphology 273

2.1.1.3 Marking of NP function in non-finite constructions

In general terms, NP functions with non-finite verbal constructions (but see 2.1.3.5 for
problems with the finite/non-finite distinction) do not differ from those with finite verbal
constructions. However, with nominalizations, the marking does differ.

2.1.1.3.1 ‘Absolute’ construction

There is no distinct ‘absolute’ construction in Maori.

2.1.1.3.2 Infinitive
The ki te infinitival construction can only be used when the like-subject constraint is met,
and its subject is compulsorily deleted under identity. Other verbal arguments take the
same form as for finite verbal constructions.

2.1.1.3.3 -4 Nominalizations
Canga nominalizations preserve the finite verbal marking for all but the subject, on most
occasions. (See also 1.10.6.)
The subjects of intransitive verbs are normally marked as O-class possessives, or as
neutral possessives. This is true of all types of intransitives: with a notionally agentive
subject as in (1117), with a non-agentive subject as in (1118), with a passive as in (1119),
with a neuter verb as in (1120), with an adjectival intransitive as in (1121):
(1117) Te tae·nga atu o ngaa tamariki, ka
the arrive·nom away gen the(pl) children T/A
whakatatanga ia, ka riro atu, ka noho anoo
arise IIIsg T/A take away T/A sit again
ia
IIIsg
‘When the children came, he got up at once and gave them up, and sat down again’ (TWh,
19)

(There is some doubt over the meaning of whakatatanga in this passage. It is unfamiliar
to my consultants. The translation which accompanies the text justifies ‘arise’, but there
appears to be some justification in Williams’s Dictionary for the sense ‘to near,
approach’.)
(1118) A te tae·nga mai o ngaa moni, ka
at(fut) the arrive·nom hither gen the(pl) money T/A
hoko mai ahau i te koha maa·u
buy hither Isg DO the gift intgen·IIsg
‘When the money arrives, 111 buy you a present’
(1119) E maumahara ana ahau ki too raaua aru·nga
T/A remember T/A Isg to sggenIIIdl chase·nom
e te kurii raa
 Maori 274

by the dog dist

‘I remember them being chased by that dog’
(1120) Naa too maatou mahue·tanga i te pahi
actgen sggenIplexcl leave·nom cause the bus
i hoki ai maatou ki te kaainga
T/A return part. Iplexcl to the home
‘Because we were left behind by the bus, we returned home’
(1121) I te pakaru·tanga o te waka nei, ka tere
at the broken·nom gen the canoe proxI T/A fast
haere ngaa kura nei
move the(pl) feather proxI
‘When the canoe broke up, these feathers drifted quickly away’ (P, 2)

As will be seen from the examples, pronominal subjects are preposed, and non-
pronominals postposed (see 1.10). Singular pronominal subjects often take the neutral
possessive form, rather than the O form, eg.
(1122) I tana puta·nga ake ki runga, hore kau
at sggenIIIsg appear·nom up to top neg intens
he waka
cls canoe
‘When he rose to the surface, there was no canoe’ (KH, 3)

After hei, agents with intransitive verbs are marked with moo, eg.
(1123) Kua aata whakarite·a hoki e raaua i raurangi
T/A clear arrange·pass. indeed by IIIdl at day
raa te waa hei haere·nga mai moo Hinemoa
dist the time prep move·nom hither intgen Hinemoa
ki a ia
to pers IIIsg
‘It had in fact clearly been arranged by them that day the time for Hinemoa to come to him’
(H, 7)

However, this may not be a subject NP, as this might be related to the actor-emphatic in
some way (compare transitive subjects below).
There are also occasional examples where agentive subjects of intransitives appear
with a when the subject’s control of the situation is emphasised. See the discussion of
(870) at the end of 1.10.6. This highlights the fact that intransitive subjects in Maori are
normally taken as non-agentive.
The subjects of canonical transitives usually take A, eg.
(1124) Kua koorero·tia mai e koe ngaa koorero moo
T/A speak·pass. hither by IIsg the(pl) talk intgen
eetahi o ngaa ika nunui—te hii·anga a
some(pl) gen the(pl) fish big the fish·nom Agen
Maaui i taana ika…; te patu·nga a Kae i
Maui DO sggenIIIsg fish the kill·nom Agen Kae DO
te ika mookai a Tinirau…
 Morphology 275

the fish pet gen Tinirau

‘You have told us about some of the big fish—Maui’s
catching of his fish; the killing by Kae of Tinirau’s pet fish…’ (TR2, 131)

However, the subjects of experience verbs take O, again underlining the fact that control,
not transitivity is the determining factor (see further 1.10.6):
(1125) Te rongo·nga atu o te tangata ki te
the hear·nom away gen the man to the
puutoorino, ka tuu
flute T/A stand
‘When the man heard the flute, he stood still’

There is no lack of examples where O is used with canonical transitives with non-passive
case-marking, eg. (869) and :
(1126) I mea mai a Tuu ki a maatou i
T/A say hither pers Tu to pers Iplexcl DO
te koohuru·tanga o Pou i a ia anoo
the kill·nom gen Pou DO pers IIIsg again
‘Tu told us about Pou’s killing of himself’

It thus appears possible that O is the unmarked form of possessive for transitive subjects,
too, and that A is used when control is stressed, which happens frequently enough for the
marked pattern to be commoner than the unmarked. As with intransitives, singular
pronominal subjects frequently use the neutral form, eg.
(1127) Naa taku horoi·nga i ngaa wini, i
actgen sggenIsg clean·nom DO the(pl) window T/A
mamae ai taku tuaraa
hurt part. sggenIsg back
‘Because I cleaned the windows, I got a sore back’

After hei, agents in transitive nominalizations are marked with maa or moo, eg.
(1128) Ka waiho tonu taatou hei tinihanga·tanga maa
T/A leave indeed Iplincl for cheat·nom intgen
too taatou whaea
sggenIplincl mother
‘We are always being left cheated by our mother’ (KM, 3)
(1129) Kaaore au i te moohio he aha hei kai·nga
neg Isg T/A know cls what for eat·nom
maa·ku
intgen·Isg
‘I don’t know what to eat’

Subjects are also frequently omitted from nominalizations:

(1130) E maumahara ana au ki te whakawhiu·tanga
T/A remember T/A Isg to the punish·nom
 Maori 276

i a koe
DO pers IIsg
‘I remember you being punished’

(Notice, however, that the Maori does not have passive morphology.) Another example:
(1131) He wiri hoki noo·na i te maaeke, i
a shake also actgen·IIIsg from the cold from
te kau·anga mai i te poo i te moana o
the swim·nom hither at the night at the sea gen
Rotorua
Rotorua
‘She was also trembling from the cold, from the swim at night in L. Rotorua’ (H, 8)

Occasionally in informal contexts, no preposition is used, eg.

(1132) I whakatuu ia i te patu·nga e Hone te
T/A boast IIIsg DO the kill·nom by John the
mango
shark
‘He boasted about John’s killing of the shark’

Other verbal arguments take the same form as in non-nominalized sentences. A few
examples are given below, and many examples already cited in this section contain
relevant phrases:
(1133) Kaaore au i ui atu, i taku
neg Isg T/A ask away from sggenIsg
maumahara·tanga ki toona ingoa
remember·nom to sggenIIIsg name
‘I didn’t ask, because I remembered her name’
(1134) I toomuri maatou i te puni·nga ai o
T/A late Iplexcl cause the block·nom part. gen
te huarahi i ngaa horo·anga
the road cause the(pl) slip·nom
‘We were late because the road was blocked by slips’

There are also stem nominalizations to consider, although these are less common. There
is variation between A and O for subject marking, but A seems commoner than O for
both transitives and intransitives (see also 1.10.6), eg.
(1135) I mea mai a Tuu ki a maatou i
T/A say hither pers Tu to pers Iplexcl DO
te koohuru a Pou i tana tuakana
the kill pers Pou DO sggenIIIsg brother
‘Tu told us about the killing of his (Tu’s) brother’

(cf. (1126).)
(1136) I te tere o te oma a Pou, ka mapu
cause the fast gen the run gen Pou T/A pant
 Morphology 277

ia
IIIsg
‘Because Pou ran so fast, he was panting’
(1137) Tino nui te aroha o te kootiro ki tana
very big the sympathy gen the girl to sggenIIIsg
tungaane
older brother
‘The girl’s love for her brother is very great’

Experience verbs take O, see (873).

As with -Canga nominalizations, singular pronominal subjects regularly use the
neutral possessive form:
(1138) I te mutu·nga o tana whakapai i te
at the finish·nom gen sggenIIIsg set in order DO the
whare, ka haere ia
house T/A move IIIsg
‘When she had cleaned and tidied the house, she went out’

After hei, m- forms are also used for agents, eg.

(1139) Ka tango ia i ngaa tahaa e ono hei
T/A take IIIsg DO the(pl) calabash num 6 for
whakatere moo·na
float intgen·IIIsg
‘She took 6 calabashes to keep herself afloat’ (H, 7)

Other case-marking in stem nominalizations follows the non-nominalized patterns.

2.1.1.4 The expression of non-local semantic functions

2.1.1.4.1 Benefactive
The overlap between benefactives and IOs has already been discussed (2.1.1.2.5).
Benefactives are usually expressed in all syntactic environments with maa or moo, eg.:
(1140) Ka haere a Rona ki te tiki wai moo
T/A move pers Rona to the fetch water intgen
ana tamariki
plgenIIIsg children
‘Rona went to fetch water for her children’ (R, 19)
(1141) Moo·ku teenei tuuru
intgen·Isg this chair
‘This chair is for me’
(1142) Kei roto i te peeke he kaakahu hou
at(pres) inside at the bag a dress new
moo tana wahine
intgen sggenIIIsg woman
 Maori 278

‘In the bag there is a new dress for his wife’

The distinction between maa and moo is the same here as elsewhere (see 1.10.6).
However, given the lack of distinction between IOs and benefactives, examples like
the following with ki should be borne in mind as possibly constituting examples of
benefactives :
(1143) Ko te koha teenei a Wairangi ki tana
top. the gift this pers Wairangi to sggenIIIsg
wahine
woman
‘This was Wairangi’s token of regard for his wife’ (W, 199)

2.1.1.4.2 Source
Adverbial sources are marked with i (unless causation is involved as well, see 2.1.1.4.12),
eg.:
(1144) Kaatahi ka tango·hia ake e Rua-kapanga he
then T/A take·pass. up by Rua-kapanga a
kura i toona keekee matau
feather from sggenIIIsg armpit right
‘Then Rua-kapanga took a feather from his right armpit’ (P, 2)

However, in predicative or attributive contexts, possessives are used. Predicatively, n-

possessives are required, eg.
(1145) Naa Kare teenei reta
actgen Kare this letter
‘This letter is from Kare’

Attributively, A or O possessives are found:

(1146) Tino roa te reta a Pani
very long the letter gen Pani
‘The letter from Pani was very long’

Coming “from” a place in the sense of identifying one’s home or birthplace is treated in
Maori as possession, and uses noo (see (849d)).

2.1.1.4.3 Instrumental
Most instrumentals are marked with ki, eg.
(1147) Kaatahi ka haere atu a Raapata ki te ihu
then T/A move away pers Rapata to the stern
ki te werowero i te ika raa ki teetahi
to the stab DO the fish dist instr a(sp)
atu raati
away harpoon
 Morphology 279

‘Then Rapata went to the stern to stab that fish with another harpoon’ (TR2, 133)
(1148) Ka haehae i ngaa ringa ki te mataa
T/A cut DO the(pl) arm instr the obsidian flake
‘[She] cut her arms with obsidian flakes’ (W, 198)

There are also a number of dubious instrumentals which take ki for their marker. A few
examples of such instances are given here to indicate the sorts of extensions which are
found:
(1149) Ka haere a Tuu ki te whakakii i te
T/A move pers Tu to the fill DO the
taraka ki te hinu
truck instr? the petrol
‘Tu went to fill the truck with petrol’
(1150) Ka mau ia ki te ringa
T/A take IIIsg instr? the hand
‘He took her by the hand’
(1151) Ko too wahine kua hara ki teetahi
top. sggenIIsg woman T/A sin instr? a(sp)
tangata ko Tupeteka te ingoa
man spec Tupeteka the name
‘Your wife has sinned with a man called Tupeteka’ (W, 198)
(1152) Me hopu ki too ringa
oblig catch instr sggenIIsg hand
‘You should catch [it] with your hand’ (TR2, 122)

However, maa is sometimes used for instrumentals, eg.

(1153) He tino tohunga teenei ki te patu i te
cls very expert this to the kill DO the
tangata maa te maakutu
people intgen the black magic
‘He was very expert at killing people with black magic’ (KWh, 1)
(1154) Ka whakaaro·tia, kia kaua e whakamate·a
T/A decide·pass. subj neg T/A kill·pass.
maa te patu
intgen the club
‘It was decided not to kill him with a club’ (KWh, 1)

These examples show that maa can be used with both concrete and non-concrete
instruments. Instruments also sometimes appear in the actor slot in the actor-emphatic,
eg.
(1155) Maa te miihini e mahi ngaa mea katoa
intgen the machine T/A work the(pl) thing all
‘The machine will do everything’
(1156) …kia riro maa te mokemoke e patu
subj taken intgen the loneliness T/A kill
 Maori 280

‘[he] would be taken to die of loneliness’ (KWh, 1)

It would appear that such uses are on the increase, probably due to English influence.
However, there seems little doubt that this is not a new feature of Maori, even if the
increased frequency is new.
Clear ki instruments do not occur as the predicates of non-verbal sentences, but the
actor-emphatic examples above are possibly to be analysed thus. They can only be
attributive in noun phrases with nominalizations, eg.
(1157) Me whakamutu te tapahi i ngaa harakeke
oblig finish the cut DO the(pl) flax
ki te kani
instr the saw
‘The cutting of flax with saws should stop’

While it is not quite clear that they are instruments, it should probably be mentioned that
means of transport use maa, eg.
(1158) Kei te tae mai ia maa runga i te pahi
T/A arrive hither IIIsg intgen top at the bus
‘She is arriving by bus’

This includes maa raro ‘on foot’ (more lit. ‘by below’).
Negative instruments do not have a simple prepositional means of expression.
Compounds with kore have already been discussed (see 1.4.2): they are joke formations
for the most part, and cannot be substituted for instrumental phrases:
(1159a) *I patu·a e ia kore·maaripi
T/A kill·pass. by IIIsg neg-knife
‘He killed it without a knife’
(1159b) *I patu·a e ia ki te kore·maaripi
T/A kill·pass. by IIIsg instr the neg-knife
‘He killed it without a knife’

These are normally rendered with negation of the instrument, eg.

(1159c) I patu·a e ia kaahore he maaripi
T/A kill·pass. by IIIsg neg a knife
‘He killed it without a knife’ (more lit. ‘He killed it and there wasn’t a knife’)

2.1.1.4.4 Comitative
The comitative preposition is me (but see the discussion of coordination, 1.3), eg.
(1160) Ka tuu ngaa tamaahine i te mataihi
T/A stand the(pl) daughter at the house front
katau o te marae me too raaua kookaa, me
right gen the marae with sggenIIIdl mother with
Hinemaurea
Hinemaurea
 Morphology 281

‘His daughters stood at the right of the front of the house, in the court, with their mother,
Hinemaurea’ (TWh, 19)
(1161) Ko Pani raaua ko Mere me ia
top. Pani IIIdl spec Mary with IIIsg
‘Pani and Mary are with her’

This can also be used attributively:

(1162) He roa rawa atu te tangata raa me Hata
cls tall intens away the man dist with Hata
‘The man with Hata is very tall’

Negative comitatives do not have a simple form. A sentence like I went without Bill could
be rendered in Maori by any of the following:
(1163a) Kaaore ahau i haere me Piri
neg Isg T/A move with Bill
‘I didn’t go with Bill’
(1163b) Kua mahue a Piri i ahau
T/A leave behind pers Bill cause Isg
‘I left Bill behind’
(1163c) Ka haere ahau, engari kaahore a Piri i
T/A move Isg but neg pers Bill T/A
haere mai
move hither
‘I went, but Bill didn’t’
(1163d) Ka haere ahau kaahore a Piri
T/A move Isg neg pers Bill
‘I went, not Bill’
(1163e) Naa·ku a Piri i whakarere iho ake
actgen·Isg pers Bill T/A abandon down away
‘I abandoned Bill’

None of these constitutes a negative comitative.

2.1.1.4.5 Circumstance
Circumstances are expressed in different ways in different syntactic environments.
Adverbially, me, the comitative preposition is used, eg.
(1164) I mahi·a e ia te wini me
T/A make·pass. by IIIsg the window with
ana ringa paru·paru
plgenIIIsg hand dirt·dup
‘He mended the window with dirty hands’

However, such me phrases cannot be used predicatively or attributively. The adjective is

used predicatively, eg.
 Maori 282

(1165) E paru ana ana ringa

T/A dirty T/A plgenIIIsg hand
‘He has dirty hands’

Attributively, a relative structure has to be used:

(1166) I kite ahau i te tangata paru ngaa ringa
T/A see Isg DO the man dirty the(pl) hand
‘I saw a man with dirty hands’

Negative circumstances, like negative comitatives, use sentence negation, eg.

(1167) He mea tino tauhou ki te kite i te
cls thing very strange to the see DO the
tamariki kaahore he ringa paru
children neg cls hand dirty
‘It is very strange to see a child without dirty hands’

2.1.1.4.6 Possessive
The possessives have already been described in 1.10. Only a brief recapitulation is given
here. Tense distinctions are marked through prepositions in non-verbal possessive
sentences. Temporary possession is expressed as location, with the locative prepositions
(see also 2.1.1.5). Permanent possession is expressed with maa, naa, moo and noo. The A
vs. O distinction which is marked both predicatively and adnominally is not clearly of the
alienable/inalienable type. Benefactives should probably be regarded as a type of
possessive, rather than as a separate function in Maori. With adnominal possessives, there
is a choice for short NPs between a pre-head and a post-head construction.

2.1.1.4.7 Possessed
Possessed items have no special marking in Maori. For a discussion of whether the A/O
distinction is alienable v. inalienable, see 1.10.2 and 1.10.6. Whatever the best
description of this contrast is, it is dear that it is relational in nature, ie. it is not a property
inherent in either possessor or possessed, but a relation between possessor and possessed.

2.1.1.4.8 Quality
Qualities are normally expressed either predicatively through classifying sentences or as
attributive adjectives, eg.
(1168) He tino whakatoi te tamaiti raa
cls very cheeky the child dist
‘That child is very cheeky’
(1169) He tamaiti tino whakatoi ia
cls child very cheeky IIIsg
‘He is a very cheeky child’
 Morphology 283

Negative qualities can only be expressed predicatively, through sentence negation:

(1170a) Eehara ia i te whakatoi
neg IIIsg at the cheeky
‘He is not cheeky’
(1170b) Eehara ia i te tamaiti whakatoi
neg IIIsg at the child cheeky
‘He isn’t a cheeky child’
(1171) Kaahore he maatauranga o teenei tangata
neg cls knowledge gen this man
‘This man has no commonsense’

However, there are occasional formations with kore, eg.

(1172) He tamaiti roro-kore koe
cls child brain-neg IIsg
‘You’re a brainless child’ (TR2, 71)

(cf Instrumentals, 2.1.1.4.3). This does not appear to be particularly productive.

Negative qualities can be made attributive through relativization:
(1173) Ko Tuu he tamaiti eehara i te whakatoi
top. Tu a child neg at the cheeky
‘Tu is a child who isn’t cheeky’

Reference qualities can be expressed either through adnominal possessives or as

attributive adjectives, eg.
(1174) Kei te titiro whakatau mai ki a ia, ki ngaa
T/A look intently hither to pers IIIsg to the(pl)
tohu o te rangatira, o te toa, e mau
sign gen the chief gen the champion T/A carry
atu ana i runga i a ia
away T/A at top at pers IIIsg
‘They gazed intently at him and the emblems of high birth and bravery which he bore with
him’ (TWh, 19)
(1175) Kei a ia ngaa tohu katoa moo te
at(pres) pers IIIsg the(pl) sign all intgen the
rangatira·tanga
chief·nom
‘He has all the attributes of the chief’
(1176) Anaa oti ko ngaa rongo toa kia nui,
Here then top. the(pl) fame champion subj big
aa ko ngaa tohu o te toa kia iti?
and top. the(pl) sign gen the champion subj small
‘How could the fame of his bravery be great and signs of that bravery be small?’ (TWh,
22)
 Maori 284

2.1.1.4.9 Quantity
Quantities are expressed attributively as modifiers:
(1177) he kapu tii
a cup tea
‘a cup of tea’
(1178) teetahi pounamu haami
a(sp) bottle jam
‘a bottle of jam’

Note that “type” or “species” expressions follow the same pattern:

(1179) he momo ika
a species fish
‘a breed of fish’
(1180) he aahua rimu
a kind rimu
‘a kind of rimu’

This means that phrases like

(1181) ngaa keene kiriimi
the(pl) can cream
‘the cream cans/ the cans of cream’

are ambiguous between a ‘type’ reading and a ‘measure’ reading.

Perhaps somewhat surprisingly, expressions like ‘ten ton truck’ are rendered by pre-
head attributives, eg.
(1182) Ko teenei he tekau tana taraka
top. this cls 10 ton truck
‘This is a ten-ton truck’

This pattern is quite probably borrowed from English. The following sentence was
proffered after rejection of he tekau taangata waka ‘a 10-man canoe’, which probably
indicates the lack of nativeness of the pattern:
(1183) Ka mahi·a he waka moo tekau taangata
T/A make·pass. cls canoe intgen 10 men
‘A ten-man canoe was made’

Reference quantities are also attributive:

(1184) I hoko mai au kotahi rita hinu
T/A buy hither Isg one litre oil
‘I bought a litre of oil’
 Morphology 285

2.1.1.4.10 Material
Materials have different means of expression depending on sentence function.
Attributively, they are expressed as adjectival modifiers, eg.
(1185) he whare raupoo
a house bulrush
‘a raupo house’
(1186) Ko teenei whare koorari tino kino te koopeke
top. this house flax very bad the cold
‘This flax house is very cold’
(1187) ngaa puna koropupuu
the(pl) spring boiling mud
‘the boiling mud pools’

They can also be expressed as non-verbal he predicates, eg.

(1188) Ka titiro a Wairangi, ko ngaa poupou o
T/A see pers Wairangi top. the(pl) post gen
te whare he koohurihuri kahikatea
the house cls sapling white pine
‘Wairangi looked and noticed that the side posts supporting the rafters were composed of
the solid trunks of white pine’ (W, 198)

Adverbially, they are marked with ki, like instruments:

(1189) Ko teenei whare mea hanga ki te raakau
top. this house thing build with the wood
‘This house was built of wood’
(1190) Mea mahi ngaa poi ki te koorari
thing make the(pl) poi with the flax
‘Pois are made of flax’

Negative materials can only be expressed with sentence negation, eg.

(1191) Ko teenei whare kaahore i mahi·a ki te
top. this house neg T/A make·pass. with the
koorari
flax
‘This house was not made of flax’

2.1.1.4.11 Manner
Manner adverbials are frequently single word verbal modifiers. Otherwise, they are
expressed predicatively, eg.
(1192) I harihari ia i too raaua tuutataki·tanga
T/A happy IIIsg at sggenIIIdl meet·nom
‘He greeted her with joy’ (lit. ‘He was glad at their meeting’)
 Maori 286

Negative manner adverbials can only be expressed through negation of predicative

adjectives, eg.
(1193) Kaahore he rangimarie i too raaua tuutataki·tanga
neg cls peaceful at sggenIIIdl meet·nom
‘He greeted her without warmth’

Adverbial modifiers expressing manner are exemplified by

(1194) E takoto whakamaa ake ana i raro i ngaa
T/A lie ashamed up T/A at below at the(pl)
tauwharenga koowhatu
overhang rock
‘[She] was lying embarrassed under the overhanging rocks’(H, 10)
(1195) I haere manuhiri mai a Tuwharetoa
T/A move visitor hither pers Tuwharetoa
‘Tuwharetoa came as a visitor’ (H, 5)

2.1.1.4.12 Cause
The expression of cause phrases is complex in Maori. There is a considerable variety of
forms of expression, some of which are dependent on verb type.
With neuter verbs, cause phrases are introduced by i:
(1196) Kei mahue taaua i te tereina
mon leave behind Idlincl cause the train
‘We might be left behind by the train’

However, it should be noted that causation by humans (which we might otherwise

interpret as agency) is also expressed with i, eg.
(1197) Kua pau i a ia ngaa kai
T/A exhausted cause pers IIIsg the(pl) food
‘He has eaten up the food’

Thus no distinction is made with neuter verbs between causation and agency. Adjectival
predicates also have i for cause phrases:
(1198) Kei kaapoo ngaa taangata i too neketai
lest blind the(pl) people cause sggenIIsg tie
‘People might be blinded by your tie’
(1199) Kua hoohaa au i te pikiniki
T/A bored Isg cause the picnic
‘I’ve got fed up with the picnic’

However, it should be noted that phrases with identical marking shade away from
causation to source (which also has i for its basic marker) and materials, compare
(1200) Maa tonu te whenua i te hukupapa
white still the ground cause the frost
‘The ground is still white with frost’
 Morphology 287

(1201) Kii tonu te ruuma i te tamariki

full indeed the room cause the children
‘The room is full of children’
(1202) Mangu·mangu katoa ana kanohi i te paru
black dup all plgenIIIsg face cause the dirt
‘His face is all black with dirt’

Such data can be interpreted as providing support for localist notions which take ‘source’
as the basic underlying relation from which causation and agency are derived. There is
certainly no distinction in the formal marking of these with adjectival predicates in
Maori.
Causation is also expressed by maa and naa+NP:
(1203) Maa te waimarie raatou e whiwhi karahipi
intgen the luck IIIpl T/A win scholarship
ai
part.
‘With luck, they will win a scholarship’
(1204) Naa te maakuu raaua i hoki mai ai
actgen the wet IIIdl T/A return hither part.
‘They returned on account of the damp’

This construction bears strong similarities to the actor-emphatic construction (see

1.11.2.1.7), where the maa/naa NPs mark the agent. However, there are certain crucial
differences. The verbs in these causation sentences are intransitive, so that the unusual
case-marking of the actor-emphatic is not found here. Also, if the T/A marker here is one
of the usual (i or e), the particle ai is required post-verbally. Following Chapin (1974), it
is likely that it marks the fronting of the adverbial. However, ai is not used by all
speakers if the T/A marker is ka, eg.
(1205) Maa te huruhuru te manu ka rere (ai)
intgen the feather the bird T/A fly part.
‘With feathers, birds can fly’

Versions without fronting are not usual, but native speakers accept them without
question:
(1206) I hoki mai raaua naa te maakuu
T/A return hither IIIdl. actgen the damp
‘They returned on account of the damp’

While these sentences are verbal, there are non-verbal sentences with maa/naa as well in
which the idea of causation appears to be present:
(1207) Naa te puru teeraa mahi
actgen the bull that work
‘That work is the bull’s; That’s the bull’s doing’
(1208) Maa ngaa kaikoorero o te marae ngaa mihi
 Maori 288

intgen the(pl) orator gen the marae the(pl) greet

‘The welcome speeches will be made by the orators of the marae’

It will thus be seen that the formal markers of Maori suggest a chain: possession—
source—causation—agency, with no clear cut-off points along the line.
In the negations of such sentences, however, the naa NPs can be replaced by i:
(1209) Eehara i te aituaa maatou i tae
neg cause the accident Iplexcl T/A arrive
toomuri ai
late part.
‘It was not because of the accident that we arrived late’

In transitive and intransitive sentences, i is also used as an alternative to maa/naa to mark

indirect causation, eg.
(1210) I te whakatakariri o Tuu, ka tangi ia
cause the vexation gen Tu T/A cry IIIsg
‘Tu cried because he was so angry’; ‘On account of Tu’s vexation, he cried’
(1211) I te koa o Aatareta, ka waiata ia ki
cause the joy gen Atareta T/A sing IIIsg to
a Hukarere
pers Hukarere
‘Because Atareta was happy, she sang to Hukarere’ (TR2, 80)
(1212) I te aataahua o te rangi ka aata
cause the beauty gen the day T/A clearly
kite·a atu a Oopootiki
see·pass. away pers Opotiki
‘Because it was a lovely day, Opotiki could be seen clearly’

However, causation in passive sentences is usually expressed as an e-NP, usually

associated with agents. The following two sentences, the first adjectival and the second
passive, occur just a few lines apart:
(1213) Ka mate i te whakamaa
T/A suffer cause the shame
‘[She] suffered from shame’ (TA, 6)
(1214) Ka tini·a ia e te whakamaa
T/A overcome·pass. IIIsg by the shame
‘She was overcome by shame’ (TA, 6)

This suggests that the marking is determined by the verb, rather than by any difference in
the semantic value of the NP.
Examples have already been given of maa/naa causatives in non-verbal sentences.
Examples with i are somewhat marginal:
(1215) ?I runga i tana aituaa·tanga ka wareware
cause top at sggenIIIsg accident·nom T/A forget
ia
 Morphology 289

IIIsg
‘His forgetfulness is because of his accident’

Naa was strongly preferred here.

In attributive contexts, cause NPs occur only in nominalizations, where i is apparently
the only marking possible: compare (1216) with (1206) above:
(1216) Ohorere taatou i too raatou hoki·nga mai
surprised Iplincl DO sggenIIIpl return·nom hither
i te kaha o te koopeke
cause the strong gen the cold
‘We were surprised at their return on account of the cold’

(Noo does not normally appear with non-verbal causatives, but the following example is
interesting:
(1217) Ko·ia i hua·ina ai te ingoa o taua
top.·that T/A name·pass. part. the name gen det aph
puke ko Puketii, noo ngaa raakau i
hill spec Puketi actgen the(pl) tree T/A
whakanoho·ia…
place·pass.
‘That is the reason why that hill was called Puketi, because of the trees placed…’(KW, 3))

Note also the following example:

(1218) He kuuware kee noo te tangata raa i
cls ignorance contr actgen the man dist T/A
wini ai koe
win part. IIsg
‘It was because of that man’s ignorance that you won’

He cannot be preceded by any preposition, so the causation is marked by the ai in the

main clause, which marks that the fronted material has an adverbial relationship to the
main clause.

2.1.1.4.13 Purpose
The commonest purpose phrases are introduced by ki:
(1219) …kia tino maarama ai toona hinengaro ki
subj very clear part. sggenIIIsg mind to
ngaa aahuatanga o te whare waananga
the(pl) activities gen the house learning
‘…so that his mind would be very clear for the activities of the University’ (TR2, 169)
(1220) Ka noho ia ki roto ki te whakamahana i
T/A sit IIIsg to inside to the warm DO
a ia
pers IIIsg
 Maori 290

‘She sat in it to warm herself (H, 8)

It will thus be seen that there is no significant difference between infinitival expressions
of purpose and nominalizations.
Hei nominalizations can also express purpose, as in the following:
(1221) Kua moohio tonu hoki a Ao-kehu i haere
T/A know indeed still pers Ao-kehu T/A move
atu a Tamaahua-rererangi ki te tiki atu
away pers Tamahua-rererangi to the fetch away
i a ia hei patu i taua taniwha
DO pers IIIsg for kill DO det aph taniwha
‘Indeed Ao-kehu had easily divined the object of Tamahuarererangi’s visit—that he had
come to fetch him to slay the taniwha’ (TP, 91)
(1222) …houhou rawa i ngaa koohao hei herehere·nga
bore intens DO the(pl) hole for tie·nom
‘…holes were also bored to tie it on’ (TP, 91)
(1223) Ka whakarite·a he tohunga hei whaangai moo
T/A appoint·pass. a tohunga for feed intgen
Tuu-whakairi-ora
Tu-whakairi-ora
‘A tohunga was appointed to feed Tu-whakairi-ora’ (TWh, 20)

(Notice that this could also be interpreted as a role.)

There is an increasing tendency (strongly disliked by some older speakers, but not all)
to use moo for this in modern Maori, probably under the influence of English:
(1224) He poo tino pai moo te hopu tuna
cls night very good intgen the catch eel
‘[It was] a very good night for catching eels’
(1225) Homai he maati hei tahu i te ahi moo te
give cls match for light DO the fire intgen the
haangii
hangi
‘Give me some matches to light the fire for the hangi’
(1226) He waahi anoo moo te taakoro, he waahi anoo
cls time again intgen the play cls time again
moo te mahi
intgen the work
‘A time for work, a time for play’

However, there are examples which fall on the purpose/benefactive borderline in older
texts, which indicate that this is not purely a modern innovation, eg.
(1227) Ka tuku·na taa raatou taurekareka kia haere
T/A order·pass. sggenIIIpl servant subj move
i roto i te iwi o Tupeteka e whawhati
at inside at the tribe gen Tupeteka T/A collect
rautao ana, e kohi koowhatu ana moo te
 Morphology 291

fern T/A T/A gather stone T/A intgen the

haakari
feast
‘Then they sent their slave to go amongst the people of Tupeteka who were gathering ferns
and stones for the feast’ (W, 198)

2.1.1.4.14 Function
The preposition used for expressing function is hei:
(1228) Ka pai te puuhaa hei kiinaki
T/A good the puha as complementary food
‘Puha is good as a complement’
(1229) Me haere koe ki te kari noke hei moounu
oblig move IIsg to the dig worm as bait
‘You’d better go and dig worms for bait’ (TR2, 13)
(1230) Kua whakairo·tia nei e Taikehu hei patu
T/A carve·pass. proxI by Taikehu as club
maa·na
intgen·IIIsg
‘It had been shaped by Taikehu as a club for himself (P, 4)

Note that hei is used for role and purpose as well in Maori, and there are examples where
it would be difficult to decide between these, eg.
(1231) Ka tango ia i ngaa tahaa e ono hei
T/A take IIIsg DO the(pl) calabash num 6 as
whakatere moo·na
float intgen·IIIsg
‘She took 6 calabashes as floats for herself/to keep herself afloat’ (H, 8)
(1232) Ka mahara kia waiho hei mokamokai maa·na
T/A think subj leave as pet intgen·IIIsg
‘[He] decided it should be left as a pet for him’ (TP, 89)

2.1.1.4.15 Reference
Reference NPs are nowadays characteristically marked by moo:
(1233) Kei te koorero a Tamahae moo tana ika
T/A talk pers Tamahae intgen sggenIIIsg fish
‘Tamahae is talking about his fish’

However, ki is also found:

(1234) Ka amuamu maatou katoa ki te hoohaa o
T/A complain Iplexcl all to the boring gen
ngaa koorero a taua tangata
the(pl) talk gen det aph man
‘We all complained about that man’s boring speech’
 Maori 292

In addition, at least with nominalizations, either -Canga or stem nominalizations, i is also

in variation with moo. Thus my consultant produced
(1235) I mea mai a Tuu ki a maatou i
T/A say hither pers Tu to pers Iplexcl DO
te koohuru·tanga o Pou i a ia anoo
the murder·nom gen Pou DO pers IIIsg again
‘Tu told us about Pou’s killing of himself’
(1236) I mea mai a Tuu ki a maatou moo
T/A say hither pers Tu to pers Iplexcl intgen
te hopu·tanga o Pou i te hooiho moona
the catch·nom gen Pou DO the horse intgenIIIsg
anoo
again
‘Tu told us about Pou’s catching of a horse for himself’
(1237) I mea mai a Tuu ki a maatou i
T/A say hither pers Tu to pers Iplexcl DO
te koohuru a Pou i tana tuakana
the murder gen Pou DO sggenIIIsg older brother
‘Tu told us about Pou’s killing of his (=Tu’s) brother’

In predicative contexts, moo is used:

(1238) Moo Maaui teenei koorero pakiwaitara
intgen Maui this story fiction
‘This story is about Maui’

Moo is also used attributively:

(1239) Ngaa whakataukii moo Tuu-whakairi-ora:…
the(pl) saying intgen Tu-whakairi-ora
‘The sayings referring to Tu-whakairi-ora:…’ TWh, 24)
(1240) …te koorero moo Hinemoa raaua ko Tutaanekai
the story intgen Hinemoa IIIdl spec Tutanekai
‘the story about Hinemoa and Tutanekai’

2.1.1.4.16 Essive
Essive NPs are marked with hei (cf. 2.1.1.2.9.3):
(1241) Hei neehi ahau i taawahi
as nurse Isg at abroad
‘I was a nurse abroad’
(1242) Kaaore au e pai hei kura-maahita
neg Isg T/A good as school-teacher
‘I will be no good as a school-teacher’ (TR2, 71)
(1243) Ka mahara kia waiho hei mokamokai maa·na
T/A think subj leave as pet intgen·IIIsg
‘[He] decided it should be left as a pet for him’ (TP, 89)
 Morphology 293

Hei is also used for personal relations eg.

(1244) Ko Tuumoana-kootore hei mokopuna anoo naa
top. Tumoana-kotore as grandchild again actgen
Porourangi, raaua tahi ko Poroumaataa
Porourangi IIIdl one spec Poroumata
‘Tumoana-kotore was also a descendant of Porourangi, he as well as Poroumata’ (TWh,
18)
(1245) Ka moe·a e tana tama, e Ngaatihau,
T/A sleep·pass. by sggenIIIsg son by Ngatihau
a Te Ataakura, te tamaahine a Poroumaataa,
pers Te Ataakura the daughter pers Poroumata
hei wahine maana
as wife intgenIIIsg
‘His son, Ngatihau, took Te Ataakura, the daughter of
‘Poroumata, as his wife’(TWh, 18)

2.1.1.4.17 Translative
Translative relations are expressed with hei, and are thus not differentiated from essive
relations:
(1246) …kua whakatuu·ria a Taa Kiingi Iihaka hei tumuaki
T/A appoint·pass. pers Sir Kingi Ihaka as head
moo te Taura Whiri i te Reo Maaori
intgen the rope plait at the language Maori
‘Sir Kingi Ihaka has been appointed as head of the Maori Language Commission’ (He
Muka, 3(4), 2)

2.1.1.4.18 Part-whole
Part-whole relations are most usually expressed with o, the O class possessive preposition
(see 1.10.6):
(1247) te ringa o Tamahae
the hand gen Tamahae
‘Tamahae’s hand’
(1248) ngaa wiira o te motokaa
the(pl) wheel gen the car
‘the wheels of the car’

However, under some circumstances, noo is possible; Wills (1960, 21) gives the example
(1249) he tinana noo te raakau
cls trunk actgen the tree
‘a tree trunk’

O is possible here too, and noo is possible if the phrase is definite as well.
 Maori 294

When local nouns are involved, there is more variety in the prepositions used. In
modern Maori, when there is an initial preposition, i appears to be the unmarked
adnominal preposition linking these nouns to the related ‘whole’, but o and occasionally
ki are also found. When local nouns are in subject position, o appears to be used
exclusively. Earlier descriptions, eg. Stowell (1911, 14–16), Harawira (1950, 36–7) give
even more variety, and suggest that the norm is for the adnominal preposition to be a
copy of the introductory preposition. Harawira also suggests that there were differences
in meaning associated with the choice of adnominal preposition. These days, most of the
variation, and any possible nuance associated therewith, has disappeared. A few
illustrations are given here; many others are scattered throughout this book particularly in
2.1.1.5.
(1250) I whakairo·tia a mua o te whare
T/A carve·pass. pers front gen the house
‘The front of the house was carved’
(1251) Kei te noho ngaa taangata i mua o te whare
T/A sit the(pl) people at front gen the house
kai
food
‘The people are sitting in front of the eating house’
(1252) Kei waho te maahita i te whare
at(pres) outside the teacher at the house
‘The teacher is at the outside of the house’
(1253) I whiu·a atu au e koe ki roto ki te
T/A throw·pass. away Isg by IIsg to inside to the
tai
sea
‘I was thrown by you into the sea’ (lit. ‘to the inside of the sea’) (KM, 2)

Note, however, that some part-whole relationships are expressed as compounds, eg.
(1254) ngaa toetoe·nga miiti
the(pl) left-over·nom meat
‘the remains of (the) meat’

2.1.1.4.19 Partitive

2.1.1.4.19.1 Partitive numeral

O marks the partitive with numerals, eg.
(1255) Toko·rima oona tuaakana, aa, kootahi o
pnum·5 plgenIIIsg older sibling and 1 gen
raatou he wahine
IIIpl a woman
‘He had five older siblings, and one of them was a girl’ (TR2, 188)
 Morphology 295

However, the partitive o can apparently be omitted, although it is possible that this textual
example represents the coalescence in speech of the partitive o and the possessive, rather
than the genuine omission of the preposition, eg.
(1256) Ka tae a Tutaanekai ki teetehi oona
T/A reach pers Tutanekai to a(sp) plgenIIIsg
kaakahu
garment
‘Tutanekai reached for one of his garments’ (H, 10)

The following were not judged to be paraphrases:

(1257a) Maa·ku e hoko mai e rua tekau pereti
intgen·Isg T/A buy hither num 2 10 plate
‘I’ll buy 20 plates’
(1257b) Maa·ku e hoko mai e rua tekau o
intgen·Isg T/A buy hither num 2 10 gen
ngaa pereti
the(pl) plate
‘I’ll buy 20 of the plates’

Note that such numerals are expressed predicatively, without the partitive where possible,
eg.
(1258) E rua ngaa tama i mate
num 2 the(pl) boy T/A die
‘Two of the boys were killed’

2.1.1.4.19.2 Non-partitive numeral

There is no special marking for non-partitive numerals. Non-partitive numerals occur in
post-head position. For the form of the numerals see 2.1.6.2.
(1259) eenei waka e rua
these canoe num 2
‘these two canoes’
(1260) Ka moe a Tuumoana-kootore i ngaa
T/A sleep pers Tumoana-kotore DO the(pl)
waahine tokorua
wahine tokorua
‘Tumoana-kotore married two women’ (TWh, 18)

Again, the preference is to express these numerals predicatively, eg.

(1261) E tekau maa rua ngaa mema o te
num 10 and 2 the(pl) member gen the
komiti ko konei a te ahiahi nei
committee at(fut) here at(fut) the afternoon proxI
‘The 12 members of the committee will be here this afternoon’
 Maori 296

2.1.1.4.19.3 Partitive quantifier

The class of quantifiers is not clear-cut in Maori (see 1.2.5.2.6). However, katoa ‘all’ and
the determiners (t)eetahi ‘a/some’, as well as mano ‘thousand, many’, tini ‘host, many’
can all be heads followed by o+NP in expressions which are the translation equivalents
of English partitive quantifiers.
(1262) Ka mahue mai eetahi o oona taangata
T/A leave hither some(pl) gen plgenIIIsg people
‘Some of the people were left behind’ (P, 1)
(1263) Kua haere te katoa o ngaa taangata
T/A move the all gen the(pl) people
‘All of the people went’
(1264) Ka whakamiiharo ia ki te mano o ngaa
T/A marvel at IIIsg to the thousand gen the(pl)
taangata
people
‘He marvelled at the great number of people’

Tini ‘numerous’ or maha ‘many’ could replace mano here. Te mano taangata is also
possible without significant change of sense.
‘Both of’ cannot be expressed partitively. E rua ‘num 2’ is used, either predicatively
or attributively, eg.
(1265) E rua ngaa tama i whiwhi ki ngaa takoha
num 2 the(pl) boy T/A receive to the(pl) gift
‘Both of the boys received presents’
(1266) Naa·ku i hoatu kootahi taara maa ngaa
actgen·Isg T/A give one dollar intgen the(pl)
tama e rua
boy num 2
‘I gave both of the boys a dollar’

‘Each of’ cannot be expressed partitively either. Takitahi can be used predicatively, or ia
attributively, eg.
(1267) Taki·tahi ngaa tama i whiwhi ki ngaa takoha
gp·one the(pl) boy T/A receive to the(pl) gift
‘Each of the boys received presents’
(1268) I whiwhi ia tama ki ngaa takoha
T/A receive each boy to the(pl) present
‘Each boy received presents’
(1269) Naa·ku i hoatu kootahi taara maa ia
actgen·Isg T/A give one dollar intgen each
tamaiti
child
‘I gave each child a present’
 Morphology 297

2.1.1.4.19.4 Non-partitiυe quantifier

There is no special marker for non-partitive quantifiers. See 1.2.5.2.6. Examples of
equivalent usage are
(1270) Eetahi tamariki i paatootoo mai i te
some(pl) children T/A knock hither DO the
tatau
door
‘Some children knocked at the door’
(1271) I hoko mai au i eetahi tiihi
T/A buy hither Isg DO some(pl) cheese
‘I bought some cheese’

2.1.1.4.19.5 Partitive negative quantifier

There are no negative quantifiers in Maori. Such expressions are rendered by sentence
negation, eg.
(1272) Kaahore ngaa tama i hoki mai
neg the(pl) boy T/A return hither
‘None of the boys returned’
(1273) Kaahore ia i hoatu i te taura ki ngaa
neg IIIsg T/A give DO the rope to the(pl)
tama tane
child boy
‘She gave ropes to none of the boys’
‘She gave none of the ropes to the boys’

2.1.1.4.19.6 Non-partitiυe negative quantifiers

Such expressions are rendered by sentence negation, eg.
(1274) Kaaore he whare o reira
neg a house gen there
‘There were/are no houses there’
(1275) Kaaore he peita o teenei whare
neg a paint gen this house
‘This house has/had no paint’
(1276) Kaaore te wahine e piirangi atu ki a
neg the woman T/A desire away to pers
Hukarere
Hukarere
‘No woman would want Hukarere’ (TR2, 72)
(1277) Kaaore anoo teetahi raruraru kootahi kia paa
neg again a(sp) problem one subj arise
mai
hither
‘Not a single problem had arisen’
 Maori 298

See also constituent negation, 1.4.2.

2.1.1.4.20 Price
Moo is normally used today, although it is not clear that this is a traditional usage:
Williams’s Dictionary says price with hoko takes ki, for instance.
(1278) Ka hoko·na e ia te kurii raa moo te rua
T/A sell·pass. by IIIsg the dog dist intgen the 2
rau taara
hundred dollar
‘He sold that dog for $200’
(1279) Maa·u teenaa moo te kotahi taara intgen·IIsg that intgen the one dollar
‘You can have it for $1’

If homai/atu ‘give’ are used in the actor-emphatic, the price appears as the unmarked NP,
and moo accompanies the item bought:
(1280) Maau e homai e toru taara moo te
intgenIIsg T/A give num 3 dollar intgen the
potae nei
hat proxI
‘You must give me $3 for this hat’

It is probably more normal in Maori to express such money matters through non-verbal
sentences, eg.
(1281) E ono taara te utu moo te whakawhitinga
num 6 dollar the price intgen the crossing
‘It costs $6 for the crossing’

2.1.1.4.21 Value
The most normal expression of value is with moo, eg
(1282) Te teepu nei ea ana moo te rima taara
the table proxI worth T/A intgen the 5 dollar
‘This table is worth $5’
(1283) I mahi·a e ia he taonga wheua moo
T/A make·pass. by IIIsg cls treasure bone intgen
te rau taara
the hundred dollar
‘He made a bone pendant worth $100’

It appears that value cannot be expressed predicatively. Attributively, value is expressed

like quantity, eg.
(1284) He rua taara tikiti teenei
cls 2 dollar ticket this
‘This is a $2 ticket’
 Morphology 299

2.1.1.4.22 Distance
The norm is to make the measure the predicate in a non-verbal sentence, eg.
(1285) Kotahi maero taku aru·nga i a ia
one mile sggenIsg chase·nom DO pers IIIsg
‘I chased him for a mile’
(1286) E rua maero te haere·nga o te one
num 2 mile the move·nom gen the beach
‘The beach extends for 2 miles’

If such expressions are adverbial, they have no special marker, eg.

(1287) I tauomaoma raatou e rua maero i te
T/A race IIIpl num 2 mile at the
one
beach
‘They had a race for two miles along the beach’

2.1.1.4.23 Extent
There is a preference, as with price and distance, to express the measure as the predicate
of a non-verbal sentence, eg.
(1288) Kotahi mita te roa o te ruuri nei
one metre the long gen the ruler proxI
‘This ruler is a metre long’
(1289) Tekau maa rua mita te roa o too raatou waka
10 and 2 metre the long gen sggenIIIpl canoe
‘They have a 12-metre long boat’
(1290) Taatou ka mahi i te tawa kia rua rau
Iplincl T/A make DO the tower subj 2 hundred
putu te roa
foot the long
‘Let’s make a tower 200 feet high’

2.1.1.4.24 Concessive
Concessive NPs are introduced by ahakoa, as are concessive clauses (see 1.1.2.4.2.8):
(1291) Ahakoa te maakuu, me te makariri, kaaore
although the damp with the cold neg
a Paki… e aaroo ake
pers Paki T/A think up
‘Despite the damp and the cold, Paki… didn’t take any notice of them’ (TR2, 118)
(1292) Ahakoa te kaha o tana mate, kaaore
although the strong gen sggenIIIsg illness neg
ia i haere ki te hoohipera
IIIsg T/A move to the hospital
‘Despite the seriousness of his illness, he didn’t go to hospital’ (TR2, 57)
 Maori 300

(1293) Ahakoa te riri o Hata, ka haere tonu a

although the anger gen Hata T/A move still pers
Hukarere ki te taraiwa taraka
Hukarere to the drive truck
‘Despite Hata’s anger, Hukarere still went truck-driving’ (TR2, 71)

They can also occur in final position, although initial position is preferred.

2.1.1.4.25 Inclusion
The comitative preposition me can be used with this function, usually supported by atu,
eg.
(1294) Ngaa taangata katoa i kata me te
the(pl) people all T/A laugh with the
rangatira atu
chief away
‘All the people, including the chief, laughed’
(1295) I whariru ia ki ngaa taangata katoa, me
T/A shake IIIsg to the(pl) people all with
Hone atu
John away
‘He shook hands with everyone, including John’
(1296) Kua tika katoa ngaa kai me ngaa miiti atu
T/A right all the(pl) food with the(pl) meat away
‘All the food, including the meat, should be ready now’

2.1.1.4.26 Exclusion
The most straightforward way to express exclusion involves the use of haaunga, which
may or may not be a nominalization related to hau ‘exceed, be in excess’. It is often
supported by anoo, eg.
(1297) I te hari katoa haaunga anoo a Hone
T/A happy all except again pers John
‘Everyone except John was pleased’
(1298) Kua maoa katoa ngaa kai haaunga ngaa
T/A cooked all the(pl) food except the(pl)
kiinaki
complement
‘Everything except the meat is cooked’
(1299) Mea wepu katoa ngaa tamariki haaunga anoo te
thing whip all the(pl) children except again the
pootiki
youngest
‘All the children except the youngest were whipped’
 Morphology 301

However, negation can also be used, for instance, as an alternative to (1298) it is possible
to say:
(1300) Kua maoa katoa kaahore anoo ngaa kiinaki
T/A cooked all neg again the(pl) complement
‘Everything except the meat is cooked’

2.1.1.4.27 Addition
Additive NPs use the form taapiri atu, ‘additional’, eg.
(1301) Taapiri atu ki a Hone, ka hiahia taatou e
add away to pers John T/A desire Iplincl num
rua taangata tino kaha
2 men very strong
‘We will need two strong men in addition to John’

The taapiri phrase can follow, but my consultant was then uncertain whether synonymy
was preserved, or whether it then means ‘We will need two strong men, and in addition,
John’.
(1302) Ko teenei hei taapiri atu ki ngaa mea i
top. this cls(fut) addition away to the(pl) thing T/A
hoatu·ngia e ahau ki a koe inanahi
give·pass. by Isg to pers IIsg yesterday
‘This one is additional to the ones I gave you yesterday’

(Note that this use of hoatu with a passive termination is modern, and not accepted by all
speakers.)
Sometimes anoo hoki or anoo or hoki alone is sufficient to indicate addition. (In some
dialects the form noki is used: it seems likely that this is a contraction.)
(1303) E haere ana noki a Hata ki te taaone
T/A move T/A again also pers Hata to the town
a te poo nei?
at(fut) the night proxI
‘Is Hata going to town tonight too?’

2.1.1.4.28 Vocatiυe
Maori has a vocative particle e which is normally described as obligatory before proper
names of two morae or less (see 3.2.3.3 and Bauer, 1981b). It is not normally used before
longer proper names, but can be for emphasis, and is variable before koe ‘IIsg’ (see
1.1.1.3.1.1, 1.1.1.3.2). Before common nouns there appears to be a choice of e+full NP,
or head of NP only. The following examples illustrate:
(1304) Whakapai·tia te teepu, e Mere!
arrange·pass. the table voc Mary
‘Set the table, Mary!’
(1305) Tango·hia oo huu, Reweti
 Maori 302

take·pass. plgenIIsg shoe David

‘Take your shoes off, David’
(1306) Kia tere, tamariki maa
subj fast children and others
‘Hurry up, children’
(1307) Haere mai, (e) ngaa toa o te
move hither voc the(pl) champion gen the
Tai Tokerau
Nth Auckland
‘Welcome, champions of North Auckland’

There are apparent exceptions like

(1308) Titiro e kawana
look voc governor
‘Look, governor’ (H, 5)

My consultant felt that the e was not optional here. In fact, there is rather more variation
in common usage than the usual rules suggest, eg. in
(1309) Whakarongo mai, e te rata
listen hither voc the doctor
‘Listen, doctor’

the addressee phrase could be e rata if the speaker was on personal terms with the
addressee. And in conventional greetings, the e is frequently omitted (nowadays, at least)
even with names of two morae, eg.
(1310) Kia ora, (e) Kare
subj well voc Kare
‘Hello, Kare’

2.1.1.4.29 Citation form

There is no special mark for citation in Maori.
(1311) Te tae·nga ki ngaa kupu whakamutu·nga o
the arrive·nom to the(pl) word finish·nom gen
te haka a Wairangi “a te”—kua mau ngaa
the haka gen Wairangi a te T/A seize the(pl)
ringa ki ngaa patu
hand to the(pl) club
‘When they came to the concluding words of Wairangi’s haka,—“a te”—their hands
grasped their clubs’ (W, 200)
(1312) Ka karanga au “Paahi!” Ko teenei kupu i
T/A call Isg pass top. this word T/A
tango·hia mai i te reo Paakehaa, araa,
take·pass. hither from the language Pakeha ie.
“Pass!”
 Morphology 303

pass
‘I called “Paahi!”. This word was taken from English, ie. “Pass!” (TR2, 133)

However, if Maori words are cited, they require the article te. The citation, if
appositional, will take the same marking as the head phrase, eg.
(1313) Ko teenei kupu, ko te “maarena” mea tango
top. this word top. the marry thing take
mai i te reo Paakehaa
hither from the language Pakeha
‘This word, maarena was borrowed from English’

In subject position, they take no marking, just like other subjects:

(1314) He kupu hou te “maarena”
cls word new the marry
‘Maarena is a new word’

2.1.1.4.30 Label form

There is no special label form in Maori. Examples can be found in book titles in the
bibliography.

2.1.1.4.31 Naming
Names in acts of naming and similar contexts are preceded by ko, which in many cases
can be analyzed as an embedded equative predicate, eg.
(1315) A, mohoa noa nei mahara·tia tonu·tia
and to the present indeed proxI remember·pass. still·pass.
e maatou te ingoa o teeraa toka ko
by Iplexcl the name gen that rock eq?
Te-rere·nga-o-Te-Aohuruhuru
the-leap·nom-gen-Te-Aohuruhuru
‘To this day that rock is known to us as the Leaping Place of Te Aohuruhuru’ (TA, 8)
(1316) Ka hua·ina e ia te ingoa ko Te Aomihia
T/A name·pass. by IIIsg the name eq Te Aomihia
‘[She] was named Te Aomihia by her’ (TWh, 18)
(1317) Kii·a iho te waahi i taapuke·a ai ko
say·pass. down the place T/A bury·pass. part. eq?
Te-ewe-o-Tuu-whakairi-ora
the-placenta-gen-Tu-whakairi-ora
‘The place where it was buried is still called Te-ewe-o-Tu-whakairi-ora’ (TWh, 18)
(1318) Ka whakapoto·a ki te karanga·tia, ko
T/A shorten·pass. to the call·pass. eq
Tuu-whakairi-ora
Tu-whakairi-ora
‘It was shortened, when they called him, to Tu-whakairi-ora’ (TWh, 18)
 Maori 304

2.1.1.4.32 Listing
In lists, articles can be omitted, but there is no special listing form eg.
(1319) …araa kee hoki he tini ko ngaa tamariki,
ie. contr also cls multitude spec the(pl) children
ngaa mokopuna, ngaa waahine, koroua,
the(pl) grandchildren the(pl) women old men
kuia, me eeraa atu…
old women with those away
‘…that is, multitudes of children, infants, women, old men, and old women, and other
things,…’ (TWh, 23)

2.1.1.4.33 Similitude
Maori uses me to express similitude, eg.
(1320) Anoo te kiri! me he Tapukoorako. Tuu, ka
again the skin like a white hawk stand T/A
tuu ki uta o te wai, me he Kootuku
stand to shore gen the water like a white heron
‘What skin!—like a white hawk. [She] stood up and stepped out of the water, like a white
heron’ (H, 10)
(1321) …ka kite·a iho hoki e ngaa paa raa e
T/A see·pass. down also by the(pl) pa dist T/A
haere ana i te one i Punaruku, i te aakau
move T/A at the beach at Punaruku at the shore
o Karakatuuwhero, me he paaraariki
gen Karakatuwhero like a sea-drift
‘And they looked down from those pas on those who were going along the beach at
Punaruku and the shore of Karakatuwhero, like the sea-drift cast up by the storm’ (TWh,
21)

2.1.1.4.34 Type/species
These are expressed as attributive modifiers, see the examples in 2.1.1.4.9.

2.1.1.5 Local semantic functions

The expression of local semantic functions is not a unitary phenomenon in Maori. It
involves prepositions, local nouns, and in some instances, verbs as well.

2.1.1.5.1 General
‘At rest’ is normally expressed through prepositions. In nonverbal sentences (equivalent
to copular complement in English), the prepositions are tense-marked: i marks past
location, kei marks present location, and any of hei, kei, ko mark future spatial location.
Some of the variation in this last set is regional, thus ko is normal in N. Auckland
dialects, while hei is normal in E. Coast dialects. Kei seems to function as a rather rare
 Morphology 305

variant regardless of dialect. In NPs, ie. attributively, ‘at rest’ is normally expressed by i
(but see 2.1.1.4.18), regardless of the tense of the sentence. I is also most commonly used
for ‘at rest’ in adverbial contexts, but it is sometimes replaced by ki under circumstances
which it is difficult to specify with precision.
For those dialects which have the set hei, kei, i, it is tempting to isolate -i as the basic
stationary locative marker, and to regard he- and ke- as tense morphs. However, the
benefits of this for the analysis as a whole are minimal, since he- and ke- do not occur
elsewhere with appropriate senses, although i, the past tense marker, could then plausibly
be linked to i, the (past) locative marker. Such an analysis does not hold for dialects with
ko, kei, i, and so these prepositional forms are glossed as monomorphemic here.
Examples illustrating these basic constructions follow:
(1322) I te kura ia
at(pt) the school IIIsg
‘She was at school’
(1323) Kei te kura ia
at(pres) the school IIIsg
‘She is at school’
(1324) Kei/Hei/Ko te kura ia
at(fut) the school IIIsg
‘She will be at school’
(1325) E mau·ria ana ngaa wuuru ki te maakete
T/A carry·pass. T/A the(pl) wool to the market
i Aakarana
at Auckland
‘The wool was taken to the market in Auckland’
(1326) He kaihoko te tangata i te kuuaha
cls salesman the man at the door
‘The man at the door is a salesman’

The use of i as the adnominal preposition with local nouns is just a particular case of this
general principle, but with local nouns, i is not the only possibility. The following
example has two adnominal ‘at rest’ locatives:
(1327) Ka titiro ia ki te parani i runga i te
T/A look IIIsg to the brand at top at the
pounamu
bottle
‘He looked at the brand on the bottle’

The adverbial use is illustrated by the following:

(1328) Kaaore te tangata e mokemoke i reira
neg the man T/A lonely at there
‘People are not lonely there’ (TR2, 56)

The problems of establishing the principles determining the choice between i and ki can
be seen in the following sentences:
 Maori 306

(1329) Kei te tahu ia i te ahi i/(*)ki te marae

T/A light IIIsg DO the fire at/to the marae
‘He is lighting the fire at the marae’

With i, consultants agreed that there was an already established fire place at the marae,
and only the lighting of the fire was in question. With ki, there was less agreement. Some
rejected it outright Some suggested that this was appropriate if there was no fixed
location for the fire. To others, it suggested that fire was brought from somewhere else, in
which case, it seems reasonable to treat ki as a goal preposition, as the gloss does. With
the following example, however, there are different considerations:
(1330) I mate ia i/ki te awa
T/A die IIIsg cause/at the river
‘He died because of/at the river’

Because mate uses i to mark cause NPs, ki must be used here for the location. The verbs
for which this situation holds are predictable. Compare (1330) with
(1331) I koohuru·tia te tangata i/ki te awa
T/A murder·pass. the man at/to the river
‘The man was murdered at the river’

Here i implies simply location, whereas ki implies that he was taken there and murdered
there, where the stationary/movement distinction is upheld. Through the distinction
between i and ki, Maori can readily distinguish between the two readings of ‘I left the
coat at your house’ : (a) “I took the coat to your house and left it there”, and (b) “The coat
was at your house, and I went away without it”, as
(1332a) I waiho ahau i te koti ki too whare
T/A leave Isg DO the coat to sggenIIsg house
‘I left my coat at your house’ (=a)
(1332b) I waiho ahau i te koti i too whare
T/A leave Isg DO the coat at sggenIIsg house
‘I left my coat at your house’ (=b)

The usual preposition for ‘motion to’ spatial locations in all syntactic environments is ki.
However, such phrases are rarely found as non-verbal predicates. The examples illustrate
adverbial, adnominal and non-verbal predicate usage:
(1333) E whakahorohoro ana au ki te kaainga
T/A hurry T/A Isg to the home
‘I’m hurrying home’
(1334) Ko teeraa te huarahi ki te ngahere
eq that the path to the bush
‘That’s the path to the bush’
(1335) Ki a Kare kee te reta nei
to pers Kare contr the letter proxI
‘This letter is to Kare’
 Morphology 307

The usual preposition for ‘motion from’ a spatial location is i in adverbial environments,
including with nominalizations, but noo predicatively and attributively other than with
nominalizations, although, again, such phrases are not particularly common as non-verbal
predicates. The examples illustrate adverbial, adnominal and non-verbal predicate usage:
(1336) Kua hoki mai raaua i Rotorua
T/A return hither IIIdl from Rotorua
‘They have returned from Rotorua’
(1337) …te kau·anga o Hinemoa i Oowhata ki
the swim·nom gen Hinemoa from Owhata to
Mokoia
Mokoia
‘Hinemoa’s swim from Owhata to Mokoia’
(1338) Noo Ingarangi mai teenei reta
actgen England hither this letter
‘This letter is from England’
(1339) Ko teenei he reta noo Ingarangi mai
top. this cls letter actgen England hither
‘This is a letter from England’

(See also Source, 2.1.1.4.2.)

There is no single general marker for ‘motion past’ in Maori. The verb pahure ‘pass
by’ (incidentally not a loan from English) is used in many contexts, eg.
(1340) Ka pahure te paa raa
T/A pass the pa dist
‘[They] passed that pa’ (TWh, 19)

Attributive and predicative examples in English are often rendered with pahure, eg.
(1341) E pahure ana koe i te kura ki toona
T/A pass T/A IIsg DO the school to sggenIIIsg
whare
house
‘His house is past the school’
(1342) Pahure atu te whare pukapuka, ko te toa
pass away the house book top. the store
hua raakau
fruit tree
‘The fruit shop is past the library’

With individual types of location, other means of indicating motion past are used.
Since many of the specific local functions to be discussed in the following sections
involve local nouns, it is perhaps worth pointing out here that, as well as the commonest
use with an adnominal prepositional phrase following, these nouns can occur without
such a phrase, eg. ki raro, ‘down’ (lit. ‘to below’), and at times with a modifying noun. In
this latter case, a lexical noun which could occur in an adnominal phrase is treated as a
modifier. This is particularly common if that NP is indefinite (compare object
 Maori 308

incorporation with verbs), eg. ki waenganui taru ‘into the middle of some weeds’ (lit. ‘to
middle weed’). All three constructions with local nouns are possible in all syntactic
environments.

2.1.1.5.2 Proximate
Maori has an intransitive stative verb tata ‘be close’ which is involved in the expression
of proximate location.
Proximate location:
(1343) Tino tata te kura ki te awa
very near the school to the river
‘The school is near the river’
(1344) Naa·ku i waiho aku huu tata tonu ki
actgen·Isg T/A leave plgenIsg shoe near still to
te kuuaha
the door
‘I left my shoes near the door’
(1345) Noo·ku te tuuru tata tonu ki te matapihi
actgen·Isg the chair near still to the window
‘The chair near the window is mine’

Movement to proximate:
(1346) Pei·a atu too nohonga kia tata ki te
push·pass. away sggenIIsg chair subj near to the
matapihi
window
‘Push your chair near the window’

Other syntactic environments appear impossible.

Movement from proximate:
Either of the following formulations is possible:
(1347a) I te matapihi tata tonu, ka haere ia ki te
at the window near still T/A move IIIsg to the
kuuaha
door
‘She walked from near the window to the door’
(1347b) E tata ana ia ki te matapihi ka haere
T/A near T/A IIIsg to the window T/A move
tonu ia ki te kuuaha
still IIIsg to the door
‘Being near the window, she walked to the door’

Movement past proximate:

This can be expressed with a combination of pahure (see 2.1.1.5) and tata:
(1348) I pahure maatou tata tonu ki te koowhao
 Morphology 309

T/A pass Iplexcl near still to the hole

‘We walked past near the hole’

Wiliams’s Dictionary gives the following example:

(1349) Numi·a tata·tia ki te whare i too
pass by·pass. near·pass. to the house at sggenIIIsg
tupuna
ancestor
‘Pass near your ancestral home’

2.1.1.5.3 Interior
The expression of interior location involves the use of the local noun roto ‘inside’ in all
syntactic environments.
Location in interior:
(1350) Kei roto ngaa pene i te pouaka
at(pres) inside the(pl) pen at the box
‘The pens are in the box’
(1351) E kaukau ana a Rewi i roto i te awa
T/A swim T/A pers Rewi at inside at the river
‘Rewi is swimming in the river’
(1352) te haereere·nga i roto o Iinia
the travel·nom at inside gen India
‘the travelling round in India’

However, the following example with ringa ‘hand’ should be noted:

(1353) …te taiaha-o-kura ki te ringa
with the taiaha-gen-feather to the hand
‘…the decorated taiaha in his hand’ (TWh, 20)

Motion to interior:
The following are two adverbial examples:
(1354) I whiu·a atu au e koe ki roto ki te
T/A throw·pass. away Isg by IIsg to inside to the
tai
sea
‘I was thrown by you into the sea’ (KM, 2)
(1355) I taka·ia ia e au ki roto ki tooku
T/A wrap·pass. IIIsg by Isg to inside to sggenIsg
tikitiki
girdle
‘I wrapped him in my girdle’ (KM, 4)
 Maori 310

Notice that, since the source of these is an older text, the adnominal preposition with the
local noun matches the introductory preposition. In modern Maori, either i or o could be
used instead of the adnominal ki. Notice also that in the second example, the English does
not make explicit the ‘motion to’ the girdle that is involved. It is normal in Maori to use
the goal marker in all such instances. The only adnominal examples found were with
nominalizations, and these follow exactly the adverbial form illustrated above. Such
phrases do not occur as non-verbal predicates.
Motion from interior:
Motion from interior can be expressed with i+roto, eg.
(1356) Ka tango·hia mai e ia tana pene i
T/A take·pass. hither by IIIsg sggenIIIsg pen at
roto i te kete
inside at the kit
‘She took her pen out of her purse’

However, it is probably commoner to express this as motion to exterior, eg. (1357),

which is a textual example, was rejected by my consultant in favour of (1358):
(1357) Haere atu i roto i teenei whare
move away from inside at this house
‘Go from this house’ (KM, 2)
(1358) E puta atu ki waho o teenei whare
imp. appear away to outside gen this house
‘Get out of this house’

Similarly, ki waho i te kete ‘to outside at the kit’ was the first choice in (1356) as well.
Adnominal and non-verbal predicate uses require noo, eg.
(1359) Ko teenei he mea noo roto i te whoounu
top. this a thing actgen inside at the phone
‘This is something from the inside of the phone’
(1360) Noo roto i te whoounu teenei mea
actgen inside at the phone this thing
‘This thing is from the inside of the phone’

Motion through interior:

No distinction is made (except in the verb) from location at interior, eg.
(1361) I haaere·ere maatou i roto i te ngahere
T/A go·dup Iplexcl at inside at the bush
‘We walked through the bush’

2.1.1.5.4 Exterior
The expression of exterior location involves the use of the local noun waho ‘outside’ in
all syntactic environments.
Exterior location:
(1362) Kei waho te kurii i te whare
 Morphology 311

at(pres) outside the dog at the house

‘The dog is outside the house’

For this word order, see 1.13.2.5.

(1363) E tau mai ana [te roonihi] i waho
T/A anchor hither T/A the launch at outside
atu o ngaa toka
away gen the(pl) rock
‘The launch was riding at anchor beyond the rocks’ (TR2, 117)
(1364) He kaihoko te tangata i waho o te whare
cls salesman the man at outside gen the house
‘The man outside the house is a salesman’

Motion to exterior:
(1365) Ka maka·ia ake e ia ki waho o te
T/A throw·pass. up by IIIsg to outside gen the
wai
water
‘It was thrown out of the water by him’ (TR2, 6)
(1366) Ka puta a Maaui ki waho, kaaore toona
T/A appear pers Maui to outside neg sggenIIIsg
whaea i piirangi ki a ia
mother T/A want to pers IIIsg
‘When Maui was born, his mother didn’t want him’ (TR2, 188)

(Note that this is not the standard expression for being born in Maori.) The only clear
examples of adnominal ‘motion to exterior’ phrases are with nominalizations, and these
follow the adverbial form just illustrated. However, examples like the following should
be noted:
(1367) Me kau hoki ki hea i te tawhiti ki
oblig swim indeed to where from the distance at
waho o te moana nui?
outside gen the sea big
‘Where could he swim to, from such a distance out in the ocean?’ (KH, 3)

‘Motion to exterior’ phrases do not occur as non-verbal predicates. Motion from exterior:
(1368) I haere mai ia i waho tonu o
T/A move hither IIIsg from outside still gen
Taupo
Taupo
‘She came here from just outside Taupo’

There is potential ambiguity with past location:

(1369) He moka peita teenei i waho o te whare
cls piece paint this from outside gen the house
 Maori 312

‘This is a piece of paint from outside the house’.

The possibility of using noo here instead of i points up the problem of deciding whether
the i is locative or source. However, the exterior is not always specified, eg.
(1370) I ahu mai au i te whare
T/A head hither Isg from the house
‘I walked away from the house’

Motion past exterior:

The following example uses the verb numi ‘pass by’:
(1371) …ka numi·a ki te pakitara maui o waho
T/A pass by·pass. to the wall left gen outside
o te whare
gen the house
‘…and led [him] past the outside of the left wall of the house’ (TWh, 20)

Pahure can also be used, and the exterior does not have to be specified, eg.
(1372) I pahure atu au i te whare
T/A pass away Isg DO the house
‘I walked past the house’

2.1.1.5.5 Anterior
The expression of anterior location involves the use of the local noun mua ‘front’ in all
syntactic environments.
Anterior location:
(1373) Kei mua te ihu i ngaa taringa
at(pres) front the nose at the(pl) ear
‘The nose is in front of the ears’
(1374) Kei te noho ngaa taangata i mua o te whare
T/A sit the(pl) people at front gen the house
kai
food
‘The people are sitting in front of the eating house’
(1375) Kaatahi raaua ka titiro atu ki te tekoteko i
then IIIdl T/A look away to the carving at
mua o te whare whakairo
front gen the house carved
‘Then they looked at the carved figure [on the gable] at the front of the meeting house’

Motion to anterior:
(1376) Karanga·tia ngaa tamariki ki mua o te kura
call·pass. the(pl) children to front gen the school
‘Call the children to the front of the school’
 Morphology 313

(1377) Kaua e haere ki mua

neg T/A move to front
‘Don’t go ahead’

The only clear examples of adnominal ‘motion to anterior’ phrases occur with
nominalizations, and follow the adverbial pattern just illustrated. Such phrases do not
occur as non-verbal predicates.
Motion from anterior:
(1378) Ka haere atu raaua i mua o te whare
T/A move away IIIdl from front gen the house
ki roto
to inside
‘They went inside from in front of the house’

Again, there is ambiguity with past location, as can be seen in

(1379) Ka kite a Tuu i te mamaha e puta
T/A see pers Tu DO the steam T/A appear
ake ana i mua o te tereina
up T/A at/from front gen the train
‘Tu saw the steam rising (from) in front of the train’

Motion past anterior:

It is possible to use pahure, eg.
(1380) I pahure atu raatou i mua o te whare
T/A pass away IIIpl at front gen the house
‘We walked past the front of the house’

However, if a sense of route is involved, maa mua can be used, eg.

(1381) I haere ia maa mua i te TV
T/A move IIIsg intgen front at the TV
‘He walked across in front of the TV’

2.1.1.5.6 Posterior
The expression of posterior location involves the use of the local noun muri ‘behind’ in
all syntactic environments.
Posterior location:
(1382) Kei muri too peeke i te kuuaha
at(pres) behind sggenIIsg bag at the door
‘Your bag is behind the door’
(1383) Ka haere raatou i muri i te kaiaarahi
T/A move IIIpl at behind at the guide
‘They walked behind the guide’
(1384) Naa Huia te reta kee i muri o te karaka
actgen Huia the letter contr at behind gen the clock
 Maori 314

‘The letter behind the clock is Huia’s’

Motion to posterior:
(1385) Ka kuhu ngaa tamariki ki muri i te toka
T/A hide the(pl) children to behind at the rock
‘The children hid behind the rock’
(1386) Ko ngaa patu poto a te hoko·whitu nei i
top. the(pl) club short gen the 20×·7 proxI T/A
huna·a ki muri i ngaa tuara
hide·pass. to behind at the(pl) back
‘The party held their short clubs concealed behind their backs’ (W, 200)

The only clear examples of adnominal ‘motion to posterior’ phrases occur with
nominalizations, and follow the adverbial pattern just illustrated. Such phrases do not
occur as non-verbal predicates.
Motion from posterior:
(1387) Ka oma mai ngaa tamariki i muri i te
T/A run hither the(pl) children from behind at the
whare
house
The children ran up from behind the house’

Again, there is ambiguity with past location:

(1388) Mau·ria mai te kii i muri i te karaka
bring·pass. hither the key from behind at the clock
‘Bring me the key from behind the clock/ at the back of the clock’

Motion past posterior:

The situation is parallel to motion past anterior, ie. either a verb like pahure is used, or
maa muri can be used to indicate a route:
(1389) I haere ia maa muri i te TV
T/A move IIIsg intgen behind at the TV
‘He walked past behind the TV’

2.1.1.5.7 Superior
The expression of superior location involves the use of the local noun runga ‘top’ in all
syntactic environments. Often, however, ake ‘up’ is required to support this.
Superior location:
(1390) I runga tonu ake i te ngaawhaa teenei waapu
at top indeed up at the boiling pool this bridge
‘This bridge did indeed go over the boiling pool’
(1391) E tari ana te rama i runga ake i te teepu
T/A hang T/A the light at top up at the table
‘The lamp hangs above the table’
 Morphology 315

(1392) Kua pakaru te rama i runga ake i te teepu

T/A broken the light at top up at the table
‘The light above the table is broken’

However, ake is not always required; the following has iho:

(1393) E rerere ana te manu i runga tata iho
T/A hover T/A the bird at top near down
i te ngata
at the snail
‘The bird is hovering just above the snail’

Motion to superior:
(1394) Kei te titiro ake ia ki runga
T/A look up IIIsg to top
‘He is looking upwards’

Alternatively, whaka·runga (lit ‘cause·top’) can replace ki runga here, but a different
order is required:
(1395) Kei te titiro whaka·runga ake ia
T/A look cause·top up IIIsg
‘He is looking upwards’
(1396) I rere te ngaro ki runga ake i te teepu
T/A fly the fly to top up at the table
‘The fly flew to above the table’

Other syntactic environments are apparently not found.

Motion from superior:
Noo can be used, eg.
(1397) I turu·turu mai te wai noo runga ake
T/A drip·dup hither the water actgen top up
‘The water dripped down from above’

However, I failed to elicit examples marked with i in this particular location. This is
probably due in part to a tendency to encode such things as ‘to inferior’, rather than ‘from
superior’ (cf. remarks in 2.1.1.5.3).
Motion past superior:
This is not distinct from location at superior or motion to above, eg.
(1398) Ka rere ngaa wakarererangi i runga ake i
T/A fly the(pl) plane at top up at
taku whare i ngaa raa katoa
sggenIsg house at the(pl) day all
‘Planes fly over my house every day’

Example (1396) above could equally well be translated as ‘The fly flew over the table’.
 Maori 316

2.1.1.5.8 Superior-contact
Superior-contact is expressed with runga ‘top’, as is ‘superior’ in general. However,
superior-contact appears to be the unmarked rather than the marked category in Maori, as
ake support is not required here.
Superior-contact location:
(1399) Kei runga i te teepu taku kapa
at(pres) top at the table sggenIsg cup
‘My cup is on the table’
(1400) Kei te moe a Hata i runga i te takapau hou
T/A sleep pers Hata at top at the mat new
‘Hata is sleeping on the new mat’
(1401) Naa wai te pukapuka i runga i te tuuru
actgen who the book at top at the chair
nei?
proxI
‘Whose is the book on this chair?’

Motion to superior-contact:
(1402) Uta·ina te hei ki runga i te taraka
load·pass. the hay to top at the truck
‘Load the hay onto the truck’
(1403) E kake ki runga i te raakau teitei raa!
imp. climb to top at the tree tall dist
‘Climb into that tall tree!’
(1404) Ka hapai raaua i ngaa tamariki ki runga i
T/A lift IIIdl DO the(pl) children to top at
te puuhara
the platform
‘They lifted the children onto the platform’

Clear adnominal examples occur only with nominalizations and follow the adverbial
form just illustrated. Non-verbal predicates of this type do not occur.
Motion from superior-contact:
(1405) E takataka ana ngaa aaporo i runga i te
T/A fall T/A the(pl) apple from top at the
whata
shelf
‘The apples are falling off the shelf’

Again, there is potential ambiguity with past location:

(1406) Kaahore ngaa kapu i runga i te taraihe
neg the(pl) cup from top at the sideboard
e paru ana
T/A dirt T/A
‘The cups from/on the top of the sideboard weren’t dirty’
 Morphology 317

Motion past superior-contact:

This is not distinct from location, eg.
(1407) I ngaoki te ngaro i runga i aku
T/A crawl the fly at top at plgenIsg
paapaaringa
cheek
‘The fly crawled over my cheeks’

2.1.1.5.8a Surface
Surfaces, like superior-contact, use runga, ‘top’.
Surface location:
(1408) He pikitia kei runga i te pakitara
a picture at(pres) top at the wall
‘There is a picture on the wall’ (TR1, 14)
(1409) Titiro ki te paanui i runga i te wini o
look to the notice at top at the window gen
te toa
the store
‘Look at the notice on the shop window’
(1410) Kei te iri te whakaahua i runga i te pakitara
T/A hang the picture at top at the wall
‘The picture is hanging on the wall’
(1411) Ka kite·a i roto i teetahi whaarua i runga
T/A find·pass. at inside at a(sp) hollow at top
i taua toka
at det aph rock
‘[It] was found in a hollow in this rock’

Motion to surface:
(1412) Ka eke atu a Te Tahi ki runga [i te
T/A embark away pers Te Tahi to top at the
tohoraa]
whale
‘Te Tahi climbed on[to the whale]’ (KWh, 2)
(1413) Te whakaeke·nga o ngaa ope ki runga i te
the arrive·nom gen the(pl) group to top at the
marae
marae
‘The groups gathered on the marae’

Non-verbal predicates of this type do not occur. Adnominal examples occur only with
nominalizations, and take the same form as the adverbial usages above.
Motion from surface:
(1414) Ka iri tanga·tanga noa iho i runga i
 Maori 318

T/A hang loose·dup indeed down from top at

oona pakihiwi
plgenIIIsg shoulder
‘[It] hung loosely from his shoulders’

In many instances there is ambiguity with past locatives:

(1415) Mukumuku·a ngaa paru i runga i te paatuu
wipe·pass. the(pl) dirt from top at the wall
‘Wipe the dirt from/on the wall’ ‘Wipe the dirt off the wall’
‘Wipe off the dirt that is on the wall’

Motion across surface:

The surface itself is usually not specified. Whiti ‘cross’ in some form is commonly
used, eg.
(1416) I te whakawhiti·nga o te katipoo i te
at the cross·nom gen the spider at the
matapihi, mea oma
window thing run
‘The spider ran across the window’

However, it is possible to specify the surface by using i runga i te matapihi ‘at top at the
window’.

2.1.1.5.9 Inferior
Inferior location is expressed by the local noun raro ‘below, underneath’.
Inferior location:
(1417) Kei raro te ngeru i te teepu
at(pres) below the cat at the table
‘The cat is under the table’
(1418) Kei raro iho te waha i te ihu
at(pres) below down the mouth at the nose
‘The mouth is below the nose’
(1419) Kei te kai te kurii i raro i te tuuru
T/A eat the dog at below at the stool
‘The dog is eating under the stool’
(1420) Ka pai ngaa kura i raro i ngaa haahi
T/A good the(pl) school at under at the(pl) church
‘Church schools are good’ (more lit. The schools under the
Churches’ [control] are good’)

Motion to inferior:
(1421) Kei te oma te kurii ki raro i te tuuru
T/A run the dog to below at the stool
 Morphology 319

‘The dog is running under the stool’

(1422) Ka tiaho iho te marama ki raro
T/A shine down the moon to below
‘The moon shone down’

Non-verbal predicate uses do not occur. Adnominal examples occur only with
nominalizations, and have the same form as the adverbial usage illustrated above.
Motion from inferior:
(1423) I oma mai te kurii i raro i te tuuru
T/A run hither the dog from below at the stool
‘The dog ran out from under the stool’

There is often ambiguity with past location, avoided here by the use of mai, and noo is a
frequent alternative to i, eg.
(1424) I oma mai te katipoo noo raro mai
T/A run hither the spider actgen under hither
i te takapau
at the mat
‘The spider ran out from under the mat’

An adnominal example showing the ambiguity is:

(1425) Ko te whaariki i raro i te koohue i
top. the mat at under at the pot T/A
pou i te ahi
burn cause the fire
‘The mat (from) under the pot was burnt’

Motion past inferior:

This is not necessarily distinct from location, eg.
(1426) Kaua koe e haere i raro i te waapu!
neg IIsg T/A move at under at the bridge
‘Don’t walk under the bridge!’

If a route is implied, naa can be used:

(1427) Haere naa raro i te teepu
move actgen under at the table
‘Go under the table’

2.1.1.5.10 Inferior-contact
Inferior-contact is also expressed by raro, ‘below, underneath’. To stress the contact,
however, it is necessary to use te taha raro, ‘the underside’.
Inferior-contact location:
(1428) Kei raro te nama o te miihini
 Maori 320

at(pres) under the number gen the machine

‘The number is underneath the machine’
(1429) Ko ngaa paru kei raro i oo huu
top. the(pl) dirt at(pres) under at plgenIIsg shoe
kia tino rahi
subj very great
‘Get the bottoms of your shoes really muddy’
(1430) Kei raro a Hata i te hiiti e moe
at(pres) under pers Hata at the sheet T/A sleep
ana
T/A
‘Hata is sleeping under a sheet’

Motion to inferior-contact:
This is expressed by ki+raro, eg.
(1431) Whakarapa·ngia atu teenei ki raro i te
stick·pass. away this to under at the
pounamu
bottle
‘Stick this on the bottom of the bottle’
(1432) Noonaa·hea noa atu eenaa kauri ngau·ngau
actgen·where indeed away these gum chew·dup
i rapa ai ki raro i te teepu naa
T/A stick part. to under at the table proxII
‘The chewing gum under the table is very old’

Other syntactic environments are not attested, except with nominalizations, which take
the form above. Motion from inferior-contact:
This can be expressed with i, but is just as likely to be locative as source, eg.
(1433) Tango·hia mai te mea naa i raro i
take·pass. hither the thing proxII from/at under at
te oko
the bowl
‘Peel off the thing from under the bowl/that is under the bowl’

My consultant said that kei ‘at(pres)’ could be substituted for i here, which underlines the
likelihood that the i is locative rather than source.
Motion across inferior-contact:
Again, this is not distinct from location:
(1434) Ko te ngaro i ngaoki haere i te taha raro
top. the fly T/A crawl move at the side under
i te oko
at the bowl
‘The fly crawled along the underside of the bowl’
 Morphology 321

2.1.1.5.11 Lateral
The expression of lateral location involves the use of the noun taha ‘side’, which is not a
local noun, unlike the majority of the forms discussed in 2.1.1.5.
Lateral location:
(1435) I te taha o te whare he puna
at the side gen the house a spring
‘There was a spring at the side of the house’
(1436) E tuu ana ia i te taha o teetahi toka
T/A stand T/A IIIsg at the side gen a(sp) rock
‘She was standing beside a rock’
(1437) I whaanau au i te taha o te moana
T/A born Isg at the side gen the sea
‘I was born beside the sea’

Motion to lateral:
(1438) Ka tuu eetahi ki tana taha waiata ai
T/A stand some(sp, pl) to sggenIIIsg side sing part.
‘Some stood beside him to sing’
(1439) Ka noho ia ki tana taha
T/A sit IIIsg to sggenIIIsg side
‘She sat down at his side’
(1440) Panga·a atu ngaa puune ki te taha o ngaa
put·pass. away the(pl) spoon to the side gen the(pl)
maaripi
knife
‘Put the spoons beside the knives’

Motion from lateral:

An expression as location is as likely as source, but as in other parallel cases, when the
preposition is i, there is ambiguity:
(1441) Ka kite·a kua puta i teetehi taha o te
T/A see·pass. T/A appear from a(sp) side gen the
paa, kua haere whaka·teetehi taha o te paa…
pa T/A move to·a(sp) side gen the pa
‘[They] were seen emerging from one side of the pa, going towards the other side…’ (W,
198)
(1442) Tiki·na atu te nohonga i te taha o
fetch·pass. away the chair from the side gen
taku moenga
sggenIsg bed
‘Fetch the chair from beside my bed /at the side of my bed’

In (1442), kei ‘at(pres)’ is possible in place of i.

Motion past lateral:
The normal expression is as location, eg.
 Maori 322

(1443) Ka haere raaua i te taha tonu o te pahi

T/A move IIIdl at the side still gen the bus
‘They walked past the bus’
(1444) I haere raaua i te taha·taha awa
T/A move IIIdl at the side·dup river
‘They walked beside the river’

Note the reduplication in this last example, which expresses the continuity of the
location.

2.1.2.5.12 Lateral-contact
Like general lateral location, lateral contact involves the use of taha ‘side’. Because of
the similarity, only a few examples are given. Comments are as in 2.1.1.5.11.
Lateral-contact location:
(1445) Ko too reta kei teeraa taha o te
top. sggenIIsg letter at(pres) that side gen the
pouaka haupapa
box ice
‘Your letter is on the side of the fridge’

Motion to lateral-contact:
This uses ki, eg.
(1446) Ko ngaa taangata o Pooneke, ka whakatuutuu
top. the(pl) people gen Wellington T/A erect
i oo raatou whare ki ngaa taha·taha o ngaa
DO plgenIIIpl house to the(pl) side·dup gen the(pl)
puke
hill
‘In Wellington, people build houses on the sides of hills’
(1447) Peita·ngia ngaa paanui ki te taha o te taraka
paint·pass. the(pl) notice to the side gen the truck
‘Paint the notice on the side of the truck’

Motion from lateral-contact:

(1448) E tiihore·hore ana ngaa peita i te taha
T/A peel·dup T/A the(pl) paint from the side
o te whare
gen the house
‘The paint is peeling off from the side of the house’

Motion past lateral-contact:

(1449) E oma haere ana te katipoo i te taha o
T/A run move T/A the spider at the side gen
te pouaka
the box
 Morphology 323

‘The spider ran across the side of the box’

2.1.1.5.13 Citerior
The expression of citerior location involves the use of teenei taha ‘this side’ in most
contexts. Occasionally, tahaki mai ‘side hither’ is possible, but this is not the norm. In
general, the patterns are like those for lateral location. In verbal sentences, mai is quite
frequently required as a post-verbal modifier with such locations. Citerior location:
(1450a) Ko Kaitaaia kei teenei taha o Te Kao
top. Kaitaia at(pres) this side gen Te Kao
‘Kaitaia is on this side of Te Kao’
(1450b) Ko Kaitaaia kei tahaki mai o Te Kao
top. Kaitaia at(pres) side hither gen Te Kao
‘Kaitaia is on this side of Te Kao’

Other syntactic environments use teenei taha, and are parallel to the lateral examples
above in 2.1.1.5.11.
Movement to citerior:
This involves the expected ki+teenei taha, eg.
(1451) Mau·ria mai ngaa toka ki teenei taha o te
bring·pass. hither the(pl) rock to this side gen the
whare
house
‘Bring the stones to this side of the house’

Movement from citerior:

(1452) Mea nuku mai ngaa raakau i teenei taha
thing move hither the(pl) tree from this side
ki teeraa taha o te mahinga
to that side gen the garden
‘The trees were moved from this side of the garden to the other’

Movement past citerior:

Some choice of preposition was possible in the following example: i or ki locatives, or
maa for route, eg.
(1453) Ko te ara hou ka haere i/ki/maa teenei
top. the path new T/A move at/to/intgen this
taha o te taaone
side gen the town
‘The new road will pass on this side of the town’

2.1.1.5.14 Citerior-contact
The expression of citerior-contact location is no different from the expression of citerior
non-contact location (2.1.1.5.13). Only a few examples are given for comparison.
 Maori 324

Citerior-contact location:
(1454) Kei teenei taha ngaa paru o te matapihi
at(pres) this side the(pl) dirt gen the window
‘The dirt is on this side of the window’

Movement to citerior-contact:
(1455) Whakapaa·ngia mai te paanui ki teenei taha o
stick·pass. hither the notice to this side gen
te matapihi
the window
‘Stick the notice on this side of the window’

Movement from citerior-contact:

(1456) Horoi·a mai ngaa peita i/ki teenei taha
clean·pass. hither the(pl) paint from/at this side
o te matapihi
gen the window
‘Clean the paint off this side of the window’

Movement past citerior-contact:

(1457) Ko teeraa katipoo e ngaoki haere ana i/maa
top. that wasp T/A crawl move T/A at/intgen
teenei taha o te matapihi
this side gen the window
‘That wasp is crawling across this side of the window’

2.1.1.5.15 Ulterior
Ulterior location is most generally expressed by teeraa taha ‘that side’. However, in
some contexts, it can be expressed by waho ‘outside’, muri ‘back’ or tua ‘far side, back’.
These are local nouns. It is also sometimes possible to use tahaki atu ‘side away’ (cf.
Citerior location, 2.1.1.5.13). Parallel to the use of mai ‘hither’ with verbs in citerior
location sentences, atu ‘away’ is frequently required in verbal sentences expressing
ulterior location.
(1458) …te rooni e teretere mai ana i waho
the launch T/A anchor hither T/A at outside
atu o ngaa toka
away gen the(pl) rock
‘…the launch riding at anchor beyond the rocks’
(1459a) A Te Kao kei teeraa taha o Kaitaaia
pers Te Kao at(pres) that side gen Kaitaia
‘Te Kao is beyond Kaitaia’
(1459b) A Te Kao kei tahaki atu o Kaitaaia
pers Te Kao at(pres) side away gen Kaitaia
‘Te Kao is beyond Kaitaia’
 Morphology 325

puke onepuu
hill sand
‘The sea is beyond the sand-dunes’

Note the following example which is not clearly ulterior in English, but is in Maori:
(1461) Kaaore kee maatou i kite i te rererangi i
neg contr Iplexcl T/A see DO the plane from
te mea i tua kee o ngaa kapua
the thing at back contr gen the(pl) cloud
‘We didn’t see the plane because it was above the clouds’

Motion to ulterior:
(1462) …ka huri ki tua o ngaa whare
T/A turn to far side gen the(pl) house
‘…and turned and went behind the houses’ (W, 198)
(1463) I hoe atu raatou ki tahaki atu i ngaa
T/A row away IIIpl to side away at the(pl)
toka
rock
‘They rowed out beyond the rocks’

The only other attested syntactic environment is adnominal to a nominalization, where

forms identical to those illustrated are found. Motion from ulterior:
The usual choice of preposition occurs (ie. i or noo), and all the above forms are
possible in some environments, eg.
(1464) Noo tahaki atu o Te Kao a ia
actgen side away gen Te Kao pers IIIsg
‘She comes from (=belongs to) beyond Te Kao’
(1465) I haere mai ia i teeraa taha o
T/A move hither IIIsg from that side gen
Te Kao
Te Kao
‘She came from beyond Te Kao’
 Maori 326

Motion past ulterior:

Similar prepositions are found as with other ‘motion past’ locations, and the forms
above are used:

2.1.1.5.16 Ulterior contact

The expression of ulterior-contact location does not differ from ulterior location (see
2.1.1.5.15). A few parallel examples are given here.
Ulterior-contact location:

(1469) Ko too raatou whare kei teeraa taha o te

top. sggenIIIpl house at(pres) that side gen the
ara
path
‘Their house is across the road’
 Morphology 327

Motion to ulterior-contact:

(1471) I whakawhiti raatou ki teeraa taha o te awa

T/A cross IIIpl to that side gen the river
‘They crossed to the other side of the river’

Motion from ulterior-contact:

Motion past ulterior-contact:

 Maori 328

2.1.1.5.17–18 Medial
There is no distinction between medial location between two and medial location among
three or more objects in Maori. Waenganui ‘middle’, a local noun, is used for all types of
medial location. However, if a large group is involved, roto ‘inside’ may be used.
Waenga can be used to indicate ‘within a large area, but not necessarily central’.
Medial location:
(1474) Kei waenganui te taaone i te ngahere
at(pres) middle the town at the bush
‘The town is in the middle of the bush’ (TR2, 40)
(1475) Kei waenga riiwai a Hata
at(pres) middle potato pers Hata
‘Hata is in the midst of the potatoes’
(1476) Ka tupu te raruraru i waenganui i ngaa
T/A grow the trouble at middle at the(pl)
tohunga nei, i a Nukutawhiti raaua ko
tohunga proxI at pers Nukutawhiti IIIdl spec
Ruanui
Ruanui
‘Trouble arose between these (two) tohungas, Nukutawhiti and Ruanui’ (KW, 1)
(1477) Ko Te Kao kei waenganui i a Kaitaaia
top. Te Kao at(pres) middle at pers Kaitaia
me Te Rangi
with Te Rangi
‘Te Kao lies between Kaitaia and Te Rangi’

Motion to medial:
(1478) Ka maka·ia atu ki waho o te wai, ki
T/A throw·pass. away to outside gen the water to
waenganui i ngaa taru i runga i te parenga
middle at the(pl) weed at top at the bank
‘[It] was thrown out of the water into the middle of the weeds on the bank’
(1479) Kaatahi ka tuku·na taa raatou taurekareka kia
then T/A send·pass. sggenIIIpl servant subj
haere i roto i te iwi o Tupeteka
go at inside at the tribe gen Tupeteka
‘Then they sent their slave to go amongst the people of Tupeteka’ (W, 198)
(1480) Me waiho ngaa putiputi ki waenganui i te
oblig leave the(pl) flower to middle at the
pukapuka kia maroke ai
book subj dry part.
‘Put the flowers in the book to dry’

Non-verbal predicate uses are not found. Adnominal uses occur only with
nominalizations and follow the adverbial pattern illustrated here.
Motion from medial:
 Morphology 329

(1481) I ngoi ake ia i waenganui i ngaa

T/A crawl up IIIsg from middle at the(pl)
raakau
tree
‘He crawled out from between the logs’

As with other types of location, there is potential ambiguity here between location and
source, although ake probably resolves it in this particular case.
Motion past medial:
(1482) I tuketuke haere atu ia i waenganui i
T/A elbow move away IIIsg at middle at
ngaa kaumaatua e rua
the(pl) elder num 2
‘He elbowed his way between the two elders’
(1483) I aata haere maatou i waenga i ngaa
T/A carefully move Iplexcl at middle at the(pl)
toka
rock
‘We picked our way through the rocks’

Maa can be used in (1483) with re-ordered constituents: Maa waenga i ngaa toka, ka
aata haere maatou.

2.1.1.5.19 Circumferential
Circumferential motion is expressed by verbs such as karapoti ‘surround’ or taiaawhio
‘encircle’, eg.
(1484) Ko te taiapa e karapoti ana i te wahanga
top. the fence T/A surround T/A at the piece
o taku whenua
gen sggenIsg land
‘The fence is round the edge of my section’
(1485) I karapoti haere maatou i te whare
T/A surround move Iplexcl DO the house
‘We walked round the house’
(1486) Karapoti·ngia te nama tika
surround·pass. the number right
‘Circle the right answer’

The following example with taiaawhio is from a hymn, Naa te Matua, tama:
(1487) Te hunga ngaakau pakeke
the group heart disbelieve
Taiaawhio noa te ao katoa
encircle indeed the world all
Ako·na mai kia moohio ai
 Maori 330

teach·pass. hither subj know part.

Kia kii ‘Kua maarama’
subj say T/A light
‘The group of disbelievers right round the world, teach them to know so that they can say
“We have seen the light’”

2.1.1.5.20 Citerior-Anterior
There is no simple means of expression of opposite location, other than anterior, although
taurite ‘opposite’ can be used in some instances. There are a number of verbs for face-to-
face personal interaction (friendly, hostile) which answer to ‘opposite’ in specific
contexts. Since it does not seem possible to make useful generalizations, I merely give
ample illustrations:
(1488) Ko too raatou whare tika tonu ki teeraa taha
top. sggenIIIpl house right still to that side
‘Their house is opposite’
(1489) I ngaa huihui·nga ka noho mai ia tika
at the(pl) meet·nom T/A stay hither IIIsg right
tonu mai i ahau
still hither at Isg
‘He sits opposite me at meetings’
(1490) Tuuhi·a mai he raina mai i te riipeka
draw·pass. hither a line hither from the cross
tika tonu ki te kokonga matau
right still to the corner right
‘Draw a line from the cross to the R.H. opposite corner’
(1491) I opa·ina mai te pooro e ia tika
T/A throw·pass. hither the ball by IIIsg right
tonu ki teetahi taha o te whiira taakoro
still to a(sp) side gen the field play
‘He threw the ball to the opposite end of the playing field’

Tika tonu could be replaced in (1491) by anganui tonu ‘right opposite still’.
(1492) Ka whiti ki teeraa taha o Waikato
T/A cross to that side gen Waikato
‘She crossed to the other side of the Waikato’ (W, 197)
(1493) E haere ana raatou i teeraa taha kee o
T/A move T/A IIIpl at that side contr gen
te rori ki ahau
the road to Isg
‘They walked down the opposite side of the road to me’
(1494) I te pito o te teepu ka whakawhiti te
at the end gen the table T/A cross the
katipoo e oma ana
spider T/A run T/A
‘The spider ran across the opposite end of the table’
 Morphology 331

The ‘opposite’ sense here is apparently included in pito.

(1495) Tuu anganui ana ngaa taangata e rua
stand opposite T/A the(pl) men num 2
‘The two men stood opposite each other’

2.1.1.5.21 Interior long object

This is generally expressed by maa (sometimes naa) with roto ‘inside’ or raro ‘beneath’
depending on the type of object, eg.
(1496) I haere raatou maa/i roto i te ana
T/A move IIIpl intgen/at inside at the cave
‘They walked through the tunnel’
(1497) I haere raatou maa raro i te
T/A move IIIpl intgen below at the
kauwhanganui
corridor
‘They walked along the corridor’
(1498) Momi·a ake te wai maa/naa roto i te
suck·pass. up the water intgen/actgen inside at the
taakakau
straw
‘Suck up the water through the straw’

2.1.1.5.22 Exterior long object

This is expressed simply as location, eg.
(1499) I haere tika raatou i te pahi
T/A move right IIIpl at the bus
‘They walked right past the bus’
(1500) I haere tonu raatou i te taha·taha o ngaa
T/A move still IIIpl at the side·dup gen the(pl)
raakau
tree
‘They walked along the row of trees’

2.1.1.5.23 Superior long object

This is expressed as location, eg.
(1501) I rere tonu te wakarererangi i runga ake i
T/A fly still the plane at top up at
te awa
the river
‘The plane flew along above the river’
 Maori 332

2.1.1.5.24 Superior-contact long object

Again, this is expressed as location, eg.
(1502) Ka kite·a iho hoki e ngaa paa raa e
T/A see·pass. down also by the(pl) pa dist T/A
haere ana i te one i Punaruku
move T/A at the beach at Punaruku
‘[They] were looked down on by those pas [as they were] going along the beach at
Punaruku’ (TWh, 21)
(1503) I haere ngaa tamariki i runga i te tinana
T/A move the(pl) children at top at the trunk
raakau
tree
‘The children walked along the tree trunk’

Naa is an alternative to i here, but implies a route.

2.1.1.5.25 Inferior-long object

Again, this is expressed as location:
(1504) I oma ngaa tamariki tika tonu i raro i
T/A run the(pl) children right still at under at
te raina kaakahu
the line clothes
‘The children ran along under the clothes line’

2.1.1.5.26 Inferior-contact long object

Again, this is expressed as location. Raro ‘below’ is sufficient to indicate inferior contact,
but it can be reinforced by taha ‘side’:
(1505) E ngoi ana te ngaro i (te taha) raro i
T/A crawl T/A the fly at the side below at
te tuuporo
the log
‘The fly was crawling along the bottom of the log’

2.1.1.5.27 Interior long object (right angles to length)

This is most readily expressed by whiti ‘cross’, and often requires mention of the sides of
the object Two examples illustrate the kinds of structure used:
(1506) I haere maatou whiti atu i te ana ki
T/A move Iplexcl cross away at the cave to
teeraa atu ara
that away path
‘We walked across the tunnel to the other footpath’
(1507) I oma te kiore tika tonu ki teetahi taha o
 Morphology 333

T/A run the rat right indeed to a(sp) side gen

te kauwhanganui
the corridor
‘The mouse ran across the corridor’

2.1.1.5.28 Superior long object (right angles to length)

This can involve whiti ‘cross’, but does not necessarily, eg.
(1508) Ka whiti ki teeraa taha o Waikato
T/A cross to that side gen Waikato
‘[She] crossed to the other side of the Waikato River’ (W, 197)
(1509) Rere tika te manu i runga o te rori
fly right the bird at top gen the road
‘The bird flew across the road’
(1510) Ko ngaa tamariki e peke·peke ana i te
top. the(pl) children T/A jump·dup T/A DO the
tuuporo
log
‘The children were jumping over the log’

(‘At the log’ is another possible interpretation of the last phrase.)

2.1.1.5.29 Superior-contact long object (right angles to length)

It does not seem possible to distinguish this from the ‘direction of length’ case, except by
specifying the start and finishing points. Thus the following could be in either direction:
(1511) I ngaoki atu te ngaro i runga i te
T/A crawl away the fly at top at the
tuuporo
log
‘The fly crawled over the log’

The addition of whiti ‘cross’ does not assist, as this can also be in either direction. To
specify the ends, it is necessary to know the location of the speaker. However, there are
cases where whiti is appropriate, eg.
(1512) I whakawhiti ia i taau ara
T/A cross IIIsg DO sggenIIsg path
‘He crossed your path’

2.1.1.5.29a Surface long object

This is not distinct from superior contact, see. 2.1.1.5.29. A rather different example is:
(1513) Ka haere tonu i te huarahi o Tupeteka
T/A move still at the path gen Tupeteka
 Maori 334

‘[She] fled along the path taken by Tupeteka’ (W, 197)

2.1.1.6 Location in time

2.1.1.6.1 General
In general terms, location in time is marked by the same prepositions as are used for
spatial location. Thus i is the commonest marker for past temporal locatives, kei, for
present and sometimes future temporal locatives, but a is the commonest marker for
future temporal location. However, noo can also be used under certain circumstances to
mark past temporal location, and hei sometimes marks future temporal location. The use
of hei is dialectally variable. Some dialects use it in place of a, while others use it before
forms which historically had a, but which are now thought of as one word, eg. aapoopoo
‘tomorrow’, aawhea ‘when (fut)’. The local nouns mua ‘front, before’, muri ‘behind,
after’ are also used in temporal expressions, and note that the second glosses indicate
their appropriate use in most temporal contexts. If a sentence contains a temporal
adverbial, it is normal to find the T/A marker is a relative tense marker, rather than an
absolute T/A marker, particularly if the adverbial precedes.
Traditionally in Maori, time was measured by sun, moon and stars. Since the arrival of
Europeans, Western-style time measurement has become the norm, but involves
borrowed vocabulary. Traditional time measurements are no longer well known to Maori
speakers. There is a recent move to replace the borrowings for months by the traditional
Maori names, but this is meeting a certain amount of resistance amongst the general
Maori-speaking population.

2.1.1.6.1.1 Time of day

Past time of day phrases are most commonly marked with i, eg.
(1514) I te whitu karaka i te poo, ka haere
at(pt) the 7 clock at the night T/A move
a Hata maa
pers Hata and others
‘At 7pm, Hata and the others left’

However, noo can also be used:

(1515) Noo te waru karaka i te poo, ka tae ki
actgen the 8 clock at the night T/A arrive to
Te Kao
Te Kao
‘At 8pm, [they] reached Te Kao’

Future time of day is commonly marked with a, eg.

(1516) A te waru karaka, ka tae mai raatou
at(fut) the 8 clock T/A arrive hither IIIpl
‘They will arrive at 8 o’clock’
 Morphology 335

These same prepositions are used both in non-verbal predicates and adnominally:
(1517) I te rua karaka te hui
at(pt) the 2 clock the meeting
‘The meeting was at 2 o’clock’
(1518) Te hui a te rua karaka, hei/ko te whare
the meeting at(fut) the 2 clock at(fut) the house
waananga
learning
‘The two o’clock meeting will be at the University’

The past time marker, i, can co-occur only with past time adverbials. Fronted future time
adverbials can be followed by verb phrases with empty T/A slots, but ai is required post-
verbally, eg.
(1519) A te whitu karaka, haere mai ai ia
at(fut) the 7 clock move hither part. IIIsg
‘He will come at 7 o’clock’

If kua is used following a time adverbial, futurity is implied, even if the temporal location
marker is i. If the temporal location marker is a, the use of kua gives a sense of certainty,
eg.
(1520) A te whitu karaka, kua haere mai ia
at(fut) the 7 clock T/A move hither IIIsg
‘He will definitely leave at 7 o’clock’

A sentence like:
(1521) I te whitu karaka, kua haere mai ia
at the 7 clock T/A move hither IIIsg
‘He will leave at 7 o’clock’

can apparently be used to check on the correctness of the information. However, if kua, is
used with a past-time adverb following, a perfective reading is obtained.
When the T/A marker is ka, if the adverbial precedes, ka takes its time reference from
the temporal location marker. However, if ka precedes the time phrase, then my
consultant selected a non-past reading regardless of the locative preposition:
(1522) Ka haere mai ia i/a te torongi·tanga o
T/A move hither IIIsg at the sink·nom gen
te raa
the sun
‘He will come at sunset’

However, this appears to be the effect of decontextualization. In narratives, for instance,

ka+i-time adverbial can be given a past interpretation, eg.
(1523) Ka hapuu anoo hoki ia i muri iho i
T/A conceive again also IIIsg at behind down at
a Koopako
 Maori 336

pers Kopako
‘She conceived again after Kopako’ (H, 5)

(This is not a time-of-day example, of course, but the point transfers.)

E…ana is compatible with time of day phrases, but with future a-marked phrases, the
adverbial has to follow, eg.
(1524) E kai ana ia a te whitu karaka
T/A eat T/A IIIsg at(fut) the 7 clock
‘He will be eating at 7 o’clock’

2.1.1.6.1.2 Period of day

The same prepositions are used as for time of day. Examples are:
(1525) I te poo ka haere
at(pt) the night T/A move
‘At night, they set out’
(1526) Naa, noo te tuuruaweepoo, ka piki a
then actgen the middle-night T/A climb pers
Tutaanekai raaua ko toona hoa ko Tiki ki
Tutanekai IIIdl spec sggenIIIsg friend spec Tiki to
too raaua atamira
sggenIIIdl platform
‘Now, in the middle of the night, Tutanekai and his friend Tiki climbed up onto their
platform’ (H, 7)
(1527) …kia mau·ria ai koorua ki te ngahere a
subj take·pass. part. IIdl to the bush at(fut)
te ata
the morning
‘…so that you can be taken to the bush in the morning’ (TR2, 41)
(1528) I te ata o te Tuurei te hui
at(pt) the morning gen the Tuesday the meeting
‘The meeting was on Tuesday morning’

Attributively, the possessive o may replace the locative preposition:

(1529) Tino kino ngaa pookaakaa i/o te ata
very bad the(pl) squall at/gen the morning
‘The morning squalls were awful’

An example of kei as the preposition marking future period of day:

(1530) Kei te ata ka eke mai ka patu
at(fut) the morning T/A arrive hither T/A kill
‘In the morning they will arrive and attack’ (W, 198)

The comments on T/A restrictions and interpretations for time of day are also valid here
(see 2.1.1.6.1.1).
 Morphology 337

2.1.1.6.1.3 Day of the week

The same prepositions are used as for time and period of day. Examples are:
(1531) I te Mane, ka haere atu raatou ki
at(pt) the Monday T/A move away IIIpl to
Rotorua
Rotorua
‘On Monday, they went to Rotorua’
(1532) Noo teetahi Paraire, kaaore ia i tae
actgen a(sp) Friday neg IIIsg T/A arrive
mai ki te mahi
hither to the work
‘One Friday, she didn’t turn up at work’
(1533) A te Raatapu, ka hoki ia ki te kaainga
at(fut) the Sunday T/A return IIIsg to the home
‘She will return home on Sunday’

However, hei can sometimes be used, eg.

(1534) Hei aapoopoo ka haere mai maatou
at(fut) tomorrow T/A move hither Iplexcl
‘We’ll come tomorrow’

(Note that aa·poo·poo almost certainly contains the future locative preposition, a, ie. it is
literally ‘at(fut)·mght·night’.)
(1535) I teeraa Tuurei te hui
at(pt) that Tuesday the meeting
‘The meeting was last Tuesday’
(1536) Hei Rotorua te hui a te Raahoroi
at(fut) Rotorua the meeting at(fut) the Saturday
‘Saturday’s meeting is in Rotorua’
(1537) Mea whakakore taku whakaritenga i te
thing cancel sggenIsg appointment at(pt) the
Taite raa
Thursday dist
‘My appointment on Thursday had to be cancelled’

The comments on T/A restrictions and interpretations for time of day (see 2.1.1.6.1.1)
also apply here.

2.1.1.6.1.4 Month of the year

The marking is the same as for the time periods already discussed. Examples are:
(1538) I Hoongongoi, ka haere atu ia ki Tonga
at(pt) July T/A move away IIIsg to Tonga
‘In July, she went to Tonga’
 Maori 338

(1539) Noo te marama o Mahuru ka mate ia

actgen the month gen September T/A die IIIsg
‘She died in the month of September’
(1540) A Hakihea tana hoki·nga mai
at(fut) December sggenIIIsg return·nom hither
‘She comes back in December’

Dates follow the pattern prep+te+No.+o+month, eg. a te 28 o Pipiri ‘on the 28th June
(coming)’.
(1541) I Haratua, te tuatahi o ngaa hararei
at(pt) May the first gen the(pl) holiday
‘The first holiday was in May’
(1542) A Haratua taku huri·tau
at(fut) May sggenIsg turn·year
‘My birthday is/will be in May’
(1543) Te hui a Whiringa-aa-nuku ko te mea
the meeting at(fut) October top. the thing
mutu·nga moo teenei tau
finish·nom intgen this year
‘The meeting in October will be the last for this year’

Most of the comments on T/A markers in 2.1.1.6.1.1 apply here. However, following a
past-time month adverbial, kua and i were both rejected:
(1544) I te Hoongongoi raa ka/*kua/*i haere te
at(pt) the July dist T/A move the
whaanau o ngaa Jones
family gen the(pl) Jones
‘The Jones family left in July’

2.1.1.6.1.5 Year
The marking is the same as for the time periods already discussed. It should perhaps be
noted that English forms for years are extremely common in spoken Maori, even with
many older speakers, ie. it is not uncommon to hear “I nineteen fifty, ka…”. The
equivalent Maori form is “I (te tau) tahi mano iwa rau rima tekau, ka…” (lit. ‘at (the
year) one thousand nine hundred five tens’). Examples are:
(1545) I haere mai ia ki Aotearoa i te tau
T/A move hither IIIsg to N.Z. at the year
1950
1950
‘She came to N.Z. in 1950’
(1546) Noo te tau 1950 taku kite·nga mutu·nga
actgen the year 1950 sggenIsg see·nom finish·nom
i a ia
DO pers IIIsg
 Morphology 339

‘The last time I saw her was in 1950’

(1547) A te tau 1995, ka haere ia ki Amerika
at(fut) the year 1995 T/A move IIIsg to America
‘She is going to America in 19950’
(1548) Noo te kore·nga mahi i te tau 1972, ko
actgen the neg·nom work at the year 1972 top.
ngaa raruraru maha i puta
the(pl) trouble many T/A appear
‘The strike in 1972 caused many problems’
(1549) Kei te tau 1990, ka tuhi ahau i teetahi
at(pres) the year 1990 T/A write Isg DO a(sp)
pukapuka
book
‘In 1990, I am writing a book’

The T/A restrictions with years do not appear to be quite the same as with other
previously discussed time periods. Kua, ka, e…ana, and i can all co-occur with past-time
year adverbials, but if the adverbial precedes and the T/A marker is i, the particle ai is
required following the verb, eg.
(1550) I te tau 1976, i whaanau ai ia
at(pt) the year 1976 T/A be born part. IIIsg
‘She was born in 1976’

With a preceding i-marked adverbial, kua gives a perfective reading, not a future reading
(cf. 2.1.1.6.1.1). With future year adverbials, if the adverbial is initial, ka, kua and e…ana
are all possible. However, if the adverbial is final, kua is not possible in the T/A slot

2.1.1.6.1.6 Festivals
The same pattern also applies to festivals:
(1551) Kei te haere koe ki hea a te Aranga?
T/A move IIsg to where at(fut) the Easter
‘Where are you going at Easter?’
(1552) Ko tana huri·tau a te Kirihimete
top. sggenIIIsg turn·year at(fut) the Christmas
‘Her birthday is at Christmas’
(1553) He mea tino rahi te hui a te
cls thing very big the meeting at(fut) the
Kirihimete
Christmas
‘The meeting at Christmas will be a big one’
(1554) Noo te Kirihimete raa ka hoki mai ia
actgen the Christmas dist T/A return hither IIIsg
ki Te Kao
to Te Kao
 Maori 340

‘She returned to Te Kao at Christmas’

With future festivals, ka, kua and e…ana are possible T/A markers. In the past time
festival example (1554), kua is not possible if the adverbial precedes, and i in the T/A slot
requires the addition of ai post-verbally. E…ana is possible with suitable verbs. If the
adverbial follows, the preposition will be i, not noo, and all the T/A markers mentioned
above are possible.

2.1.1.6.1.7 Seasons
The same markers are used for seasons, eg.
(1555) I te ngahuru i hauhake·tia e raatou ngaa
at(pt) the autumn T/A harvest·pass. by IIIpl the(pl)
riiwai
potato
‘In the autumn, they harvested the potatoes’
(1556) A/I te koanga te waa pai moo te
at(fut)/at(pt) the Spring the time good intgen the
tirotiro i teeraa waahi
look DO that place
‘The best time to see that area is in Spring’
(1557) Ko ngaa hau o te hootoke he makariri
top. the(pl) wind gen the winter cls cold
‘The winds in winter are cold’
(1558) Ka hoki mai ia ki te kaainga a te
T/A return hither IIIsg to the home at(fut) the
raumati
summer
‘She will return home in the summer’

With future seasons, ka and kua are both possible, but there was some doubt about
e…ana. In past-time contexts, ka, kua, e…ana, and i are all possible, and kua has a
perfective reading.

2.1.1.6.2 Frequentative
Frequentatives are frequently indicated by using a plural determiner with the noun.
However, this cannot apply to those time expressions which are used without
determiners, in which case the frequentative is not overtly marked in the adverbial, but is
indicated through some post-verbal particle, such as tonu ‘still’, eg.
(1559) I ngaa ata, haere ai ngaa kaimahi… ki
at the(pl) morning move T/A the(pl) worker to
te ngahere
the bush
‘In the mornings, the workers go…to the bush’ (TR2, 40)
 Morphology 341

(1560) I eetahi tau, he maha ngaa kooura;

at(pt) some(sp,pl) year cls many the(pl) crayfish
i eetahi atu tau, he ruarua noa
at(pt) some(sp,pl) away year cls few indeed
‘Some years there are plenty of crayfish; other years hardly any’
(1561) Kore rawa ia e kai hikareti i ngaa
neg intens IIIsg T/A eat cigarette at the(pl)
ahiahi
evening
‘He never smokes in the evenings’
(1562) I Hanuere ka whakataa tonu maatou
at January T/A pause still Iplexcl
‘We always take our holidays in January’
(1563) Ko aa raatou hui o ngaa Mane tino roroa
top. plgenIIpl meeting gen the(pl) Monday very long
‘Their meetings on Mondays are always very long’

The T/A restrictions in such sentences depend on whether or not the frequentative is
marked with a plural. If so, all T/A markers are possible, although ka is much more
probable if the adverbial precedes, and i if the adverbial follows, eg.
(1564) I ngaa ata ka hia·moemoe tonu ia
at the(pl) morning T/A desire-sleep still IIIsg
‘She always feels sleepy in the mornings’

(Kua or e…ana are also possible here.) Compare:

(1565) I hia·moemoe tonu ia i ngaa ata
T/A desire-sleep still IIIsg at the(pl) morning
‘She was always sleepy in the mornings’

(Kua or e…ana here give non-past readings.)

However, if the frequentative is not marked by a plural, i or kua do not give a
frequentative reading, eg. in (1562) above. E…ana is possible in this sentence if tonu
does not appear, and can still have a frequentative reading.

2.1.1.6.3 Punctual-future
The futurity is often marked by a ‘at(fut)’ but sometimes i is used if there is a T/A marker
which marks futurity, eg.
(1566) Ko te hui a te rua haora
top. the meeting at(fut) the 2 hour
‘The meeting is in 2 hours’
(1567) Ko too taaua tuutaki·tanga a te rua haora
top. sggenIdlincl meet·nom at(fut) the 2 hour
ka tautohetohe
T/A quarrelsome
 Maori 342

‘Our meeting in two hours will be quarrelsome’

(1568) Ka hoki mai au i te rua haora
T/A return hither Isg at the 2 hour
‘I’ll be back in 2 hours’

The only other T/A marker possible in this context is kua, which adds certainty in an
example like the last.

2.1.1.6.4 Punctual-past
In punctual-past expressions, the pastness is expressed by noa atu raa ‘extend away
there’, and by the past locative preposition if appropriate, and the time phrase takes no
article, eg.
(1569) E rua haora noa atu raa, i konei ahau
num 2 hour extend away dist at here Isg
‘I was here two hours ago’
(1570) Rua haora noa atu raa te hui
2 hour extend away dist the meeting
‘The meeting was two hours ago’
(1571) Tino kino te tautohetohe o taa taatou hui
very bad the quarrel gen sggenIplincl meeting
rua haora noa atu raa
2 hour extend away dist
‘Our meeting two hours ago was very quarrelsome’
(1572) I riingi mai ia i te waenganui poo
T/A ring hither IIIsg at the middle night
‘She phoned at midnight’

If the adverbial is final, ka and e…ana are possible, but kua is not. If the adverbial is
initial, ka, kua, and e…ana are possible, but i requires the addition of ai post-verbally.

2.1.1.6.5 Duration
Duration phrases are introduced by moo, eg.
(1573) I reira ahau e noho ana moo te rua tau
at there Isg T/A stay T/A intgen the 2 year
‘I lived there for two years’
(1574) Ka noho au moo te wiki
T/A stay Isg intgen the week
‘I’ll stay for a week’
(1575) Moo te kotahi tau anake te karahipi
intgen the one year only the scholarship
‘The scholarship is for just one year’
(1576) Noo te karahipi moo te kotahi tau anake
 Morphology 343

actgen the scholarship intgen the one year only

i haere ai raatou ki Amerika
T/A move part. IIIpl to America
‘A scholarship for a year has taken them to America’

If the adverbial is final, there are no restrictions on T/A markers, but kua gives a future
reading. If the adverbial precedes, kua was rejected by my consultant, and i requires ai,
eg.
(1577) Moo te wiki i noho ai ahau
intgen the week T/A stay part. Isg
‘I stayed for a week’

2.1.1.6.6 Anterior-duration-past
Sometimes it is sufficient to add noa ‘indeed’ to the time phrase to indicate anterior-
duration-past, but on other occasions, noo…raa anoo is required. No generalization is
apparent from the data available. Some examples are:
(1578) Haere tonu te hui, aa, waenganui poo noa
move still the meeting and middle night indeed
‘The meeting went on until midnight’

Such phrases cannot be either attributive or predicative.

(1579) Noo te Mane raa anoo kaatahi anoo
actgen the Monday dist again then again
tana reta ka tae mai
sggenIIIsg letter T/A arrive hither
‘He didn’t get the letter until Monday’
(1580) Tino kore ia i tae·a atu e ahau
very neg IIIsg T/A reach·pass. away by Isg
noo muri i te awateatanga raa anoo
actgen behind at the midday dist again
‘I didn’t get hold of her until after midday’

Such phrases can only be attributive via relative clauses, and can apparently not be
predicative. No other tense markers are possible in the last two examples, but e…ana is
compatible with contexts like (1578).

2.1.1.6.7 Anterior-duration-future
As with anterior-duration-past, sometimes noa is sufficient, and sometimes a…raa anoo
is required.
(1581) Ka haere tonu te hui, aa waenganui poo
T/A move still the meeting and middle night
noa
extend
‘The meeting will go on until midnight’
 Maori 344

(1582) A te Mane raa anoo, kaatahi anoo

at(fut) the Monday dist again then again
tana reta ka tae mai
sggenIIIsg letter T/A arrive hither
‘He won’t get my letter until Monday’
(1583) E kore ahau e kite i a ia anoo,
T/A neg Isg T/A see DO pers IIIsg again
a Oketopa raa anoo
at(fut) October dist again
‘I won’t see her again until October’

Attributive and predicative examples do not occur, and apart from e…ana in contexts like
the first, other T/A markers cannot occur.

2.1.1.6.8 Posterior-duration-past
Mai i…raa is used to express posterior duration past in all contexts, eg.
(1584) Mai i te Kirihimete raa, kaahore anoo ahau
hither at the Christmas dist neg again Isg
i kite noa i a ia
T/A see indeed DO pers IIIsg
‘I haven’t seen her since Christmas’
(1585) Tino roroa ngaa hui mai i te Kirihimete
very long the(pl) meeting hither at the Christmas
raa
dist
‘The meetings since Christmas have been very long’

Such time phrases cannot be predicative. Note that the negations in this type of example
require the form kaahore anoo, and the verb is normally supported by noa or noa atu.
(1586) Mai i te Mane raa, i taakaro tiitiitorea
hither at the Monday dist T/A play stick-game
ngaa tamariki
the(pl) children
‘From that Monday, the children played stick games’

In the negative main clause, there is no choice of T/A marker. In particular, both kua and
ka were rejected. In the positive main clause, only e…ana is a possible alternative to i.
Kua was rejected.

2.1.1.6.9 Posterior-duration-future
Mai a…(nei) is used to express posterior duration future, eg.
(1587) Mai a te waenganui poo, ka kake te
hither at(fut) the middle night T/A increase the
utu o te hinu
 Morphology 345

price gen the petrol

‘The petrol price increase will take effect from midnight tonight’

If the adverbial is in final position, nei is needed after poo in (1587).

(1588) Mai a teeraa tau ko Pou te heamana
hither at(fut) that year top. Pou the chairman
‘Pou will be the chairman from next year’
(1589) Ko Pou te heamana mai a teeraa tau
top. Pou the chairman hither at(fut) that year
‘The chairman from next year on is Pou’

Such phrases cannot be predicative. Kua and e…ana are possible in place of ka.

2.1.1.6.10 Anterior-general
Anterior general phrases are expressed by mua (noa) atu i…, which is sometimes
preceded by noo or i, but in informal speech these prepositions are frequently omitted, eg.
(1590) I konei ahau mua atu i te Mane raa
at here Isg before away at the Monday dist
‘I was here prior to Monday’
(1591) I/Noo mua atu i te Kirihimete taa raatou
at/actgen before away at the Christmas sggenIIIpl
whakaritenga
appointment
‘Their appointment was before Christmas’
(1592) I haere ahau ki reira i mua atu i te
T/A move Isg to there at before away at the
waenganui poo
middle night
‘I went there before midnight’
(1593) Kua hoohaa ahau ki ngaa huihuinga o mua
T/A fed up Isg to the(pl) meeting gen before
atu i te Mane raa
away at the Monday dist
‘I got fed up with the meetings prior to Monday’
(1594) Teenaa pea te aituaa noo mua atu i
that perhaps the accident actgen before away at
te toru karaka
the 3 clock
‘The accident must have been before 3pm’

Kua cannot occur as the T/A marker in sentences which have anterior-general adverbials,
but e…ana or ka is possible in examples like (1592).
 Maori 346

2.1.1.6.11 Pos terior-general

(Hei/a) muri o/i…(nei) is used for posterior-general. Hei appears to be possible only if the
adverbial is in initial position, but this may vary from dialect to dialect, eg.
(1595) Ko konei tonu au a muri o te
at(fut) here still Isg at(fut) behind gen the
Mane
Monday
‘I’ll be here after Monday’
(1596) A muri i te waenganui poo nei, kua
at(fut) behind at the middle night proxI T/A
tae mai raatou
arrive hither IIIpl
‘They will arrive after midnight’
(1597) Hei/A muri o te Kirihimete taa raatou
at(fut) behind gen the Christmas sggenIIIpl
whakaritenga
appointment
‘Their appointment is after Christmas’
(1598) Ko te marae aa raatou hui a muri
at(fut) the marae plgenIIIpl meeting at(fut) behind
o te Mane
gen the Monday
‘Their meetings after Monday will be at the marae’

Ka and e…ana are possible alternatives to kua in (1596).

2.1.1.6.12 Point in period past

Noo/I+roto i+a clause to express ‘last/past’ if necessary can be used for point in period
past, eg.
(1599) I konei ia i roto i ngaa rua haora kua
at here IIIsg at inside at the(pl) 2 hour T/A
pahure ake nei
pass away proxI
‘He’s been here within the last two hours’
(1600) Ko teenaa karanga noo roto i te haora kua
top. that call actgen inside at the hour T/A
pahure ake nei
pass away proxI
‘That call was within the last hour’
(1601) He whakamaamaa·tanga ngaa hui i roto i
cls enlighten·nom the(pl) meeting at inside at
te marama kua pahure ake nei
the month T/A pass away proxI
 Morphology 347

‘The meetings within the last month have been useful’

(1602) I tae mai raatou i roto i te haora
T/A arrive hither IIIpl at inside at the hour
kua pahure ake nei
T/A pass away proxI
‘They arrived within the last hour’

compare:
(1603) I roto i te haora e rima tangi·nga o
at inside at the hour num 5 ring·nom gen
taku whoounu
sggenIsg phone
‘I’ve had five phone calls in an hour’

There do not appear to be any restrictions on T/A markers, apart from those imposed by
the past time context, ie. the non-past absolute marker kei te is not compatible.

2.1.1.6.13 Point in period future

This is usually expressed (like point in period past) with i roto and, most usually, a clause
to indicate futurity. However, sometimes phrases involving mua can be used to indicate
futurity, eg.
(1604) Ka tae mai ia i roto i ngaa rua
T/A arrive hither IIIsg at inside at the(pl) 2
haora kei te haere mai
hour T/A come hither
‘He’ll arrive within the next two hours’
(1605) Ka tuu te hui i roto i te wiki e
T/A stand the meeting at inside at the week T/A
haere mai nei
move hither proxI
‘The meeting will be within the next week’
(1606) Ko ngaa karanga o te haora e tuu mai
top. the(pl) call gen the hour T/A stand hither
nei e kore e whakautu·a
proxI T/A neg T/A answer·pass.
‘The calls within the next hour won’t be answered’
(1607) E toru aku hui i roto i ngaa raa
num 3 plgenIsg meeting at inside at the(pl) day
e rima kei te haere mai nei
num 5 T/A move hither proxI
‘I’ve got three meetings in the next 5 days’

In this last sentence, A mua ake nei ‘at before away here’ could be used sentence initially,
in place of the final clause, and in the first example, ki mua ‘to before’ could replace the
clause (and remain in final position).
 Maori 348

Kua and i are not possible with such adverbials, but e…ana is possible with suitable
verbs.

2.1.1.7 Double case marking

Maori does not display double case marking.

2.1.1.8 Number in nouns

2.1.1.8.1 Number-marking in nouns

There are a few nouns in Maori which show a singular-plural number distinction. They
are all terms referring to humans, and more particularly expressing close familial ties.
The forms involved can be found in 2.1.1.8.6. However, the majority of nouns in Maori
do not change form to show number.

2.1.1.8.2 Obligatoriness of number marking

The group of nouns with a distinct plural obligatorily have the appropriate number form.

2.1.1.8.3 Number marking other than in nouns

Number is regularly marked in the determiner in Maori. All determiners other than he
and generic te show a singular-plural distinction. Since a determiner is obligatory with
non-local nouns in almost all circumstances, number is marked in the noun-phrase in a
very high proportion of instances in Maori. The forms of the determiners are given in
1.2.5.2.4 (with some comments on use in singular-plural contexts). It should also be
noted that te forms complex determiners with demonstratives and possessives which are
also marked for plural by the deletion of the initial t-, and this is also true for the
interrogative (t)eewhea ‘which’.

2.1.1.8.4 Collective υs. distributive plural

No such distinction is made in either nouns or determiners in Maori. However, some
adjectives show a special form of reduplication for distributive plurals, see 2.1.4.3.1.

2.1.1.8.5 Collective nouns

Maori does not have a distinct class of collective nouns. However, there are reduplicated
forms which sometimes have a collective type of function. Such reduplicated forms can
normally occur with either singular or plural determiners. Thus there are sets of NPs like
(from paru ‘dirt, dirty’) te paru, ngaa paru, te paruparu, ngaa paruparu. The
reduplicated forms can be used in contexts like the following:
(1608) Te paru·paru o teeraa tangata!
the dirt·dup gen that man
 Morphology 349

‘What a dirty man!’

(1609) Kua hoohaa ahau ki ngaa paru·paru e puta
T/A tired Isg to the(pl) dirt·dup T/A appear
ana ki te moana
T/A to the sea
‘I hate the pollution of the sea’

Compare:
(1610) Pango tonu tana kanohi i te paru
black indeed sggenIIIsg face from the dirt
‘His face is all black with dirt’
(1611) Ko ngaa paru e rere ana ki waho o te
top. the(pl) dirt T/A flow T/A to outside gen the
moana
sea
‘The sewage is flowing out to sea’

Because reduplicated forms are often lexicalized, it is not possible to make

generalizations about their use.

2.1.1.8.6 Realizations of number distinctions

The complete list of nouns showing number distinction is:
Singular Plural Gloss
matua maatua parent
tangata taangata man
teina teeina younger sibling (same sex)
tuahine tuaahine sister (of male)
tuakana tuaakana older sibling (same sex)
tupuna tuupuna ancestor
wahine waahine woman

and for some,

whaea whaaea mother

All of these form the plural by reduplication of the antepenultimate vowel. In addition
tamaiti tamariki child

has an irregular singular-plural relationship.

 Maori 350

2.1.1.8.7 Foreign words

Foreign words are integrated into the Maori system. English is the only major source of
borrowings, and such words are borrowed in their stem form, and take whatever
determiner would be appropriate in Maori, eg.
te kau ‘the cow’ ngaa kau ‘the cows’
te raihi ‘the rice’ ngaa raihi ‘the rice’

I have not been able to trace any clearcut modern borrowings from other Polynesian
languages. It appears that such forms can be reduplicated, although examples are not
readily found, partly because of the types of words borrowed. Note putiputi ‘flower’
(from ‘pretty-pretty’) and pukapuka ‘book’, which Williams suggests in his Dictionary is
not purely a borrowing from English.

2.1.1.9 Noun classes/genders

Nouns are not divided into genders in Maori, although there is a distinction between local
and non-local nouns (see 1.16.1). These two classes are not morphologically distinct, but
they behave rather differently syntactically. Proper names also constitute a separate class
in syntactic terms (see 1.16.1). Maori does not have classifiers, although the prefix toko-
used with numbers for people bears some resemblance to one (see 2.1.6.3). No loan
words except place names belong to the class of local nouns.

2.1.1.10 The marking of definiteness in NPs

The standard view is that definiteness is marked in NPs in Maori. However, the situation
seems to me much less clearcut than previous descriptions suggest (see eg. Williams,
1862, 19ff; Biggs, 1969, 48; Hohepa, 1967, 23).

2.1.1.10.1 Locus of definiteness marking

The standard view is that definiteness is marked in determiners. in Maori; te, ngaa, and
forms including t- as a prefix are usually described as definite (or definitives), while he is
described as indefinite. Clark points out (1976, 47) that the notion of definiteness is not
the same as in English. He says the articles described as definite are “used whenever the
speaker has a particular individual in mind, whether or not the addressee is expected to be
able to identify the individual”. He suggests that the distinction is better described as an
opposition between specific and non-specific To Clark’s comments it is necessary to add
that in narrative texts, whether the narrator has a particular object in mind, rather than
whether the protagonists in the story have, is often crucial (cf. Hohepa’s comments on the
use of directional particles and the narrator (Hohepa, 1981, 45ff)). However, even with
these refinements, it is not clear that there is a consistent distinction in Maori. The
problem is discussed in the next paragraphs.
 Morphology 351

2.1.1.10.2 Obligatoriness of definiteness marking

Two major noun classes must be distinguished here. Under almost all circumstances local
nouns have no determiner, and are not marked for definiteness. However, as sentence
subject, they take the personal article, a, as do personal nouns when they are subjects and
following prepositions ending in -i. Since no change in definiteness occurs from the
omission of a, it seems likely that this determiner does not form part of the system of
definiteness marking. Thus it seems fair to conclude that local nouns and personals are
not marked for definiteness, though many personal names and place names include the
article te.
With general nouns, the situation is far less clear. Firstly, it must be pointed out that
under almost all circumstances, general nouns in Maori are preceded by a determiner.
Lists provide occasional exceptions, eg.
(1612) …pau katoa i a ia te kai—kaakahu
exhaust all cause pers IIIsg the food clothes
atu; meremere atu; parawai atu; aha atu…
away mere away fine mat away what away
‘He finished up all the food—clothes, meres, fine mats and all…’ (TP, 90)

The other major exception involves the process of object incorporation, whereby
indefinite direct objects become post-verbal modifiers, eg.:
(1613) …kia haere i roto o te iwi o Tupeteka
subj move at inside gen the tribe gen Tupeteka
e whawhati rautao ana e kohi koowhatu
T/A gather fern T/A T/A gather stone
ana moo te haakari
T/A intgen the feast
‘…to go amongst the people of Tupeteka who were gathering ferns and stones for the
feast’ (W, 198)

Secondly, he which is usually described as the indefinite article is extremely restricted in

syntactic position: it cannot follow any preposition (except me in the sense ‘like’) which
excludes it from non-subject NPs, and it cannot occur in the subject phrase of transitive
sentences (see eg. Chung, 1978, 73–4). This means that he does not provide a contrast
with definite determiners in many syntactic environments, although object incorporation
provides a means of marking the distinction in DOs.
There are thus two problems to consider: firstly whether the definite/indefinite
distinction is supported in intransitive subjects, and secondly, whether the contrast is
made in other ways in the types of NP from which he is excluded. These are discussed in
turn below.
There are three determiners which need to be considered in intransitive subjects: he, te
(ngaa) and (t)eetahi, the last (constructed from te+tahi ‘one’) often glossed as ‘a certain’.
The following textual examples illustrate the problems:
(1614) Kai·nga i waho, ka whakarite·a he tohunga
eat·pass. at outside T/A appoint·pass. a tohunga
 Maori 352

hei whaangai moo Tuu-whakairi-ora

for(fut) feed intgen Tu-whakairi-ora
‘They ate out of doors, and a tohunga was appointed to feed Tu-whakairi-ora’ (TWh, 20)

Compare:
(1615) Ka whaangai·tia ringa·ringa·tia hoki a
T/A feed·pass. hand·dup·pass. also pers
Ruataupare e teetahi tohunga
Ruataupare by a(sp) tohunga
‘Ruataupare was also fed by hand by a tohunga’ (TWh, 20)

It is not clear that the first is any less definite than the second, although teetahi could
have been used in the first. The following example has what appears to be a maximally
indefinite NP with teetahi in the subject of an intransitive sentence:
(1616) Ki te kite·a noa·tia e teetahi tangata teetahi
to the see·pass. indeed·pass. by a man a
mea…
thing
‘If somebody just sees something…’ (P, 3)

While examples of this kind are not common, they appear to undermine the
definite/indefinite distinction. Perhaps more importantly, it seems to me that the use of
the labels ‘definite’ and ‘indefinite’ obscures the crucial factor in determining the choice
between he and teetahi in those places where choice can occur: he is used when the type
of object is crucial, and teetahi is used when the number of individuals present is
significant (Recall that he is not marked for singular/plural, and note that teetahi contains
tahi ‘one’.) Thus in (1614), it is important that a tohunga (and not some other class of
person) is the feeder. Clearly, it was one specific (though not identified to narrator or
hearer) individual, which is why teetahi (and not te) was used in the parallel instance in
(1615). The relevance of this can be seen more clearly in examples like the following
from Hinemoa. The story reports the conversation in which Tutanekai and Hinemoa
arrange to meet Tutanekai says:
(1617) “E tangi he puu i ngaa poo katoa, ko
T/A sound a flute at the(pl) night all top.
ahau teenaa, hoe mai”
Isg that row hither
“‘A flute sounds every night; it’s me; row to me’” (H, 7)

There is no doubt that the speaker has in mind a specific flute, and that the narrator and
hearer know of the flute. But what matters to the story is the type of object which is to be
Hinemoa’s guide. This is by no means an isolated example. This conclusion about the
function of he is the same sort of conclusion reached by Polinsky 1991, although the
terminology there is rather different.
Now it is necessary to consider whether a distinction between definite and indefinite
can be supported in environments from which he is excluded. Examples like the
following show teetahi in indefinite contexts: in the story of Maui’s search for his
 Morphology 353

parents, when Maui disguised as a bird is throwing berries onto the people below, the text
runs:
(1618) Ka mea teetahi tangata, “Naa te manu pea”
T/A say a man actgen the bird perhaps
‘One man said, “Perhaps the bird was responsible”’ (KM, 3)

There seems no reason to suppose that the individual is identified by the narrator.
There seems little doubt that definite and indefinite direct objects can be distinguished
by the process of object incorporation, eg.
(1619) …ka koorero a Toto, ki ana kootiro,
T/A speak pers Toto to plgenIIIsg daughter
mehemea e pai raaua ki te moe taane
if T/A good IIIdl to the sleep husband
‘…Toto asked his daughters if they wanted husbands’ (KH, 2)

Compare:
(1620) Ka moe a Hera i tana taane a
T/A sleep pers Hera DO sggenIIIsg man at(fut)
te Aaranga
the Easter
‘Hera is to marry her man at Easter’

However, it is not uncommon to find that te (and not object incorporation) is used.
Considering the following:
(1621) …me ngaa iwi o te ngahere, o ngaa
with the(pl) tribe gen the bush gen the(pl)
maunga, e mau mai ana i te hinu me
mountain T/A carry hither T/A DO the game with
eeraa atu kai
those other food
‘…and the tribes of the forest and mountains, who brought game and other foods’ (TWh,
21)

It is not difficult to see why object incorporation was not used: only one object at a time
can be incorporated. Nevertheless, the example shows that te in a DO is not necessarily
clearly definite. There is a good deal of similar evidence. Examples like this also make it
clear that it is not the case that (t)eetahi takes over the indefinite functions of he in those
environments where he is excluded. It has often been claimed in my hearing that teetahi
is abbreviated to te in such environments, and this statement appears in print in Foster,
1987, 36. Plainly, the effect of any such ‘abbreviation’ is the neutralization of any
distinction. But I do not believe that there is any sound evidence for regarding this as a
process of abbreviation. Examples like the one cited above could not arise from such a
process, since teetahi would be inappropriate. Rather, I believe, in such environments
teetahi is used when number is important, and te is the ‘default’ article, used without
necessary implication of definiteness when the specific senses of other articles are
inappropriate. This also happens in other syntactic environments. Some textual examples
 Maori 354

follow. The first one consists of two parallel sentences a few lines apart, one with teetahi
and one with te :
(1622) Ka tango ia i teetahi hua o taua raakau,
T/A pick IIIsg DO a(sp) fruit gen det aph tree
ka pangaa iho… Ka tango anoo taua kereruu
T/A throw down T/A take again det aph pigeon
i te hua raakau
DO the fruit tree
‘He took a berry from that tree and threw it down…That pigeon again picked a berry…’ (KM, 3)

It is clearly not the case that the two berries differ in definiteness here.
(1623) Ka tango te tangata raa i te koowhatu hei
T/A take the man dist DO the stone for
kuru i a ia
pelt DO pers IIIsg
‘That man grabbed a stone to pelt him with’ (KM, 3)

The te in bold accompanies a noun which again seems maximally indefinite.

(1624) …ka tuku·a e te paa te karere ki roto
T/A send·pass. by the pa the messenger to inside
o Hauraki
gen Hauraki
‘…the pa sent a messenger to Hauraki’ (W, 198)

Example (1624) shows a parallel use in subject position, where he might be expected if
the indefinite/definite distinction had validity in Maori, although if a tribe had just one
messenger, there is a possibility here that the messenger might be specific, though not
identified to the hearer.
While the distinction is perhaps better motivated for plural forms, there are still
instances where ngaa does not seem to be definite, eg.
(1625) Noo te tae·nga mai hoki o ngaa
actgen the arrive·nom hither also gen the(pl)
tauhou ki konei, ka arahi·na raatou e maatou
stranger to here T/A guide·pass. IIIpl by Iplexcl
ki te toka nei kia kite
to the rock proxI subj see
‘And when visitors come to our district we lead them to this spot so they can see it
themselves’ (TA, 8)
(1626) He kotahi te kuumara i roto i te rourou
cls one the kumara at inside at the food basket
maa ngaa taangata toko·rua
intgen the(pl) men pnum·2
‘There was one small round basket containing one kumara to two men’ (W, 198)

Eetahi can be definite, at least as a substantive, eg. (note that the dialectal variant eetehi
is used):
 Morphology 355

(1627) Ka puukai·a iho teetehi o ngaa kaakahu,

T/A place·pass. down a(sp) gen the(pl) clothes
ka hari·a eetehi
T/A carry·pass. some(sp,pl)
‘She left one of her cloaks, taking the others with her’ (W, 197)

It should perhaps also be pointed out that of the forms discussed here, only (t)eetahi is
possible substantivally, and the distinction between definite and indefinite cannot be
made in that environment
I thus conclude that there is considerable doubt as to whether definite vs. indefinite is
an appropriate oppostion for the description of the use of these articles in Maori, and even
if it is, in the majority of syntactic environments the distinction is not made. Even in
those where it can be made, it is not obligatory.

2.1.1.10.3–4 Definiteness marker in different spatial relationships

The determiner te can combine with the deictic particles nei, naa, raa to form complex
determiners, the plurals of which drop the initial t-. (Alternatively, and equivalently, the
deictic particles can be placed after the head noun, see 2.1.2.5.4 and 2.1.2.5.8.) These
forms are not obligatory with the definiteness marker, ie. the spatial relationship of an
object is not necessarily indicated. The spatial relationships involved are the same as for
demonstrative pronouns (see 2.1.2.5.1).

2.1.1.10.5 Definiteness with various noun types

2.1.1.10.5.1 Proper names

It has already been suggested that the personal article does not mark definiteness (see
2.1.1.10.2). However, it must be pointed out that many proper names in Maori contain the
determiner te, though by no means all do. However, as the discussion in 2.1.1.10.2
indicates, it cannot be concluded that such names contain an indication of definiteness.
(They may also contain the plural ngaa, eg. in tribal names with Ngai- or Ngati-
prefixes.)
In contrasting environments, proper names can be preceded by te, and other
determiners.
(1628) Ko te Hone e koorero nei, kaahore e roa
top. the John T/A talk proxI neg T/A tall
‘The John I’m talking about isn’t tall’
(1629) He koretake taaku Hone
cls useless sggenIsg John
‘My John is useless’

Under these circumstances, proper names with te are interesting: they co-occur with
possessives, but not with te: ie. taaku Te Aotakii ‘my Te Aotaki’ could be substituted for
taaku Hone in (1629), but Ko Te Aotakii (not *Ko te Te Aotakii) is appropriate in (1628).
 Maori 356

2.1.1.10.5.2 Abstract nouns

Since general nouns (including abstract nouns) must have a determiner, it follows that in
those environments where he is not possible, abstract nouns occur with te or teetahi, or
some other form based on te, or the corresponding plurals, eg.
(1630) Kaua e pooheehee he huarahi ngaawari
neg T/A think wrongly cls path easy
noa iho te huarahi ki te maatauranga
indeed down the path to the knowledge
‘Don’t kid yourself that the path to knowledge is an easy one’ (TR2, 166)
(1631) E rapu ana ia i teetahi rangimarie·tanga
T/A search T/A IIIsg DO a(sp) peaceful·nom
‘He is looking for peace of mind’

In contrastive environments, te is used, eg.

(1632) E kore e kite·a te rangimarie·tanga e
T/A neg T/A find·pass. the peaceful·nom T/A
rapu·a nei e ia
seek·pass. proxI by IIIsg
‘The peace of mind he’s looking for can’t be found’

2.1.1.10.6 Optionality of definiteness marking

In contrastive environments, the definiteness marker is required for both classes (but note
the comments on proper names containing Te). In non-contrastive environments, proper
names cannot take definiteness markers, but require either the personal article a or Ø in
the determiner slot, depending on the preposition (see 2.1.1.10.2). Abstract nouns can
occur with he in syntactic environments where he is generally possible. In all other
environments, they must have a determiner which is based on te, or the plural ngaa.

2.1.1.11 The marking of indefiniteness in NPs

It is not clear that the category indefinite is marked in Maori. The issues involved have
been discussed in 2.1.1.10.2, and the reader is referred to the discussion there.

2.1.1.11.1 Method of marking indefiniteness

If indefiniteness is marked, then it is marked with he or (t)eetahi or object incorporation.
The following set of contrasting sentences shows the potential marking:
(1633) I hoko mai ahau i teetahi pukapuka
T/A buy hither Isg DO a(sp) book
‘I bought a book’
(1634a) E hiahia pukapuka ana ahau
T/A want book T/A Isg
‘I want a book’ (but I don’t know what one)
(1634b) Ko taku hiahia he pukapuka
 Morphology 357

top. sggenIsg wish cls book

‘A book is what I want’

Thus a distinction can be made in some syntactic environments between specific and
non-specific indefiniteness.

2.1.1.11.2 Optionality of indefiniteness marking

See 2.1.1.10.2.

2.1.1.11.3–4 Indefiniteness marker in different spatial relationships

Neither of the potential indefiniteness markers, he or (t)eetahi, changes form depending
on the spatial relationship of the entity and the participants. My consultant judged it
impossible for he to co-occur with the deictics nei or naa, though it can co-occur with
raa, eg.
(1635) He pukapuka raa teenei hei ako i a koutou
cls book dist this for teach DO pers IIpl
‘This is a book that will teach you’

This could be used by a teacher holding a book up (hence teenei ‘near speaker’) and
showing it to a class, whose distance from the book is marked with raa. However, there
are no contrasts possible.
Examples with teetahi co-occurring with post-posed deictics were all rejected. The
plural eetahi, however, was accepted, but only in a partitive sense:
(1636) Eetahi pukapuka raa tino pai
some(sp,pl) book there very good
‘Some of those books are good’

2.1.1.11.5 Indefiniteness with various noun types

2.1.1.11.5.1 Non-singular nouns

There are only eight non-singular nouns in Maori. He can cooccur with them, eg.
(1637) Ka roa e rere ana, aa, ka noho ia
T/A long T/A fly T/A then T/A sit IIIsg
i teetahi raakau e noho raa he taangata i
at a(sp) tree T/A sit dist some people at
raro
under
‘He flew on for some time and then came to rest in a tree under which some people sat’
(KM, 3)

Teetahi, of course, cannot co-occur with non-singular nouns, but the corresponding plural
can, eg.
 Maori 358

(1638) Ka tae ki Whakatuu, kaatahi raatou ko

T/A arrive to Nelson then IIIpl spec
eetahi taangata o reira ka haere ki te
some(sp,pl) people gen there T/A move to the
moana ki te hii ika
sea to the catch fish
‘He got to Nelson, and then he and some people from there went to sea to catch fish’ (TP,
89)

Plural nouns do not occur as incorporated objects.

2.1.1.11.5.2 Mass nouns

It is not at all clear that mass nouns form a distinct class in Maori. Nouns which refer to
masses behave like nouns which refer to individual objects, although they are not
normally the heads of subjects of predicates like tini ‘multitude’, and some are rejected
with maha ‘many’. Thus it is normal to speak of ngaa raihi ‘the(pl) rice’ ngaa tote
‘the(pl) salt’, ngaa onepu ‘the(pl) sand’, ngaa wai ‘the(pl) water’, eg.
(1639) He maa ngaa onepu i te taha·taha o te
cls white the(pl) sand at the side·dup gen the
moana
sea
‘The sand beside the sea is white’
(1640) Kei hea ngaa tote?
at(pres) where the(pl) salt
‘Where is the salt?’
(1641) Ka haehae ngaa ringa ki te mataa kia
T/A slash the(pl) arm with the obsidian flake subj
heke iho ai te toto ki runga i a
run down part the blood to top at pers
Wairangi, kia tapu ai i ana toto, kia
Wairangi subj sacred part. from plgenIIIsg blood subj
kore e kai·nga
neg T/A eat·pass
‘[She] slashed her arms with obsidian flakes so that the blood would trickle down on Wairangi
and render him sacred that he might not be eaten’ (W, 198)

Such nouns are no different from non-mass nouns with respect to indefiniteness, eg.
(1642) “Tiki·na he wai moo·ku.” Ka haere tana
fetch·pass. a water intgen·Isg T/A move sggenIIIsg
taurekareka ki te tiki wai
servant to the fetch water
“‘Fetch me some water”. His servant went to fetch water’ (H, 8)

However, it appears that with (t)eetahi, containers of the substance are implied, because
these forms indicate countability.
 Morphology 359

2.1.1.12.1–2 Referential vs. non-referential indefiniteness

Pairs of examples like the following appear to support such a distinction:
(1643) I hoko mai ahau i teetahi pukapuka
T/A buy hither Isg DO a(sp) book
‘I bought a book’
(1644) Ko taku hiahia he pukapuka
top. sggenIsg wish a book
‘I want a book’ (but I don’t know which one)

However, because of the restrictions in the syntactic positions for he, such a distinction is
impossible in all but intransitive subject positions. See also the discussion in 2.1.1.10.2.

2.1.1.12.3–4 Referentiality with particular noun types

2.1.1.12.3.1 Non-singular nouns

In theory, since he can co-occur with either singular or plural nouns, the distinction could
be made. However, since most nouns are not marked for number except in the
determiner, he+noun is read as singular, and if plurality is significant, then eetahi is used
in non-referential as well as referential contexts, eg.
(1645) I hoko mai ahau i eetahi pukapuka
T/A buy hither Isg DO some(sp,pl) book
‘I bought some books’
(1646) E hiahia ana ahau ki eetahi pukapuka
T/A desire T/A Isg to some(sp,pl) book
‘I want some books’ (but I don’t know which ones)

Alternatively, object incorporation, which does not indicate number, is used for non-
referential instances:
(1647) E hiahia pukapuka ana ahau
T/A want book T/A Isg
‘I want books’ (more lit ‘I am book-wanting’)

Thus it is probably fair to conclude that the distinction is not made with non-singular
nouns.

2.1.1.13 Genericness in noun phrases

2.1.1.13.1 Marking of genericness

Genericness is often marked with the article te+the plural of the noun if there is a distinct
plural, eg.
(1648) Ko te utu moo te tamariki kotahi taara
 Maori 360

top. the price intgen the children one dollar

‘$1 is the price for children’

However, in the vast majority of instances, the noun has no distinct plural, so a generic
NP is not distinct from a definite singular NP (but see 2.1.1.10.2). However, various other
determiners also occur in generic contexts. The plural determiner ngaa can be so used,
eg.
(1649) Tino reka te/ngaa ika ki te/ngaa tori
very sweet the/the(pl) fish to the/the(pl) cat
‘Cats love fish’/‘The cat loves the fish’/The cats love the fishes’

He occurs in contexts like:

(1650) …te ngaru anoo he ngaru moana
the wave again a wave sea
‘…the waves were like sea waves’ (TP, 90)

and there are instances where te with the singular of a noun with a distinct plural also
appears generic:
(1651) Naa te wahine anake teeraa mahi
actgen the woman alone that work
‘That task is performed only by women’

2.1.1.13.2 Optionality of generic marking in NPs

If an NP is generic, it must have te, he or ngaa. NPs with other determiners do not have
generic readings. However, only te+plural noun is exclusively generic.

2.1.1.14 Relative importance of NPs

Maori has no means of marking the relative importance of NPs. However, it should be
borne in mind that in 1.13 and 1.14 it was suggested that animacy might be a factor in
determining constituent order in Maori.

2.1.2 Pronouns

2.1.2.1 Personal pronouns

2.1.2.1.1 Free pronouns

Free pronouns occur in Maori in all syntactic positions with the exception of singular
pronouns in all three persons following these prepositions: maa, naa, moo, noo, a, o. The
special forms which cliticize to these possessive prepositions are given in 2.1.2.4. In
general, the personal pronouns refer to humans, but on occasions they are extended to
 Morphology 361

other animates, but not normally to inanimates. However, ia ‘IIIsg’ has demonstrative-
type uses in which it refers, for instance, to preceding discourse.

2.1.2.1.1.1–2 Obligatoriness of free pronouns

In all persons, pronouns can be deleted anaphorically, although this is commoner with
3rd person pronouns (see 1.5). Second person subjects are not normally included in
imperatives except for emphasis (see 1.1.1.3). It is not normal to use pronouns in
apposition to the NPs for which they stand, except in instances of dislocation (see
1.11.2.1.6).

2.1.2.1.1.3 Pronouns and emphasis

2.1.2.1.1.3.1 Non-contrastiυe non-emphatic contexts

Free pronouns are normal in this environment.

2.1.2.1.1.3.2 Emphatic contexts

Free pronouns are normal in this environment.

2.1.2.1.1.3.3 With unemphatic imperatives

Second person subjects and passive agents are normally omitted in unemphatic
imperatives (see 1.1.1.3).

2.1.2.1.1.3.4 With imperatives for emphasis

Second person agents can be included as subjects or addressees for emphasis, see
1.1.1.3.1. In transitive imperatives, they take the form of passive agents.

2.1.2.1.1.3.5 Answers to questions

Free pronouns occur in answers to questions of appropriate types, eg.
(1652) Ka mau ia ki te ringa, ka mea, “E!
T/A seize IIIsg with the hand T/A say Hey
Ko wai teenei?” Ka mea a Hinemoa. “Ko
eq who this T/A say pers Hinemoa eq
ahau, e Tutaanekai.” Ka mea a Tutaanekai.
Isg voc Tutanekai T/A say pers Tutanekai
“Ko wai koe?” Ka mea ia. “Ko au, ko
eq who IIsg T/A say IIIsg eq Isg eq
Hinemoa.”
Hinemoa
‘He seized her by the hand and said, “Hey, who is this?” Hinemoa said, “It’s me,
Tutanekai.” Tutanekai said “Who are you?” She said, “It’s me, Hinemoa.”’ (H, 10)
 Maori 362

2.1.2.1.1.3.6 Cleft and pseudo-cleft constructions

Free pronouns can occur in both types of construction, eg.
(1653) Ko raatou e opaopa koowhatu ana
eq IIIpl T/A throw stone T/A
‘They’re throwing stones’
(1654) Ko ahau te mea moohio ki te whakautu
eq Isg the thing know to the answer
‘I’m the one who will know the answer’

2.1.2.1.1.4 Emphatic vs. unemphatic pronouns

In emphatic contexts, pronouns receive emphatic stress just as other types of constituent
do. There are no reduced forms in unemphatic contexts.

2.1.2.1.1.5 Reduced pronouns

There are no reduced pronouns, except possibly the clitic forms of singular pronouns
which occur with possessive prepositions. (However, they do not occur in other forms,
and are not clearly related to the free pronouns, so this is rather unlikely.) The occurrence
of these is determined by the occurrence of the possessive forms a and o, see 2.1.2.4.

2.1.2.1.2 Person distinctions

Maori distinguishes three persons, 1st, 2nd and 3rd, glossed here as I, II, III.

2.1.2.1.3 Inclusion υs. exclusion

Maori has distinct forms for inclusive and exclusive in 1st person dual and plural.
Inclusive forms contain t-, exclusive forms contain m-.

2.1.2.1.4 Number
Maori pronouns show a three-term number system, singular vs. dual vs. plural.

2.1.2.1.4.2 Optionality of number marking

Number marking is obligatory in pronouns.

2.1.2.1.4.3–4 Reference of number distinctions

Dual forms always imply two, and plural is used for three or more. If there is uncertainty
about the number involved, the plural is used, eg. the following is appropriate if one is in
doubt:
(1655) E haere ana ahau kia kite i a raatou
T/A move T/A Isg subj see DO pers IIIpl
 Morphology 363

i te ahiahi
at the afternoon
‘I’m going to see them this afternoon’

2.1.2.1.4.5 Paucal
Maori does not have paucal forms.

2.1.2.1.4.6 Pronouns with numerals

Pronouns can be associated in noun phrases with numerals with or without head nouns, in
a considerable variety of constructions eg.
(1656a) Maaua (e) toko·rua e whakaae ana
Idlexcl num pnum·2 T/A agree T/A
‘We two agree’
(1656b) Toko·rua maaua e whakaae ana
pnum·2 Idlexcl T/A agree T/A
‘We two agree’
(1657a) Ko raaua ngaa tamariki toko·rua tino kino te
top. IIIdl the(pl) children pnum·2 very bad the
tutu
naughtiness
‘Those two children are very naughty’
(1657b) Ko raaua toko·rua ngaa tamariki tino kino te
top. IIIdl pnum·2 the(pl) children very bad the
tutu
naughtiness
‘Those two children are very naughty’
(1657c) Ko raaua te toko·rua tino kino te tutu
top. IIIdl the pnum·2 very bad the naughtiness
‘Those two are very naughty’

Such forms apparently serve merely to emphasise the information about number
deducible from the pronouns. It appears that all reasonable pronoun-number
combinations are possible with duals and plurals, but combinations of singular pronouns
and kotahi are deemed odd, the norm being to use anake ‘only’. However, if anake is
present, then the numeral+pronoun combination is possible eg.
(1658a) *Ko ahau kotahi kaahore e whakaae ana
top. Isg one neg T/A agree T/A
‘*I, one, do not agree’
(1658b) Kotahi ahau anake kaahore e whakaae ana
one Isg only neg T/A agree T/A
‘I alone do not agree’
 Maori 364

The possessive pronoun can also be used:

(1658c) Ko taku kotahi anake kaahore e whakaae
top. sggenIsg one only neg T/A agree
ana
T/A
‘I alone do not agree’

2.1.2.1.4.7 Collective and distributive pronouns

No such distinction is made in pronouns in Maori.

2.1.2.1.4.8 Exhaustive non-singulars

Maori does not indicate in its pronouns whether the entire set of possible referents is or is
not included.

2.1.2.1.5.1–2 Relative importance of referent

Maori pronouns do not mark importance or obviation.

2.1.2.1.6 Proximity to participants

Personal pronouns do not contain an indication of the proximity to participants. Pronouns
can only marginally be combined with deictic particles to mark proximity, eg.
(1659) ?Hoatu ki a raaua raa
give to pers IIIdl dist
‘Give [it] to them there’

My consultant was very doubtful about this but felt it might be possible if, for instance, it
was necessary to resolve confusion over possible referents for raaua. It is not a standard
construction. However, the following textual example with ia (which might be a
demonstrative rather than a pronoun in this context) indicates that there may be occasions
when it is possible:
(1660) Koia nei raanei a Ruataupare, ara raanei
top.·IIIsg proxI or pers Ruataupare then or
ko teeraa raa?
top. that dist
‘Is this Ruataupare’s, or is that?’ (TWh, 19)

The same distinctions apply as are found in other uses of the deictic particles (see
2.1.2.5).

2.1.2.1.7 Special anaphoric third person pronouns

There are no special anaphoric third person pronouns. (There is an anaphoric determiner,
but it cannot be used substantivally.) But see locative pronouns 2.1.2.8.
 Morphology 365

2.1.2.1.7.1 Gender clashes

Since there is no grammatical gender in Maori, the problem does not arise.

2.1.2.1.8 Gender/class distinctions in pronouns

The pro-forms which refer to places are totally different from the set of personal
pronouns. They are discussed in 2.1.2.8. Otherwise, as has been mentioned, the personal
pronouns refer almost exclusively to personal nouns, but on occasion extend to animate
general nouns, and ia ‘IIIsg’ occasionally refers to abstractions. No changes in form
result from this extension.

2.1.2.1.8.1 Sex of speaker/hearer

No sex distinctions are marked in Maori pronouns.

2.1.2.1.9 Pronouns for tribal affiliation

There are no pronouns indicating whanau, hapu or iwi affiliations of either referent or
speaker/hearer, although dialectal variation in pronoun form may in some cases
inadvertently identify the tribal affiliation of the speaker. See 2.1.2.1.10.

2.1.2.1.10 Forms of pronouns

singular dual plural
I inclusive au ~ ahau taaua taatou (~taatau)
exclusive maaua maatou (~maatau)
II koe koorua koutou
III ia raaua raatou (~ raatau)

The variant forms for the plural are dialectal variants. The variation between au and ahau
seems to be found in all dialects, although the two forms may differ in frequency from
dialect to dialect. Ahau almost certainly derives from the personal article a+au, with
epenthetic insertion of h. Ahau cannot follow the personal article: compare ki a au, ki
ahau, but *ki a ahau. Some dialects have awau for ahau.
The singular forms cliticized to possessive prepositions are:
I -ku
II -u
III -na

See further 2.1.2.4.

 Maori 366

2.1.2.1.11 Verbal categories in pronouns

No verbal categories are marked in pronouns.

2.1.2.1.12 Status distinctions

There are no status distinctions in Maori pronouns. There do not appear to be any
instances where a title or other noun is preferred to a pronoun. (However, my consultant
felt a capital letter might be used for a pronoun in the written form if it referred to an
eminent person, such as the Queen.)

2.1.2.1.13.1 Non-specific indefinite pronouns

Maori does not have any non-specific indefinite pronouns.

2.1.2.1.13.2 Use of personal pronouns for non-specific indefinite reference

The plural 2nd person pronoun appears to serve this purpose:
(1661) A koutou nei naa, kia tika aa koutou
pers IIpl proxI proxII subj right plgenIIpl
tika·nga
right·nom
‘You should always be polite’
(1662) Kaua koutou e taahae
neg IIpl T/A steal
‘You shouldn’t steal’

2.1.2.1.13.3 Nouns as non-specific indefinite pronouns

Mea ‘thing’ can be used in this way, as in pseudo-clefts. See, for example, (1654) in
2.1.2.1.1.3.6. However, it is not possible to use te mea in place of koutou in (1662), for
instance:
(1663) *Kaua te mea e taahae
neg the thing T/A steal
‘You shouldn’t steal’

2.1.2.1.14 Specific indefinite pronouns

Forms equivalent to ‘someone’ and ‘somewhere’ consist of the appropriate interrogative
pro-form+raanei ‘or’, eg.
(1664) Maa wai raanei e aawhina atu
intgen who or T/A help away
‘Someone should help’
(1665) Kei hea raanei!
 Morphology 367

at(pres) where or
‘[It] must be somewhere!’

However, sometimes the specific indefinite determiner (t)eetahi is used substantivally

with an equivalent sense:
(1666) Teeraa pea teetahi e moohio ana ki te
that perhaps a(sp) T/A know T/A to the
roa·nga ake o te koorero nei, e
long·nom away gen the story proxI T/A
tae·a pea e ia te whakamaarama ki
arrive·pass. perhaps by IIIsg the explain to
te hunga e noho ware ana
the group T/A live forget T/A
‘Perhaps there is somebody who knows the rest of this story and who perhaps will be able
to explain it to those of us who live in ignorance’ (KH, 3)

Mea ‘thing’ is the usual equivalent for ‘something’, especially if qualifications are
attached, eg.
(1667) E hiahia ana ahau ki teetahi mea pararahi
T/A desire T/A Isg to a(sp) thing flat
‘I need something flat’

However, aha raanei can be used if there are no qualifications, eg.

(1668) He aha raanei taa Pita i hoatu ai ki
a what or sggen Peter T/A give part. to
a Hone
pers John
‘Peter gave John something’

‘Sometime’ in future contexts is normally rendered with aakuanei ‘presently’, eg.

(1669) Aakuanei kua tae mai ia
presently T/A arrive hither IIIsg
‘She’ll be here sometime’

However, aahea raanei is possible for full indefinites, eg.

(1670) Aa·hea raanei ka haere au
at(fut)·when or T/A move Isg
‘I’m going sometime’

In past contexts, it does not appear possible to specify sometime, eg.

(1671) I tae mai ia inanahi
T/A arrive hither IIIsg yesterday
‘He arrived (sometime) yesterday’
 Maori 368

Other potential parallel combinations (teehea raanei, e hia raanei, peehea raanei) do not
appear to be used as indefinite pronouns, but only in questions. Senses like ‘whichever’,
‘however many’ are rendered with ahakoa, ‘although’, eg.
(1672) Ahakoa te tini, me whaangai
although the numerous oblig feed
‘No matter how many there are, we will feed [them]’

See also 2.1.6.6.1.

2.1.2.1.15 Emphatic pronouns

Maori has no special emphatic pronouns.

2.1.2.1.16 Complex pronouns

Maori has no complex pronouns.

2.1.2.1.17 Pronoun-noun constructions

Pronoun-noun constructions are possible in all persons and numbers, eg.
(1673) E kaha rawa atu maatou ngaa kaiako naa
T/A strong very away Iplexcl the(pl) teacher proxII
ki te pata·patai
to the dup·ask
‘We teachers ask a lot of questions’

(Parallel singulars require te kaiako, of course.)

2.1.2.1.18 Coordination of pronouns and other NPs

This occurs in both dual and plural numbers in Maori, with the exception of Idlincl,
which is, of course, already fully specified.
(1674a) maaua ko te rata
Idlexcl spec the doctor
‘the doctor and I’
(1674b) koorua ko te rata
IIdl spec the doctor
‘you (the hearer) and the doctor’
(1674c) raaua ko te rata
IIIdl spec the doctor
‘he/she and the doctor’
(1674d) maatou ko Tuu
Iplexcl spec Tuu
‘Tu, the speaker, and at least one other (not the hearer)’
 Morphology 369

(1674e) taatou ko Tuu

Iplincl spec Tuu
‘Tu, the speaker, the hearer (and possibly others)’

(though maa would be likely to be included following Tuu if more were intended)
(1674f) koutou ko Tuu
IIpl spec Tu
‘Tu, the hearer, and at least one other’

(with the general expectation that there are two hearers; otherwise maa is likely)
(1674g) raatou ko Tuu
IIIpl spec Tu
‘Tu and at least two others’

(some speakers prefer raatou me Tu in the last instance).

2.1.2.1.18.1 Inclusive pronoun+pronoun combinations

The phenomenon of the type ‘we with you=you and I’ does not occur with pronouns.
Maaua me koe ‘Idlexcl with IIsg’ is possible, but refers to three people. *Taaua ko/me
koe ‘Idlincl spec/with IIsg’ was rejected, as was *maaua ko koe ‘Idlexcl spec IIsg’. Other
combinations were also rejected.

2.1.2.1.19 Secondary pronoun system

Maori does not have a secondary pronoun system for further specification of pronominal
reference.

2.1.2.1.20 Case in pronouns

Case is not marked in pronouns in Maori.

2.1.2.2 Reflexive pronouns

Maori does not have special reflexive pronouns, nor common nouns used as reflexives.
Personal and possessive pronouns are used, supported by anoo or (an)ake in certain
circumstances. For examples of the expression of reflexivity, see 1.6.

2.1.2.3 Reciprocal pronouns

Maori does not have special reciprocal pronouns, nor common nouns used as reciprocals.
Dual and plural personal and possessive pronouns are used, and are not distinct from
reflexives. For exemplification and discussion, see 1.7.
 Maori 370

2.1.2.4 Possessive pronouns

2.1.2.4.1 Special possessive pronouns

Maori has a complex set of possessive pronouns which distinguish not only the same
distinctions as the personal pronouns in terms of person and number, but which also
distinguish the A/O possessive categories (see 1.10), and the number of the object
possessed.

2.1.2.4.2 Types of possession in pronouns

2.1.2.4.2.1 Alienable/inalienable
I argued in 1.10.2 that the A/O distinction in Maori is probably not of this type. See
1.10.6 and 2.1.2.4.2.5.

2.1.2.4.2.2 Temporary /permanent

Maori does make a distinction of this type, but only permanent possession uses
possessive pronouns. Temporary possession is encoded as location, and the personal
pronouns are used following locative prepositions. See 1.10.3.

2.1.2.4.2.3 Persons/animals/things
The possessive pronouns in Maori are constructed from personal pronouns, which are
primarily reserved for persons in Maori. However, possessive pronouns are fairly readily
extended to animals and inanimates. Thus it is possible to say
(1675) He pai teenei whare. He nunui oona ruuma,
cls good this house cls big plgenIIIsg room
e hou ana toona peita
T/A new T/A sggenIIIsg paint
‘This house is nice. Its rooms are big and its paint is new’.
(1676) Ka wehe atu a Tainui i Whangaparaaoa,
T/A depart away pers Tainui from Whangaparaoa
ka mahue mai eetahi o oona taangata
T/A leave hither some(sp,pl) gen plgenIIsg people
‘When the Tainui left Whangaparaoa, some of its people were left behind’ (P, 1)

(Note that Tainui was a fleet of canoes, whence the plural possessive.)
(1677) Koia te whakataukii nei, ‘He kura pae
this the proverb proxI a feather come ashore
naa Maahia’. Ko toona whakamaoritanga teenei…
actgen Mahia top. sggenIIIsg meaning this
‘Hence this proverb, “He kura pae naa Maahia”. Its meaning is this:…’(P, 3)
 Morphology 371

Examples of possessives extended to animals are common, eg.

(1678) Kaatahi ka tangi te taniwha nei ki toona
then T/A mourn the taniwha proxI to sggenIIIsg
ariki
chief
‘Then this taniwha mourned for its master’ (TP, 89)

2.1.2.4.2.4 Present/past
Possessive pronouns are not marked for tense distinctions. However, the compound
prepositions maa/moo, naa/noo which contain possessive prepositions (aa, oo), combine
with personal pronouns to give a set of possessive forms where a tense-like distinction is
marked. See 1.10.5 and 2.1.2.4.10.

2.1.2.4.2.5 A/O
Maori has two parallel sets of possessive pronouns, one containing the preposition oo,
and the other containing aa. The nature of the distinction has been discussed in 1.10.6.
The remarks there apply to pronouns as well. In addition, in the singular only, there is a
set of possessive pronouns which neutralizes the distinction. The conditions for the use of
these neutral forms are not at all clear, partly because there seems to be considerable
variation from speaker to speaker, and partly because in texts which fail to mark vowel
length (and that includes all early texts), it is not always clear whether a given form is
neutral or not. The neutral forms are probably more common than the A/O forms for first
and second person singular possessors, although in most instances, there is a choice. The
choice does not seem to be affected by the syntactic position of the possessed NP, but
rather by the type of object possessed. Thus with tamaiti ‘young son’, my consultant
often rejected the A form, but with tama ‘son’, the A form was generally an acceptable
alternative. Similarly, with maaripi ‘knife’, the A form was rejected in both 1st and 2nd
persons in favour of the neutral form. It appears that this is a matter of familiarity.
Besides familiarity, it seems that the choice of the neutral form with a possessed object
normally accorded respect can indicate disapproval. Thus my consultant felt it more
appropriate to say
(1679) E moorikarika ahau ki taku kaiako
T/A abominable Isg to sggenIsg teacher
‘I’m cheesed off with my teacher’

than to use tooku. (Katarina Mataira, personal communication, described this use of the
neutral form as ‘belittling’.) However, these factors do not fully account for the variation
found. My consultant said, for instance, that it would be appropriate to say
(1680) Kei hea taku pootae?
at(pres) where sggenIsg hat
‘Where is my hat?’

if the hat was in sight, and about to be donned, but that if one were hunting for one’s hat,
and unable to find it, (1681) would be used:
 Maori 372

(1681) Kei hea tooku pootae?

at(pres) where sggenIsg hat
‘Where is my hat?’

Such a distinction is not clearly accounted for by either of the principles enunciated
above.

2.1.2.4.3 The forms of possessive pronouns

Below, the forms referring to a singular possessed item are listed. The forms for plural
possessed items are derivable by deletion of the initial t-.
Neutral forms (Singular possessees)
Singular possessor
I taku
II too
III tana

A Forms (Singular possessees)

Possessors
Singular Dual Plural
I taaku taa taaua taa taatou
taa maaua taa maatou
II taau taa koorua taa koutou
III taana taa raaua taa raatou

O Forms (Singular possessees)

Possessors
Singular Dual Plural
I tooku too taaua too taatou
too maaua too maatou
II toou too koorua too koutou
III toona too raaua too raatou

By comparison with 2.1.2.1.10, it will thus be clear that possessive pronouns for dual and
plural possessors are derived by a regular process from the personal pronouns: these
forms all consist of
t-aa/oo+personal pronoun
 Morphology 373

(See also Biggs, 1966, Biggs 1969, 46–7.) The singular pronominal forms -ku ‘1st
person’, -u ‘2nd person’, -na ‘3rd person’ are found only in combination with the
possessive prepositions in Maori. The singular forms can be seen to have the same form
as the duals and plurals, but with special pronominal allomorphs.
The neutral possessive forms are always attributive. The A/O forms may be attributive
or substantival. None of the above forms can occur in post-posed genitive phrases, nor
predicatively. Note that all these forms are definite.
A few illustrative examples follow; many others are scattered throughout the text.
(1682) Ka hoki a Tupeteka ki toona kaainga
T/A return pers Tupeteka to sggenIIIsg home
‘Tupeteka returned to his home’ (W, 197)
(1683) Ko aana kupu eenei:…
top. plgenIIIsg word these
‘These were her words…’ (W, 198)
(1684) Kaatahi ka tuku·na taa raatou taurekareka…
then T/A send·pass. sggenIIIpl slave
‘Then their servant was sent…’ (W, 198)
(1685) Ko te koha teenei a Wairangi ki tana
eq the gift this gen Wairangi to sggenIIIsg
wahine
woman
‘This was Wairangi’s gift to his wife’ (W, 199)
(1686) Ko too tane teeraa?
eq sggenIIsg man that
‘Is that your husband? (W, 200)
(1687) Ka kii hei taana ka huaki
T/A say at(fut) sggenIIIsg T/A attack
‘[He] said at his [haka] [they] should attack’ (W, 199)
(1688) E rima noa iho ooku paraikete
num 5 indeed down plgenIsg blanket
‘I’ve got five blankets’

2.1.2.4.4 Case marking in possessive pronouns

Possessive pronouns in Maori do not show case.

2.1.2.4.5 Pronominal possession without possessive pronouns

Not applicable.

2.1.2.4.6 Alternative means of expressing possession

There are no alternative means of expressing possession with pronouns. Note however
that the expression of temporary possession in Maori involves the use of locative
 Maori 374

prepositions+ordinary personal pronouns. The two means of expression are in

complementary distribution.

2.1.2.4.7 Reflexive possessive pronouns

Maori does not have any special reflexive possessive pronouns. The possessive pronouns
listed above are used. 1.6 contains ample exemplification.

2.1.2.4.8 Reciprocal possessive pronouns

Maori does not have any special reciprocal possessive pronouns. The dual and plural
possessive pronouns listed above are used. Exemplification can be found in 1.7.

2.1.2.4.9 Emphatic possessive pronouns

Maori does not have any special emphatic possessive pronouns. Note, however, that in
predicative contexts (where English, for instance, makes use of emphatic possessive
pronouns), Maori uses the compound possessive prepositions described in the next
section.

2.1.2.4.10 Compound prepositional possessive pronouns

These forms are poly-morphemic, with the following structure:

In the dual and plural, such forms can be seen simply as compound (possessive)
preposition+personal pronoun, and do not look as if they are possessive pronouns at all,
compare
e raatou ‘by them’
maa raatou ‘for them’ (intgen IIIpl)

However, in the singular, the normal personal pronouns do not occur, but are replaced by
the same forms as are used in the singular possessive pronouns (see 2.1.2.4.3), giving
forms like
maa·ku ‘for me’ (intgen·Isg)
naa·na ‘belonging to her’ (actgen·IIIsg)

As discussed in 1.10.5 and 1.10.6, the m-/n- distinction appears to be between intended
and realized possession, while the aa/oo distinction hinges on matters of control.
Besides their use in functions such as benefactives and causation (see 2.1.1.4.1,
2.1.1.4.12), these compound possession prepositions are used predicatively in statements
of possession (see 1.10.1, 1.10.5), eg.
(1689) Naa·ku ngaa kura raa!
 Morphology 375

actgen·Isg the(pl) feather dist

‘Those feathers are mine’ (P, 3)
(1690) Maa·ku teenei
intgen·Isg this
‘This one is for me/This one is to be mine’ (KH, 1)

2.1.2.4.11 Possessive pronouns as noun modifiers

Some information on this topic has already been given in 1.10.1 and in 2.1.2.4.3. The
forms discussed in 2.1.2.4.10 are primarily predicative or adverbial, ie. not noun
modifiers. However, by relativization, the prepositional forms can appear as post-head
modifiers, eg.
(1691) …kua whakairo·tia nei e Taikehu hei patu
T/A carve·pass. proxI by Taikehu as club
maa·na
intgen·IIIsg
‘…[it] had been carved by Taikehu as a club for himself (P, 4)

However, since the possessive determiners are definite, the expression of indefinite
possession cannot make use of the possessive determiners. These prepositional forms are
thus used in the expression of indefinite possession, (see also Williams, 1862, pp.21–22)
eg.
(1692) Noo·u teeraa tungaane? Aae, he tungaane
actgen·IIsg that brother yes a brother
noo·ku
actgen·Isg
‘Is that your brother? Yes, he’s a brother of mine’

If determiners other than indefinite he or definite te are required, the possessive

determiners are also precluded from appearing. Williams (1862, 22) states that aa and oo
are used and gives the example
(1693) teenei taha oo·ku
this side gen·Isg
‘this side of me’

Biggs, however, calls this class of forms ‘Zero-class’ possessives (1966, 347). He gives
the examples (glosses altered to conform to the present work):
(1694) I whaanau au i teetahi tamaiti aaku
T/A give birth Isg DO a(sp) child A;Isg
‘I gave birth to a child of mine’
(1695) Ka tae a Manaia ki taua ika aa·na
T/A arrive pers Manaia to det aph fish gen·IIIsg
‘Manaia reached that fish of his’
 Maori 376

Such forms appear to consist of the possessive preposition (aa, oo)+ personal pronoun,
which in the singular takes the special clitic form. In dual and plural, the structure is
much clearer:
(1696) I tuutaki ahau ki teetahi hoa o raaua
T/A meet Isg to a(sp) friend gen IIIdl
‘I met a friend of theirs’

2.1.2.5 Demonstrative pronouns

Maori has a three-term demonstrative system.

2.1.2.5.1 Parameters for demonstratives

2.1.2.5.1.1 Relative distance from speaker

There appear to be two terms here, nei, ‘near speaker (within reach)’ and distant from
speaker, but that second distinction is further differentiated in relation to the hearer.

2.1.2.5.1.2 Relative distance from hearer

There is a term naa, ‘near hearer’, but far less physical proximity is required than is
required for ‘near speaker’. This term also contrasts with the ‘distant’ member of the
tripartite system.

2.1.2.5.1.3 Relative distance from speaker and hearer

There is a term raa, ‘out of reach of speaker and out of reach of hearer’, which contrasts
with the proximate terms described in the previous two sections.

2.1.2.5.1.4 Equidistance from speaker and hearer

If an object is equidistant from speaker and hearer, but within reach of both, there is a
choice between nei, ‘near speaker’ and naa, ‘near hearer’. If an equidistant object is out
of reach, and between speaker and hearer, it can be marked ‘near hearer’ or raa, ‘distant’.
Equidistance is thus not a relevant parameter.

2.1.2.5.1.5 Contact with speaker

‘This is marked as nei, ‘near speaker’.

2.1.2.5.1.6 Contact with hearer

‘This is marked as naa, ‘near hearer’.
 Morphology 377

2..1.2.5.1.7 Behind speaker

This is not a relevant parameter: an object behind the speaker, but within reach is nei,
‘near speaker’.

2.1.2.5.1.8 Behind hearer

This is not a relevant parameter: an object behind the hearer, but within the hearer’s reach
can be marked naa, ‘near hearer’. However, if the hearer is between the speaker and the
object, so that the speaker could be in doubt as to its precise location, the ‘distant’ form
raa could be used.

2.1.2.5.1.9 Between speaker and hearer

This is not a relevant parameter. See also 2.1.2.5.1.4 above.

2.1.2.5.1.10 On the other side of hearer from speaker

This is not a relevant parameter, but ‘distant’ seems to be more likely than ‘near hearer’
where those two terms are in potential conflict (see 2.1.2.5.1.8).

2.1.2.5.1.11 On other side of speaker from hearer

This is not a relevant parameter. The choice of form depends on proximity or otherwise
to the speaker.

2.1.2.5.1.12–18 Position in relation to some object

The choice of demonstrative in Maori is not determined in relation to any object in the
communication situation other than the speaker and hearer.

2.1.2.5.1.19 Vertical orientation in relation to speaker

Relation in the vertical dimension is similar to the horizontal dimension. The primary
distinction is ‘within reach’ vs. ‘out of reach’. As with the horizontal dimension, if an
object is out of reach of both speaker and hearer, but about the same distance from both,
it can be marked either as ‘near hearer’ or ‘distant’.

2.1.2.5.1.20 Other spatial relations or reference points

No other features are relevant to Maori.

2.1.2.5.1.21 Visible/inυisible
The visibility or otherwise of an object is irrelevant A smudge on the speaker’s cheek is
still ‘near speaker’. However, note the remarks in 2.1.2.5.1.8.
 Maori 378

2.1.2.5.1.22 Known/unknown
This is not a relevant parameter.

2.1.2.5.1.23 Referred to in previous discourse

The demonstrative pronouns and the deictic particles from which they are derived are an
important means of achieving anaphoric reference in Maori (see 1.5). It is not easy to sort
out the factors that are involved in the choice of form. However, it is clear that the ‘near
hearer’ form is rare in narratives (except in direct speech): for the most part a two-term
system operates. Most of the examples which follow are not independent pronouns, but
the same principles apply to their use as to the use of the particles.
Sometimes it is clear that relative distance of mention is the crucial factor, eg.
(1697) …piri ana te kookoowai i te paapaaringa o
stick T/A the clay at the cheek gen
Tupeteka, noo Pare-whete. Kua moohio a
Tupeteka actgen Pare-whete T/A know pers
Puuroku kua tae·a te wahine raa e
Puroku T/A reach·pass. the woman dist by
Tupeteka
Tupeteka
‘…the day sticking to the cheek of Tupeteka belonged to Parewhete. Thus Puroku knew
that that woman had yielded to Tupeteka’ (W, 197)

Here the distant form is used to select the further back of two potential referents.
Sometimes it seems that the proximate form is used to select the party with whom
sympathy lies, or the party central to the story, eg.
(1698) Ka tae ki a Ngaati-Raukawa ka koorero·tia
T/A arrive to pers Ngati-Raukawa T/A say·pass.
“Kei te ata ka huaki. Ko te kai e
at(fut) the morning T/A attack top. the food T/A
mahi·a mai nei maa teeraa kee, maa
make·pass. hither proxI intgen that contr intgen
Ngaati-Maru.”
Ngati-Maru
I roa te iwi nei e nohopuke ana
T/A long the tribe proxI T/A silent T/A
kaaore he hamumu, kaaore he aha
neg cls sound neg cls what
‘When he got back to the Ngati-Raukawa, he told them, “In the morning they attack. This
feast that is being prepared is for the others, for Ngati-Maru.”
For a considerable time the people remained silent; they didn’t speak or stir’ (W, 199)

The nei does not refer to the nearest possible antecedent, but to Ngati-Raukawa.
However, the unmarked form appears to be the distant form, which is considerably
more frequent in texts; the following is an example where other factors are probably
absent:
(1699) “…kei te ata ka eke mai ka patu.”
 Morphology 379

at(fut) the morning T/A land hither T/A fight

Heoi anoo, ka rongo a Pare-whete i te
well T/A hear pers Pare-whete DO the
koorero raa, kaatahi ka puta te aroha o
talk dist then T/A appear the sympathy gen
te wahine raa ki a Wairangi me toona
the woman dist to pers Wairangi with sggenIIIsg
iwi
tribe
‘“…in the morning they will arrive and attack.” Then it was that Parewhete heard these
words, and her love revived for Wairangi and his people’ (W, 198)

However, it is just possible that both distant forms here are accounted for on ‘sympathy’
grounds. Both the words quoted and Parewhete are on the ‘other’ side from Wairangi,
from whose point of view the story is told. This could be another manifestation of the
presence of the narrator’s viewpoint (see Hohepa, 1981, 44–45).
One of the rare examples of the ‘near hearer’ form is the following:
(1700) Pee·naa tonu; ka haere o teenaa hapuu, ka
like·proxII still T/A move gen that sub-tribe T/A
pau anoo i taua nanakia nei te kai
exhausted again cause det aph monster proxI the food
‘Thus it continued; other sub-tribes came down, they were all eaten by the monster’ (TP,
90)

2.1.2.5.1.24 Time dimension

The demonstrative pronouns are relatively seldom used for time reference, although the
deictic particles are frequently part of temporal phrases, either attached to the determiner
or in post-head position. As with anaphoric reference in discourse, the ‘near hearer’ form
is relatively rare, and a two-way opposition ‘near/distant’ operates for the most part. The
‘near’ form is used for periods including the present, while the distant form is used for
periods either past or future. The following examples illustrate the basic uses.
(1701) I haere maatou i teeraa wiki
T/A move Iplexcl at that week
‘We left last week’
(1702) A teeraa wiki ka haere maatou
at(fut) that week T/A move Iplexcl
‘We will go next week’
(1703) Ka tae mai raatou a te wiki nei
T/A arrive hither IIIpl at(fut) the week proxI
‘They will arrive this week’
(1704) I eenei raa maa te mihiini e mahi te
at these day intgen the machine T/A do the
nui·nga o ngaa mahi
big·nom gen the(pl) work
‘These days, machines do most of the work’ (TR2, 59)
 Maori 380

(1705) He whare too maatou i eeraa raa…

cls house sggenIplexcl at those day
‘We had a house in those days…’

In narrative, the proximate form is used for ‘contemporaneous with the events in the
narrative’, eg.
(1706) Noo teenei waa, ka tae mai ngaa kootiro
actgen this time T/A arrive hither the(pl) girl
a Toto
gen Toto
‘Just at this moment, Toto’s daughters arrived’ (KH, 1)

Note that a deictic particle is included in temporal adverbs like inaianei ‘now’, akuanei
‘later, presently, today’. One of the rare examples of a ‘near hearer’ form is the
following:
(1707) Ka haere ngaa tuumau o too Poroumaataa
T/A move the(pl). attendant gen sggen Poroumata
paa ki ngaa awa ki te tiki i ngaa ika
pa to the(pl) landing to the fetch DO the(pl) fish
i teenaa raa, i teenaa raa
at that day at that day
‘The attendants of Poroumata’s pa went to the landing places to fetch the fish day by day’
(TWh, 17)

2.1.2.5.1.25 Other parameters

There do not appear to be other parameters involved.

2.1.2.5.1.26 Neutral demonstrative

There is no neutral demonstrative pronoun.

2.1.2.5.2 Forms of demonstrative pronouns

Demonstrative pronouns are the same in form as demonstrative determiners. They consist
of the determiner te+deictic particle in the singular, with the plural being formed as
elsewhere by the dropping of the initial t-. The three deictic particles are nei, ‘near
speaker’, naa ‘near hearer’, raa ‘distant. Thus the resulting forms are:
singular plural gloss
teenei eenei ‘near speaker’
teenaa eenaa ‘near hearer’
teeraa eeraa ‘distant’

A few examples are given.

 Morphology 381

(1708) Ka hoki eeraa ki Rurunui

T/A return those to Rurunui
‘They returned to Rurunui’ (W, 197)

(Williams notes (1862, 23) the use of teeraa as an emphatic 3rd person singular pronoun,
but the use is not restricted to teeraa.)
(1709) He whare koohuru teenei
cls house murder this
‘This is a house for murder’ (W, 198)
(1710) Maa teeraa kee te kai e mahi·a nei,
intgen that contr the food T/A make·pass. proxI
maa Ngaati-Maru
intgen Ngati-Maru
‘The feast we are preparing is for them, for Ngati-Maru’ (W, 199)
(1711) He take pai noa iho teenaa
cls matter good indeed down that
‘That’s no trouble’
(1712) He tino tohunga teenei ki te karakia
cls very tohunga this to the incant
‘He was a great expert at incantation’ (KWh, 1)
(1713) Ko te hoki·nga teenei o Rua ki Hawaiki ki
top. the return·nom this gen Rua to Hawaiki to
te tiki i a Pou
the fetch DO pers Pou
‘So Rua returned to Hawaiki to fetch Pou’ (P, 1)
(1714) Ko te riri·nga teenei o Pou
top. the angry·nom this gen Pou
‘This angered Pou’ (P, 3)

Note that teeraa is used to introduce new protagonists into narrative (see 1.12.1).

2.1.2.5.3 Iconic demonstratives

The demonstratives in Maori do not change their phonological form to show iconically
different degrees of distance.

2.1.2.5.4 Number in demonstratives

Demonstratives show a singular-plural number distinction, see 2.1.2.5.2. Number is
marked in the same way in NPs, except that NPs have the alternative of post-posing the
deictic particle, and then ngaa is used for plural. Compare:
teenei whare te whare nei teenei
this house this house this
eenei whare ngaa whare nei eenei
 Maori 382

these houses these houses these

The loss of initial t- in the plural is in common with all other t-compound determiners in
Maori, both those that can be used substantivally, and those that cannot.

2.1.2.5.5 Class/gender in demonstratives

These demonstratives are not used for local nouns (see 2.1.2.8), but are not in themselves
marked for noun class (which is not in any case a major feature of Maori).

2.1.2.5.6 Case in demonstratives

Demonstratives are not marked for case. They follow the same prepositions as other NPs
to indicate syntactic function.

2.1.2.5.7 Other grammatical categories in demonstratives

No other grammatical categories are marked in demonstratives.

2.1.2.5.8 Attributive use of demonstratives

As indicated in 2.1.2.5.4, the demonstrative pronouns can also be used attributively, but
they then alternate with the semantically equivalent construction ‘Det+N head+deictic
particle’. There would seem to be good reason for deriving the demonstrative determiners
from this latter construction, and for regarding the demonstrative determiner as prior to
the substantival use. The evidence involves the fact that the deictic particles also form
other series of compounds (eg. with ko, pee, aa); and that the possessive determiners also
have parallel variant forms (te whare o Pou ~ too Pou whare); and the fact that the deictic
particles can co-occur with other determiners, eg. taua nanakia nei ‘that monster here’
(Tutae-poroporo, many times).
The choice between eg. teenei whare and te whare nei ‘this house’ appears to be
purely stylistic; rhythmic considerations often appear to influence the choice. In texts, the
non-compounded forms far outnumber the compounded forms. (In the story Wairangi,
there are 18 non-compound attributives (ie. Det+N+deictic), and only one compound
attributive (ie. Det·deictic+N); there are 10 substantival demonstratives (Det·deictic).

2.1.2.6 Interrogative pronouns and other question words

2.1.2.6.1 Interrogative pronouns

2.1.2.6.1.1.1 General question words

Maori has the following interrogative proforms:
wai for persons (and names)
 Morphology 383

aha for non-persons (and actions)

(w)hea for locations (spatial and temporal)
hia for numbers, quantities
peewhea for like what’ and diverse other adverbial functions

Note that not all these are pro-nouns exclusively. Aha can be a proverb, and pee(w)hea is
a pro-adverb. Examples can be found in 1.1.1.2.2.

2.1.2.6.1.1.2 Selective from a group

(T)ee(w)hea is selective from a group. It is a compound of the determiner te+(w)hea, and
can be used either as a determiner or substantivally, eg.
(1715) Ko teehea poo teenei?
top. which night this
‘Which night is this?’
(1716) Ko teehea te roto tuarua i kite te roopuu
top. which the lake second T/A see the group
nei?
proxI
‘Which was the second lake this group saw?’

Note the following, where the selection is presumed rather than explicit:
(1717) Noo teehea raa i tuhi·a ai teenei reta
actgen which dist T/A write·pass. part. this letter
‘Where is this letter from?’

However, a selective form is not necessarily used, eg.

(1718) Ko wai o koutou kaahore anoo kia rongo ki
top. who gen IIpl neg again subj hear to
ngaa koorero moo te tangata nei moo Maaui?
the(pl) talk intgen the man proxI intgen Maui
‘Which of you have not heard of the deeds of this man, Maui?’ (KM, 1)

2.1.2.6.1.2 Number in interrogative pronouns

Of the Maori interrogative words, only (t)ee(w)hea is directly marked for number, with
tee(w)hea (sg) and ee(w)hea (pl), using the process common to other compound
determiners.
However, with wai, the addition of maa ‘and others’ serves the purpose of a plural, eg.
(1719) Ko wai teenei?
eq who this
‘Who is this?’
(1720) Ko wai maa eenei?
eq who and others these
 Maori 384

‘Who are these?’

If wai is used in a pre-posed possessive phrase, number relating to the possessed object is
then marked in the determiner:
(1721) Ko taa wai pukapuka teenei?
eq. sggen who book this
‘Whose book is this?’
(1722) Ko aa wai pukapuka eenei?
eq. plgen who book these
‘Whose books are these?’

This does not mark the number of the interrogative possessor. However, these are not
normal questions in modern Maori. (The normal question is Naa wai (t)eenei pukapuka?)
When aha is a pro-noun (as opposed to a pro-verb) it takes a determiner, which can be
he or ngaa:
(1723) Ko ngaa aha eenei?
eq the(pl) what these
‘What are these?’

However, this is not the most normal question: the normal question is
(1724) He aha teenei/eenei?
cls what this/these
‘What is this?/What are these?

The other question words cannot be marked for number:

(1725) *Kei hea maa too koti me
at(pres) where and others sggenIIsg coat with
oo huu?
plgenIIsg shoe
‘Where are your coat and shoes?’

2.1.2.6.1.3 Class/gender in interrogative words

None of the interrogative words is marked for class or gender. However, note that hea,
like the local nouns for which it substitutes, follows locative prepositions directly, with
no article:
(1726) Kei hea tooku pootae?
at(pres) where sggenIsg hat
‘Where’s my hat?’

ie. while not marked for class or gender, different interrogative words exhibit syntactic
behaviour consonant with the class for which they substitute.
 Morphology 385

2.1.2.6.1.4 Case marking in interrogative words

Interrogative words in Maori are not marked for case, although compound forms for
temporal questions such as aa(w)hea ‘when(fut)’, noona(w)hea, ina(w)hea ‘when(past)’
contain prepositions, which serve the purpose of marking syntactic function in Maori.
However, sometimes an additional preposition is used with them, as hei with aawhea.

2.1.2.6.1.5 Other grammatical categories in interrogative words

No other grammatical categories are marked in interrogative words.

2.1.2.6.1.6 Attributive use of interrogatives

The possibilities for tee(w)hea have already been illustrated (see 2.1.2.6.1.1.2).
Wai can be used in possessive phrases preposed as illustrated in 2.1.2.6.1.2, but not
post-posed as in:
(1727) *Ko teenei te pukapuka a wai?
eq this the book gen who
‘Whose book is this?’

The normal way of asking this question uses maa/naa+wai, which makes the
interrogative word predicative, not attributive eg.
(1728) Naa wai teenei pukapuka?
actgen who this book
‘Whose book is this?’

Aha can be used attributively to a local noun as in:

(1729) Kei runga aha a Hata?
at(pres) top what pers Hata
‘What is Hata on?’

Other interrogative words cannot be used attributively.

2.1.2.6.2 Other question words and their meanings

See also 1.1.1.2.2 and 2.1.2.6.1.1.1.
Aha as a pro-verb is exemplified by
(1730) E aha ana raatou inaianei?
T/A what T/A IIIpl now
‘What are they doing now?’

(W)hea for temporal locations (always in compound forms) is illustrated by

(1731) A·whea koe haere ai?
at(fut)·when IIsg move part.
 Maori 386

‘When are you going?’

(1732) Ina·whea koe i haere ai?
at(pt)·when IIsg T/A move part.
‘When did you go?’

Hia for numbers is illustrated by

(1733) E hia ngaa tuuru?
num num-Q the(pl) stool
‘How many chairs are there?’

Peehea has a variety of uses, many of which answer to English questions with how. A
few are illustrated here. (Note that pee- also forms compounds with the deictics, see
2.1.2.9.)
(1734) Pee·hea te nui o te puurere?
like-how the big gen the flour
‘How much flour [is needed]?’
(1735) Kia pee·hea te hoohonu o te rua nei?
subj like-how the deep gen the hole proxI
‘How deep should this hole be?’
(1736) Me pee·whea raa?
oblig like-how dist
‘How should it be done?’
(1737) I pee·hea te hoki·nga o Te Tahi i
T/A like-how the return·nom gen Te Tahi from
Whakaari?
White Island
‘How did Te Tahi get back from White Island?’

Direct and indirect questions use the same question forms.

2.1.2.7 Relative pronouns and other relative words

2.1.2.7.1 Special relative pronouns

Maori does not have special relative pronouns, although the part played in relativization
by the particle ai and the deictics nei/naa/raa is still open to question. (For some
discussion, see Bauer, 1981a, 391ff and Bauer, 1983, 329ff and references therein.) The
problem is outlined briefly here. In certain cases of relativization, a personal pronoun
remains behind. In other cases, sometimes as an alternative to the personal pronoun
strategy, ai or one of the deictics appears post-verbally in the relative clause, clearly
introduced by the relativization process. Chapin (1974) argued that in Proto-Polynesian,
ai was an oblique NP pro-form. However, there are a number of problems in adopting
this analysis for modern Maori. Firstly, there are cases where ai appears with subject
relativization, and some speakers regularly use it in that context. Secondly, there are a
 Morphology 387

variety of tense restrictions on ai, and certain consequences for tense of the choice
between the deictics. Thirdly, if the T/A marker is e…ana, none of these forms appears,
apparently because ana occupies the “slot” in which these post-verbal particles can
appear. All of these things suggest that ai has been reinterpreted (or is in the process of
reinterpretation) as a general marker of subordination. Consequently, I believe that Maori
does not have relative pronouns. (Contrast the view in Reedy, 1979).
In case these forms are subsequently shown to be relative pronouns, it should be noted
for the record that they are not marked for number, noun class, case (but note the possible
restriction to obliques), or any other grammatical category, and they are not used
adjectivally as relative pronouns.

2.1.2.7.2 Other relative words

No words other than those discussed in 2.1.2.7.1 are involved in the process of
relativization in Maori.

2.1.2.7.3 Relative pronouns for ‘place’ and ‘time’

Neither relative pronouns nor the forms corresponding to ‘where’ and ‘when’ are
involved. The forms discussed in 2.1.2.7.1 are used.

2.1.2.8 Locative proforms

The discussion of types of pronoun above leaves various proforms for local nouns
unaccounted for. These are treated below.

2.1.2.8.1 Parameters for locative proforms

The two major sets of locative proforms are compound forms with deictics for their
second element. The deictics show the same distribution as in other contexts (ie. a three-
way distinction ‘near speaker’, ‘near hearer’, ‘distant’), see 2.1.2.5.1. In addition there is
a special form reira for anaphoric reference to locations. (Although this may historically
contain the deictic raa, it is not part of a set with a three-way contrast, but is used for all
anaphoric locative reference.) For instance:
(1738) …aa ka puta ki mua tonu ki te aroaro o
and T/A appear to front still to the front gen
te haka. Takoto tiiraha ana i reira ki te
the haka lie face up T/A at there to the
maatakitaki
watch
‘…[He] came quite close to the front of the haka. Here he lay down upon his back to
watch’ (W, 200)
 Maori 388

2.1.2.8.2 Forms of locative proforms

Apart from reira illustrated above, there are two other sets of locative proforms. The two
sets are constructed from a(a)- and ko- + nei, naa, raa, to give
aanei konei
aanaa konaa
aaraa koraa

The a(a)- forms are predicative demonstratives, the ko- forms are used in subjects and
oblique NPs. Examples are
(1739) Ka pai a konei
T/A good pers here
‘Here would be good’
(1740) Noo te tae·nga mai hoki o ngaa
actgen the arrive·nom hither also gen the(pl)
tauhou ki konei ka arahi·na raatou e maatou
stranger to here T/A lead·pass. IIIpl by Iplexcl
ki te toka nei kia kite
to the rock proxI subj see
‘When visitors come here, we lead them to this rock to see’ (TA, 8)
(1741) Aanei teetahi rua
here a(sp) hole
‘Here’s a hole’ (TR2, 119)
(1742) Aanei ko aku koowhaiwhai
here top. plgenIsg rafter patterns
Aaraa ngaa tukutuku
there the(pl) lattice wall patterns
‘Here are my rafter patterns, Here are the lattice work patterns on the walls’ (WhWh, 1)

The a(a)-forms have a few derived usages. According to Williams’s Dictionary, anei can
be used for ‘in this case’, but this usage was not familiar to my consultant. Anaa is used
to draw attention, eg.
(1743) Anaa oti, ko ngaa rongo toa kia nui,
there then top. the(pl) hear champion subj big
aa ko ngaa tohu o te toa kia iti?
and top. the(pl) sign gen the champion subj small
‘How could the fame of this bravery be great and the signs of that bravery be small?’
(TWh, 22)
(1744) Anaa, mau·ria atu!
here take·pass. away
‘Here you are: take it!’ (pressing a gift on someome)
(1745) Anaa naa!
here proxII
 Morphology 389

‘Oh, wow!’

Araa is used for ‘namely, in other words, ie.”, eg .

(1746) “…mehemea e kore e tae·a e ia te
if T/A neg T/A reach·pass. by IIIsg the
haere mai ki te patu i taua taniwha,”
come hither to the kill DO det aph taniwha
-araa, i a Tuutae-poroporo
ie. DO pers Tutae-poroporo
‘“…if he couldn’t come and kill that taniwha”, ie. Tutaeporoporo’ (TP, 91)

The first two ko-forms are also used in farewells, eg.

(1747) Hei konei raa moo teenei waa
at here dist intgen this time
‘Goodbye for now’

(said by the person leaving; konaa is also possible).

2.1.2.8.3 Iconicity
These forms are not variable in length iconically.

2.1.2.8.4–7 Grammatical categories marked in locative proforms

These proforms are not marked for number, class (except in so far as they can only stand
for local nouns), gender, case or any other grammatical category.

2.1.2.8.8 Adjectival use

These proforms cannot be used adjectivally, except through relativization.

2.1.2.9 Further deictic proforms

Maori has a further set of deictic proforms, consisting of peelike’+a deictic. (The
interrogative proform pee(w)hea contains the same first element.) These forms are verbal
pro-forms, eg.
(1748) Naa pee·raa tonu te tikanga a Tutaanekai raaua
now like·dist still the custom gen Tutanekai IIIdl
ko Tiki, i ngaa poo katoa
spec Tiki at the(pl) night all
‘Now, it was Tutanekai’s and Tiki’s custom to do that every night’ (H, 6)
(1749) He pee·nei tonu au ki te mahi
cls like·proxI still Isg to the work
‘I always work like this’
(1750) Kei te pee·raa tonu hoki a muri i a ia
 Maori 390

T/A like·dist still also pers behind at pers IIIsg

Those behind him were doing the same’ (TWh, 23)

These forms are not marked for any other grammatical categories.

2.1.3 Verb Morphology

2.1.3.1 Voice

2.1.3.1.1 Passive
The Maori passive has occasioned a good deal of discussion, both on account of its
morphology, and on account of its prevalence in narrative. The subject came to general
linguistic attention with the publication of Hohepa’s thesis in 1967, and Hale’s
subsequent review of it (Hale, 1968). As an aid to the discussion of the debate and the
problems, the basic facts are outlined here first.
There is general agreement that Maori has a passive, eg.
(1751) I koohete·tia a Pani e Huia
T/A scold·pass. pers Pani by Huia
‘Pani was scolded by Huia’

Compare the active:

(1752) E koohete ana a Huia i a Pani
T/A scold T/A pers Huia DO pers Pani
‘Huia is scolding Pani’

These two sentences would suggest that the passive in Maori is essentially the same as
the passive in, say, English: the underlying DO of the active becomes the subject in the
passive (it has Ø marking); the subject of the active becomes an oblique NP (marked with
the preposition e); the verb in the passive sentence acquires a suffix.

2.1.3.1.1.1 Passive verb morphology

The suffix for the passive is traditionally analyzed as having a variety of forms. The
attested forms vary considerably in frequency. They are (based on Williams, 1862, 39)
with an example of each:
suffix active stem gloss passive
-tia koohete ‘scold’ koohetetia
-a horoi ‘clean’ horoia
hanga ‘build hangaa
-ngia paa ‘touch’ paangia
-hia ruku ‘dive’ rukuhia
 Morphology 391

-mia inu ‘drink’ inumia

-ria mau ‘carry’ mauria
-kia hopu ‘catch’ hopukia
-ina aroha ‘love’ arohaina
-na ako ‘learn’ aakona
-nga kai ‘eat’ kainga
-whia, whao ‘fill’ whaowhia ~
-whina whaowhina ~
whaoria ~
whaongia ~
whaona
-hina *roko ‘be overtaken’ rokohina
-kina taataa ‘break’ taataakina

There are a number of points to note. Firstly, with forms like hangaa in texts not marking
vowel length, the passive appears to have a Ø allomorph, but this is misleading.
Secondly, the last four suffixes are very rare. Thirdly, there is considerable variation of
suffix in the case of certain lexical items, some of which is dialectal: in particular, some
dialects regularly have -ngia where others have -tia. Williams records different passives
for different senses of aroha in his Dictionary. Some speakers have free variation
between two or more suffixes for individual lexical items. The sort of variation which is
found in some individual instances is exemplified by whao above, but it should not be
taken that only whao of the above forms shows variation. Fourthly, the number of forms
showing any ending after the first three is relatively small. Fifthly, the initial consonant of
these suffixes is normally also found in the nominalization of that stem, but in some
cases, non-matching variants also occur:
stem gloss passive nominalization
hopu catch hopukia hopukanga ~ hoputanga ~ hopuranga
mau carry; mauria maunga ~mauranga
seize
inu drink inumia inutanga ~ inuranga ~ inumanga

With both hopu ‘catch’ and inu ‘drink’ my consultant was unfamiliar with the
nominalization matching the passive. (Notice also that there are no suffixes -pia, -ipa, -
pina; -wia (etc.); -nia, and compare the information on labial consonants in 3.2 and 3.4.)
There are also a few irregular passives, such as heuea < (he)heu ‘separate’, meinga <mea
‘make’ (sometimes meingatia), and homai, hoatu which do not change form.
These alternations have received considerable attention from linguists since Hale’s
review (Hale, 1968) of Hohepa 1967. There is little doubt that, historically, these
 Maori 392

consonants belonged to the stem (see eg. Hale, 1968, 87). Hohepa, however, derived
these forms by generating them all with the suffix -tia and then having a set of obligatory
rules to change this suffix to the required form (1967, 106). Hale suggested that if the
consonant was seen as part of the base, then the suffix could be seen as having two basic
forms, -ia after consonants and -a after vowels (1968, 86). (There would still be some
exceptions, for which Hale proposes further rules, some of which also provide an
explanation for the absence of certain potential suffixes from those actually attested.) The
debate over the appropriateness of these two analyses still rages in the linguistic
literature, see eg. Sanders (1990, 1991) (but note that there are some errors in the Maori).
Sanders argues forcibly for the correctness of the stem-consonant hypothesis.
A few further points relating to this debate must be noted:
(i) Stems which are principally nominal are passivized with -tia. (Sanders cites counter-
examples (1990, 160) from Williams and Hohepa, but all the counter-examples are
forms which have non-passive verbal uses, as well as nominal uses.)
(ii) Agreeing adverbials take -tia.
(iii) Borrowings from English most frequently take -tia, but this is not without exception,
eg. kuki, kukia ‘to cook’.
(iv) In some dialects, -tia can be used in cases of doubt as to the correct ending, but
others use -ngia as the default
Before leaving the topic of the form of the passive verb, it is necessary to point out that
certain other modifications to the stem are made in individual instances. If the stem
contains an initial reduplication, the passive form is usually not reduplicated:
titiro > tirohia ‘look’
tatari > taaria ‘wait for’

but note
hoehoe > hoehoea ‘ferry’
mukumuku > mukumukua ‘wipe’.

Changes to vowel length, as in tatari above, also occur irregularly. Harlow (1991)
suggests that these apparent irregularities may in fact be regular results of a repeated-
consonant deletion process.
Note in addition that those few verbs which traditionally take no passive suffix are
quite commonly suffixed nowadays. Also, following the mood particle me ‘prescriptive’,
no verb traditionally takes a passive suffix, though in all other respects the sentence may
be passive; this rule is also increasingly frequently broken.

2.1.3.1.1.2 Passive distribution

The other major consideration with respect to the Maori passive concerns its widespread
use, especially in past-time contexts. This raises the question of the relationship between
the active and passive in Maori. Again, some facts are outlined before the arguments are
reviewed.
Native speakers frequently reject sentences like
 Morphology 393

(1753) *I patu a Rewi i te kurii

T/A beat pers Rewi DO the dog
‘Rewi beat the dog’

They correct to either the passive (1754) or the actor-emphatic (1755):

(1754) I patu·a te kurii e Rewi
T/A beat·pass. the dog by Rewi
‘The dog was beaten by Rewi’
(1755) Naa Rewi te kurii i patu
actgen Rewi the dog T/A beat
‘Rewi beat the dog’

(Note that Hohepa (1967, 33–4) regards these two as transforms of each other.) However,
with other T/A markers the active is entirely acceptable:
(1756) E patu ana a Rewi i te kurii
T/A beat T/A pers Rewi DO the dog
‘Rewi is beating the dog’

The unnaturalness of the past active is reflected in narratives, where a rather high
proportion of transitive verbs are passive. Some examples are (with rather literal
translations to highlight the passives):
(1757) Ko tana wahine he tuutuuaa, he mea tango
top. sggenIIIsg woman cls commoner cls thing take
mai e ia i te tangata i aroha·tia
hither by IIIsg from the man T/A love·pass.
nui·tia e teenei wahine. Ko te take i
big·pass. by this woman top. the reason T/A
tango·hia ai e ia te wahine nei, he pai…
take·pass. part. by IIIsg the woman proxI cls good
‘His wife was a commoner, someome taken by him from the man who was dearly loved by
this woman. The reason this woman was taken by him was that she was good…’ (TA, 6)
(1758) Kaatahi ka tahu·na e ia te ahi, ka kaa
then T/A light·pass. by IIIsg the fire T/A burn
te ahi, ka tiro·hia e ia ngaa pakikau, ka
the fire T/A see·pass. by IIIsg the(pl) garment T/A
takoto kau ia
lie naked IIIsg
‘Then the fire was lit by him, and when it was alight, the clothes were seen by him, she was
lying bare’ (TA, 8)

However, not all transitives are passive, eg.

(1759) Kaatahi te tamaahine ka tahuri ki te
then the girl T/A turn oneself to the
taatai i a ia naa ka heru i a
arrange DO pers IIIsg now T/A comb DO pers
ia, naa ka raakei i a ia . ki
 Maori 394

IIIsg now T/A adorn DO pers IIIsg with

oona kaitaka, ka tia hoki i toona
plgenIIIsg cloak T/A stick in also DO sggenIIIsg
maahunga ki te raukura…
hair with the feather

‘Then the girl set about getting herself ready; she combed her tresses, adorned herself
with her ornamental cloak, and stuck feathers in her hair…’ (TA, 7)
Note the following from the same text:
(1760) Kaatahi ka kite mai taua koroheke raa i
then T/A see hither det aph old man dist DO
a ia ka rere i te pari. I kite·a
pers IIIsg T/A fly from the cliff T/A see·pass.
mai e ia ki ngaa kaakahu ka maa i
hither by IIIsg with the(pl) clothes T/A white at
toona rere·nga ai
sggenIIIsg fly·nom part
‘Then that old man saw her as [she] leaped from the cliff. [She] was seen by him with her
clothes gleaming white in her flight’ (TA, 8)

In addition to this “unexpected” distribution of passives in narratives (unexpected from

the point of view of a language like English), there is the fact that the passive is the verb
form of transitive imperatives, eg.:
(1761) Huri·hia too ika!
turn·pass. sggenIIsg fish
‘Turn your fish over!’ (P, 3)

These are true imperatives, and non-passive forms are not possible.
Hohepa (1967, 100) has both active and passive as alternative rewrites of VP. Hale
(1968, 88ff) questioned the motivation for this, and concluded that it was motivated by
certain facts about pronominalization. Hale proposed that passives in Maori be derived
from actives in a standard generative fashion, but suggested that pronominalization had to
be ordered before passive in Maori (unlike in English). Clark’s paper (1973) to the LSA
meeting in San Diego shows that Hale’s arguments fail. Clark proposes, on the basis of a
study of the passive in texts, that “passives are completive or perfective; actives are
incomplete, imperfective or hypothetical” (1973, 10), although this does not seem to
provide any explanation for the use of passives in imperatives. Chung (1978, 75ff) re-
examines the textual data, and argues that this does not support Clark’s hypothesis, since
she finds many passives which are not perfective (while agreeing that perfectives are
almost always passive). She proposes an alternative explanation, namely that passive
applies to affected DOs (1978, 80). She claims that reflexive objects (like those in
(1759)) do not refer, and hence are not affected. However, there seem to be instances of
the passive (eg tirohia in (1758)) which are still not well accounted for by this
hypothesis. Other examples of this sort can be adduced, like the following two, where the
first has two passives unaccounted for (since the events are projected, not completed),
and the second has an active which is not overtly reflexive:
 Morphology 395

(1762) …ka tuku·a e te paa te karere ki roto

T/A send·pass. by the pa the messenger to inside
o Hauraki kia tiki·na mai kia patu·a
gen Hauraki subj fetch·pass. hither subj kill·pass.
a Ngaati-Raukawa
pers Ngati-Raukawa
‘…the pa sent a messenger to Hauraki that they should come and kill Ngati-Raukawa’ (W,
198)
(1763) …ka haehae i ngaa ringa ki te mataa
T/A slash DO the(pl) arm with the obsidian flakes
‘…She slashed her arms with obsidian flakes’ (W, 198)

While this is notionally reflexive, there is no overt marking of the reflexivity to which a
rule like the one Chung proposes could respond. The following examples raise the
question of the interpretation of “affected”:
(1764) Ka tango ia i teetahi hua o taua raakau,
T/A take IIIsg DO a(sp) fruit gen det aph tree
ka panga·a iho
T/A throw·pass. down
‘He picked a berry from the tree and threw it’ (KM, 3)

I find it rather difficult to see how picking a berry affects it less than throwing it, although
this would be expected if Chung’s proposal is correct The following two sentences occur
within a few lines of each other:
(1765) Kaatahi ka ruku·hia e Hotu
then T/A dive for·pass. by Hata
‘Then [it] was dived for by Hotu’ (KH, 3)
(1766) Ka hoki te koorero ki a Hotu i ruku
T/A return the story to pers Hotu T/A dive
raa i te punga o too raaua waka
dist DO the anchor gen sggenIIIdl canoe
‘The story returns to Hotu who dived for the anchor of their canoe’ (KH, 3)

It is difficult to see how the anchor is affected differently in the two instances, although
the first is passive and the second active.
(1767) Ka moe a Rongo i a Turi; ka moe
T/A sleep pers Rongo DO pers Turi T/A sleep
hoki a Kura i a Hotu
also pers Kura DO pers Hotu
‘Rongo married Turi and Kura married Hotu’ (KH, 2)

Explaining the active here involves the claim that marrying someone affects that person
less than, for instance, sending them on an errand (compare (1762)). The matter thus
remains in need of further study. I suspect that, in addition to constraints like those
suggested by Clark and Chung, packaging for textual cohesion also plays a part, and that
the matter is far from simple.
 Maori 396

2.1.3.1.1.1 Personal passive

The personal passive has been described above.

2.1.3.1.1.1.1 DO of active as subject of passive

This is the commonest passive in Maori. The DO of the active takes case-marking in
the passive; the active subject takes e, and the verb takes the appropriate passive
allomorph, eg. (the passive (a) can be compared with the related active (b)):
(1768a) Kei te poowhiri·tia raatou e Hata
T/A welcome·pass. IIIpl by Hata
‘They are being welcomed by Hata’
(1768b) Kei te poowhiri a Hata i a raatou
T/A welcome pers Hata DO pers IIIpl
‘Hata is welcoming them’
(1769a) E puuhi·a ana ngaa kuuaka e Rewi
T/A shoot·pass. T/A the(pl) godwit by Rewi
‘The godwits are being shot by Rewi’
(1769b) E pupuhi ana a Rewi i ngaa kuuaka
T/A shoot T/A pers Rewi DO the(pl) godwits
‘Rewi is shooting the godwits’

2.1.3.1.1.1.2 IO of active as passive subject

Chung (1978, 173–4) claims that the passive in Maori is restricted to DOs, citing Mark
1970 as her main source of evidence. However my consultants accepted the following:
(1770) I paatai·tia te tamaiti ki te paatai e te
T/A ask·pass. the child to the question by the
kaiako
teacher
‘The child was asked a question by the teacher’

alongside
(1771) I paatai·tia e te kaiako he paatai ki te
T/A ask·pass. by the teacher a question to the
tamaiti
child
‘A question was asked the child by the teacher’

cf. the active:

(1772) I paatai te kaiako i te paatai ki te
T/A ask the teacher DO the question to the
tamaiti
child
‘The teacher asked the child a question’
 Morphology 397

Note that, in (1770), te paatai requires the marker ki (which it cannot take in the active).
This might suggest that the underlying IO is promoted to DO, and the underlying DO
demoted to IO before passivization, but there is no independent evidence for this.
However, not all underlying IOs can become passive subject, eg.
(1773) *I hoatu te kaumatua (k)i te taare e
T/A give the elder DO the doll by
tana mokopuna
sggenIIIsg grandchild
‘The old man was given a doll by his grandchild’

In fact, it is true to say that in general terms, notional IOs cannot become passive
subjects, and it is not clear under precisely what circumstances it is possible.

2.1.3.1.1.1.3 Other constituent of active as passive subject

Chung’s claim (see 2.1.3.1.1.1.2 above) excludes such possibilities. However, there are
many instances of passive subjects in Maori which are not clearly DOs. Locatives may be
subjects, eg.
(1774) He maha ngaa waahi o Niu Tiireni i
cls many the(pl) place gen NZ T/A
haere·a e teenei roopuu
move·pass. by this group
‘This group went to many places in NZ’ (more lit. Many are the places in NZ which were
gone to by this group’)
(1775) Kaua e taka·hia te taha o te awa o
neg T/A tread·pass. the side gen the river gen
Whanga-nui
Whanganui
‘Do not tread on the banks of the Whanganui River’ (TP, 91)

It is possible to classify such examples as DOs, of course, especially as they take the
preposition i in the active, although i is not necessarily a DO marker in Maori. But they
are not clear-cut instances of the category. The following is also possible:
(1776a) Ka koorero·tia tana ika e Rewi
T/A talk·pass. sggenIIIsg fish by Rewi
‘His fish was talked about by Rewi’

compare:
(1776b) Ka koorero a Rewi moo tana ika
T/A talk pers Rewi intgen sggenIIIsg fish
‘Rewi talked about his fish’

With respect to other properties shared by clear-cut DOs such as relativization strategies,
these types of NP behave as obliques, so it seems to me doubtful that the Maori passive is
restricted to DOs. Note also that it is quite common to find passives of verbs which are
usually intransitive where native speakers reject the corresponding active, eg.
 Maori 398

(1777a) Ko te mate o toona tipuna, o

top. the death gen sggenIIIsg grandfather gen
Poroumaataa, ka haere·a e Tuu-whakairi-ora
Poroumata T/A move·pass. by Tu-whakairi-ora
‘The death of his grandfather, Poroumata, was the occasion for Tu-whakairi-ora’s going’
(TWh, 19)
(1777b) *Ka haere a Tuu-whakairi-ora i te mate o
T/A move pers Tu-whakairi-ora DO the death gen
toona tipuna, o Poroumaataa
sggenIIIsg grandfather gen Poroumata
‘Tu-whakairi-ora went on account of the death of his grandfather, Poroumata’

The first shows topicalization of the subject (ko-fronting does not apply to adverbials);
the second was rejected with or without topicalization. It was accepted with matenga for
mate, but it is then clear that the phrase is adverbial.

2.1.3.1.1.2 Impersonal passive

Maori does not have an impersonal passive, although it should be noted that subjects of
personal passives are frequently omitted in narrative by the process of anaphoric deletion
(see 1.5), eg.
(1778) Ka mea atu a Kupe, “Ruku·hia te punga
T/A say away pers Kupe dive·pass. the anchor
o too taaua waka.” Kaatahi ka ruku·hia e
gen sggenIdlincl canoe then T/A dive·pass. by
Hotu
Hotu
‘Kupe said “Dive for the anchor of our canoe.” Then [it] was dived for by Hotu’ (KH, 3)

2.1.3.1.1.3 Expression of the subject of the active

It is always possible for the subject of the active to be expressed in passive sentences in
Maori. It is normally expressed in a prepositional phrase with the preposition e. The e-
marking is normal whether or not the subject of the active is agentive. The examples
illustrate:
(1779) Kua kauhoe·tia e ia te awa
T/A swim·pass. by IIIsg the river
‘He has swum the river’
(1780) Kei paa·ngia ia e te ruumaatiki!
mon touch·pass. IIIsg by the rheumatism
‘He might get rheumatism!’ (TR2, 6)
(1781) I pooki·a te rangi e ngaa kapua
T/A cover·pass. the sky by the(pl) cloud
The sky was covered over with clouds’
(1782) Ka piirangi·tia e ia ngaa mea katoa
T/A want·pass. by IIIsg the(pl) thing all
 Morphology 399

‘All the things are wanted by him’

(1783) Ka rapu te iwi raa, kite·a iho e
T/A search the tribe dist see·pass. down by
taa raatou rapu, he koohuru
sggenIIIpl search cls murder
‘The people consulted, and as a result of their consultation, suspected murder’ (W, 198)

It should perhaps be noted that e is occasionally used in sentences with no other

indications of passivity, eg.
(1784) Ko ngaa waka he mea toto ki uta; pee·raa
top. the(pl) canoe cls thing pull to shore like·dist
tonu i ngaa raa katoa, i ngaa poo hoki, e
still at the(pl) day all at the(pl) night also by
te iwi
the tribe
‘The canoes were drawn ashore; it was like this every day and at nights too, the tribe saw
to it’ (H, 7)

2.1.3.1.1.4.1 Tenses and aspects with the passive

The passive and active have the same range of possibilities, but the passive verb form
does not follow me ‘weak imperative’. However, a textual study of the frequencies of the
T/A markers would, I think, reveal certain differences of patterning. Perhaps the major
difference occurs with kua, which does not occur freely with actives, although it can in
questions, for instance. The following judgements were given by my consultant:
(1785) ______ aawhina/koohuru a Pani i a Tuu
help/murder pers Pani DO pers Tu
‘Pani HELP/MURDER Tu’

Kei te, i ie, e…ana, i, ka and ana were all acceptable. Kua was accepted only as a
question, and was less acceptable with koohuru than with aawhina. Me was rejected with
koohuru, presumably through semantic incongruence, rather than any general
ungrammaticality.
(1786) ______ aawhina·tia/koohuru·tia a Tuu e Pani
help·pass./murder·pass. pers Tu by Pani
Tu HELP/MURDER by Pani’

All T/A markers except me were accepted. Me was accepted with aawhina if the passive
termination was removed from the verb.
Note that the two verbs used have different object types, and that past tense markers
were not rejected with actives in either instance, which is somewhat surprising, given the
general tendency towards this.

2.1.3.1.1.4.2 Dynamic υs. static passive

 Maori 400

Basically, the Maori passive is dynamic, and in many cases the passive cannot be used
for states. However, consider the following examples:
(1787) E hanga·a ana te whare
T/A build·pass. T/A the house
‘The house is being built’
(1788) Kua hanga·a kee te whare
T/A build·pass. contr the house
‘The house is already built’

In the second, kee is required along with kua to capture the stativity. But it appears that
this is not a general pattern. The following two sets indicate two other common ways of
capturing this type of distinction in Maori.
(1789a) Mea koohuru a Pani (naa Tuu)
thing murder pers Pani actgen Tu
‘Pani was murdered (by Tu)’ (stative)
(1789bi) Ko Pani naa Tuu i koohuru
top. Pani actgen Tu T/A murder
‘Pani was murdered by Tu’ (active)
(1789bii) Ka koohuru·tia a Pani (e Tuu)
T/A murder·pass. pers Pani by Tu
‘Pani was murdered (by Tu)’ (active)
(1790a) Kua maoa ngaa kai
T/A cooked the(pl) food
‘The dinner was cooked’ (stative)
(1790bi) Naa Pani i whaka·maoa ngaa kai
actgen Pani T/A cause·cooked the(pl) food
‘The dinner was cooked by Pani’ (active)
(1790bii) I whaka·maoa·tia ngaa kai (e Pani)
T/A cause·cooked·pass. the(pl) food by Pani
‘The dinner was cooked by Pani’ (active)

The stative sentences in these sets are not passive. (1789a) uses a type of relative
construction, and (1790a) uses the corresponding adjectival intransitive verb. Thus the
way in which the static/dynamic distinction is made in any particular case is
unpredictable, since it depends on such lexical factors as whether there happens to be a
corresponding adjectival intransitive verb.

2.1.3.1.2 Means of decreasing the valency of a verb

There are various means of decreasing the valency of a verb in Maori, some of which
appear to be less than totally productive.

2.1.3.1.2.1 Intransitive from transitive by not specifying the subject

There seem to be occasional examples which show this, compare
 Morphology 401

(1791) E whaka·maoa ana a Hone i ngaa kai

T/A cause·cooked T/A pers John DO the(pl) food
‘John is cooking the dinner’
(1792) E whaka·maoa ana ngaa kai
T/A cause·cooked T/A the(pl) food
‘The dinner is cooking’

However, the pattern is not widely available. The DO of the transitive becomes the
subject of the intransitive, but similarities to the passive end there.
English examples of the type That dress washes well’ cannot be paralleled in Maori. A
non-verbal construction is used:
(1793) He pai teeraa kaakahu moo te horoi
cls good that dress intgen the wash
‘That dress washes well’

2.1.3.1.2.2 Intransitive from transitive by not specifying the DO

Cognate or predictable objects can be omitted, and regularly are in Maori, eg.
(1794) E kai ana a Hone
T/A eat T/A pers John
‘John is eating’

The ‘understood’ object is general, eg. i ngaa kai ‘the food’.

(1795) E karanga ana ia
T/A call T/A IIIsg
‘She is calling in welcome’

Note that the understood object here is of a specific kind, too: it is the ritual welcome
calls.

2.1.3.1.2.3 Reciprocal intransitive

With some examples of this type, my consultant had no doubt about the possibility of
creating an intransitive by conjoining underlying subject and DO as subject, eg.
(1796a) E kihi ana a Hone i a Mere
T/A kiss T/A pers John DO pers Mary
‘John is kissing Mary’
(1796b) E kihi ana a Hone raaua ko Mere
T/A kiss T/A pers John IIIdl spec Mary
‘John and Mary are kissing’

However, with other examples, there was uncertainty, not as to the grammaticality of the
sentence, but as to its meaning. Consider
(1797a) E whawhai ana a Hone i a Piri
 Maori 402

T/A fight T/A pers John DO pers Bill

‘John is fighting Bill’
(1797b) E whawhai ana a Hone raaua ko Piri
T/A fight T/A pers John IIIdl spec Bill
‘John and Bill are fighting’

My consultant was in considerable doubt as to the equivalence of these two. It is clear

that one possible reading of (1797b) is that John and Bill are fighting (on the same side)
some unspecified opponent (either mentioned before, or deducible from the real world
context). My consultant was very uncertain whether it could also have the same
interpretation as (1797a). I was unable to determine what factors might be involved, but
suspect that pragmatic expectation is involved, at least in the rejection of a non-reciprocal
reading as a possibility in (1796b).

2.1.3.1.2.4 Other means of decreasing verb valency

Three further possibilities must be mentioned here.

(i) The first is object incorporation, eg.
(1798) E tuhi·tuhi reta ana ia
T/A write·dup letter T/A IIIsg
‘She is letter-writing’

This creates a compound verb (note the position of the post-verbal particle ana) by the
incorporation of an indefinite DO. The fact that such constructions are intransitive can be
demonstrated by the possibility of using he in their subjects, eg.
(1799) He tamaiti e ruku·ruku kooura ana
a child T/A dive·dup crayfish T/A
‘A child is diving for crayfish’

(but note that my consultant found this odd if the subject was not preposed). See further
2.1.3.6.12.
(ii) The second construction (which decreases verb valency on some analyses only) is
the actor-emphatic (see 1.11.2.1.7), eg.
(1800) Naa Rewi i kai ngaa pea
actgen Rewi T/A eat the(pl) pear
‘Rewi ate the pears’

Both NPs remain in the sentence, of course, but the verb kai is considered by some to be
intransitive, with the underlying DO as its subject. There are no other similarities to the
passive, but note that Hohepa (1967, 33–4) treats this as a variant of the passive.
(iii) The third way in which the valency of a verb can be decreased on the surface is by
anaphoric deletion, eg.
(1801) Ko te mahi hoki a toona ariki e haere
top. the work also gen sggenIIIsg chief T/A move
 Morphology 403

tonu i ngaa raa katoa ki te titiro, ki te

still at the(pl) day all to the look to the
whaangai, ki te karakia i ngaa karakia taniwha
feed to the incant DO the(pl) incant taniwha
‘Its master used to visit it every day to look at [it], to feed [it], and to recite the taniwha
incantations’ (TP, 89)

2.1.3.1.3 Means of increasing the valency of a verb

Maori has a morphological process for increasing the valency of intransitives, but uses
periphrastic means to increase the valency of transitives.

2.1.3.1.3.1.1 Intransitive verbs made causative

The prefix whaka- is the major means of making an intransitive verb causative. Thus
from haere ‘move’, the form whakahaere ‘to run’ (eg. a meeting, a function, a machine)
can be derived. However, some of these derived forms have become lexicalized, and so
no longer function as straightforward causatives. Whaka- can be prefixed to neuter verbs
and adjectives, as well as more canonical intransitives, eg.
(1802) Kua mutu te hui
T/A finished the meeting
‘The meeting has ended’
(1803) Naa Hone i whaka·mutu te hui
actgen John T/A cause-finished the meeting
‘John has ended the meeting’

But the following pair shows a less direct relationship:

(1804) E haere ana a Hone i runga i te ara
T/A move T/A pers John at top at the path
‘John walked along the road’
(1805) E whaka·haere·ere ana a Hone i tana
T/A cause·move·dup T/A pers John DO sggenIIIsg
kurii i runga i te ara
dog at top at the path
‘John walked his dog along the road’

Whakahaere could not be used here, because of the specialization of sense mentioned
above. It is probably more natural to use a comitative to express the last:
(1806) E haere ana a Hone me tana kurii
T/A move T/A pers John with sggenIIIsg dog
i runga i te ara
at top at the road
‘John walked along the road with his dog’

Some further examples indicating the sorts of relationships between bases and whaka-
derivatives are given here:
 Maori 404

hoki ‘return’ whakahoki ‘reply;

‘take back’
maarama ‘clear, bright, lucid whakamaarama ‘explain’
oho ‘wake up, be awake’ whakaoho ‘waken’;
‘rouse’
pono ‘be true’ whakapono ‘believe’;
‘admit as true’

Where a whaka- form is not possible, because of lexicalization, a periphrastic causative

with mea ‘make, cause’ is often possible, eg. with whara ‘be hurt’:
(1807) Kua whara taku koonui
T/A hit accidentally sggenIsg thumb
‘I hurt my thumb’ (lit. ‘My thumb has been hurt’)
(1808) Naa Hone i mea kia whara taku koonui
actgen John T/A make subj hurt sggenIsg thumb
‘John made my thumb get hurt’

Whakawhara is associated with a different sense of whara, ‘be pressed’, so whakawhara

means ‘press, touch’, and is thus not available for ‘to cause to be hurt’.

2.1.3.1.3.1.2 Transitive verbs made causative

The only regular means of making transitive verbs causative is the verb mea ‘cause,
make’ (which has an irregular passive, meinga, or meingatia). Compare the following:
(1809) Naa Pita i whaka·maoa ngaa kai
actgen Peter T/A cause·cooked the(pl) food
‘Peter cooked the dinner’
(1810) Naa Pani a Pita i (tino) mea ki te
actgen Pani pers Peter T/A very make to the
whaka·maoa kai
cause·cooked food
‘Pani made Peter cook the dinner’
(1811) I meinga a Pita e Pani ki te
T/A make·pass. pers Peter by Pani to the
whaka·maoa kai
cause·cooked food
‘Peter was made to cook the dinner by Pani’

There are also a number of instances where transitive verbs are prefixed by whaka-, but
this does not clearly create a causative in all cases. Thus rongo ‘hear’, whakarongo listen,
cause to hear, inform’; ako ‘learn, teach’, whakaako ‘teach’. However, whaka- cannot
generally be prefixed to transitive verbs.
 Morphology 405

2.1.3.1.3.1.3 Ditransitive verbs made causative

The only available means of making ditransitive verbs causative is the use of mea, ‘cause,
make’, eg.
(1812) Naa Hone i hoatu ki a Pita te maaripi
actgen John T/A give to pers Peter the knife
‘John gave Peter the knife’
(1813) Naa Pou i mea ki a Hone kia hoatu
actgen Pou T/A make to pers John subj give
te maaripi ki a Pita
the knife to pers Peter
‘Pou made John give Peter the knife’
(1814) I meinga a Hone e Pou kia hoatu te
T/A make·pass. pers John by Pou subj give the
maaripi ki a Pita
knife to pers Peter
‘Hone was made to give the knife to Peter by Pou’

My consultant varied in the marking assigned to the NPs in these causative constructions.
Compare the above with
(1815) Ka paatai·a e te rooia ki a Pita ko wai
T/A ask·pass. by the lawyer to pers Peter eq who
tana ingoa
sggenIIIsg name
‘Peter was asked his name by the lawyer’
(1816) Naa te kaiwhakawaa i mea ki te rooia
actgen the judge T/A make to the lawyer
kia paatai·a ki a Pita ko wai tana ingoa
subj ask·pass. to pers Peter eq who sggenIIIsg name
‘The judge made the lawyer ask Peter his name’

On a subsequent occasion, the following was given:

(1817) Naa te kaiwhakawaa i mea ki te rooia
actgen the judge T/A make to the lawyer
kia paatai ki a Pita ko wai tana ingoa
subj ask to pers Peter eq who sggenIIIsg name
‘The judge made the lawyer ask Peter his name’
(1818) I meinga e te kaiwhakawaa ki te rooia
T/A make·pass. by the judge to the lawyer
kia paatai·a ki a Pita ko wai tana ingoa
subj ask·pass. to pers Peter eq who sggenIIIsg name
‘The judge made the lawyer ask Peter his name’

I suspect that the variation is the result of stretching the system beyond its normal limits,
ie. such causatives are not usual, and some indirect paraphrase, eg. ‘When the judge
 Maori 406

instructed him to, the lawyer asked John his name’, is a much more likely means of
encoding such messages.

2.1.3.1.3.3 Omission of causee

The causees in periphrastic constructions with mea can only be omitted from the passive,
and not from the actor-emphatic formulations. The omission of the passive agent does not
lead to ambiguity, eg.
(1819) I meinga ahau ki te whaka·maoa kai
T/A cause·pass. Isg to the cause·cooked food
‘I was made to cook the dinner’

In non-passive, non-A-E constructions with mea, the only way to omit the causee appears
to be to pseudo-cleft. Compare the following:
(1820) E mea ana a Pani i a Pou ki te
T/A make T/A pers Pani DO pers Pou to the
horoi i te papa
clean DO the floor
‘Pani is making Pou clean the floor’
(1821) Mea mea a Pou ki te horoi i te papa
thing make pers Pou to the clean DO the floor
‘Pou is being made to clean the floor’ (more lit. ‘Pou is the
one being made to clean the floor’)

With the whaka- causatives from intransitives, the causee cannot simply be omitted. It is
possible to use, eg.
(1822) Mea whaka·mutu teenei hui
thing cause-finished this meeting
‘This meeting is over’

which preserves the causative form of the verb, and does not specify the causee, but the
non-causative form (1802) is equally likely.

2.1.3.1.4 Special reflexive or reciprocal verb forms

Maori does not have any special reflexive or reciprocal verb forms.

2.1.3.1.5 Neuter verbs

While these verbs do not clearly function as a separate voice in Maori, they nevertheless
are a voice-like phenomenon, and thus are treated briefly here. This group of verbs has
been recognized as requiring special treatment by all previous scholars, although there
has been considerable disagreement as to the most approriate label for them. The most
extensive treatments are Hooper, 1982, 1984a, 1984b.
There are relatively few neuter verbs in Maori, the common ones being ea ‘be
avenged’, hinga ‘be defeated’, mahue ‘be left behind’, mau ‘be caught’, mutu ‘be ended’,
 Morphology 407

oti ‘be finished’, pakaru ‘be broken’, pau ‘be consumed’, riro ‘be seized’, whati ‘be
broken’. The criteria which distinguish them from other verbs are (i) neuter verbs cannot
be made passive; (ii) they have as their subject a non-agentive NP; (iii) if an agent is
included, it occurs in an oblique phrase marked with i, eg.
(1823) Kua pau i a ia te kai
T/A exhausted cause pers IIIsg the food
‘She has finished the food’ (more lit. The food is finished because of her’)
(1824) Kua mau i a Tuu teetahi manu
T/A caught cause pers Tu a(sp) bird
‘Tu has caught a bird’

It will thus be seen that there are certain parallels with passives, compare (1799) with
(1825) Kua hopu·kia teetahi manu e Tuu
T/A catch·pass. a(sp) bird by Tu
‘Tu has caught a bird’

However, there is no formal resemblance to the passive; the marking of the NPs is
ergative in type. Neuter verbs also share some features with adjectival predicates, but
unlike adjectival predicates, they are not notionally stative. Hohepa (1969, 8–17) reviews
the similarities and differences, the most important being that neuter verbs cannot be used
as modifiers to other verbs, and while adjectives have the proform peeraa, neuter verbs
do not.
(Note that it is not possible for the same lexeme to be both a neuter verb and a non-
neuter verb, although certain orthographic forms eg. mau occur in both classes, but with
different senses.)
Neuter verbs also allow the following construction:
(1826) Kua oti kee i a Pou te taarai
T/A finished contr cause pers Pou the adze
toona waka
sggenIIIsg canoe
‘Pou had already finished making his canoe’ (P, 1)

There are, unusually for Maori, two zero-marked NPs in this construction. Its derivation
is discussed in Hooper 1984a and Waite 1989.

2.1.3.2 Tense

2.1.3.2.1 Formal distinction of tense

There are many problems associated with the analysis of the tense-aspect-modality
(T/A/M) area in Maori. Many of these problems cut across the analytical categories
implied by the Questionnaire, and so the problems are discussed here first, before an
attempt is made to supply the analysis.
 Maori 408

Firstly, there is not complete agreement on the list of particles which can function as
T/A/M markers in Maori, especially in regard to tense. The following list contains those
which are considered by at least some analysts to be involved in the encoding of tense:
ka e i te
kua e…ana kei te
i ana ina(a)

There are other markers which occupy the same slot, but which are exclusively aspectual
or modal. Of the above, only three seem to me not to require further comment: ka, i and
e. Kua requires comment only in that it is not entirely clear that it marks tense, rather than
aspect. It has usually been described as perfect(ive) (eg. Williams, 1862, 33; Biggs, 1969,
34; Hohepa, 1967, 19), although I have argued that it is inchoative (Bauer, 1981a, 64ff.)
rather than perfective on many occasions.
The two markers kei te and i te are labelled by Biggs “pseudo-verbal continuous”
markers (1969, 86). The label refers to the fact that these are transparently formed from a
locative preposition + determiner te, so that they superficially maintain non-verbal
characteristics. However, the constructions in which they occur show no other nominal
characteristics, and so they are treated here as unitary T/A markers, on a par with ka and
i, for instance. However, they are largely excluded from subordinate constructions, which
may be a reflection of their origin as overtly tense-marked forms. ‘Superficially nominal’
might be a preferable label from the point of view of their function.
The problem raised by e…ana must be considered in conjunction with ana and e. This
has been treated as a discontinuous imperfect/continuous marker by all the sources cited,
and I believe that this is a reflection of the intuitions of native speakers with respect to
this marker. The acceptance of this analysis is complicated by the ability of each part of
this complex marker to occur independently. Harlow, however, (1989, 198) asserts the
independence of e…ana from its constituents. One fact not in doubt is that e…ana is a
relative tense marker: it can be past, present or future in appropriate contexts.
E occurs principally in subordinate constructions in modern Maori, and it is usually
described as either future (Williams, 1862, 33) or non-past (Biggs, 1969, 34, Hohepa,
1967, 19). However, there are textual examples where it appears in past-time contexts,
eg.
(1827) Heoi, kaatahi ka moohio·tia ko·ia nei anoo
thus then T/A know·pass. top.·IIIsg proxI again
e patu nei i ngaa ope tuatahi e
T/A kill proxI DO the(pl) group first T/A
ngaro nei
lost proxI
‘Thus it became known who it was who had killed the first groups that had vanished’ (TP,
90)

This is particularly often the case for habitual past events, eg.:
(1828) Ko te mahi hoki a toona ariki e haere
top. the work also gen sggenIIIsg chief T/A move
tonu i ngaa raa katoa ki te titiro…
 Morphology 409

still at the(pl) day all to the look

‘Its master used to visit it every day to look at it…’ (TP, 89)

The following example is actor-emphatic:

(1829) Maa·na e whaka·piko te ara o te ika ki
intgen·IIIsg T/A cause·bend the path gen the fish to
taana waahi e hiahia ai
sggenIIIsg place T/A desire part.
‘He could make a fish go wherever he wanted it to’ (KH, 1)
(1830) …kua kite·kite noa ake hoki raaua i a raaua,
T/A see·dup indeed up also IIIdl DO pers IIIdl
i ngaa waa e hui·hui ai ngaa taangata
at the(pl) time T/A meet·dup part. the(pl) people
o Rotorua
gen Rotorua
‘…they had seen each other a lot at the times when the people of Rotorua gathered
together’ (H, 6)

Such examples occur alongside present and future (or hypothetical) examples, both
habitual and non-habitual:
(1831) Kaati, tuku·a mai ki Hikurangi, ki te
enough receive·pass. hither to Hikurangi to the
maunga e tau·ria i te huka
mountain T/A adorn·pass. by the snow
‘Enough; let him come hither to Hikurangi, to the mountain crowned with snow’ (TWh,
19)
(1832) E tangi he puu i ngaa poo katoa
T/A sound a flute at the(pl) night all
‘A flute plays every night’ (H, 7)
(1833) E mau nei anoo aua ingoa
T/A carry proxI again det aph(pl) name
Those names still remain’ (TWh, 17)
(1834) Kaati, ko au anake e haere
enough eq Isg alone T/A move
‘Enough; I alone will go’ (TWh, 19)
(1835) Me·i noho atu ia i te paa, e roa te
if·T/A stay away IIIsg at the pa T/A long the
kawe·nga
persevere·nom
‘If he had remained in the pa, we would have had a long job’ (TWh, 22–3)

This data suggests that, rather than being an absolute non-past marker, e is a relative
marker, which takes its temporal reading from the context. (Notice that the last 4
examples are not subordinate, and (1832) and (1834) are counter-examples to Chung’s
claim that e occurs with non-stative verbs only in embedded clauses (Chung, 1978, 177).)
It will be recalled that e…ana is also a relative tense marker. Thus it is conceivable that
 Maori 410

the e of e…ana and independent e are the same morph. If that is the case, then it is
necessary to consider whether there is independent support for ana as a continuative
(aspectual) particle. Evidence for that would provide support for a bimorphemic analysis
of e…ana as a tense particle e+and aspectual particle, ana.
Ana alone is used in narratives, often in contexts where it appears equivalent to
e…ana, eg.
(1836) Koowatawata ana ngaa uru maawhatu i te
gleam T/A the(pl) hair curly at the
hana o te ahi; ko toona tinana, ngangana
flame gen the fire top. sggenIIIsg body glow
ana; ko toona kiri karengo kau ana; ko
T/A top. sggenIIIsg skin smooth intens T/A top.
te kanohi anoo he rangi raumati paruhi kau
the face again a sky summer calm intens
ana;…
T/A
‘Her curly tresses were gleaming in the firelight; her body flowing; her skin silky smooth;
her face like a beautiful summer’s day…’ (TA, 6)

ie. in contexts like this, it is continuative. However, on other occasions, it appears to mark
punctual events, eg.
(1837) Awatea kau ana, ka haere te koroheke raa…
day intens T/A T/A move the old man dist
‘When morning came, that old man left…’ (TA, 6)
(1838) Whakautu·a ana e Kura ki tana matua…
reply·pass. T/A by Kura to sggenIIIsg parent
‘Kura answered her father,“…”’ (KH, 2)

Previous scholars have differed somewhat in their treatment of this particle. Williams
describes it as a “narrative past” (Williams, 1862, 36); most later writers merely follow
this, although Biggs is not specific as to its status: it is not included in his list of T/A
markers (1969, 34), but its post-verbal position would be sufficient to exclude it, given
Biggs’ criteria for this list. However, he says such phrases are “unambiguously marked as
verbal” (1969, 71) and “common in animated narrative in the past tense” (1969, 72).
Hohepa 1967 does not handle this particle at all. All these accounts agree on the
description “narrative” and “past”. However, the following example, while from a
narrative, is not past:
(1839) …ko teenei ka puta ia ki waho, ka mate
top. this T/A appear IIIsg to outside T/A die
aakuanei, aa taui ana too ringa i
presently and strained T/A sggenIIsg hand cause
te patu·nga
the kill·nom
‘…but now he has come forth, he will soon succumb, and your hand will be strained by the
slaughter’ (TWh, 20)
 Morphology 411

An example of this sort might be explained as having an e elided, but this supposes that it
is possible to distinguish two uses of ana which are clearly not formally distinct Thus I
think a question-mark must remain over the association of ana and past time.
In a small corpus of 33 textual examples of ana alone, (ie. without e), 18 are clearly
continuative, but many of the remainder are clearly non-continuative. This does not
provide clear support for the analysis of ana as a continuative particle. (Neither does it
confirm the “narrative past” analysis.) Only 4 of the 33 are in subordinate structures, so it
is clearly not a particle dependent on another verb form. However, there is one instance
of ana co-occurring with kei te, rather than e:
(1840) …ko te tane raatou ko ngaa hoa kei te
top. the husband IIIpl spec the(pl) friend T/A
hoe mai ana ki uta
row hither part. to shore
‘…her husband and friends were paddling to shore’ (TA, 7)

This confirms the independence of ana from e.

The evidence available on the subject of ana is thus inconclusive. Further support for
its analysis as an aspectual-type particle may, however, come from a consideration of
other post-verbal particles. There are many textual examples of other post-verbal particles
occurring with no preceding tense marker, eg.
(1841) Haere tonu, ka tae ki Aaniwaniwa
go still T/A arrive to Aniwaniwa
‘[She] went on and reached Aniwaniwa’ (W, 197)
(1842) Mahi noa, naawai raa, ka puta mai te
work extend presently dist T/A appear hither the
mate moe ki a Te Tahi
lack sleep to pers Te Tahi
‘They worked on, and then presently Te Tahi felt sleepy’ (KWh, 2)
(1843) Hua·ina iho ki te tau·nga ko Kamokamo
name·pass. down to the anchor·nom eq? Kamokamo
‘The fishing ground was called Kamokamo’ (TWh, 17)

(Note however, that the presence of such post-verbal particles does not preclude the use
of a T/A marker, eg.
(1844) Ka tino kaha rawa toona tangi ki toona
T/A very strong intens sggenIIIsg cry to sggenIIIsg
mamae
pain
‘Her lamenting at her injury continued strongly’ (TWh, 18)

In the cases I checked, a wide range of T/A markers was judged possible with the
verb+particle combinations that occurred in textual examples.) Harlow (1989, 196–7)
calls this T/A marker “adverbial”, because such clauses frequently have “adverbial
senses”. However, in my data, there are many like (1843) where the sense is not clearly
adverbial, and the clause is independent.
 Maori 412

In a corpus of 75 examples of verbs without a preceding particle, only 4 examples had

no particles following. (For some discussion of these particles see 1.2.1.3.1.1). 46 (almost
two-thirds) had a manner particle (tonu: 27; rawa: 11; noa: 8); 17 had a directional
particle; 3 had ai; 1 had anake, an emphatic particle, and 4 had katoa (a raised
quantifier). (These figures account for the first particle only: 13 of those with manner
particles also had directional particles; 3 of those with initial directional particles also had
emphatic particles.) None had deictic particles. However there is evidence from
relativization that ai and ana go in the same slot as deictic particles. If the examples with
ana are included, of 108, 36 would then have ‘deictic’ particles, and 8 of these also
include a directional particle. Figures for the first particle only are then:
Manner Directional ‘Deictic’ Emphatic katoa
46 25 28 1 4 4

Two of the three manner particles have senses which could be aspectual, tonu ‘still’ and
noa ‘extend’ (see Mutu-Grigg, 1982). I have also argued elsewhere (Bauer, 1981a, 67)
that ai has an aspectual function in at least some dialects. While all of these things lend
some credence to the analysis of ana as a post-posed aspectual particle, none of them
provides firm proof. Thus the most suitable analysis for the particle ana remains
uncertain. However, it seems that when it occurs independent of e, it is not a marker of
tense, any more than, say, tonu.
It is just possible that there are two homophonous ana’s post-verbally. While other
particles appear to have fixed order relative to each other, ana has somewhat variable
position. While it most commonly follows mai, it can also precede it, eg.
(1845) Anoo! toorino kau ana mai i runga i te
again glide swim part. hither at top at the
kare o te wai…
ripple gen the water
‘Now gliding in on the ripples of the water…’ (TA, 7–8)

However, there is no evidence that the differing positions are associated with difference
in sense, so it is not possible to support two ana’s in this way. We are thus left with a
post-verbal particle which is sometimes punctual and sometimes continuative.
There is one further particle in the list to discuss, and that is inaa, which is listed as a
verbal particle by Biggs (1969, 34) and Harlow (1989, 197). One of the major problems
posed by this particle is its relative rarity—there is a paucity of data on which to base
conclusions. It precedes the verb, ie. it occupies the same slot as the undisputed tense
markers, eg.
(1846) I hiahia a Himi hei taakuta a Piri
T/A desire pers Jim cls(fut) doctor pers Bill
inaa tangata ia
T/A? adult IIIsg
‘Jim wants Bill to be a doctor when he grows up’
(1847) Inaa tae mai koe, ka kai ai taatou
T/A? arrive hither IIsg T/A eat part. Iplincl
‘When you get here well have a meal’
 Morphology 413

Inaa occurs only in subordinate time clauses referring to the future. Biggs says it means
‘if and when’ (1969, 34) and this is accepted by Harlow (1989, 197). It is not clear from
this whether it is a T/A marker, or whether it is a conjunction which is followed by a non-
tense-marked verb. The latter analysis is espoused by Williams in the entry in his
Dictionary, and is accepted here.
As a result of all these deliberations, I conclude that the markers relating to tense in
Maori are: ka, i, e, e…ana, kei te, i te, and kua.
It is perhaps also worth pointing out that the descriptions of e and ka which I have
suggested overlap. This perhaps provides an explanation for the fact that e is rare in main
clauses in modern Maori, and is largely replaced by ka, while ka is largely (although not
exclusively) confined to main clauses. There are sufficient examples of main clause e in
older texts to warrant the assumption that e was possible in main clauses in former times.
If the two particles had the same function, then they were in direct competition (as
relative markers without aspectual features), and subsequent functional differentiation is
a predictable outcome. It appears likely that e which frequently occurred with ana has
become almost exclusively restricted to that co-occurrence in main clauses, so that the
two do not compete, while in those subordinate clauses where ka does not occur, e has
retained a broader range of functions. Harlow (1989), however, argues that ka is an
injunctive, ie. that it has no tense, mood, or aspectual value, but merely marks the phrase
in which it occurs as verbal. (The parallel to my suggestion that te constitutes a ‘default’
article, marking the phrase as nominal, is interesting, see 2.1.1.10.2.)

2.1.3.2.1.1 Universal
No single tense marker can be associated with this function in Maori. Firstly, many such
statements are made in non-verbal constructions which have no tense marker (note that
Waite (1990, 403) suggests that e before numerals is a tense marker, but see the
discussion in 1.1.1.3.1), eg.
(1848) E rua ngaa pari·nga o te tai i te raa
num 2 the(pl) flow·nom gen the tide at the day
‘The tide comes in twice a day’
(1849) Ko te raumati te waa kauhoe
eq the summer the time swim
‘Summer is the time for swimming’
(1850) He tino rongonui ngaa ngaawhaa o Rotorua
cls very famous the(pl) spring gen Rotorua
‘The hot springs of Rotorua are very famous’

Note that such statements are not necessarily universal:

(1851) E rua ngaa whakawhitiwhiti·nga i teenei raa
num 2 the(pl) cross·nom at this day
‘There are two crossings today’

The commonest verbal particle in verbal universal sentences is ka, eg.

(1852) Ka mataku ngaa kararehe i te whatitiri
 Maori 414

T/A frightened the(pl) animal cause the thunder

‘Thunder frightens animals’
(1853) I te koanga, ka horo te tupu o te puuhaa
at the spring T/A fast the grow gen the puha
‘Puha grows fast in Spring’

In other contexts, these would not necessarily make universal statements: the latter could,
for instance, be a statement about a particular spring in the past, although in this case, raa
would probably be added to the adverbial. It is also possible to use e…ana on some
occasions for general statements, eg.
(1854) E raranga ana ngaa katipoo i oo raatou
T/A weave T/A the(pl) spider DO plgenIIIpl
puungaawere
web
‘Spiders spin webs’

‘This could also mean The spiders are spinning their webs’. In negative universals, e is
usual, eg.
(1855) E kore ngaa kiwi e rere
T/A neg the(pl) kiwi T/A fly
‘Kiwis don’t/can’t fly’

This could also be used to make a prediction about particular kiwis. The initial e in
examples like (1855) is frequently omitted. Some dialects use ai, which is probably an
aspectual particle in such instances, for certain types of universal statement, eg.
(1856) Pee·hea ai taa te ngeru tangi? Miao ai
like·how T/A sggen the cat sound miaow T/A
‘What kind of sound do cats make? They miaow’

2.1.3.2.1.2 Present
There are four main possibilities for marking present tense in verbal contexts in Maori,
ka, e…ana, kei te and e. Kei te is much commoner in Eastern dialects than elsewhere. The
middle two are not pure tense markers: they also mark the action as on-going:
(1857) E moe ana te peepi
T/A sleep T/A the baby
‘The baby is sleeping’
(1858) Kei te moe te peepi
T/A sleep the baby
‘The baby is sleeping’

These two are equivalent, and all Maori speakers are familiar with both, although they
differ in frequency (and in some cases form, eg. ko te for kei te) from dialect to dialect.
The kei te form is not possible in all embedded constructions, and e…ana replaces it
under these circumstances. However, e…ana is a relative tense marker, and takes its
 Morphology 415

reading from any time adverbial present. Its default reading is present Kei te, on the other
hand, is an absolute tense marker, and is marked as non-past. Its default reading is present
rather than future. However, it can also occur in narrative as a historic present, eg.
(1859) Te tae·nga atu o ngaa tamariki, ka
the arrive·nom away gen the(pl) children T/A
whakatatanga ia, ka riro atu, ka noho anoo
stand up IIIsg T/A take away T/A sit again
ia. Kei te kaakahu ngaa waahine raa, kei te titiro
IIIsg T/A dress the(pl) women dist T/A look
whakatau mai ki a ia… Kei te mea hoki ia
intently hither to pers IIIsg T/A say also IIIsg
ki toona kore i paatai ki ngaa tamariki
to sggenIIIsg neg T/A ask to the(pl) children
raa ko teewhea a Ruataupare
dist eq which pers Ruataupare
‘When the children came, he got up at once and gave them up, and sat down again. While
the women were putting on their clothes, they were gazing intently at him… He was asking
himself why he had not questioned the children as to which was Ruataupare’ (TWh, 19)

The other possibility is ka, which is not progressive. It is used in acts of naming, for
instance:
(1860) Ka tapa au i tana ingoa ko Huia
T/A name Isg DO sggenIIIsg name eq Huia
‘I name this child Huia’

It can also be used in running commentary, for instance:

(1861) Ka panga·a atu te paaoro e Hone ki a
T/A throw·pass. away the ball by John to pers
Pita
Peter
‘John passes the ball to Peter’

(Either e…ana or kei te is also possible here.). The default reading for ka is non-past, but
it is not entirely clear whether a present or future reading is more normal in the absence
of any specific time adverbial. Ka is also excluded from some embedded contexts, where
its place is often taken by e, which I argued above was a relative tense marker. Except
before the negative with kore, e occurs only in residual main clauses in modern Maori,
and is almost in complementary distribution with ka. An example of e as a present tense
marker in a subordinate clause is:
(1862) …tae noa ki uta ki te waahi e tuu nei
arrive indeed to shore to the place T/A stand proxI
te whare nei, a Wairaka
the house proxI pers Wairaka
‘…and arrived at the shore, at the place where the house named Wairaka now stands’
(KWh, 3)
 Maori 416

There are also examples where no tense marker is used before the verb when the sense is
present, eg.
(1863) Aa mohoa noa nei mahara·tia tonu·tia
and to the present indeed proxI remember·pass. still·pass.
e maatou te ingoa o teeraa toka ko
by Iplexcl the name gen that rock spec
Te-Rere·nga-O-Te-Aohuruhuru. Aa mahara·tia
The-jump·nom-gen-Te-Aohuruhuru and remember·pass.
tonu·tia hoki e maatou ngaa kupu o taana
still·pass. also by Iplexcl the(pl) word gen sggenIIIsg
waiata
song
‘To this day that rock is known to us as Te Rerenga O Te Aohuruhuru, and we still
remember the words of her song’ (TA, 8)

This normally only happens in formal contexts if there is an adverb associated with the
verb which indicates the temporal reference. However, in informal Maori, initial tense
markers are frequently omitted. Tonu ‘still’ is especially frequent in such T/A-markerless
verb phrases.
It should perhaps be noted also that there are occasional instances where i, normally a
past marker, apparently has a present reading, eg.
(1864) “Kei whea rawa te waahi i noho ai
at(pres) where intens the place T/A live part.
te nanakia?”
the monster
‘“Where is the place where the monster dwells?”’ (TP, 91)

This might be explained by the time-marked adverbial preceding, but e is usual in such
contexts. No such explanation exists for cases like the following provided by my
consultant:
(1865) I hiahia a Himi hei taakuta a Piri
T/A desire pers Jim cls(fut) doctor pers Bill
inaa tangata ia
when adult IIIsg
‘Jim wants Bill to be a doctor when he grows up’

Neither of these can be written off as mere slips. In (1864), if the T/A marker was e, the
particle ai following the verb would have to be replaced by a deictic. I questioned my
consultant about the use of i in examples like (1864), and was assured that a present tense
translation was appropriate. From a different consultant I received the following:
(1866) Naana i whakamaatau te wai i moohio
actgen T/A test the water T/A know
ai ia he makariri te wai
part. IIIsg cls cold the water
‘My brother tested the water, and he can feel that the water is cold’
 Morphology 417

I have no explanation to offer, but point out that apart from the example from Tutae-
poroporo ((1864)), all the examples I have involve experience verbs. Discrepant data of
this kind can be adduced almost everywhere in the field of tense/aspect marking in
Maori, and it is not at all clear what significance it is to be accorded. The generalizations
in all sections are thus true only in the majority of instances.
In non-verbal sentences which are equative or classifying, the unmarked reading is
present tense, eg.
(1867) Ko au te kaiako
eq Isg the teacher
‘I am the teacher’
(1868) He maa tana motokaa
cls white sggenIIIsg car
‘Her car is white’

In locative prepositional sentences, present tense is marked by kei, eg.

(1869) Kei Rotorua ia
at(pres) Rotorua IIIsg
‘She is in Rotorua’

In possessive prepositional sentences, naa indicates present possession, eg.

(1870) Naa·ku ngaa kura raa!
actgen·Isg the(pl) feather dist
Those feathers belong to me’ (P, 3)

However, it should be noted that the actor-emphatic with naa does not have a present
reading.
It is perhaps necessary to spell out separately the system in various types of
subordinate clause. Firstly, ka is normal in noun clauses either as subject or as DO, eg.
(1871) Ka mea mai ia, ka pai tana kaiako
T/A say hither IIIsg T/A good sggenIIIsg teacher
ki te waiata
to the sing
‘He says that his teacher is good at singing’

However, ka does not occur as the T/A marker with negatives, and nor does kei te. Ka is
replaced by e, and kei te by i te or e…ana, eg.
(1872) Kaaore ia e moohio ki te waiata
neg IIIsg T/A know to the sing
‘He doesn’t know how to sing’
(1873) Kaaore te hau i te pupuhi
neg the wind T/A blow
‘The wind is not blowing’
(1874) Kaaore ahau e pai ana ki te kai hikareti
neg Isg T/A good T/A to the eat cigarettes
‘I don’t like smoking cigarettes’
 Maori 418

There is also a special ‘not yet’ negative construction, eg.

(1875) Kaahore anoo ia kia haere
neg again IIIsg subj move
‘He has not yet gone’

Note that this contains the subjunctive kia, and is thus probably a modal rather than a
tense negation.
In relative clauses, ka and kei te are extremely rare, and e or e…ana are the norm.
However, the first two are attested occasionally. An example of each is given:
(1876) E rua eenei ka hoatu e ahau ki a koe,
num 2 these T/A give by Isg to pers IIIsg
ko Kupe raaua ko Hoturapa
spec Kupe IIIdl spec Hoturapa
‘There are these two whom I offer to you, Kupe and Hoturapa’ (KH, 2)
(1877) Ko te toko·rua anake nei ngaa mea kei te
eq the pnum·2 only proxI the(pl) thing T/A
maatakitaki·tia
watch·pass.
‘These two are the only ones being watched’ (TR2, 152)
(1878) E rima ngaa iwi e noho mai nei i
num 5 the(pl) tribe T/A live hither proxI at
te Tai Tokerau
the Northland
‘There are 5 tribes who live in North Auckland’ (KW, 1)
(1879) He kau anake ngaa mea e haere ana i
cls cow only the(pl) thing T/A move T/A at
runga i ngaa rori pee·nei i Te Kao
top at the(pl) road like·proxI compar Te Kao
‘Cows are the only things that use roads like this in Te Kao’ (TR1, 141)

Thus it is clearly not entirely accurate to say, as Chung does (1978, 21) that ka is not an
embedded clause particle. This statement appears to be true only of negatives and actor-
emphatics. Actor-emphatics are rarely present, but the following example uses e with
maa:
(1880) I eenei raa, maa te miihini e mahi te
at these day intgen the machine T/A do the
nui·nga o ngaa mahi
big·nom gen the(pl) work
‘These days, machines do most of the work’ (TR2, 59)

2.1.3.2.1.3 Past
Past tense is usually marked by i or i te (both absolute tense markers), or by ka or e…ana
(relative tense markers). I te and e…ana are both progressive; i and ka are unmarked
aspectually. Like kei te, i te is commoner in Eastern dialects, but seems less widely used
than kei te by speakers of other dialects. It is excluded from many embeddings, and
 Morphology 419

replaced by e…ana. Ka is also excluded from some embeddings, and i is used. To obtain
past readings for ka and e…ana, a preceding adverbial is required, eg.
(1881) I pupuhi te hau
T/A blow the wind
‘The wind blew’
(1882) Inapoo nei ka pupuhi te hau
last night proxI T/A blow the wind
‘Last night, the wind blew’
(1883) Inapoo nei, e pupuhi ana te hau
last night proxI T/A blow T/A the wind
‘Last night, the wind was blowing’
(1884) I te pupuhi te hau
T/A blow the wind
‘The wind was blowing’
(1885) I tihewa ia i te poo roa nei
T/A sneeze IIIsg at the night long proxI
‘She sneezed all night’
(1886) I te poo roa nei, ka tihewa ia
at the night long proxI T/A sneeze IIIsg
‘She sneezed all night’

As has been mentioned in 2.1.3.2.1, there are also examples with no initial T/A marker
which have past interpretations. Two further examples are:
(1887) Tae ana te taurekareka, hoki mai ana
arrive part. the servant return hither part.
‘The slave accomplished his object and returned’ (W, 199)
(1888) Kai·nga i waho…
eat·pass. at outside
‘They ate out of doors…’ (TWh, 20)

After the introducer kaatahi, ka is required, eg.

(1889) Kaatahi anoo ia ka kite he raakau kee
then again IIIsg T/A see cls tree contr
‘It was then that he saw it was a stick’ (TR2, 6)

In non-verbal locative contexts, past time is indicated by i, eg.

(1890) I taku whare te hui
at(pt) sggenIsg house the meeting
The meeting was at my place’

In possessive prepositional sentences, past time is indicated by naa/noo, eg.

(1891) Noo Whanga-nui a Tamaahua
actgen Whanga-nui pers Tamaahua
‘Tamahua belonged to Whanganui’ (TP, 91)
 Maori 420

In the actor-emphatic construction, naa combines with i to give a past reading. In past
contexts, other non-verbal sentences can have past readings, but contain no overt marking
of the tense, eg.
(1892) He tohunga a Kupe…
cls expert pers Kupe
‘Kupe was an expert…’ (KH, 1)
(1893) Toko·toru ngaa tohunga e hahau ana i
pnum·3 the(pl) expert T/A hew T/A DO
ngaa waka nei. Ko Kupe teetahi, ko Kauika
the(pl) canoe proxI eq Kupe a(sp) eq Kauika
teetahi, ko Turi-ua-nui teetahi
a(sp) eq Turi-ua-nui a(sp)
‘There were three experts hewing these canoes. One was Kupe, one was Kauika, one was
Turi-ua-nui’ (KH, 2)

In many subordinate constructions, i te is extremely rare, and e…ana is normal. There are
also certain constructions from which ka is excluded, and its place is normally filled by i.
Note also the examples already adduced of e with past readings (2.1.3.2.1, examples
(1827) -(1830)).
In negatives, i te is possible, and varies with e…ana. I is used for events not
aspectually marked as continuous; ka is not possible, eg.
(1894) Kaaore a Hata i te whakarongo
neg pers Hata T/A listen
‘Hata wasn’t listening’

(Note that, out of context, such negatives are normally interpreted as present, rather than
past.)
(1895) I ngaa raa o mua, kaaore teenei mahi e
at the(pl) day gen before neg this work T/A
mahi·a ana e ngaa Maaori o Te Kaha
do·pass. T/A by the(pl) Maori gen Te Kaha
‘Formerly, this activity was not performed by the Te Kaha Maoris’ (TR2, 132)
(1896) Kaaore ia i kite i te toka
neg IIIsg T/A see DO the rock
‘He didn’t see the rock’

The negator kiihai occurs exclusively with i, eg.

(1897) Kiihai i piirangi toona whaea ki a ia
neg T/A want sggenIIIsg mother to pers IIIsg
‘His mother did not want him’ (KM, 1)

In embedded noun clauses, ka, i and e…ana are all found, eg.
(1898) Ka whakahoki a Rewi ka haere kee a
T/A reply pers Rewi T/A move contr pers
Tamahae ki te hii
Tamahae to the fish
 Morphology 421

‘Rewi replied that Tamahae had gone fishing instead’

(1899) I tautohe ia i hee oo maatou hoariri
T/A contend IIIsg T/A wrong plgenIplexcl enemy
‘He contended that our opposition was wrong’
(1900) Noo too raaua tae·nga, ka tuupono ia
actgen sggenIIIdl arrive·nom T/A light upon IIIsg
e ruku·ruku kooura ana raaua
T/A dive·dup crayfish T/A IIIdl
‘When they arrived, he realized that they were diving for crayfish’

In relative clauses, i and e…ana are normal. However, given the attested example of ka in
2.1.3.2.1.2, (1876), and my consultant’s judgements recorded in 1.1.2.5.2, ka cannot be
excluded outright, although I have no examples of ka with past reference. I te is
extremely rare. The normal markers are illustrated in the following:
(1901) I kite raatou i te ara i haere ai a
T/A see IIIpl DO the path T/A move part. pers
Hongi
Hongi
‘They saw Hongi’s track’ (lit. ‘the track along which Hongi went’) (TR2, 106)
(1902) …ka tata ngaa poti o Te Kaha me Maungaroa
T/A near the(pl) boat gen Te Kaha with Maungaroa
ki te waahi e kau ana te ika raa
to the place T/A swim T/A the fish dist
‘The boats of Te Kaha and Maungaroa neared the place where that fish was swimming’
(TR2, 133)

For possibilities in various types of adverbial clause, see 1.1.2.4.

2.1.3.2.1.3.1 Degrees of past remoteness

Maori does not distinguish degrees of past remoteness in its tense markers.

2.1.3.2.1.3.2 Tenses relative to a point in the past

One of the functions of kua is to mark events relative to a point in the past; kee is often
required post-verbally, eg.
(1903) I taku tae·nga atu ki te kaainga, kua
at sggenIsg arrive·nom away to the home T/A
maoa kee i a ia ngaa kai
cooked contr cause pers IIIsg the(pl) food
‘When I got home, he had already cooked the dinner’

However, events which are past relative to another past event can also be rendered by ka
or i, ie. they are not necessarily distinguished from past tenses, eg.
(1904) Teeraa noo te ahiahi ka kii·a e ia
 Maori 422

that actgen the evening T/A say·pass. by IIIsg

kia ho·horo he kai
subj fast·dup a food
‘He had already, in the evening, given orders that the preparation of food should be
hastened’ (TWh, 20)
(1905) Ka hoki te koorero ki a Hotu i ruku
T/A return the story to pers Hotu T/A dive
raa i te punga o too raaua waka
dist DO the anchor gen sggenIIIdl canoe
‘The story returns to Hotu who had dived for the anchor of their canoe’ (KH, 3)
(1906) Tae rawa atu, kaatahi tonu ia ka haere
arrive intens away then still IIIsg T/A move
‘When we arrived, he had just left’

Negative pluperfects use kaahore anoo...kia, (as do present perfects), eg.

(1907) Kaaore anoo kia hua·ina teenei waahi ko
neg again subj name·pass. this place eq?
Waimarama
Waimarama
‘This place had not yet been named Waimarama’

Future-in-the-past, like the pluperfect, can be marked by kua, eg.

(1908) I mea atu raa hoki ahau ki a koe kua
T/A say away dist also Isg to pers IIsg T/A
whakapuare·tia e ia te keeti ki a taatou
open·pass. by IIIsg the gate to pers Iplincl
‘I told you he would open the gate for us’

However, not all futures-in-the-past are marked with kua. They can also be marked with
T/A markers which do not distinguish them from past or future tenses, eg.
(1909) I moohio ahau e haere mai ana koe
T/A know Isg T/A move hither T/A IIsg
i teenei raa
at this day
‘I knew you would come today’
(1910) Ka hee ai te koorero a taku tuakana,
T/A wrong part. the talk gen sggenIsg brother
ka makariri te wai a/i te rangi nei
T/A cold the water at(fut)/at the day proxI
‘My brother’s prediction that the water would be cold today was wrong’
(1911) Kia taka te tau, kia pirau, ka kawe ai
subj pass the year subj rotted T/A carry part.
i ngaa iwi ki taua toma
cause the(pl) tribe to det aph resting-place
‘When a year had passed, and the flesh decomposed, they would carry away the bones to that
 Morphology 423

resting-place’ (TWh, 18)

2.1.3.2.1.4 Future
Future tense is marked by ka, e…ana, e or kei te, and in the presence of certain time
adverbials, without a tense marker, or with kua. Kei te and e…ana mark the action as on-
going, eg.
(1912) Ka tuhi·tuhi reta ia
T/A write·dup letter IIIsg
‘He will write/be writing’
(1913) Mehemea ki te kore e mutu too
if to the neg T/A stop sggenIIsg
makamaka i te paaoro, ka tango·hia e au
play DO the ball T/A take·pass. by Isg
‘If you don’t stop playing with that ball, 111 take it away’
(1914) Kei te haere koe ki hea a te Aranga?
T/A move IIsg to where at(fut) the Easter
‘Where are you going at Easter?’
(1915) E haere ana taatou ki te toa aapoopoo
T/A move T/A Iplincl to the store tomorrow
‘We will be/are going to the store tomorrow’

In modern Maori, e is relatively rare in main clauses, and usually co-occurs with neuter
verbs, adjectival predicates, or passives, although these restrictions are not absolute:
(1916) Kaati; ko au anake e haere
enough eq Isg alone T/A move
‘Enough; I alone will go’ (TWh, 19)
(1917) A te wiki tua·tahi o Tiihema, mutu ai
at(fut) the week ord·1 gen December finished part.
te kura
the school
‘School will finish in the third week of December’ (TR2, 176)
(1918) Pakeke rawa ake au, kua tuu kee au hei
adult intens up Isg T/A stand contr Isg as
Mema-Paaremata!
member-parliament
‘When I eventually grow up, I’m going to stand as an MP’ (TR2, 166)

Kua appears to add certainty to the statement, ie. has a modal component.
In non-verbal classifying sentences, futurity is indicated by hei rather than he, eg.
(1919) Hei kai kee te miiti naa, kaaore hei
cls(fut) food contr the meat proxII neg cls(fut)
moounu
bait
 Maori 424

‘That meat is for food, not for bait’

Equative sentences can have future readings in the presence of appropriate adverbials, eg.
(1920) I toona tikanga ko Pita tana ingoa
from? sggenIIIsg intention eq Peter sggenIIIsg name
‘It is her intention that his name will be Peter’

Future locational sentences use hei/ko/kei/a, depending on dialect, and whether the
location is temporal or spatial, eg.
(1921) Ko/kei/hei Tauranga te hui a Oketopa
at(fut) Tauranga the meeting at(fut) October
‘The meeting in October will be at Tauranga’

Future possession is marked with maa/moo, eg.

(1922) Maa·ku teenei
intgen·Isg this
‘I will have this one’ (KH, 1)

In subordinate clauses, a different set of options is found. Ka and e…ana are usual in
noun clauses:
(1923) Naa wai i kii e maarena ana too
actgen who T/A say T/A marry T/A sggenIIsg
tuakana
brother
‘Who said your brother’s getting married?’ (TR2, 72)
(1924) Kua kite noa atu ahau ka ngaro te
T/A see extend away Isg T/A lost the
reo Maaori
language Maori
‘I saw long ago that the Maori language will disappear’
(1925) E moohio ana ahau, e tae·a e ia
T/A know T/A Isg T/A arrive·pass. by IIIsg
te whakamaarama i te whakataukii nei
the explain DO the proverb proxI
‘I know that she will be able to explain this proverb’

In negatives, e is used with both kaahore and kore while ka is impossible, eg.
(1926) Kaahore maatou e maakuu!
neg Iplexcl T/A wet
‘We won’t get wet!’
(1927) Ki te kore koe e haere, kaahore hoki ahau
to the neg IIsg T/A move neg also Isg
e haere
T/A move
‘If you won’t go, I won’t go either’
 Morphology 425

(1928) E kore ahau e wareware ki a ia

T/A neg Isg T/A forget to pers IIIsg
‘I will never forget her’

E also occurs with negative imperatives:

(1929) …kaua e moe roa…!
neg T/A sleep long
‘...don’t sleep in…!’ (TR2, 41)

E…ana can also have a future reading in negations with kaahore, eg.
(1930) Kaaore taatou e haere ana aapoopoo
neg Iplincl T/A move T/A tomorrow
‘We will not be going tomorrow’

In actor-emphatics, e co-occurs with maa to mark future, eg.

(1931) Maa te Aatua koe e manaaki, e tiaki
intgen the God IIsg T/A guide T/A care
‘God will guide and protect you’

In relative clauses, e and e…ana are frequently found:

(1932) Ka kimi·hia teetahi tikanga, e kite·a ai
T/A seek·pass. a(sp) plan T/A find·pass. part.
he moni
some money
‘A plan will be sought whereby money will be found’ (TR2, 88)
(1933) Kotahi te poaka e puuhi·a ana a te
one the pig T/A shoot·pass. T/A at(fut) the
Mane
Monday
‘One pig will be shot on Monday’

Ka is also possible:
(1934) Ko koe te mea ka mahue i te pahi
eq IIsg the thing T/A leave behind cause the bus
‘You’re the one who’ll be left behind by the bus’

For possibilities in various types of adverbial clause, see 1.1.2.4.

2.1.3.2.1.4.1 Modal/aspectual values of future forms

The only future marker which clearly has both modal and aspectual values is kua, which
adds certainty to future statements, and has aspectual values of inchoativity and
perfectivity (see 2.1.3.3.1). Volition may well be implied in many future statements, but
the forms for future cannot be used independently, and so this is not easy to demonstrate.
 Maori 426

2.1.3.2.1.4.2 Degrees of future remoteness

Maori does not subdivide the future in its tense markers. However, for an event in the
distant future, tonu is sometimes used to modify the verb, compare:
(1935) Ka haere ia aapoopoo
T/A move IIIsg tomorrow
‘He is going tomorrow’
(1936) Ka haere tonu ia a teeraa tau
T/A move still IIIsg at(fut) that year
‘He will go next year’

But tonu can also be used for immediacy:

(1937) Ka haere tonu atu nei ahau
T/A move still away proxI Isg
‘I’ll go right now’

It can thus not be regarded as a marker of distant future.

2.1.3.2.1.4.3 Tenses relative to a point in the future

Kua can be used for future perfect, eg.
(1938) Kia hoki mai ahau, kua mutu teenei mahi
subj return hither Isg T/A finished this work
‘When I return, you are to have finished this job’

Alternatively, these two events can occur in the other order, in which case kua is not
used:
(1939) Kia oti i a koe teenei mahi a te
subj finished cause pers IIsg this work at(fut) the
waa e hoki mai ai ahau
time T/A return hither part. Isg
‘This job should be finished by you at the time when I return’

The first clause is a weak imperative. However, sometimes e is used, eg.

(1940) Me·i noho atu ia i te paa, e roa te
if·T/A stay away IIIsg at the pa T/A long the
kawenga
business
‘If he had remained in the pa, we would have had a long job’ (TWh, 22)

There does not seem to be any special marker for future-in-the-future. Ka is the normal
marker, eg.
(1941) Kia pakeke au, ka hoko whare nui au,
subj grow up Isg T/A buy house big Isg
moo·ku
intgen·Isg
 Morphology 427

‘When I grow up, I’ll buy a big house for myself’

If the actor-emphatic is used, the marker is e:

(1942) Inaa poro·a e ahau te taura, maa·u e
when cut·pass. by Isg the rope intgen·IIsg T/A
hopu mai te kete
catch hither the basket
‘When I cut the rope, you catch the basket’

2.1.3.2.2 Tense distinctions in other moods and non-finite forms

Other moods are marked in Maori by the use of mood markers in the slot which is
occupied by the tense marker in the indicative. This means that tense is not marked in
moods other than the indicative. For instance, compare the following pairs. In (1943), the
speaker knows that the boy concerned used to receive occasional sums of money:
(1943) Kia whiwhi te tamaiti taane ki te moni, ka
subj receive the child boy to the money T/A
hoko mai ia i te koha maa te kootiro
buy hither IIIsg DO the gift intgen the girl
‘When the boy got the money, he used to buy a present for the girl’
(1944) Kia hoki mai koe, kua oti taku
subj return hither IIsg T/A finished sggenIsg
reta
letter
‘When you return, I will have finished my letter’
(1945) I tono ahau me hui taatou inanahi
T/A order Isg oblig meet Iplincl yesterday
‘I ordered that we should meet yesterday’
(1946) Me hui taatou aapoopoo
oblig meet Iplincl tomorrow
‘Let’s meet tomorrow’

The definition of non-finiteness is problematic in Maori, see 2.1.3.5. However, in

constructions which might be non-finite, no verbal-type tense distinctions are marked,
because the appropriate slot for these markers is filled by a mutually exclusive form.
However, -Canga nominalizations can be introduced by time-indicating prepositions.
Thus they can show time reference, without using the tense markers associated with finite
verbs. Compare the following:
(1947) I te whiwhi·nga o te tamaiti taane ki te
at(pt) the receive·nom gen the child boy to the
moni, ka hoko·na mai e ia he koha maa
money T/A buy·pass. hither by IIIsg a gift intgen
te kootiro
the girl
‘On the boy’s receiving the money, he bought a present for the girl’
 Maori 428

(1948) A te whiwhi·nga o te tamaiti taane ki

at(fut) the receive·nom gen the child boy to
tana moni, ka hoko mai ia i te
sggenIIIsg money T/A buy hither IIIsg DO the
koha maa te kootiro
gift intgen the girl
‘On the boy’s receiving the money, he will buy a present for the girl’

Notice that the T/A marker of the main clause is ka in both cases, although the use of the
passive in (1947) but the active in (1948) is associated with the difference in the temporal
reading of these clauses. These temporal prepositions have been discussed in 2.1.1.6.
There do not appear to be stem nominalizations which are time-marked
prepositionally, although note that hei nominalizations, which use the stem form of the
verb refer to not-realized events:
(1949) …ki te tiki atu i a ia hei patu i
to the fetch away DO pers IIIsg cls(fut) kill DO
taua taniwha
det aph taniwha
‘…to fetch him to kill that taniwha’ (TP, 91)

2.1.3.2.3 Absolute υs. relative tense

Most of the material relevant to this distinction has been discussed in the preceding
sections. E…ana is clearly a relative tense marker. The markers which are clearly not
relative, ie. which are absolute are i and kei te. I te is probably also absolute, but the
picture is clouded a little by its use as a neutralized form in negatives, corresponding to
either past or non-past. This is probably fairly described as neutralization; there seems to
me good evidence that the locative preposition from which this derives is used as a
neutral preposition in contexts where there is already a tense marker (see Bauer, 1981a,
179f for some discussion, and compare the remarks on neutralization in tense systems in
Comrie, 1985, 102ff). I suggested in 2.1.3.2.1 that e might be a relative marker. Ka is
clearly not an absolute tense marker. I treated it as a relative marker in Bauer 1981 a.
However, Harlow has since argued that it is an injunctive (Harlow, 1989), ie. that it does
not mark tense, mood or aspect, but merely marks the phrase as verbal. There is one
further marker, kua, which is an absolute-relative marker in the sense of Comrie, 1985,
44ff.

2.1.3.2.3.1 Mood
The above system is found only in the indicative. No tense distinctions are made in other
moods.

2.1.3.2.3.2 Finiteness
The problems surrounding the definition of finiteness in Maori are discussed in 2.1.3.5.
All the most probable candidates for non-finiteness lack tense marking totally, so it is
 Morphology 429

fairly safe to say that the absolute/relative distinction is irrelevant except in finite
constructions.

2.1.3.2.3.3 Main υs. subordinate clause

Chung has suggested (1978, 177, 21) that i, e and e…ana are the characteristic markers of
subordinate clauses in Maori, and that ka, in particular, is not a subordinate clause
marker. While there are certain subordinate constructions where this is true, this is a
considerable over-simplification. However, whether or not it is true, T/A marking is
neither exclusively absolute nor exclusively relative in either main or subordinate clauses.
All T/A markers, including can occur in main clauses. The possibilities for
subordinate clauses vary from type to type. An attempt is made to summarize the use of
T/A markers in subordinate clauses here.
The actor-emphatic (note that not all scholars agree that it involves subordination)
presents the simplest system: i is used in clauses co-occurring with naa, and e is used in
clauses co-occurring with maa. In addition, there is the co-occurrence of noo and e…ana
in the partially parallel subordinate construction described by Biggs (1969, 74) (see
1.11.2.1.7).
In negatives, the following tense markers are found: i te, e…ana, i, e. Kei te, ka,
and kua are excluded. Note that this cuts across absolute/relative divisions.
On the topic of embedded object clauses, it must firstly be noted that all tense markers
can occur in direct speech. In non-quotes, the following markers are found: kua (though
none of these in my data are clearly absolute-relative tense instances—they are
aspectual), i, e…ana, ka, e. E occurs in my data only in negatives with kore in such
embeddings. does not occur in such embeddings.
The data on tense markers in relative clauses was presented in 1.1.2.5.2. There is no
tense marker which cannot occur in at least some relative clauses, although it is not clear
from my data whether kua occurs as a relative-absolute tense marker, or if it is restricted
to aspectual uses. Nevertheless, it must be noted that some markers are much more
frequent than others. Thus of 147 examples in a small corpus, 58 used i, 47 used e, and 12
used e…ana. For the record, an example from that corpus for each of the other tense
markers is given (see also (1876) and (1877) above):
(1950) He aha ngaa kupu kei te ngaro?
cls what the(pl) word T/A missing
‘What are the missing words?’ (TR2, 38)
(1951) Toko·rua ngaa neehi i te tiaki i a ia
pnum·2 the(pl) nurse T/A care DO pers IIIsg
‘There were two nurses who were looking after him’ (TR2, 57)
(1952) Aanei ngaa moni ka pau moo te tuku
here the(pl) money T/A exhausted intgen the send
i a koe ki te Kura-tua·rua a teenei tau
DO pers IIsg to the school-ord·2 at(fut) this year
‘Here is the money it will cost to send you to secondary school this year’ (TR2, 164)
(1953) Ko eenaa eetahi o ngaa keemu taakaro·tia ai
eq those some(pl) gen the(pl) game play·pass. part.
 Maori 430

i reira
at there
‘Those are some of the games that are played there’ (TR2, 176)
(1954) …ka whakakite i ngaa mea hou kua
T/A show DO the(pl) thing new T/A
ako·na e ia
learn·pass. by IIIsg
‘…[he] showed the new things he had learned’ (TR2, 167)

It may not be accident that all but the last of these is embedded in a non-verbal sentence.
The data on T/A markers in adverb clauses does not show general patterns. The reader
is referred to the sections on adverb clauses (1.1.2.4).

2.1.3.3 Aspect

2.1.3.3.1 Perfect aspect

2.1.3.3.1.1 Form of the perfect aspect

Kua can be used to indicate a past event with present relevance. However, this is not the
only function of kua, which is also an absolute-relative tense marker (see 2.1.3.2.1.3.2
and 2.1.3.2.1.4.3). Typical examples of kua as perfect aspect marker are (with the context
specified): (1955 context) I want to give your brother a book to read, but I don’t know
which. Are there any of these books that he has already read?
(1955a) Aae, kua koorero ia i teenei pukapuka
yes T/A read IIIsg DO this book
‘Yes, he has read this book’
(1955b) Kua oti i a ia te koorero teenei
T/A finished cause pers IIIsg the read this
pukapuka
book
‘He has read this book right through’

(1956 context) Is the king still alive?

(1956) Kaaoo, kua mate
no T/A died
‘No, he has died’

(1957 context) Can I go now?

(1957) Kua horoi·a e koe oo niho?
T/A clean·pass. by IIsg plgenIIsg teeth
‘Have you cleaned your teeth?’
 Morphology 431

These examples are from the Göteborg Questionnaire on tense, mood and aspect, the
results of which are presented in Dahl, 1985. These are some of Dahl’s “prototypical”
perfects; the Maori consultant with whose help I completed the Questionnaire used kua in
all Dahl’s prototypical perfects (Dahl, 1985, 131–2). Note, however, that negatives of
sentences with kua in the positive have kaahore anoo ‘neg again’ for the negator, and kia
‘subjunctive’ in the T/A slot, see (606a, b) in 1.4 and (1875) in 2.1.3.2.1.2.

2.1.3.3.1.2 Tenses of perfect aspect

The particle kua is used in the same slot as the tense markers, and since only one marker
is possible, kua excludes other tense markers, and perfect aspect is thus invariable for
tense.

2.1.3.3.1.3 Subdivisions of perfect aspect uses

2.1.3.3.1.3.1 Present result of past situation

Kua in Maori can express the present result of a past situation, eg.
(1958) Kua toto tana ihu
T/A blood sggenIIIsg nose
‘His nose is bleeding’
(1959) Kua oti te maara i a Hata te parau
T/A finished the garden cause pers Hata the plough
‘Hata has finished ploughing the garden’

2.1.3.3.1.3.2 A situation that has held at least once in the period leading up to the present
Kua can be used for this function, eg.
(1960) Kua tuutaki koe ki taku tuakana?
T/A meet IIsg to sggenIsg brother
‘Have you ever met my brother?’

2.1.3.3.1.3.3 A situation that began in the past and is still continuing

Kua is used in this type of situation, eg.
(1961) Kua maha kee hoki ngaa tau e noho ana
T/A many contr also the(pl) year T/A stay T/A
ki a au
to pers Isg
‘He has stayed with me for many years (KH, 2)

However, note that the T/A marker for the verb noho ‘stay’ is not kua, but e…ana. Thus
it is not clear whether examples of this kind genuinely show kua with this function. Kua
is not always found in this situation: the following were given as alternatives by my
consultant:
 Maori 432

(1962 context) The child is still coughing. The doctor asks how long the child has
been coughing for.
(1962a) Kua pau te haaora e mare·mare ana
T/A exhausted the hour T/A cough·dup T/A
‘He has spent an hour coughing’/‘He has been coughing for an hour’
(1962b) Ka kotahi haaora e mare·mare ana
T/A one hour T/A cough·dup T/A
‘He has been coughing for one hour’

Sometimes i is used, eg.

(1963) E rua tekau ngaa tau i mahi ai ahau
num 2 10 the(pl) year T/A work part. Isg
i teenei pukapuka
DO this book
‘I have been working on this book for twenty years’

2.1.3.3.1.3.4 Others
The marker kua is used in several situations not accounted for above.
The first I have labelled inchoative. This has certain obvious links with the ‘present
result of a past situation’ category of examples, but there are many examples where the
idea of a result seems absent, and kua seems to mark the beginning of a state, eg.
(1964) Kua moe a Tamahae i runga i te teepu
T/A sleep pers Tamahae at top at the table
‘Tamahae has gone to sleep on the table’

This sort of use is particularly common with verbs of perception, so that moohio ‘know’
with kua can have the sense ‘realize’, and rongo ‘hear, smell, feel’ can have the sense
‘perceive’ in combination with kua, eg.
(1965) Kua moohio ahau i hee ia
T/A know Isg T/A wrong IIIsg
‘I know/I’ve realized she was wrong’

although to stress the realization, other forms are used, eg.

(1966) Moohio rawa ake ahau kua hee a ia
know intens up Isg T/A wrong pers IIIsg
‘I’ve just realized she was wrong’

Note also the following:

(1967) Kua tata mutu ia ki te peita i tana
T/A near finished IIIsg to the paint DO sggenIIIsg
whare
house
‘He is about to finish painting his house’
(1968) E kore e roa kua tiimata ia ki te
 Morphology 433

T/A neg T/A long T/A start IIIsg to the

raranga i tana kete
plait DO sggenIIIsg kit
‘She will soon start to make her kit’

Secondly, kua is sometimes used for the immediate past, eg. (1969 context) Is your
brother at home? No, we are very unlucky,…
(1969) Kua haere kee ia
T/A move contr IIIsg
‘He’s just gone’

However, kaatahi…ka more usually has this function, eg.

(1970) Kaatahi anoo ka mutu te hui
then again T/A finished the meeting
‘The meeting has just finished’

Kua is also used for ‘hot news’ (Dahl, 1985, 132), eg. (1971 context) Have you heard the
news? No. Tell me.
(1971) Kua mate te Pirimia
T/A dead the P.M.
‘The Prime Minister has died’

Thirdly, kua is used for most of these functions even in past-time narrative contexts,
where the description ‘present’ in the characterization is not strictly applicable. Thus
compare the examples in 2.1.3.3.1.3.1 and 2.1.3.3.1.3.4 with
(1972) Ka hoki mai a Kupe, kua moohio ia
T/A return hither pers Kupe T/A know IIIsg
kua mate a Hoturapa, kua waatea a Kura
T/A dead pers Hoturapa T/A free pers Kura
maa·na
intgen·IIIsg
‘Kupe returned, and he knew that Hoturapa was dead and that Kura was free for him’ (KH,
3)
(1973) Ka hongi te ihu ki te hauaauru; kua rongo
T/A sniff the nose to the west wind T/A smell
i te hau·nga o toona ariki
DO the wind·nom gen sggenIIIsg chief
‘He sniffed the winds of the west, and recognized the scent of his master’ (TP, 90)

Lastly, there are cases where kua appears to be used simply for vividness in narrative.
This may be the past context parallel to the ‘hot news’ perfect, eg.
(1974) Te tae·nga ki ngaa kupu whakamutu·nga o
the arrive·nom to the(pl) word finish·nom gen
te haka a Wairangi “aa te”—kua mau ngaa
the haka gen Wairangi “a te” T/A seize the(pl)
ringa ki ngaa patu, “aa ta” kua maunu mai -
 Maori 434

hand to the(pl) club “a ta” T/A be drawn hither

“aa tau” kua huaki te ope, kua patu i te
“a tau” T/A attack the group T/A kill DO the tangata
man
‘When they came to the concluding words of Wairangi’s haka,—“a te”—their hands
grasped their clubs; “a ta” the clubs were drawn forth; “a tau” the party attacked and began
to kill’(W, 200)

2.1.3.3.1.4 Similarities in expression between perfect aspect and recent past time
While kua can be used for recent past (see (1969) above), kaatahi…ka is probably more
common if the immediate past is stressed. However, it is not necessary in Maori to mark
events as recent past. Thus the narration of events which have just occurred will not
necessarily differ from the narration of those same events the following day, or even
several decades later (except in time adverbials, of course): the first marker is usually i,
then other sequential events use ka. The forms used for recent past are also used in
narrative contexts for ‘recent past in relation to the narrative time’, eg. (1975 context) Did
you find your brother at home? No, we did not, we were very unlucky.
(1975a) Kua haere kee i mua tata o taku
T/A move contr at before near gen sggenIsg
tae·nga atu
arrive·nom away
‘He left just before we arrived’

or
(1975b) Tae rawa atu, kaatahi tonu ia ka haere
arrive intens away then still IIIsg T/A move
‘When we arrived, he had just left’

2.1.3.3.2 Aspect as ways of viewing the duration of a situation

2.1.3.3.2.1 Formal marking of aspects

2.1.3.3.2.1.1 Perfectiυe (aoristic) aspect

The tense markers ka and i are most commonly used for this function. Of Dahl’s
examples of prototypical perfective aspect (1985, 78), however, none were translated by
my consultant with i as T/A marker; 4 (possibly 5) had ka, two had no T/A marker but
rawa ‘intensifier’ following the verb, two were non-verbal sentences, and two used
nominalizations. This gives a rather strange picture for Maori. The two non-verbal
sentences contained numerical frequency adverbials once and twice. These have no direct
equivalent in Maori, and so were translated as eg.
(1976) Kotahi tana mare·nga
one sggenIIIsg cough·nom
‘He gave one cough’/‘His cough numbered one’
 Morphology 435

A parallel instance where a verbal sentence could be used is

(1977) I titiro atu ia ki te raa, ka tihewa
T/A look away IIIsg to the sun T/A sneeze
‘He looked at the sun and sneezed’

The narrative texts in the Göteborg Questionnaire contained two of Dahl’s prototypical
examples (in bold):
(1978) I a au e haere ana i te ngahere, ka
at pers Isg T/A move T/A at the bush T/A
taka·hia noa·tia e au teetahi ngarara. Kaatahi
step·pass. sudden·pass. by Isg a(sp) snake then
ka ngau·a mai taku waewae. Ka mau
T/A bite·pass. hither sggenIsg leg T/A seize
au ki te koohatu, ka kuru·a e au te
Isg to the stone T/A throw at·pass. by Isg the
ngarara, mate rawa
snake die intens
‘I was walking in the forest, and suddenly I stepped on a snake. It bit me in the leg. I took a stone
and threw it at the snake. It died’

Kaatahi ka mate ‘then T/A die’ is an alternative to mate rawa. Ka is thus possible for
both, and is also used for the intervening actions. It will be noted that three of the verbs in
this passage are passive (see the discussion of the relation between passive and perfective
in Maori narrative in 2.1.3.1.1).
The use of rawa following the stem (as in the text above) is not restricted to clearly
perfective examples. Thus we find both
(1979) Roa rawa, kaatahi anoo a Mahia ka mea, “…”
long intens then again pers Mahia T/A say
‘[This] went on for ages, and then Mahia said “…”’ (P, 3)

and
(1980) …houhou rawa i ngaa koohao hei here·here·nga
bore intens DO the(pl) hole for tie·dup·nom
‘…and the holes to tie it on were bored’ (TP, 91)

Thus this particle does not serve to mark perfectivity.

Perfective examples with i are particularly common in the actor-emphatic, eg.
(1981) Naa·na i pupuhi te manu
actgen·IIIsg T/A shoot the bird
‘He shot the bird’
(1982) Ko te ingoa nei naa Te Aotakii i tapa
top. the name proxI actgen Te Aotaki T/A name
‘This name was given by Te Aotaki’ (TWh, 24)
 Maori 436

They are relatively rare in narrative, except in subordinate clauses, because ka is the
normal marker once the time reference is established. However, isolated perfective
sentences elicited from consultants regularly used i. Both ka and i readily occur with
situations which are not prototypical perfectives, eg.
(1983) Ka tangi ngaa tamaahine ki too raatou paapaa
T/A weep the(pl) daughter to sggenIIIpl father
‘The daughters mourned for their father’ (TWh, 17)

However, such events are presented as unanalyzed wholes by the use of these markers.
Similarly with i:
(1984) I tae pai katoa raatou ki uta
T/A arrive good all IIIpl to shore
‘They all reached the shore safely’ (P, 2)

The following example is instructive:

(1985) I poo, i poo, aa, ka maarama
T/A night T/A night and T/A light
‘It was dark for a long time, and then at last it became light’ (KM, 2)

The durative sense is obtained by repetition of the perfective situation.

Without the repetition, there would not be a durative sense here.

2.1.3.3.2.1.2 Imperfective aspect

Certain types of imperfective aspect are marked by the ‘pseudo verbal continuous’ forms
and by e…ana, but there are no general markers for all types of imperfectivity in the
sense of Comrie (1976, 25). In the Göteborg questionnaire, a sentence which was
frequently an instance showing imperfective marking was the following, where my
consultant offered two alternatives: (1986 context) What was your brother’s reaction
when you gave him the medicine yesterday?
(1986a) Kootahi haaora e mare ana
one hour T/A cough T/A
‘He coughed for an hour’

or
(1986b) Ka mare moo te haaora kootahi
T/A cough intgen the hour one
‘He coughed for an hour’

Thus events which are imperfective in many languages are not necessarily marked
differently from perfective events in Maori. Events which are scene-setting, and used as a
frame within which some other event is ‘placed’ are marked with e…ana, eg. the first
clause in (1978) above.
 Morphology 437

2.1.3.3.2.1.3 Habitual aspect

The expression of habitual aspect appears to be subject to more dialectal variation in
Maori than most other areas of the grammar. I am not certain that the following account
covers all the dialectal variation found.
Firstly, consider Dahl’s examples of prototypical habitual aspect (Dahl 1985, 97). In
all instances a nominal construction was one of the alternatives provided by my
consultant, and there is certainly a strong tendency in all dialects to prefer non-verbal to
verbal constructions in the area of habitual statements, eg. (1987 context) What does your
brother usually do after breakfast?
(1987) He tuhi·tuhi reta
cls write·dup letter
‘[He] writes letters’

This is also appropriate for past habitual contexts, eg.

(1988) I mua raa, ko tana mahi hoki he
at(pt) before dist top. sggenIsg work also cls
puhi·puhi puukeko
shoot·dup pukeko
‘He used to shoot pukeko’

For some dialects only, it is possible to use a verbal construction with no T/A marker
before the verb, but with the particle ai following the verb, eg. (context as for (1987)):
(1989) Tuhi·tuhi reta ai ia
write-write letter part. IIIsg
‘He writes letters’

The dialects for which this is possible include E. Coast dialects like Ngati-Porou, but
exclude northern dialects like Te Aupouri where e…ana is the normal marker for non-
past habituals, eg.
(1990) I ngaa ata katoa e ara ana ia i te
at the(pl) morning all T/A rise T/A IIIsg at the
ono karaka
6 clock
‘He gets up at 6 every morning’

In some Taranaki dialects, at least, e…ai rather than just ai is used for non-past habituals.
The consultant from whom this information came was unsure how past habituals were
rendered in that dialect (see also Bauer, 1981a, 67–8).
In Dahl’s habitual-generic examples (1985, 99) similar constructions were used as for
prototypical habituals, with one exception:
(1991 context) What do your cats do when they are hungry?
(1991) Ka miao
T/A miaow
‘They miaow’
 Maori 438

A further example of ka for habitual generics is

(1992) …kua koorero·tia atu hoki ki a raatou, ka
T/A say·pass. away also to pers IIIpl T/A
aanini te maahunga o te tangata, ki te roa
giddy the head gen the man to the long
rawa e kaukau ana i roto i te wai
intens T/A swim T/A at inside at the water
ngaawhaa
hot spring
‘…They had been told that a person’s head becomes giddy if they swim for too long in the
hot pools’ (TR2, 99)

In Dahl’s prototypical past habituals (1985, 101), again the same nominal constructions
occurred (or post-head ai), but there was one exception:
(1993 context) The boy used to receive a sum of money now and then.
(1993) Kia whiwhi te tamaiti taane ki te moni, ka
subj receive the child boy to the money T/A
hoko mai ia i te koha maa te kootiro
buy hither IIIsg DO the present intgen the girl
‘When the boy got the money, he bought a present for the girl’

Note that these translations were provided by a consultant who in other contexts used ai
for habituals. In dialects where ai is not possible, past habituals can be marked with i, eg.
(1994) I mua i whakapono ngaa taangata ki
at(pt) before T/A believe the(pl) people to
ngaa taniwha
the(pl) taniwha
‘People used to believe in taniwhas’

However, these are not the only markers used for habituals in texts. E…ana can be used,
and is particularly common in negatives, where ka and ai are impossible.
(1995) E whaa tau ia i reira e kura ana
num 4 year IIIsg at there T/A school T/A
‘He went to school there for 4 years’
(1996) Kaaore e au ana te ngeru
neg T/A bark T/A the cat
‘Cats don’t bark’

E is also found, again especially, but not exclusively, in negatives. Thus a consultant
from a different dialect preferred the following to (1996):
(1997) E kore te tori e tautau
T/A neg the cat T/A bark
‘Cats don’t bark’

(Note the considerable lexical differences, typical of this sort of vocabulary.) Other
examples are:
 Morphology 439

(1998) E tangi he puu i ngaa poo katoa

T/A sound a flute at the(pl) night all
‘A flute plays every night’ (H, 7)
(1999) Ko te mahi hoki a toona ariki e haere
eq the work also gen sggenIIIsg chief T/A move
tonu i ngaa raa katoa…
still at the(pl) day all
‘Its master used to visit it every day…’ (TP, 89)
(2000) I ngaa raa o mua, maa te hooiho kee
at the(pl) day gen before intgen the horse contr
te moua e too
the mower T/A pull
‘In former times, a horse pulled the mower instead’ (TR2, 59)

(Note that this is actor-emphatic.)

(2001) Kore rawa ia e kai hikareti i ngaa poo
neg intens IIIsg T/A eat cigarette at the(pl) night
‘He never smokes in the evenings’

Note also the i in the following habitual:

(2002) Kore rawa au i moohio i tiimata mai
neg intens Isg T/A know T/A start hither
te pepa i te raakau
the paper at the tree
‘I didn’t know that paper came from wood’

Sometimes no T/A marker is used, especially with peeraa (etc.) like that’:
(2003) Pee·raa tonu i ngaa raa katoa, i ngaa poo
like·dist still at the(pl) day all at the(pl) night
hoki e te iwi
also by the tribe
‘The tribe did that by day and at night’ (H, 7)

Thus habituality is not at all a unified area with respect to its marking in Maori. Even in
individual dialects, there is considerable variation found.

2.1.3.3.2.1.4 Continuous aspect

Like habituality, continuous aspect is not marked consistently in Maori. Both markers
indicating imperfectivity and markers which typically occur in perfective situations are
found.
The commonest markers are e…ana and kei te, i te. Williams comments that e…ana is
not used with neuter verbs (1862, 49), (= Williams’s participles and Biggs’s statives).
However, that rule is not held to absolutely in modern Maori. Some examples of
continuous aspect are:
 Maori 440

(2004) E aroha ana ahau ki a Hotu

T/A love T/A Isg to pers Hotu
‘I love Hotu’ (KH, 2)
(2005) I too raatou tae·nga atu, e poouri tonu ana
at sggenIIIpl arrive·nom away T/A dark still T/A
te hooro
the hall
‘When they arrived, the hall was still dark’ (TR2, 85)
(2006) Ki te kore koutou e haere mai inaianei
to the neg IIpl T/A move hither now
e mahue ana koutou i a maatou
T/A leave behind T/A IIpl cause pers Iplexcl
‘If you don’t come at once, we will leave you behind’

(The last is a neuter verb.) E…ana can also occur with experience verbs:
(2007) Teeraa pea teetahi e moohio ana ki te
that perhaps a(sp) T/A know T/A to the
roa·nga ake o te koorero nei, e
long·nom away gen the story proxI T/A
tae·a pea e ia te whakamaarama ki
arrive·pass. perhaps by IIIsg the explain to
te hunga e noho ware ana
the group T/A stay forget T/A
‘Perhaps there’s somebody who knows the rest of this story and who perhaps will be able
to explain it to those of us who live in ignorance’ (KH, 3)

Kei te and i te occur with both adjectival predicates and other verb types:
(2008) Kei te marino tonu te moana
T/A calm still the sea
‘The sea is still calm’
(2009) I te moe a Rewi
T/A sleep pers Rewi
‘Rewi was asleep’

Beside these examples with imperfective markers, we find several other markers. Ka
occurs with adjectival predicates:
(2010) Ka nui ngaa toenga miiti kei roto i te
T/A big the(pl) remains meat at(pres) inside at the
kaapata
cupboard
‘There’s a lot of leftover meat in the cupboard’

(Note the unexpected use of kei rather than i here, for which I can offer no explanation.)
E is common, especially in subordinate clauses, eg.
(2011) Teenaa koe, e noho mai naa i Aakarana
that IIsg T/A live hither proxII at Auckland
 Morphology 441

‘Hello to you who live in Auckland’

(2012) E mau nei anoo aua ingoa
T/A carry proxI again det aph (pl) name
‘Those names still remain’ (TWh, 17)

Ana post-verbally can be continuous, eg.

(2013) Koowatawata ana ngaa uru maawhatu i te
gleam T/A the(pl) hair curly at the
hana o te ahi
glow gen the fire
‘Her curly tresses were gleaming in the firelight’ (TA, 6)

No T/A marker is also extremely common with adjectival predicates, but is not restricted
to them:
(2014) Kii tonu te whare karakia
full still the house chant
‘The church was full’
(2015) Tapa·ia tonu·tia atu taua maunga nei ko
call·pass. still·pass. away det aph mountain proxI spec
Tihirau
Tihirau
‘That mountain is still called Tihirau’ (P, 2)

Non-verbal constructions can also be used, eg.

(2016) Tana mea pai teenaa he kooura
sggenIIIsg thing good that cls crayfish
‘She likes crayfish’

2.1.3.3.2.1.5 Progressive aspect

The normal markers for progressive aspect are e…ana and kei te/i te. Of Dahl’s
prototypical progressives (1985, 92), all were translated by my consultant with e…ana
and the appropriate ‘pseudo-verbal’ as alternatives, eg.
(2017 context) Please do not disturb me.
(2017a) Kei te tuhi·tuhi reta kee ahau
T/A write·dup letter contr Isg
‘I’m writing a letter’
(2017b) E tuhi·tuhi reta kee ana ahau
T/A write·dup letter contr T/A Isg
‘I’m writing a letter’

However, it should be noted that of the characteristics Dahl suggests for distinguishing
progressive from imperfective, the e…ana form is also used for habituals under some
circumstances, and is also used for states, (1985, 92–3) as are kei te and i te, eg.
 Maori 442

(2018) E moohio ana ia he porotaka te ao

T/A know T/A IIIsg cls round the earth
‘He knows that the earth is round’

Occasionally, in subordinate clauses, e rather than e…ana is used, eg.

(2019) Mahue mai ana ngaa waka e hari raa i
leave hither T/A the(pl) canoe T/A carry dist DO
ngaa taangata, naa·na nei i tinihanga a
the(pl) people actgen·IIIsg proxI T/A trick pers
Te Tahi
Te Tahi
‘[It] left behind the canoes which were carrying the men who had tricked Te Tahi’ (KWh,
3)

2.1.3.3.2.1.6 Ingressive aspect

It is necessary here to draw attention to the fact that ka has been labelled ‘inceptive’ by
various previous writers on Maori. However, I do not believe that this form is a marker of
ingressive aspect. The history of the descriptions of ka is outlined in Bauer (1981a, 53–
6), and some arguments against the label ‘inceptive’ are given on pp. 61–3 of that work.
To the extent that ingressive aspect is marked in Maori, it is marked by kua, rather than
ka. Some relevant examples were adduced in 2.1.3.3.1.3.4, and the following may be
added:
(2020) Kua tangi te piana
T/A sound the piano
‘The piano has begun to play’
(2021) …naawai raa i pai te tiki, kua kino
for a while dist T/A good the fetch T/A bad
‘…for some time, all went well with the fetching, then it got bad’ (TWh, 17)

However, it is not the case that a sentence like the following has an ingressive
interpretation:
(2022) Kua horoi raaua i te whare
T/A clean IIIdl DO the house
‘They’ve washed the house’

To indicate ingressive aspect with a verb like this, it is necessary to use some periphrasis,
eg. with tiimata ‘start’:
(2023) Ka tiimata raaua ki te horoi i te whare
T/A start IIIdl to the clean DO the house
‘They started to wash the house’
(2024) Tiimata ana ki te raranga i tana kete
start T/A to the plait DO sggenIIIsg kit
harakeke
flax
 Morphology 443

‘She is starting to make a flax kit’

It appears that kua marks ingressive aspect only with semantically stative predicates,
particularly with adjectival predicates and experience verbs. (Kua can, of course, be used
with tiimata to give ‘have started’.)

2.1.3.3.2.1.7 Terminative aspect

Terminative aspect is not marked by any of the T/A markers of Maori. However, there
are three neuter verbs which are commonly used for this purpose: mutu ‘finished’, oti
‘finished’, pau ‘exhausted, used up’ (the latter is perhaps marginal as a contender for the
expression of terminative aspect).
(2025) Ka mutu ngaa karakia a Te Aotakii ka
T/A finished the(pl) chant gen Te Aotaki T/A
wero·hia e ia a Rangipopo
invoke·pass. by IIIsg pers Rangipopo
‘Te Aotaki finished his incantations and then he invoked Rangipopo’ (TWh, 20)

This use in a coordinate clause is extremely common.

(2026) Kua oti kee i a Pou te taarai
T/A finished contr cause pers Pou the adze
toona waka
sggenIIIsg canoe
‘Pou had already finished adzing out his canoe’ (P, 1)
(2027) …pau katoa i a ia te kai
exhausted all cause pers IIIsg the food
‘…he finished up all the food’ (TP, 90)

Note also the following:

(2028) E whaka·mutu ana ia i tana
T/A cause-finished T/A IIIsg DO sggenIIIsg
horoi·nga
wash·nom
‘She is finishing the washing’

Mutu cannot occur with e…ana, so whakamutu is used.

2.1.3.3.2.1.7.1 Completion of one situation prior to another

As will be clear from the previous paragraph, this is one of the chief uses of mutu. Such
constructions take the form of two coordinate clauses in Maori although it is often more
natural to translate the first one as a ‘when’ clause in English:
(2029) Ka mutu te kai, ka kii ia…
T/A finished the food T/A say IIIsg
‘When the meal was over, he said…’ (TWh, 21)
 Maori 444

(2030) Ka mutu oona tohu, ka noho ki raro…

T/A finished plgenIIIsg sign T/A sit to below
‘When his instructions were ended, he sat down…’ (TWh, 23)

In general terms, a sequence of actions with the marker ka is taken as sequential, ie. as
implying the completion of the first before the start of the next, eg.
(2031) Ka takai·a, ka kawe·a, ka whaka·iri·a
T/A wrap·pass. T/A carry·pass. T/A cause·hang·pass.
ki runga ki te kauere
to top to the puriri
‘They wrapped him up and took him and suspended him in a puriri tree’ (TWh, 18)

However, strings of events with ka are not necessarily sequential, compare Comrie, 1985,
27–8:
(2032) Ka pupuhi te hau, ka hinga·hinga ngaa
T/A blow the wind T/A fall·dup the(pl)
raakau, ka rere ngaa tiini o te whare
tree T/A fly the(pl) tin gen the house
‘The wind blew, the trees blew over, the roofing iron blew off’

2.1.3.3.2.1.8 Iterative aspect

There is no special T/A marker for iterative aspect. This is often indicated in Maori by
the use of a reduplicated verb form. However, it cannot be predicted which verbs will
have reduplicated forms in this context. Compare the following two sentences:
(2033) I tihewa ia i te poo roa nei
T/A sneeze IIIsg at the night long proxI
‘She sneezed all night long’
(2034) Ka mare·mare tonu ia
T/A cough·dup still IIIsg
‘He coughed many times’

In the judgement of my consultant, reduplication with tihewa has the sense of periodic
sneezes:
(2035) E tihewa·hewa ana ia i te poo roa nei
T/A sneeze·dup T/A IIIsg at the night long proxI
‘She sneezed off and on all night’

My consultant judged that (2035) implied fewer sneezes than (2033). For many verbs
expressing activities of brief duration, reduplication expresses iterativity, eg.
(2036a) E meke ana a Hone i a Piri
T/A punch T/A pers John DO pers Bill
‘John is punching Bill (once now)’
(2036b) E meke·meke ana a Hone i a Piri
T/A punch·dup T/A pers John DO pers Bill
 Morphology 445

‘John is punching Bill’

The following pair shows the contrast clearly:

(2037) I peke te tamaiti i te taiapa
T/A jump the child DO the fence
‘The child jumped the fence (once)’
(2038) I peke·peke te tamaiti i te taiapa
T/A jump·dup the child DO the fence
‘The child jumped backwards and forwards over the fence’

Iterative reduplication usually involves the reduplication of two morae of the stem
(although not all such formations are iterative, see 2.2.6.4). With stems of more than two
morae, the most common form of reduplication involves the final two morae (usually
accompanied by lengthening of the vowel of the first mora), but sometimes the first two
morae are reduplicated. In a few cases, both forms are attested with the same base, eg.
paataitai and patapatai from paatai ‘ask’. Occasionally other forms of reduplication have
an iterative sense, eg. hahaka ‘frequentative of haka ‘dance” (Williams’s Dictionary).
There are also verbs with iterativity as part of the meaning of their stem, eg. paerangi ‘to
stick (feathers) into’. In other cases, iterativity is indicated by an adverb like tonu ‘still’,
see the examples in (2039) below.

2.1.3.3.2.1.9 Semelfactiυe aspect

There is no special T/A marker for semelfactive aspect. Often the lack of reduplication in
a commonly reduplicated form indicates a single occurrence of an event, see eg. (2036a)
above. In a few cases, there is a special reduplicated form for semelfactive aspect (see
Biggs, 1969, 107). The reduplications with this sense reduplicate the first mora only.
(Note that this is not the commonest semantic effect of this form of reduplication, for
which see 2.1.4.3). Biggs’s examples are kikimo ‘shut the eyes’ from kimo ‘wink’, and
papaki ‘slap or clap once’ from paki ‘pat, clap’. However, this is not a productive
process. In the majority of instances, iterativity is marked, rather than semelfactivity,
compare the following:
(2039a) I paatootoo ia i runga i te teepu
T/A knock IIIsg at top at the table
‘She knocked on the table’ (a description of a single knock)
(2039b) I paatootoo tonu ia i runga i te teepu
T/A knock still IIIsg at top at the table
‘She kept knocking on the table’/ ‘She knocked on the table several times’

2.1.3.3.2.1.10 Punctual aspect

There is no marker of punctual aspect in Maori; the same markers are used as for
perfective situations.
 Maori 446

2.1.3.3.2.1.11 Durative aspect

There is no special marking to indicate situations which necessarily last in time.
However, certain reduplicated forms may show this type of relationship, compare
iterative reduplication above: it is largely a matter of the sense of the stem whether these
reduplicated forms have an iterative or a durative interpretation, eg. awhi ‘embrace’,
awhiawhi ‘cuddle’ (rather than ‘give a series of distinct embraces’); whakaeke ‘attack,
assault’ whakaekeeke ‘invade’; haere ‘come, go’ haaereere ‘stroll about’. Again, this is a
process which applies sporadically: it cannot be guaranteed that any stem will have a
reduplicated form of the appropriate sense. Note also that durativity is sometimes
encoded in other parts of the sentence in Maori. Compare
(2040a) I rongo ia i te turituri naa
T/A hear IIIsg DO the noise proxII
‘He heard a noise’
(2040b) Ka rongo tonu ia i ngaa turituri
T/A hear still IIIsg DO the(pl) noise
‘He heard a prolonged noise’

where the change from singular to plural is an integral part of the marking of durativity.

2.1.3.3.2.1.12 Simultaneous aspect

There is no special marker for simultaneous events in Maori. For the expression of
simultaneity, see 1.1.2.4.2.1.1. Sometimes the second of two simultaneous events is
expressed by the use of the particle ai, eg.
(2041) Kei te noho ia i runga i te tuuru, koorero
T/A sit IIIsg at top at the chair read
pukapuka ai
book part.
‘He’s sitting in a chair reading a book’

This construction is most usual for expressing ‘dependent action’, ie. where the second
action depends upon the first. The actions are frequently sequential, rather than
simultaneous. Thus this construction is not primarily a means of indicating simultaneity.

2.1.3.3.2.1.13 Other aspects

Maori has no formal markers of other aspects.

2.1.3.3.2.1.14 Telic aspect

Maori does not distinguish formally in the verb between telic and atelic situations.
However, transitive atelic situations are often expressed with object incorporation, while
telic situations are not. Compare the following pairs:
(2042a) Inanahi ka haere atu ia i runga i te
 Morphology 447

yesterday T/A move away IIIsg at top at the

huarahi
path
‘Yesterday, he walked along the road’
(2042b) Inanahi ka haere atu ia mai i
yesterday T/A move away IIIsg hither from
Rotorua ki Pooneke
Rotorua to Wellington
‘Yesterday, he went from Rotorua to Wellington’
(2043a) E inu ana ia i te kapu wai
T/A drink T/A IIIsg DO the cup water
‘She is drinking a cup of water’
(2043b) E inu wai ana ia
T/A drink water T/A IIIsg
‘She is drinking water’
(2044a) E poouto ana ia i te raakau naa
T/A chop T/A IIIsg DO the tree proxII
‘He is chopping down the tree’
(2044b) E poouto·uto wahie ana ia
T/A chop·dup firewood T/A IIIsg
‘He is chopping firewood’

The only way of indicating that the logical end of a telic situation is reached is by using
either a perfective T/A marker, or by using a verb expressing termination (see
2.1.3.3.2.1.7). Consider the following: (2045 context) A: ‘It seems that your brother starts
reading a lot of books, but he never finishes them’. B: That’s not quite true.’
(2045a) Kua koorero·hia e ia teenei pukapuka
T/A read·pass. by IIIsg this book
‘He has read this book right through’
(2045b) Kua oti/pau i a ia te koorero
T/A finished/exhausted cause pers IIIsg the read
teenei pukapuka
this book
‘He has read this book right through’

2.1.3.3.2.2.1 Possibilities for combining aspectual values

The aspectual particles which fill the T/A slot before the verb stem are mutually
exclusive. Thus the only possibilities for combined aspectual values occur where one of
the aspectual values is marked by lexical means, as with terminative and iterative aspect.
There do not appear to be any restrictions on such combinations, nor any unpredictable
senses.
 Maori 448

2.1.3.3.2.2.2 Restrictions on combinations of aspects and other verbal categories

2.1.3.3.2.2.2.1 Aspect+voice combinations

There is a major restriction placed on the combination of the perfective markers ka and i
with active voice. However, it has been suggested following Chung (see 2.1.3.1.1) that
this is probably not principally a voice-aspect restriction, since these combinations do
occur with, for instance, reflexive predicates (see 2.1.3.1.1.4.1).
Kua is particularly common with passives and statives, but is not prevented from
occurring with actives. E…ana is very rare with neuter verbs, and most such examples
are rejected outright. It occurs readily with both actives and passives. Reduplicated forms
can be passive as well as active.

2.1.3.3.2.2.2.2 Aspect+tense combinations

Most of the markers of Maori have combined tense/aspect values. Since only one marker
can occur with a verb, combinations of markers are impossible. Those aspectual
distinctions which are made by lexical means combine freely with the tense markers.

2.1.3.3.2.2.2.3 Aspect+mood combinations

Mood markers occupy the same slot as the tense/aspect markers, and thus no
combinations are possible, except with lexically marked aspectual values. There do not
appear to be any restrictions on such combinations.

2.1.3.3.2.2.2.4 Combination of aspects with finite and non-finite forms

The problems of making the finite/non-finite distinction in Maori are discussed in 2.1.3.5.
Following the discussion, non-lexical aspectual values (ie. T/A markers) are possible
only with finite forms. There are no restrictions on the occurrence of lexically marked
aspectual distinctions with non-finite forms.

2.1.3.3.2.2.2.5 Combination of aspect markers with verb classes

The sense of the markers kua and e…ana varies according to the type of predicate
involved. The following sets of examples are indicative of the interaction that occurs; (a)
shows a canonical intransitive, (b) an adjectival intransitive, (c) a (transitive) passive, (d)
and (e) neuter verbs, (f) and (g) transitive active, and (h) an experience verb:
(2046a) Kua oma raatou
T/A run IIIpl
‘They have run’
(2046b) Kua pai ia
T/A good IIIsg
‘He’s well again’
 Morphology 449

(2046c) Kua puuhi·a ngaa kereruu

T/A shoot·pass. the(pl) pigeon
‘The pigeons have been shot’
(2046d) Kua pau ngaa kai
T/A exhausted the(pl) food
‘The food’s all finished’
(2046e) Kua whati te raakau
T/A broken the stick
‘The stick has broken’
(20461) Kua horoi ia i te motokaa
T/A wash IIIsg. DO the car
‘He has washed the car’
(2046g) Kua koohuru ia i a Tuu
T/A murder IIIsg DO pers Tu
‘He has murdered Tu’
(2046h) Kua hiahia ia ki te hooiho raa
T/A desire IIIsg DO the horse dist
‘He wants that horse’
(2047a) E oma ana raatou
T/A run T/A IIIpl
‘They are running’
(2047b) E pai ana ia
T/A good T/A IIIsg
‘He’s well’ (a continuing state)
(2047c) E puuhi·a ana ngaa kereruu
T/A shoot·pass. T/A the(pl) pigeon
‘The pigeons are being shot’
(2047di) *E pau ana ngaa kai
T/A exhausted T/A the(pl) food
(2047dii) E pau haere ana ngaa kai
T/A exhausted move T/A the(pl) food
‘The food is in the process of being eaten up’
(2047e) E whati ana te raakau
T/A broken T/A the stick
‘The stick is breaking’ /The stick is broken’
(2047f) E horoi ana ia i te motokaa
T/A wash T/A IIIsg DO the car
‘He is washing the car’
(2047g) E koohuru ana ia i a Tuu
T/A murder T/A IIIsg DO pers Tu
‘He is murdering Tu’
(2047h) E hiahia ana ia ki te hooiho raa
T/A desire T/A IIIsg DO the horse dist
 Maori 450

‘He wants that horse’ (≡2046h)

2.1.3.4 Mood
Maori has three mood markers, kia, me and kei.

2.1.3.4.1 Indicative
The indicative mood is marked in general by the presence of a T/A marker, rather than a
mood marker. However, the absence of a marker in the pre-verb slot can be associated
with either indicative or imperative.

2.1.3.4.2 Conditional
Conditional clauses in Maori are introduced by one of a variety of conjunctions, of which
mehemea and ki te are the most common (see 1.1.2.4.2.5). With mehemea, the verb takes
a T/A marker, and all are possible. In conditional clauses referring to non-past conditions,
the verb is marked with ki te. (See the examples in 1.1.2.4.2.5, and Williams, 1862, 38.)
With certain conditional conjunctions, no overt T/A/M marker occurs, eg.
(2048) Mei noho atu ia i te paa, e roa te
if stay away IIIsg at the pa T/A long the
kawenga
burden
‘If he had stayed in the pa, we would have had a long job’ (TWh, 22)

However, it is entirely possible that the -i of mei is in fact the T/A marker, so that this
form is a combined conjunction and T/A marker. This may also be true of meinaa, where
the -naa is possibly related to the deictic form (it recurs in other conjunctions). This may
also be true of inaa, although the temporal reference of inaa does not coincide with that
of i.
While me is a mood marker in some contexts, it normally occupies the same slot as the
T/A marker when it is a mood marker, so it is unlikely that the me in these conjunctions
is a mood marker.

2.1.3.4.3 Imperative
The imperative mood is marked by the absence of a T/A marker, except possibly where
the verb has two morae or fewer, when e precedes the verb, see the discussion in 1.1.1.3.
Transitive imperatives take the passive verb form (see 1.1.1.3 for details and examples).

2.1.3.4.3.1 Person and number imperative forms

The imperative form is not affected by person and number.
 Morphology 451

2.1.3.4.4 Optative
Kia, which has been glossed ‘subjunctive’ here is the normal marker for optative mood. It
occurs with adjectival and other notionally stative predicates, eg.
(2049) Kia ora!
subj well
‘Health be with you’ (a conventional greeting)
(2050) Kia tere te kai
subj fast the food
‘Eat quickly’
(2051) Kia taka pea koe i runga i te hooiho!
subj fall perhaps IIsg from top at the horse
‘I hope you fall off the horse!’
(2052) Kia pee·hea te hoohonu o te rua nei?
subj like-how the deep gen the hole proxI
‘How deep should this hole be?’

(Further examples are found in 1.1.1.3.1.) Note the use of pea in (2051) and (2053): it
appears to indicate uncertainty that the wish will be fulfilled.
After predicates of wishing and hoping, kia is used in variation with ki te, depending
on the like-subject constraint, but the verb of wishing etc. is not necessarily expressed:
(2053) Kia tae mai pea ia
subj arrive hither perhaps IIIsg
‘I hope she comes’
(2054) E tuumanako ana maatou kia whakaae mai
T/A hope T/A Iplexcl subj agree hither
te Kawanatanga ki taa maatou pitihana
the Government to sggenIplexcl petition
‘We are hoping the Government will agree to our petition’
(2055) Ka hiahia aua tamariki ki te haere a
T/A desire det aph(pl) children to the move at(fut)
te Tuurei
the Tuesday
‘The children want to go on Tuesday’

Even more commonly, the ‘wish’ is expressed nominally, in which case kia introduces
the clause:
(2056) Ko taku hiahia kia mau·ria mai e koe
eq sggenIsg wish subj bring·pass. hither by IIsg
he kuuao tori
a baby cat
‘I wish you would bring me a kitten’
 Maori 452

2.1.3.4.5 Intentional mood

There is no special marker for intentional mood in Maori. Nonpast tense markers are used
in main clauses, sometimes supported by tino ‘very’ or some other intensive. In
subordinate clauses, the choice between kia and ki te is determined by the like-subject
constraint, eg.
(2057) E tino mutu ana i ahau teenaa pukapuka
T/A very finished T/A cause Isg that book
‘I will finish that book’
(2058) Maa·ku raapea e utu
intgen·Isg indeed T/A pay
‘I will pay (the bill)’
(2059) Ko taku hiahia teenaa kia ohorere ia
eq sggenIsg desire that subj surprise IIIsg
‘I intend him to get a fright’

Negative intentions use kore, eg.

(2060) Kore rawa ahau e haere
neg intens Isg T/A go
‘I won’t go’

Sometimes me seems to suggest intention to comply with obligation, eg. in the story of
Maui, his mother orders him to leave and he replies :
(2061) Aae, me haere kee atu au
yes M go contr away Isg
‘Yes, 111 go’ (KM, 2)

2.1.3.4.6 Debitive mood

Obligation to do something is most often expressed by me (glossed ‘oblig’ here), often
described as the ‘weak imperative’.
(2062) Me tuhi koorero koutou inaianei
oblig write story IIpl now
‘You should write a story now’
(2063) Me taapuke te tuupaapaku i roto i te toru
oblig bury the body at inside at the 3
raa
day
‘The body must be buried within three days’ (TR2, 30)
(2064) Me aata noho koe i runga i teenaa tuuru!
oblig slowly sit IIsg at top at that chair
‘You’d better sit carefully on that chair’

Me is used for a wide range of degrees of obligation, from advice to legal requirements.
 Morphology 453

Sometimes, however, kia is used, eg.

(2065) Kia tino horo tooku ara a te ata
subj very early sggenIsg rise at(fut) the morning
‘I have to get up early in the morning’

The following were provided as alternatives to each other:

(2066a) Kia tino horo atu koe i teenaa
subj very fast away IIsg compar that
‘You must run faster than that’
(2066b) Me tino horo atu koe
oblig very fast away IIsg
‘You must run faster’

Negative debitives use kaua, the negator for imperatives:

(2067) Kaua rawa atu koe e hoko i teenaa pootae
neg intens away IIIsg T/A buy DO that hat
‘You mustn’t buy that hat’

2.1.3.4.6.1 Moral υs. physical obligation

Moral obligation is expressed with me, eg.
(2068) Me tino maarena e koe taku mokopuna
oblig indeed marry by IIsg sggenIsg grandchild
naa·u hoki ia i hapuu ai
actgen·IIsg indeed IIIsg T/A pregnant part.
‘You must marry my granddaughter since you got her pregnant!’
(2069) Me tino haere ngaa tamariki ki te kura
oblig very move the(pl) children to the school
‘Children have to go to school’

Physical obligation does not use a mood marker. The following translations were offered
for examples expressing physical obligation:
(2070) Ka maremare rawa atu ahau
T/A cough intens away Isg
‘I had to cough’
(2071) Ko te tihewa ahau!
T/A sneeze Isg
‘I’ve got to sneeze’/‘I’m going to sneeze!’
(2072) I te kaha o te raa ki te whiti, ka
from the strong gen the sun to the shine T/A
poo·ngia iho ahau
dark·pass. down Isg
‘The sun was so bright I had to close my eyes’
 Maori 454

2.1.3.4.6.2 Degrees of obligation

The degree of obligation can be indicated by the use of an intensifier, eg.
(2073a) Me haere koe ki te hui
oblig move IIsg to the meeting
‘You should go to the meeting’
(2073b) Me tino haere koe ki te hui
oblig very move IIsg to the meeting
‘You must go to the meeting’

2.1.3.4.7 Potential mood

There is no special mood marker for the ability to do something. Very often where ability
is expressed in English, it is simply not expressed in Maori, eg.
(2074) E kore ngaa kiwi e rere
T/A neg the(pl) kiwi T/A fly
‘Kiwis can’t fly’/ ‘Kiwis don’t fly’
(2075) I toa anoo ahau meina i wheeke ahau
T/A win again Isg if T/A press Isg
‘I could have won if I’d tried’

However, if ability is stressed, then a paraphrase with aahei ‘be able’ (a neuter verb) or
taea ‘arrive-pass.’ can be used (see also Wills, 1960,118), eg.
(2076) Kaaore ia e aahei ki te haere
neg IIIsg T/A able to the move
‘She will not be able to go’

(Wills, loc. cit., deplores the use of ki te in this construction, but in my experience it is
commoner than the te+stem construction.)
(2077) …mehemea e kore e tae·a e ia te
if T/A neg T/A reach·pass. by IIIsg the
haere mai…
move hither
‘…if he could come…’ (TP, 91)

There are also non-passive uses of tae ‘arrive’ which at least border on the expression of
ability, eg.
(2078) Ka tae au ki te mate o tooku tipuna
T/A arrive Isg to the death gen sggenIsg ancestor
‘I will be able to [avenge] the death of my grandfather’ (TWh, 21)

In many cases, the verb moohio ‘know’ is used:

(2079) E moohio ana ahau ki te kaukau
T/A know T/A Isg to the swim
‘I can swim’
 Morphology 455

2.1.3.4.7.1 Physical ability υs. permission

Tae(a) is normally related to physical ability. Aahei is also used for physical ability, but
can be used for permission, eg.
(2080) Kaaore ahau e aahei te hiikoi maa raro
neg Isg T/A able the step intgen below
‘I can’t walk’
(2081) Kaaore e tae·a e ahau te hiikoi naa
neg T/A reach·pass. by Isg the step actgen
raro
below
‘I’m not able to walk’
(2082) Kaahore ahau e aahei ana ki te kai rare
neg Isg T/A able T/A to the eat sweet
‘I’m not allowed to eat sweets’
(2083) E kore e tae·a e ahau ki te kai
T/A neg T/A reach·pass. by Isg to the food
‘I can’t eat’ (= not physically able)

However, the most normal method of asking and granting permission does not involve
these verbs. The following exchange is typical:
(20841) Maa, e hiahia ana ahau ki te haere ki te
Mum T/A desire T/A Isg to the move to the
kaukau
swim
‘Mum, I want to go for a swim’/‘Mum, can I go for a swim?’
(2084ii) Aae, haere
yes move
‘Yes, go’/‘Yes, you can go’

or
(2084iii) Kaua koe e haere
neg IIsg T/A move
‘Don’t go’/‘You can’t go’

For giving permission without a request being made, either a direct imperative or a ‘weak
imperative’ with me is used:
(2085) (Me) haere koe ki te kaukau
oblig move IIsg to the swim
‘You (should) go for a swim’/‘You can go for a swim’

2.1.3.4.7.2 Learned ability

Learned ability is usually expressed with moohio ‘know’, see (2079) above and
(2086) Kua moohio ia ki te koorero pukapuka
 Maori 456

T/A know IIIsg to the read book

‘She can read now’

However, abilities which are ‘learned’ as natural physical development do not use
moohio, eg.
(2087) Kua haaere·ere te mokopuna
T/A move·dup the grandchild
‘The grandchild can walk now’
(2088) Kua koorero
T/A talk
‘She’s talking now’

2.1.3.4.8 Certainty
It is not possible to mark the degree of certainty through morphological verbal means. It
can only be indicated adverbially with adverbs like pea ‘perhaps’. Some statements
which in English use ‘might’ are expressed in Maori with the monitory marker kei.
Compare the following set of examples:
(2089a) E koorero pukapuka ana ia
T/A read book T/A IIIsg
‘She’s reading a book’ (I just peeped in and saw her)
(2089b) Teenei naa, e koorero pukapuka ana ia
this proxI T/A read book T/A IIIsg
‘She must be reading a book’ (to judge by what I can hear)
(2089c) Teenaa pea, e koorero pukapuka ana ia
that perhaps T/A read book T/A IIIsg
‘She might be reading a book’ (I’m guessing)
(2089d) E koorero pukapuka pea ana ia
T/A read book perhaps T/A IIIsg
‘Perhaps she’s reading a book’ (I’m guessing)

(2089c) and (2089d) are more or less equivalent They can be made more tenuous by
stressing pea. Maori does not make a distinction between probability and possibility.
An example with kei is
(2090) Kei whai·a koe e te kurii
mon chase·pass. IIsg by the dog
‘You might be chased by the dog’

Such instances clearly involve a warning, as well as stating a possibility. In Maori they
are coded as warnings.
 Morphology 457

2.1.3.4.9 Authority for assertion

Authority for assertion does not have to be marked in Maori. Thus an account of
something one has witnessed oneself, and an account second-hand do not differ in their
forms.

2.1.3.4.10 Hortatory
Encouragement is normally expressed with kia, and thus is equivalent in form to optative
expressions (see 2.1.3.4.4), eg.
(2091) Kia kaha ki te reo Maaori
subj strong to the language Maori
‘Support the Maori language’
(2092) Kia mutu taatou i te waa tika, ne?
subj finished Iplincl at the time right Q
‘Please let’s try to finish on time, shall we?’

However, plain imperatives can also be used, with tone-of-voice indications of the
communicative force, eg.
(2093) Teenaa kai·nga e koe teenei waahi
that eat·pass. by IIsg this piece
‘Why don’t you try this piece?’

2.1.3.4.11 Monitory
Maori has a monitory particle kei which is used for positive warnings, eg.
(2094) Kei mate koe i ngaa motokaa raa!
mon die IIsg cause the(pl) car dist
‘You might be killed by those cars’
(2095) …kei paa·ngia ia e te ruumaatiki
mon touch·pass. IIIsg by the rheumatism
‘…he might get rheumatism’ (TR2, 6)
(2096) Kei moomona koe!
mon fat IIsg
‘You’ll get fat!’

Negative warnings use the same particle with the negator kore:
(2097) Kei kore koutou e horo ki te oma
mon neg IIpl T/A fast to the run
‘You might not be able to run fast’

Warnings can also be expressed through imperatives, eg.

(2098) Kia tuupato!
 Maori 458

subj care
‘Take care’/‘Look out!’
(2099) Kaua e tuku·na te taiapa
neg T/A drop·pass. the fence
‘Don’t let go of the fence’

2.1.3.4.12 Narrative mood

There is no special narrative mood. The indicative is used.

2.1.3.4.13 Consecutive mood

There is no special marker for consecutive mood.

2.1.3.4.14 Contingent mood

Contingency does not differ in expression from possibility, and uses pea ‘perhaps’.

2.1.3.4.15 Others
There are no other mood markers in Maori (kia, kei and me are the only mood markers in
the system).

2.1.3.5 Finite and non-finite

The distinction between finite and non-finite forms, is not entirely clear as it relates to
Maori. Finiteness in a language like English can be defined in terms of forms which
‘respond’ to changes in other parts of the sentence—finite forms change with change of
time adverbial and/or change of person in the subject. Maori does not show subject-verb
agreement, and relative tense markers and mood markers do not change in response to
changes in temporal adverbials.
Another major characteristic of non-finite constructions in languages like English is
that they are normally subordinate constructions. However, Maori uses both stem and -
Canga nominalizations as independent constructions (see 1.15), although these appear
likely candidates for non-finite status on intuitive grounds. Such nominalizations do not
show variation with time and do not have subjects (underlying subjects are expressed as
possessors). It thus appears that independent status is irrelevant to considerations of
finiteness in Maori.
It has already been shown that lack of variation with change of time reference does not
lead to an intuitively satisfactory classification in Maori. The criterion of having a subject
has been used to define finiteness in English (eg. Joos, 1968, 14). However, if this
criterion is applied to Maori, there are a variety of unsatisfactory classifications. Firstly,
there are many anaphoric deletions of subjects in Maori (see 1.5), even across sentence
boundaries. It does not seem appropriate to classify such sentences as non-finite, as they
exhibit no other behavioural differences from finite sentences. Secondly, instances of
 Morphology 459

relativization on subjects automatically cause the deletion of the subject in the relative
clause, but it does not seem satisfactory to call all relative clauses with subject
relativization non-finite, but all other relative clauses finite. It is not possible to escape
from this dilemma by altering the criterion to ‘having an underlying subject which may
be deleted’, because of the ki te construction, see 1.1.2.2.6. This also has an underlying
subject which is obligatorily deleted under co-reference with a like subject in the main
clause, and therefore appears exactly parallel to the case of subject relativization. But the
ki te construction bears many resemblances to other stem nominalizations, and it is not at
all clear that it should be classified as finite. On the other hand, it also bears a number of
resemblances to noun clauses introduced by kia, which would appear to be finite.
Further problems for the ‘having a subject’ criterion are posed by imperatives. Joos
(1968, 14) specifically excludes second person pronominal subjects from his definition
for English in order to exclude imperatives from the category finite in English. Whether
or not such an exclusion is made for Maori, imperatives in Maori cannot all be classified
in the same way with respect to finiteness if the criterion ‘having a subject’ is used.
Imperatives of intransitives in Maori, like English imperatives, take the stem form of the
verb (see 1.1.1.3). With the possible exception of the second person pronouns, the
addressee, if included, is clearly not a subject, but a vocative, compare:
(2100) E ara, e Tama!
imp. wake voc Tama
‘Wake up, Tama!’
(2101) Ka ara a Tama
T/A wake pers Tama
‘Tama woke up’

By the suggested criterion, these imperatives are then non-finite. However, the
imperatives of transitive verbs do have subjects, because these verbs are passive in form,
and these subjects are not second person:
(2102) Aawhina·tia a Mere!
help·pass. pers Mary
‘Help Mary!’

(Note that this subject has the same form as the subject in (2101).) Transitive imperatives
are thus finite by the definition. Note, however, that no clearcut tense markers are
possible in either type of imperative. It does not seem satisfactory to claim that these two
classes of imperative differ in finiteness.
Another possibility is to take some combination of these criteria as definitional for
Maori. The nominalizations which seem the clearest candidates for non-finiteness have
neither subject nor tense marker. It is immediately clear that demanding that both these
criteria be fulfilled will not resolve all the problems raised: the two types of imperative
will still be classified differently. However, by this criterion, the cases of subject
relativization would be classified as finite, while the ki te construction would be non-
finite, which seems intuitively satisfactory. However, there are further problematic cases
to consider. One of the major ones is raised by those instances discussed in 2.1.3.2.1 of
verbs which have no T/A markers. Plainly, these fail on one of the two counts, and they
can fail on both, as the following shows:
 Maori 460

(2103) Haere tonu, ka tae ki Aaniwaniwa

move still T/A arrive to Aniwaniwa
‘[She] went on and reached Aniwaniwa’ (W, 197)

By the criterion suggested, the first clause would be non-finite, while parallel cases
without anaphoric subject deletion would be finite. This does not seem satisfactory, since
a construction like this could potentially have both a T/A marker and a subject, unlike the
nominalizations. Similarly awkward cases can be found among time expressions.
Consider the following:
(2104) Poo iho anoo ka hoki mai a Taranga,
night down again T/A return hither pers Taranga
aa, ao ake te raa, ka haere
and day up the day T/A move
‘When night fell again, Taranga returned, and when day broke, she left’ (KM, 2)

The second time phrase has a subject, and thus counts as finite under the definition.
However, the first expression, which is intuitively parallel, does not have a subject, and
my consultant was a little uncertain about the possibility of including one:
(2105i) ?Poo iho anoo te raa
night down again the day
‘When the day became night’
(2105ii) ?Poo iho anoo te poo
night down again the night
‘When the night fell again’

The only other possible definition I can conceive of is to define non-finiteness as

involving the presence of overt nominal characteristics. However, even this leads to
certain problems, eg. with structures like the two modifying structures in the following:
(2106) Ka whakamiiharo anoo a Rewi ki te kaimahi
T/A wonder at again pers Rewi to the worker
whakahaere i te miihini uta i ngaa raakau
operate DO the machine load DO the(pl) tree
‘Rewi marvelled again at the worker operating the machine loading the trees’ (TR2, 43)

Neither whakahaere nor uta shows any overt nominal characteristics, though neither has
a subject (because of the process of subject deletion through relativization), and neither
has a T/A marker, and a classification as non-finite seems intuitively required. Problems
in the other direction are posed for this criterion by the ‘pseudo-verbal continuous’ (see
2.1.3.2.1), which has overt nominal characteristics, but which behaves in all other
respects like constructions without nominal characteristics.
In conclusion, it seems that the joint requirement for nonfiniteness of having no
subject and having no T/A marker gives the fewest unsatisfactory results, but it should be
remembered that this criterion classifies transitive and intransitive imperatives differently
with respect to finiteness, and that it classifies as non-finite those constructions which
lack T/A markers and which have their subjects deleted by anaphoric reference deletion
processes. It seems to me that this raises the question as to whether the distinction
 Morphology 461

between finite and non-finite has any relevance for the description of Maori. It would be
possible to account for the differing behaviour of various types of construction by making
a distinction between verbal, nominal and attributive constructions, such that nominal
constructions would be those with te or some other nominal marker, and attributive
constructions would have no marker in the ‘construction type’ slot and occur in post-head
position in some other construction. All the other constructions considered here would be
verbal, and not specified with respect to finiteness. The ‘pseudo-verbal continuous’
construction would have to be excluded from the nominal category, but since it is fairly
clearly an exception, and the markers are listable, this does not seem to pose a problem.
However, I am unsure which category hei nominalizations belong to, despite the fact that
the label appears to categorize. This matter will have to wait for further examination of
the theoretical implications of this proposal.
The characteristics of the de-verbal forms can be found in other sections: ki te
(1.1.2.2.6); -Canga nominalizations (1.1.2.2.6); hei nominalizations (1.1.2.2.6); non-
finite relative clauses (1.1.2.3.9); plain stem nominalizations (1.1.2.4.3). Information on
the uses of these forms can be found as follows:
ki te: noun clause (1.1.2.2.6); purpose clause (1.1.2.4.2.3); condition clause(1.1.2.4.2.5)
- noun clause (1.1.2.2.6); time adverbials (1.1.2.4.2.1.1–3); cause clause (1.1.2.4.2.4);
Canga: comparatives (1.8.2–4); equatives (1.9.2–4); form of underlying subject (1.10.6); minor
sentence (1.15)

stem: form of underlying subject (1.10.6); minor sentence (1.15) The use of non-finite
constructions in adverb clauses is summarized in 1.1.2.4.3.

2.1.3.5.1–4 Verbal categories in non-finite forms

Given the above definition of non-finiteness, it follows that tense, mood, and aspect are
not marked in non-finite forms. In general, voice is not marked in these forms overtly.
Thus in -Canga nominalizations, only the marking of the associated non-subject NPs
allows the determination of whether the nominalization is active or passive in sense,
compare
(2107) te patu·nga a Tuu i a Pou
the kill·nom gen Tu DO pers Pou
‘Tu’s killing of Pou’
(2108) te patu·nga o Pou e Tuu
the kill·nom gen Pou by Tu
‘Pou’s being killed by Tu’

(See also Williams, 1862, 45, 24; Biggs, 1969, 80–82).

Ki te is normally followed by the active form (Chung (1978, 68) claims that the
passive never occurs after ki te Equi). However, the following, at least superficially
contradictory, example must be noted:
(2109) Ka whakapoto·a, ki te karanga·tia, ko
T/A shorten·pass. to the call·pass. spec
Tuu-whakairi-ora
 Maori 462

Tu-whakairi-ora
‘[This] was shortened to the calling form, Tu-whakairi-ora’ (TWh, 20)

This appears to be the use of the passive form as a noun. (It is certainly not a case of
Equi.) Williams cites maa te matapihitia ‘by the window-pass.’ ‘let it be passed through
the window’ (1862, 40). This phenomenon is rather rare in modern Maori. Neither hei
nor plain stem nominalizations allow the passive verb form.

2.1.3.6 Person/Number etc.

The form of the verb is not affected by the person or number of the subject or any other
category.

2.1.3.6.7 Identity between subjects of consecutive verbs

The identity of the subject of a verb with that of a preceding verb is frequently (but not
compulsorily) indicated by deletion of the subject of the second (and subsequent) verbs.
See 1.5.

2.1.3.6.8–9 Reflexive and reciprocal verb forms

There are no reflexive or reciprocal verb forms. For reflexivity see 1.6, for reciprocity,
see 1.7.

2.1.3.6.10 Marking of discourse participants

Maori has four directional particles which follow the verb and indicate the direction of
action in relation to the speaker. Sometimes one of these is so closely associated with a
particular verb that verb and particle are written as one word. The four particles with their
basic meanings are
mai ‘hither, towards speaker’
atu ‘away from speaker’
ake ‘upwards from speaker’
iho ‘downwards towards speaker’

However, there are many uses which are not readily explained in terms of these basic
senses: the details of the appropriate use of these particles are not entirely clear. There are
many occasions when native speakers feel that these particles are required to make a
sentence acceptable, but there are other occasions where movement in relation to the
speaker is clearly involved, and yet these particles are rejected. There is frequently no
actual movement involved when these particles are used: they can indicate mental
orientation. The use of the last three of these particles in comparative structures (see 1.8,
2.1.1.4.2) is probably to be seen in that light. Williams (1862, 55–6) points out that, in
addition to their speaker-oriented senses, they can also be used in pairs to indicate
 Morphology 463

relative location of objects. Biggs, 1969, 65–8 contains some useful discussion and
illuminating examples. Hohepa, 1981, 44–6 suggests that, in narrative, they often indicate
position relative to the narrator, rather than to the participants in the story. The
information that follows does not represent a complete account of their use.
The commonest verbs incorporating these particles are:
homai ‘give to speaker’
hoatu ‘give away from speaker’
hoake ‘give or bring to some place connected with the
speaker, but not the speaker’s current location
haramai ‘come, arrive’

Straightforward directional examples of these particles are:

(2110) Haere mai!
move hither
‘Welcome’ (a conventional greeting, especially on ritual occasions)
(2111) He kura pae aaku, noo reira kaare
cls feather come ashore plgenIsg actgen there neg
e whakahoki·a atu
T/A return·pass. away
‘Feathers washed ashore are mine, and so I’m not going to return them’ (P, 3)
(2112) Ka tango ia i teetahi hua o taua raakau,
T/A take IIIsg DO a(sp) fruit gen det aph tree ka panga·a iho
T/A throw·pass. down
‘He took a berry from that tree and threw it down’ (KM, 3)
(2113) …ka hari·a ake ki te rangi
T/A carry·pass. up to the sky
‘…and [he] was taken up to the sky’ (KM, 4)

Examples which do not involve physical motion, but where the directionality of the
action is clear are:
(2114) Kaua e paatai mai ki a au!
neg T/A ask hither to pers Isg
‘Don’t ask me!’
(2115) Kua awhiawhi atu a Moana
T/A cuddle away pers Moana
‘Moana has given [him] a cuddle’
(2116) He tino aataahua ngaa whetuu ki te titiro ake
cls very beautiful the(pl) star to the look up
‘The stars were very beautiful to look up at’ (TR2, 6)
(2117) Hua·ina iho ki te taunga ko
name·pass. down to the fishing ground spec
Kamokamo
Kamokamo
‘The fishing ground was called Kamokamo (knowing glances)’ (TWh, 17)
 Maori 464

Examples of use in pairs to indicate relative location are

(2118) …ka rongo atu raaua i te kani miihini e
T/A hear away IIIdl DO the saw machine T/A
haruru mai ana
roar hither T/A
‘They heard the machine saw roaring’ (TR2, 43)
(2119) I toona koorero·tanga atu ki a Hata, ka
at sggenIIIsg talk·nom away to pers Hata T/A
kohete mai a Pani
scold hither pers Pani
‘When she talked to Hata, Pani scolded [him]’

In (2118) the particles indicate the physical distance between the referents of raaua and
the machine saw; in (2119) the particles indicate the emotional distance that separates
Pani and Hata. These particles are frequently used thus in the narration of dialogue, eg.
(2120) Ka paatai atu ia… Ka kii mai raatou… Ka
T/A ask away IIIsg T/A say hither IIIpl T/A
mea atu ia ki a raatou… Ka kata mai
say away IIIsg to pers IIIpl T/A laugh hither
oona tuaakana ki a ia
plgenIIsg brothers to pers IIIsg
‘He asked… They replied… He said to them… His brothers laughed at him’ (KM, 1)

Sometimes the use of a particle affects the sense of the verb rather more radically, eg.
(2121) Ko wai te tangata e tuu atu raa?
eq who the man T/A stand away dist
‘Who is the man who has just stood up and gone?’

The sorts of examples which are common in discourse are illustrated by the following
passage. Tu-whakairi-ora has reached a beach where there is a group of women and
children. He has inquired who the women are, and has sat down on the women’s clothes:
(2122) Kua eke ia ki runga o ngaa kaakahu
T/A get in place IIIsg to top gen the(pl) clothes
noho ai. Kai te riri mai ngaa tamariki raa,
sit part. T/A angry hither the(pl) children dist
kai te titiro mai ngaa waahine raa. Ka haere
T/A look hither the(pl) women dist T/A move
ngaa tamariki, ka koorero atu, ka kii mai
the(pl) children T/A talk away T/A say hither
raaua, “Teenaa koa, kii atu, kia mau·ria
IIIdl That extend say away subj bring·pass.
mai e koutou o maaua kaakahu.” Te
hither by IIpl plgenIdlexcl clothes the
tae·nga atu o ngaa tamariki, ka
arrive·nom away gen the(pl) children T/A
whakatatanga ia, ka riro atu, ka noho anoo
stand IIIsg T/A give away T/A sit again
 Morphology 465

ia. Kei te kaakahu ngaa waahine raa, kei te titiro

IIIsg T/A clothe the(pl) women dist T/A look
whakatau mai ki a ia, ki ngaa tohu o
intently hither to pers IIIsg to the(pl) sign gen
te rangatira, o te toa, e mau atu
the chief gen the champion T/A bring away
ana i runga i a ia
T/A at top at pers IIIsg
‘He had taken his seat upon the clothes. The children expressed their disapproval, while the
women looked on. The children went and told them, and they said, “Well, tell him that you
must bring us our clothes”. When the children came, he got up at once, and gave them up,
and sat down again. While the women were putting on their clothes, they gazed intently at
him and the emblems of high birth and bravery which he bore with him’ (TWh, 19)

At the start of the passage, Tu-whakairi-ora appears to be the reference point for the use
of mai and atu. However, when the action shifts to the location of the women, they
become the reference locus, and apparently remain so until Tu-whakairi-ora is mentioned
again. Note that, while in many cases the preceding sentence subject is the reference
locus, this is not the case for the last mai in this excerpt. Thus it may be that part of the
function of these particles is to identify the participant of the story who is the chief point
of focus in episodes where there are several participants. When Mary Boyce (personal
communication) elicited native-speaker commentary on the use of these particles in
Maori texts, one of the recurrent comments was their “affective” use: mai is used with
participants with whom there is emotional rapport, while atu marks emotional distance.
However, in Tu-whakairi-ora, when the women leave, the text runs
(2123) Ka ahu mai ki te pito ki te tonga
T/A head hither to the end to the south
‘They made their way to the South end [of the bay]’ (TWh, 19)

No participant in the story can be the locus of the reference point for this mai if it is
understood to be directional. However, it might also be understood as ‘from where the
participant is’, as is necessary sometimes with iho and ake (see below). Note that mai can
be used in instances like
(2124) I haere ia mai i Te Rere·nga Wairua ki
T/A move IIIsg hither from The jump·nom spirit to
Muri Hiku
Bluff
‘He went from North Cape to Bluff’

which suggests a sense like ‘from where the participant is’. However in instances like
these, an unknown factor is the relevance of the location of the speaker, as opposed to the
participants. No obvious solution suggests itself, and the matter will have to await further
study.

2.1.3.6.10.1 Actions towards the speaker

Mai is added in many instances, eg.
 Maori 466

(2125) Naa Hone i hoomai te paaoro ki ahau

actgen John T/A give the ball to Isg
‘John gave the ball to me’
(2126) I haere mai a Mere ki te kite i ahau
T/A move hither pers Mary to the see DO Isg
‘Mary came to see me’

(Notice that the use of mai does not render ki/i ahau redundant.)
(2127) I kata mai a Mere ki ahau
T/A laugh hither pers Mary to Isg
‘Mary smiled at me’
(2128) Naa te kurii i mau mai te raakau
actgen the dog T/A bring hither the stick
‘The dog brought me a stick’

However, compare the following:

(2129) I riri ia ki ahau
T/A angry IIIsg to Isg
‘She growled at me’

My consultant said that the addition of mai in (2129) would give the sense ‘She vented
her anger upon me’.

2.1.3.6.10.2 Actions away from the speaker

Atu is added in many instances, eg.
(2130) Naa·ku i hoatu te paaoro ki a Hone
actgen·Isg T/A give the ball to pers John
‘I gave John the ball’
(2131) I kata atu ahau ki a Mere
T/A smile away Isg to pers Mary
‘I smiled at Mary’

However, compare (2126) with (2132), where atu was judged not appropriate:
(2132) I haere ahau ki te kite i a Mere
T/A move Isg to the see DO pers Mere
‘I went to see Mary’

Similarly compare (2128) with

(2133) Naa·ku i panga te raakau maa te kurii
actgen·Isg T/A throw the stick intgen the dog
hei tiki
for fetch
‘I threw the stick for the dog to fetch’
 Morphology 467

(2134) Naa·ku ia i riri

actgen·Isg IIIsg T/A angry
‘I growled at her’

It seems likely that actions such as throwing and moving are taken as normally involving
movement from the speaker’s location, and so this is not marked unless a contrast is
already implicit in the situation.

2.1.3.6.10.3 Actions towards the hearer

In some cases atu is added because the action is at the same time away from the speaker.
At other times atu is used when the speaker is not involved. There are also occasions
when mai can be used. Compare the examples in the previous two sections with the
following:
(2135) Naa·ku i whakahoki atu too pukapuka
actgen·Isg T/A return away sggenIIsg book
‘I returned your book’
(2136) E kata atu ana te mokopuna ki a koe
T/A smile away T/A the grandchild to pers IIsg
‘My grandchild is smiling at you’

(This may be a matter of speaker empathy.)

(2137) I haere mai maatou ki te kite i a koe
T/A move hither Iplexcl to the see DO pers IIsg
‘We came to see you’
(2138) E mau mai ana te kurii i te raakau ki
T/A bring hither T/A the dog DO the stick to
a koe
pers IIsg
‘The dog’s bringing you a stick’

These last two examples are indicative of the difficulties of accounting for these forms in
terms of speaker orientation in any literal sense. However, they make it clear that atu
cannot be regarded as hearer oriented.

2.1.3.6.10.4 Actions away from the hearer

Atu is sometimes used under such circumstances, eg.
(2139) Naa·u i hoatu te tikiti ki a ia
actgen·IIsg T/A give the ticket to pers IIIsg
‘You gave her the ticket’

(Ho- ‘give’ is a bound form requiring a directional particle.)

(2140) I kata atu koe ki a ia
 Maori 468

T/A smile away IIsg to pers IIIsg

‘You smiled at her’

Contrast:
(2141) Naa·u i panga te raakau ki te kurii
actgen·IIsg T/A throw the stick to the dog
‘You threw the stick to the dog’

Again, the principles governing the use of these particles are not clear to me.

2.1.3.6.10.5 Towards a third person

Atu is used if the speaker is involved as the source, but can also be used if some other
participant is the source, eg.
(2142) Ka whakahoki·a atu e maatou te kaakahu o
T/A return·pass. away by Iplexcl the garment gen
Hone
John
‘We’ll give John back his coat’
(2143) I kata atu ia ki a Hone
T/A smile away IIIsg to pers John
‘She smiled at John’

Cases like the last may depend upon the speaker’s spatial or emotional relationship to the
participants.

2.1.3.6.10.6 Away from a third person

Again, the speaker’s relationship to the other participants seems to determine the form
used.

2.1.3.6.10.7 Other kinds of directionals

Iho and ake have already been discussed in 2.1.3.6.10. There are as many problems
associated with these as with mai and atu. In temporal contexts, iho is commonly used to
mean from time past towards the present, eg.
(2144) Hua·ina iho ki te taunga ko
name·pass. down to the fishing ground spec
Kamokamo
Kamokamo
‘The fishing ground was called Kamokamo (knowing glances)’ (TWh, 17)

Ake is used from the future back to the present, eg.

(2145) Aapoopoo koe i te ata haapara te rongo
tomorrow IIsg at the morning dawn the hear
 Morphology 469

ake ai
away part.
‘Tomorrow at break of day you’ll hear of it’ (TWh, 17)

In some cases, it is clear that iho means ‘from a superior position down towards speaker,
and ake ‘from the speaker to a superior position’, eg.
(2146) Ka whakahuu ake raatou, ka akiaki iho
T/A shout up IIIpl T/A complain down
te waha“…”
the voice
‘They shouted, and the voice protested from above, “…”’(TWh, 18)

However, in other instances, although such an interpretation cannot be totally excluded, it

is much less clear; consider the examples in bold below:
(2147) Ka tae a Taranga ki teetahi puu wiiwii
T/A arrive pers Taranga to a(sp) clump rushes
ka tango·hia ake e ia. He rua i raro. Te
T/A take·pass. up by IIIsg cls hole at below the
heke·nga iho o Taranga ki roto i taua
descend·nom down gen Taranga to inside at det aph
rua, ngaro tonu atu
hole vanish quite away
‘Taranga reached a clump of bulrushes and pulled it up. Below it was a hole. Taranga
descended into this hole and vanished’ (KM, 3)
(2148) “Ko aku maatua tonu eenei e noho ake
eq plgenIsg parents indeed these T/A sit up
nei”. Ka tango ia i teetahi hua o taua
proxI T/A take IIIsg DO a(sp) fruit gen det aph
raakau, ka panga·a iho
tree T/A throw·pass. down
‘“These are my parents sitting beneath me.” He picked a berry from that tree and threw it down’
(KM, 3)

The speaker, Maui, disguised as a bird, is sitting up in a tree. When examples like these
are compared with examples like (2123) in 2.1.3.6.10, it will be seen that a somewhat
similar use is involved. Since Maui is the speaker here, it is difficult with ake, at least, to
argue that the position of the narrator is relevant. Thus it seems that we must accept
apparently contradictory uses of these particles, and hope that further research will
produce generalizations which make more sense of the data.

2.1.3.6.11 Body orientation

Body position is not coded on the verb, either morphologically or with particles.
 Maori 470

2.1.3.6.12 Incorporation processes

The only clear case of incorporation involves indefinite direct objects of canonical
transitive verbs. In this process, only the noun stem is incorporated; there is no case
marker or determiner, and thus no number marking. The incorporated object follows the
verb directly, and any post-posed verbal particles follow the incorporated object, eg.
(2149)) E tuhi·tuhi reta ana ia
T/A write-write letter T/A IIIsg
‘She is writing letters/a letter’ ‘She is letter-writing’

However, there are cases where lexical adverbs might be considered to be incorporated.
The types in question appear to be manner adverbials. Some examples follow:
(2150) I haere manuhiri mai a Tuwharetoa
T/A move visitor hither pers Tuwharetoa
‘Tuwharetoa came as a visitor’ (H, 5)
(2151) E wawata puku ana teetehi me teetehi ki
T/A desire stomach T/A a(sp) with a(sp) to
a raaua ake
pers IIIdl up
‘Each ardently desired the other (H, 6)
(2152) Kei te titiro whakatau mai ki a ia
T/A look intently hither to pers IIIsg
‘[They] gazed intently at him’ (TWh, 19)

It is clear from the position of the particles and the lack of phrase-type markers that there
are some similarities between these constructions and the incorporated direct objects.
However, manner adverbs are subject to passive agreement, eg.
(2153) Ka moe·a tahae·tia a Rangiuru e
T/A sleep·pass. thief·pass. pers Rangiuru by
Tuwharetoa
Tuwharetoa
‘Rangiuru was slept with illegitimately by Tuwharetoa’ (H, 5)

Incorporated direct objects, naturally, cannot accompany passive verbs, and constructions
like the following are ungrammatical:
(2154) *E hii ika ana ia i ngaa moki
T/A catch fish T/A IIIsg DO the(pl) moki
‘He is fish-catching the moki’

In both cases, the formation appears to create a compound verb. However, while the
incorporation of direct objects is completely productive, and obligatory for indefinite
non-specific objects, the creation of compound verbs from manner adverbials is not
particularly productive. It is perhaps important to note that Maori does not have a totally
synonymous syntactic paraphrase for these incorporated objects (compare Mithun, 1984,
 Morphology 471

847–8). In fact, in several respects, this process in Maori contravenes generalizations

proposed by Mithun. She suggests that only expected combinations of verb and object
will be encoded this way, but that is not true for Maori: sentences of the following kind
are perfectly possible:
(2155) E whakairo ngeru ana ia
T/A draw cat T/A IIIsg
‘She is drawing cats’
(2156) E hii kahawai ana ia
T/A catch kahawai T/A IIIsg
‘He is fishing for kahawai’

Such incorporated objects can also have modifiers, both nominal and adjectival:
(2157) E kohi hua raakau ana ia
T/A gather fruit tree T/A IIIsg
‘She is berry-picking’
(2158) E ruku·ruku kooura nu·nui ana ia
T/A dive·dup crayfish big·dup T/A IIIsg
‘He is diving for big crayfish’

2.1.3.7 Strings of verbs

There are instances like the following which may constitute strings of verbs:
(2159) …whai haere tonu ia i ngaa waahine raa
follow move still IIIsg DO the(pl) women dist
‘…he still walked on, following the women’ (TWh, 19)
(2160) Kei te takahi haere atu i ngaa tapuae
T/A tread move away at the(pl) footprint
‘He was walking, treading in their footsteps’ (TWh, 19)

However, it is impossible to be certain that the second form in such a string is a verb,
since it is only possible to determine in most instances whether a given base is a verb by
the accompanying particle, and the second item in such constructions has no particle,
since it functions as a modifier. This construction is probably not different from the
manner adverbs discussed in the preceding section. These formations are probably thus
best regarded as compound verbs, rather than as strings of verbs.
Other than this, strings of verbs occur only through coordination or subordination, in
which case the markers are determined by the type of construction involved. In
coordination, each verb retains its T/A marker, and any post-posed particles (see further
1.3.1.3.1), eg.
(2161) Kei te ata poo, ka eke, ka patu·a
at(fut) the morning night T/A arrive T/A kill·pass.
‘In the early morning they will arrive and [the others] will be killed’ (W, 199)
(2162) Ka mutu, ka haere atu, ka tae ki te
T/A finished T/A move away T/A arrive to the
 Maori 472

one
beach
‘When that was finished, they left and arrived at the beach’

Several types of subordination involve the replacement of the T/A marker with ki te or
kia, in which case there is no tense or aspect marked on the subordinate verb. However,
these processes can only produce verb chaining if the subject (which would otherwise
intervene) is omitted, either by anaphoric deletion or some other process, eg.
(2163) Ka hiahia ia ki te haere kia kite i ngaa
T/A desire IIIsg to the go subj see DO the(pl)
kaakaapoo
kakapo
‘He wants to go to see the kakapo’

2.1.4 Adjectives
It must be borne in mind in this section that it is questionable whether Maori has a class
of adjectives distinct from verbs (see the discussion in 1.16.3).

2.1.4.1 Predicative υs. attributive adjectives

There are no morphological changes in the form of the adjective associated with
attributive vs. predicative use. However, adjectives when predicative are the heads of
either verbal or non-verbal predicates, and are accordingly accompanied by appropriate
phrase-type particles, whereas attributive adjectives do not have associated particles of
their own independent of the head word to which they are attributive. Compare the
following:
(2164) He maha ngaa raakau nu·nui
cls many the(pl) tree big·dup
‘There are/were many big trees’ (lit. The big trees are/were many’)
(2165) He nu·nui ngaa raakau
cls big·dup the(pl) tree
‘The trees are big’
(2166) Ka nu·nui ngaa raakau
T/A big·dup the(pl) tree
‘The trees are getting big’

2.1.4.2 Absolute and contingent state

Attributively, permanent and temporary attributes are not different in form, eg.
(2167) He puke teitei teenaa
cls hill tall that
‘That’s a high hill’
(2168) He wai paru teenaa
cls water dirty that
 Morphology 473

‘That’s dirty water’

although predicative forms are normally used for such statements, eg.
(2167a) E teitei ana teenaa puke
T/A tall T/A that hill
That’s a high hill
(2167b) E paru ana teeraa wai
T/A dirty T/A that water
‘That’s dirty water’

However, predicatively, only absolute states can be expressed as classifying predicates,

while contingent states are expressed as verbal predicates. Note, however, that what
counts as absolute includes many states of lengthy, but not permanent duration.
Sometimes the sense of the adjective changes with predicate type, eg.
(2169) He maa taku whare
cls white sggenIsg house
‘My house is white’
(2170) E maa ana taku whare
T/A clean T/A sggenIsg house
‘My house is clean’

Thus there are certain adjectives which are not found as the head in he predicates, eg. ora
‘well’
(2171) *He ora teenei tamaiti
cls well this child
‘This is a healthy child’

but note
(2171a) Tino ora teenei tamaiti
very well this child
‘This child is very healthy’
(2171b) He peepi ora teenei
cls baby well this
‘This is a healthy baby’

2.1.4.3.1 Adjective-Noun agreement

Adjectives do not agree with nouns in person, gender, case or definiteness. They may
appear to agree in number, but this is probably best described as marking distributiveness
(ie. the possession of the quality by every individual member of the group), rather than
number.
Compare the following:
(2172) Ka kite ahau i teetahi raakau nui
T/A see Isg DO a(sp) tree big
 Maori 474

‘I saw a big tree’

(2173a) *Ka kite ahau i ngaa raakau nui
T/A see Isg DO the(pl) tree big
‘I saw the big trees’
(2173b) Ka kite ahau i ngaa raakau nu·nui
T/A see Isg DO the(pl) tree big·dup
‘I saw the big trees’
(2174a) He poto ngaa raakau
cls short the(pl) tree
‘The trees are short’
(2174b) He po·poto ngaa raakau
cls short·dup the(pl) tree
‘The trees are short’

For some speakers, at least, certain attributive adjectives must always be reduplicated in
an NP marked as plural (but see below, 2.1.4.3.2.1). As will be seen, the marking of the
distribution of the quality across all the members of the set takes the form of
reduplication of the initial mora of the base (ˆ indicates the division into morae):
nuˆi > nuˆnuˆi
paˆi > paˆpaˆi

Only a small number of adjectives show this form of reduplication. Those that do include
kino ‘bad’, nohinohi (>nonohi) ‘small’, nui ‘big’, pai ‘good’, poto ‘short’, rahi ‘big’, riki
‘small’, roa ‘long’, but I cannot guarantee that this is a complete list. Some common
adjectives which do not have distributive reduplication include iti ‘small’, kawa ‘sour’,
koi ‘sharp’, mate ‘unwell’, ora ‘well’, reka ‘sweet’, tika ‘right’. It will be clear from these
lists that there are not obvious phonological constraints determining the occurrence of
reduplication.

2.1.4.3.2 Factors affecting the use of the distributive

2.1.4.3.2.1 Relative position of noun and adjective

The use of the distributive form occurs whether the adjective is predicative or attributive,
eg.
(2175) He kai pa·pai eenei
cls food good·dup these
‘This is good food’
(2176) He pa·pai aa maatou kai
cls good·dup plgenIplexcl food
‘Our food is good’
(2177) Ka pa·pai hoki aa koutou kai
T/A good·dup also plgenIIpl food
 Morphology 475

‘Your food is especially tempting’

(Note that kai ‘food’ in all these is marked as plural elsewhere in the sentence.) As noted
above in 2.1.4.3.1, some consultants reject sentences with an unreduplicated adjective
attributive to a plural noun. Other descriptions of Maori differ in their comments on this
matter, which suggests that there is considerable variation among Maori speakers. Biggs,
1969, 107 implies that reduplication is optional, and Foster, 1987, 18 states that it is.
Hohepa, 1967, 43 implies that it is obligatory for those adjectives that reduplicate, as
does Harawira, 1950, 17. Wills, 1960, 48 says “the longer form is more pleasant to the
Maori ear”. Certainly younger speakers frequently do not use the reduplicated forms.

2.1.4.3.2.2 Overt expression of the noun

The noun does not have to be overtly expressed to trigger the agreement, eg.
(2178) He nu·nui eenei
cls big·dup these
‘These are big’

However, note that the plurality of the understood noun is marked in the pro-form.

2.1.4.3.3 Factors affecting distributive agreement

2.1.4.3.3.1 Conflict between grammatical and semantic categories

This does not arise.

2.1.4.3.3.2 Agreement with coordinated nouns

The use of the distributive with coordinated nouns shows variability. If both nouns are
singular, both distributive and nondistributive are possible, but the non-distributive was
the first choice, eg.
(2179) Te matua me te tamaiti tino roa/ro·roa
the parent with the child very tall/dup·tall
‘The father and the boy are tall’

With plural nouns, both are possible, but the distributive was the first choice:
(2180) Ka pa·pai/pai ana koorero·tanga me
T/A dup·good/good plgenIIIsg read·nom with
ana tuhituhi·nga
plgenIIIsg write·nom
‘Her reading and writing are good’

Where there is a singular coordinated with a plural, there is attraction to the nearest item:
(2181a) Ka pai/*pa·pai te whare me ngaa kai
T/A good/dup·good the house with the(pl) food
 Maori 476

‘The house and the food are good’

(2181b) Ka pa·pai/pai ngaa kai me te whare
T/A dup·good/good the(pl) food with the house
‘The food and the house are good’

However, note the asymmetry in the judgements provided about the other alternative in
these two cases.

2.1.4.4 Expression of comparison

Comparison in Maori is not expressed by changes in the form of adjectives, but by
particles (see also 1.8, 1.9). The particles in question are manner particles, and most can
modify several other grammatical classes in addition to adjectival predicates.

2.1.4.4.1 Equality
Equality is expressed through the use of rite ‘like’, usually with tonu ‘still, indeed’. The
standard of equality is expressed nominally, eg.
(2182) He/Ka rite tonu te roa o Piri ki te roa o
cls/T/A like indeed the tall gen Bill to the tall gen
Hone
John
‘Bill is as tall as John’

2.1.4.4.2 Comparative
Comparison uses the adjectival predicate in either type of predicative construction, with
the adjective followed by a directional particle, eg.
(2183) He/Ka roa atu a Piri i tana
ds/T/A long away pers Bill compar sggenIIIsg
paapaa
father
‘Bill is taller than his father’
(2184) E nui ake teenei
T/A big up this
‘This one is bigger’

Attributive comparatives take the same form:

(2185) Te mea nui ake teenei
the thing big up this
‘This is a bigger one’
 Morphology 477

2.1.4.4.3 Superlative
Maori does not have a morphological superlative. Superlatives can be translated as
definite NPs, often with intensifiers such as tino or rawa accompanying the quality.

2.1.4.4.3.1 Compared to other entities

Often no intensives are added, eg.
(2186) Ko Waikato te awa whaanui
eq Waikato the river wide
‘The Waikato is the widest river’
(2187) Ko ngaa rere o te Sutherland te mea tiketike
eq the(pl) fall gen the Sutherland the thing high
(o te motu)
gen the island
‘The Sutherland falls are the highest (in the country)’

It appears that such superlatives can only be expressed attributively; mea ‘thing’ is
introduced as a prop if required. (Note that these constructions provided by my consultant
differ from those in Head, 1989, 42 in not having the article repeated before the
adjective.)
The addition of intensives serves to emphasize the great degree to which the quailty is
found:
(2188) Ko teehea te mea pai rawa?
eq which(sg) the thing good intens
‘Which is the best one?’
(2189) Ko teehea te mea tino pai rawa?
eq which(sg) the thing very good intens
‘Which is the very best one?’

Sometimes such forms are made overtly comparative by the addition of atu, eg.
(2190) He pai rawa atu te mea whero
cls good indeed away the thing red
‘The red one is the best’

Notice that this is predicative.

2.1.4.4.3.2 Compared to itself at other points/times

Such matters are usually expressed as overt comparatives, with atu and often an
intensifier to capture the high degree to which the quality is present, eg.
(2191) Ko teenei te waahi whaanui atu o te awa
eq this the place wide away gen the river
nei
 Maori 478

proxI
‘This is the widest place on this river’
(2192a) I tekau maa ono oona tau, tino tuuai rawa
at 10 and 6 plgenIIIsg year very thin intens
atu ia
away IIIsg
‘She was thinnest when she was 16’, ‘She was extremely thin
when she was 16’

The last example does not imply, as the English superlative does, that there was no time
in her life at which she was thinner. It does not appear possible to capture that except
through some overt verbal expression, eg.
(2192b) I tekau maa ono oona tau, ka tae ia
at 10 and 6 plgenIIIsg year T/A arrive IIIsg
ki toona tuuai·tanga
to sggenIIIsg thin·nom
‘When she was sixteen, she reached her thinnest’

2.1.4.5 Degrees of quality

2.1.4.5.1 In large measure

This is expressed by intensifiers of which tino and rawa are the principal ones, eg.
(2193) Tino roa a Hata
very tall pers Hata
‘Hata is very tall’
(2194) Pakeke rawa atu teenaa paatai
difficult intens away that question
‘That question is very difficult’

The initial particles are very commonly omitted from such predicates. Both verbal and
non-verbal predicates can express ‘large measure’, and such phrases can be attributive,
although there is a strong tendency to prefer predicative forms:
(2195) He paatai pakeke rawa teenaa
cls question difficult intens that
‘That’s a very difficult question’

2.1.4.5.2 In superabundance
There is no way of expressing ‘in superabundance’ distinct from ‘in large measure’,
although it appears that rawa (not tino) is the appropriate intensifier if superabundance is
intended:
(2196) He moomona rawa a Hone
cls fat intens pers John
 Morphology 479

‘John is very fat’/‘John is too fat’

Sometimes atu is added, eg.

(2197) He roa rawa atu a Hone
cls tall intens away pers John
‘John is too tall’

Both predicative types are found. One way of indicating superabundance is to use kua as
the T/A marker:
(2198) Kua pakeke rawa ngaa paatai
T/A hard intens the(pl) question
‘The questions have got too hard’

2.1.4.5.3 In small measure

Aahua ‘somewhat’ is used for the expression of ‘in small measure’. It precedes the
adjective, eg.
(2199) He aahua tini tonu te kai
cls somewhat big still the food
“Food is still moderately plentiful’
(2200) He aahua nu·nui rawa ngaa kai
cls somewhat big·dup intens the(pl) food
‘This is rather too big a helping’

Frequently, ake ‘up’, a directional particle , is added:

(2201) Kua aahua roa ake a Hone
T/A somewhat tall up pers John
‘John is quite tall’

Such phrases can also be attributive:

(2202) He paatai aahua pakeke ake teenaa
cls question somewhat difficult up that
‘That is a rather difficult question’

2.1.4.5.4 Others
There are a few other intensifiers which have more specific senses, eg. raapea ‘indeed’,
which implies agreement with the judgement of one’s interlocutor, eg.
(2203) Te nui raapea o te pahi!
the big indeed gen the bus
‘Yes, the bus is indeed big’

However, such forms do not measure different degrees to which a quality is possessed.
 Maori 480

2.1.4.6 Verbal morphology with predicative adjectives

Predicative adjectives are indistinguishable from other predicates in terms of the T/A
markers they take (see 1.16.3). Like other predicates, they have -Canga nominalizations.
Since they are intransitive, passive voice does not apply. There are no copular verbs in
Maori.

2.1.5 Prepositions
Maori has only prepositions, and no post-positions.

2.1.5.1 Prepositional usages

The prepositions of Maori are listed in 1.16.5. The discussion of their uses is scattered in
various places throughout this work. An attempt is made here to summarize the main uses
of each, and to cross-refer to the principal sections where those uses are discussed and
illustrated. Note the general discussion of the preposition as a part of speech in 1.16.5.
The list below is alphabetical.
A(a) (i) future temporal location: 2.1.1.6
A(a) A-class possessive
(ii) in nominalizations: 2.1.1.3.3–4
possession: 1.10, esp. 1.10.6
source: 2.1.1.4.2
E (i) vocative: 1.1.1.3; 2.1.1.4.28
E (ii) passive agent: 2.1.1.2.13
passive causer: 2.1.1.4.12
†
Hei classifying predicates: 1.2.1.1.1.2; 2.1.1.2.9.3
†
essive: 2.1.1.4.16
†
function: 2.1.1.4.14
future spatial location: 1.2.1.1.4.1–2; 1.10.3; 2.1.1.5
future temporal location: 1.1.2.4.2.1.2; 2.1.1.6
nominalizations: 1.1.2.2.6
†
object complement: 2.1.1.2.11
†
subject complement: 2.1.1.2.10
†
In those instances daggered hei is not unambiguously a preposition. It behaves in certain
respects like he, which is clearly in some instances an article. For some discussion see
Johansen, 1948, 14–15.
I cause: 1.1.2.4.2.4; 2.1.1.2.13; 2.1.1.4.12
comparison: 1.1.2.4.2.7.1; 1.8; 2.1.1.2 6
DO: 1.2.1.2.2; 2.1.1.2.4
equatives: 1.1.2.4.2.7.2; 1.9; 2.1.1.2.7
part-whole (local nouns): 2.1.1.4.18
reference: 2.1.1.4.15
source: 2.1.1.4.2
spatial locative: 1.2.1.1.4.1–2; 1.10.3; 2.1.1.5
temporal locative: 1.1.2.4.2.1.1; 2.1.1.6
 Morphology 481

Kei part-whole (local nouns): 2.1.1.4.18

spatial locative: 1.2.1.1.4.1–2; 1.10.3; 2.1.1.5
temporal locative: 2.1.1.6
Ki addition: 2.1.1.4.27
benefactive: 2.1.1.4.1
†
complement of adjective: 1.2.2.1; 2.1.1.2.12
†
condition: 1.1.2.4.2.5
DO: 1.2.1.2.2.; 2.1.1.2.4
equatives: 1.1.2.4.2.7.2; 1.9; 2.1.1.2.7
instrumental: 2.1.1.4.3
IO: 1.2.1.2.3; 2.1.1.2.5
†
noun clauses: 1.1.2.2.2.3; 1.1.2.2.6
part-whole (local nouns): 2.1.1.4.18
purpose: †1.1.2.4.2.3; 2.1.1.4.13
reference: 2.1.1.4.15
†
In uses marked with a dagger, the construction following ki always begins with te. Such
uses are not unambiguously prepositional. Ki te appears to function together as a
conjunction. There are other instances below which suggest that the functions prepositon
and conjunction are not always clearly discrete in Maori.
Ko clefting: 1.11.2.1.4
equative predicates: 1.2.1.1ff; 2.1.1.2.9.2
spatial locative (only some dialects): 1.2.1.1.4.1–2; 2.1.1.5
specifier: 1.3.1.3.1; 2.1.1.4.31
topic switch marker: 1.12.1
Maa actor-emphatic: 1.11.2.1.7; 2.1.1.2.13
agent after hei: 2.1.1.3.3–4
benefactive: 2.1.1.4.1
cause: 2.1.1.4.12
instrumental: 2.1.1.4.3
IO: 2.1.1.2.5
possession: 1.10; 2.1.2.4.10
route: 2.1.1.5.5, 6, 17-18,21
transport: 2.1.1.4.3
Me circumstance: 2.1.1.4.5
comitative: 2.1.1.4.4
†
condition: 1.1.2.4.2.5
coordination: 1.3.1.3.1; 1.3.1.4
inclusion: 2.1.1.4.25
†
manner: 1.1.2.4.2.2
similitude: 2.1.1.4.33
†
In the constructions marked with a dagger, it is not clear that me is a preposition rather
than a conjunction.
Moo agent after hei: 2.1.1.3.3–4
benefactive: 2.1.1.4.1
10: 2.1.1.2.5
possession: 1.10; 2.1.2.4.10
 Maori 482

price: 2.1.1.4.10
purpose: 2.1.1.4.13
reference: 2.1.1.4.15
temporal duration: 2.1.1.6.5
value: 2.1.1.4.21
Naa actor-emphatic: 1.11.2.1.7; 2.1.1.2.13
cause: 2.1.1.4.12
possession: 1.10; 2.1.2.4.10
†
result clause: 1.1.2.4.2.6
route: 2.1.1.5,9,21
source: 2.1.1.4.2
†
It is unclear in this context whether naa is a preposition or conjunction.
Noo part-whole: 2.1.1.4.18
possession: 1.10; 2.1.2.4.10
†
result clause: 1.1.2.4.2.6
source: 2.1.1.4.2
spatial locative: 2.1.1.5
temporal locative: 2.1.1.6.1.1
†
It is unclear in this context whether noo is a preposition or conjunction.
Oo in nominalizations: 2.1.1.3.3–4
partitive numeral: 2.1.1.4.19.1
partitive quantifier: 2.1.1.4.19.3
part-whole: 2.1.1.4.18
possession: 1.10.1–2; 1.10.6
source: 2.1.1.4.2
Whaka goal: 2.1.1.5.7

There are also two forms which, although not generally classified as prepositions, appear
to have prepositional usages:
ahakoa concession: 2.1.1.4.14
haunga exclusion: 2.1.1.4.26

A large corpus of prepositional usages reveals only one not adequately covered in 2.1.1,
and that concerns the expression of a source of opinion. There appears to be some
dialectal variation in this. Some dialects use ki, eg.
(2204) He pai ki a Pou ngaa mahi o te kura
cls good to pers Pou the(pl) work gen the school
‘Pou likes schoolwork’
(2205) Ki tooku whakaaro, tino turituri te miihini raa
to sggenIsg think very noise the machine dist
‘In my opinion, that machine is very noisy’

However, other dialects use i, eg.

(2206) He maahita ia, i taa te tuakana o Hone
 Morphology 483

cls teacher IIIsg from sggen the brother gen John

‘He’s a teacher, according to John’s brother’

It must be borne in mind that many of the more semantically specific English
prepositions are rendered in Maori by the use of preposition+ local noun+preposition.
(Note also that Maori has many combinations of preposition+local noun which are not
equivalent to English prepositions; these are not characteristically followed by an
adnominal prepositional phrase. See also 2.1.5.5.)

2.1.5.2 Agreement between prepositions and nouns

Prepositions in Maori do not show agreement with the nouns they govern.

2.1.5.3 Combinations of prepositions and personal pronouns

Singular personal pronouns combine with possessive prepositions (a, o, naa, noo, maa,
moo)’, however, the pronoun combining forms are not the same as the free pronouns:
Free Combining
au -ku
koe -u
ia -na

Information about the use of these forms is found in 1.10, 2.1.2.4.3, 2.1.2.4.10–11.
There is also a form koia, usually written thus as one word, which probably derives
from ko+ia. This form appears to be used in quite a wide variety of ways, from
expressing agreement (≡‘that’s so’) to emphasis, and sometimes the construction appears
to be a type of clefting:
(2207) Ko·ia ahau i haere mai ai
eq·IIIsg Isg T/A move hither part.
‘That’s why I came’

In these instances ia appears to be used as a sentential pro-form:

(2208) Ko·ia anoo a Te Aotakii i waawaahi ai
eq·IIIsg again pers Te Aotaki T/A break part.
i toona maunga
DO sggenIIIsg mountain
‘So that was the reason why Te Aotaki rent his mountain’ (TWh, 21)
(2209) Ko·ia nei hoki te ingoa o te kaainga o
eq·IIIsg proxI also the name gen the home gen
Pou i Hawaiki
Pou at Hawaiki
‘This was also the name of Pou’s home in Hawaiki’ (P, 1)
 Maori 484

There are also cases where koia is written as one word when -ia is clearly a IIIsg
anaphoric personal pronoun, eg.
(2210) Oho rawa ake ia, ko·ia e maatakitaki·na
wake intens up IIIsg eq·IIIsg T/A watch·pass.
ana e te tini koroheke o roto i te
T/A by the many old men gen inside at the
whare raa
house dist
‘She awoke and found herself being stared at by a crowd of old men from that house’ (TA,
6)

2.1.5.4 Prepositions combining with articles

Prepositions do not combine with articles in Maori. However, the status of the forms he
and hei in constructions like the following must be raised:
(2211) He kaiako ia
cls teacher IIIsg
‘She is a teacher’
(2212) Hei kaiako ia
cls(fut) teacher IIIsg
‘She is to be a teacher’
(2213) Haere hei tauira ki ngaa iwi o eeraa waahi
move as example to the(pl) people gen those place
‘Go as an example to the people of those places’ (TR2, 97)

He is unmistakably an article in instances like:

(2214) E tangi he puu i ngaa poo katoa
T/A play a flute at the(pl) night all
‘A flute sounds every night’ (H, 7)

However, he cannot follow any preposition except me ‘like’ (see 2.1.1.4.33). Hei is a
preposition with a future sense in examples like
(2215) Hei taku whare te hui
at(fut) sggenIsg house the meeting
‘The meeting will be at my place’

It will thus be seen that some arguments exist for classifying he and hei as articles (since
phrases like the first in (2211) and (2212) would then conform to the normal Maori
structure for noun phrases). But it might also be argued that both forms are prepositions.
(For some discussion see Johansen, 1948, 14–15, where he concludes that hei is both, but
does not extend that conclusion to he.) It is clearly a logical possibility that they are
forms which combine preposition+article, although I cannot see any way to argue for it
 Morphology 485

2.1.5.5 Combinations of prepositions and local nouns

The purpose of this section is to draw attention to the fact that many of the more specific
prepositions of English are rendered in Maori by combinations of preposition+local
noun+preposition. Maori has an open-ended set of local nouns, since they include place
names. The majority characteristically occur in constructions of the form
‘preposition+local noun’, eg. ki uta ‘to shore’. However, there is a small group of what
might be called ‘relational local nouns’ which occur at least as often in the construction
‘preposition+local noun + preposition+NP’. The chief of these relational local nouns with
references to the most relevant sections discussing them are:
runga superior: 2.1.1.5.7,8,8a; 2.1.1.5.23,24,29
raro inferior: 2.1.1.5.9,10; 2.1.1.5.26
waho exterior: 2.1.1.5.3,4; 2.1.1.5.15–16
roto interior: 2.1.1.5.3; 2.1.1.5.21; 2.1.1.6.12–13
mua anterior: 2.1.1.5.5; 2.1.1.6.10
muri posterior: 2.1.1.5.6; 2.1.1.5.15–16; 2.1.1.6.11
waenga(nui) medial: 2.1.1.5.17–18

2.1.6 Numerals/quantifiers

2.1.6.1 Numerals for counting

The basic numbers of Maori are as follows:
1 tahi 6 ono
2 rua 7 whitu
3 toru 8 waru
4 whaa 9 iwa
5 rima 10 tekau

The form used for ‘zero’ in mathematical contexts is kore.

The numbers 11–19 have the form ‘10+maa+digit’, eg.
11 tekau maa tahi
19 tekau maa iwa

The multiples of ten from 20–90 have the form ‘(e)+digit+10’, eg.
20 (e) rua tekau
80 (e) waru tekau
 Maori 486

If the number is used in isolation (eg. as a total), the particle e is normally used. In a list
of numbers, however, it is usually omitted. Numbers intermediate between the multiples
of ten have the form ‘(e) +multiple of ten+maa+digit’, eg.
21 (e) rua tekau maa tahi
99 (e) iwa tekau maa iwa

Round hundreds are expressed as ‘(e)+digit+rau’ with the exception of

100 kotahi rau

Numbers intermediate between round hundreds take the form for the round hundred
followed by the form for the two figure number, eg.
525 (e) rima rau, (e) rua tekau maa rima

Round thousands are expressed as ‘(e)+digit+mano’ with the exception of

1000 kotahi mano

Numbers intermediate between round thousands take the form for the round thousand
followed by the form for the three figure number, eg. 3150 (e) toru mano, kotahi rau,
rima tekau
When counting objects, the digits can be preceded by ka ‘T/A’, and typically are when
the process of counting is in focus, as when teaching a small child, ie. under these
circumstances it would be usual to say:
Ka tahi, ka rua, ka toru…ka iwa, (ka) tekau, tekau maa tahi…
Speakers vary as to whether they use ka with tekau ‘10’ or not. However, my
consultants would not use ka if they were counting a pack of cards, for instance. There it
would be normal to use unprefixed numbers, and to abbreviate forms after 10, eg.
tekau, maa tahi, maa rua…(e) rua tekau, maa tahi…
The same forms are used for the counting process regardless of the type of object
being counted.

2.1.6.2 Attributive forms of cardinal numerals

There are two possible constructions for attributive numerals: either they precede the
noun, in which case the noun has no determiner, or they follow it, in which case the noun
has a determiner, eg.
 Morphology 487

The numerals 2–9 and all forms beginning with these digits require e; tahi ‘1’ always
takes the form kotahi; numerals from 10–19 take no prefix, thus:
(2217) te maaripi kotahi
the knife one
‘one knife’
(2218) ngaa waka e rua tekau
the(pl) canoe num 2 10
‘the twenty canoes’

There does not appear to be any semantic difference between the pre-head and post-head
constructions. Both are found in older texts, eg.
(2219) Ka tango ia i ngaa tahaa e ono hei
T/A take IIIsg DO the(pl) calabash num 6 for
whaka·tere moo·na
cause-float intgen·IIIsg
‘She took six calabashes as floats for herself’ (H, 5)
(2220) E whaa waewae i kite atu ai ahau i
num 4 feet T/A see away part. Isg at
roto
inside
‘I saw four feet inside’ (H, 10)

It should be noted, however, that expressions containing numerals in Maori are of ten
predicative, rather than attributive, eg.
(2221) E rua tekau, e rua tekau maa tahi raanei,
num 2 10 num 2 10 and 1 or
ngaa hooiho i reira
the(pl) horse at there
‘Twenty or twenty-one horses were there’/‘The horses that were there numbered 20 or 21’
(2222) E ono ngaa matapihi
num 6 the(pl) window
‘There are six windows’ (lit ‘The windows are 6’)

Even when they are not clearly predicative, they are often fronted, leaving ai behind, as
in (2216) and (2220) above.

2.1.6.3 Counting different types of object

When counting, no distinction is made between types of object. However, when a number
of people is mentioned from 2–9, the prefix toko- is traditionally used, rather than e (but
this rule is increasingly of ten ignored), eg.
(2223) Toko·rima oona tuaakana
pnum·5 plgenIIIsg brother
‘He had 5 older siblings’ (KM, 1)
(2224) He kotahi te kuumara i roto i te rourou
 Maori 488

cls one the kumara at inside at the basket

maa ngaa taangata toko·rua
intgen the(pl) people pnum·2
‘There was one small round basket containing one kumara to two men’ (W, 198)
(2225) He kotahi kuumara maa te tangata kotahi
cls one kumara intgen the man one
‘There was one kumara for each man’ (W, 198)
(2226) I haere mai e rua rau taangata
T/A move hither num 2 hundred people
‘Two hundred people came’

These forms can be used substantivally, eg.

(2227) Noo whea too koutou toko·rima?
actgen where sggenIIpl pnum·5
‘Where is your fifth one from?’ (KM, 2)

2.1.6.4 Ordinal numerals

There are two methods of creating ordinal numerals in Maori. Ordinal numerals can be
formed with the prefix tua-. However, such forms are possible only as far as iwa ‘9’.
Such ordinal numerals follow the noun, eg.
(2228) Te tangata tua·tahi tonu ko Tupeteka
the man ord·1 still eq Tupeteka
‘The first man was Tupeteka’ (W, 200).
(2229) Ko te tohu tua·rua teenei i tae mai ki
eq the sign ord·2 this T/A arrive hither to
a ia
pers IIIsg
‘This was the second sign that came to him’ (W, 198)

These forms can also be used as substantives, eg.

(2230) Naa, ka mau ki te tahaa, i te tua·rua…
then T/A seize to the calabash DO the ord·2
‘Then [he] took the calabash, the second one…’ (H, 9)

Alternatively, a definite article with a cardinal numeral can be used, eg.

(2231) Ko te toru o ngaa manu, he kuuaka
eq the 3 gen the(pl) bird cls godwit
‘The third bird was a godwit’

This is the only form for numbers above 9, eg.

(2232) I te whaa tekau o ngaa tau o te
at the 4 10 gen the(pl) year gen the
kiingitanga o Rawiri…
reign gen David
 Morphology 489

‘In the 40th year of the reign of David…’ (PT, I Ngaa Whakapapa 26, 31)

The following o phrase is not obligatory, eg.

(2233) te raa whitu
the day 7
‘the seventh day’ (PT, Ekoruhe 20,11)

2.1.6.5 Other derivatives of numerals

There are a number of other derivatives to describe.
There is a further numeral prefix, taki- which means ‘in groups of. It can be applied to
the numerals 1–9, eg.
(2234) Kua uru ia ki roto o ngaa maatua a
T/A enter IIIsg to inside gen the(pl) battalion gen
teeraa, taataa haere ai taki·rua, taki·toru, ki
that beat down move part. gp·2 gp·3 to
roto i te rapa o toona taiaha
inside at the blade gen sggenIIIsg taiaha
‘He had made his way into the centre of the enemy’s battalions, striking down as he went 2
or 3 with each stroke of his taiaha’ (TWh, 23)

Takitahi means ‘a scattering’ as well as the predictable ‘one at a time’, eg.

(2235) Taki·tahi noa maatou i te hui
gp·1 indeed Iplexcl at the meeting
‘There were just a few of us at the meeting’

Takiruarua (with reduplicated ‘2’) has a similar derived usage, eg.

(2236) E taki·rua·rua haere ngaa tupu
T/A gp·2·dup move the(pl) grow
‘The plants were growing scattered in pairs’

Rua ‘2’ and toru ‘3’ have reduplicated forms, ruarua and torutoru with the sense ‘few’,
but most dialects use one to the exclusion of the other, eg.
(2237) He rua·rua ngaa waahi e kite·a ai teenei
cls 2·dup the(pl) place T/A see·pass. part. this
manu
bird
‘There are few places where this bird is found’ (TR2, 117)

Whaka+numeral is used in the Bible for translating fractions, eg. whakatekau for ‘tenth
part’ or ‘tithe’, eg.
(2238) …he waahi whaka·tekau o aua whaka·tekau
cls part cause·10 gen det aph(pl) cause·10
‘…even a tenth part of the tithe’ (PT, Tauanga 18, 26)
 Maori 490

However, this form does not seem to be in common use nowadays. See below for modern
mathematical formulations.
The prefix hoko- can be used with numbers from 1–9 to mean ‘× 20’. Thus hokowhaa
‘eighty’. Of the numbers produced with this prefix, hokowhitu appears to be used not
only literally for 140, but also for a party (often in connection with war-parties), and
sometimes simply for a large number. In the following example, it is not clear that the
party consisted literally of 140:
(2239) Ka whao·na te wharau nei e te hoko·whitu
T/A enter·pass. the shed proxI by the 20×·7
raa
dist
‘The party entered the construction’ (W, 198)

With the present impetus to restore the Maori language, there has been a need for means
of expressing mathematical concepts. The Maori Language Commission has accepted the
following schema for the expression of fractions (he hau) (Bill Barton, personal
communication):
number+hau+number (the preferred form)
number+number·nom
number wehenga number, eg.
(2240) e rua hau toru
num 2 portion 3
or e rua toru·tanga
num 2 3·nom
‘Two-thirds’
e rua wehe·nga toru
num 2 divide·nom 3
‘two over three’

It will be seen that the second of these involves -Canga suffixation of numerals, and thus
means that an entire set of derivatives from numbers of this form is potentially in use. For
the record, decimals (tau-aa-ira ‘number-link-spot’) are expressed simply as ‘number ira
number’, eg.
(2241) toru ira whitu
3 spot 7
‘three point seven’, ‘3·7’

2.1.6.6 Quantifiers
Only katoa ‘all’ and the numerals are clear-cut quantifiers in Maori. Certain other items
which are quantifiers in other languages are determiners in Maori. Some examples are
given below of the equivalents in Maori for the commonest English quantifiers. Further
examples are given in 1.2.5.2.6. The first group includes ‘some’, ‘each’, ‘other’ and
‘any’; the Maori equivalents are determiners.
‘Some’ is rendered by he or (t)eetahi, eg.
 Morphology 491

(2242) Mau·ria mai he miraka maa·ku

bring·pass. hither a milk intgen·Isg
‘Bring me some milk’

He is restricted in the syntactic environments in which it can occur (see 2.1.1.10.2). He is

not marked for number. (T)eetahi can be used in the environments from which he is
excluded. However, te or ngaa are often used. (T)eetahi appears usual if ‘some’ is
stressed:
(2243) E tika ana koe kia moohio ki eetahi kupu
T/A right T/A IIsg subj know to some(pl) word
Maaori
Maori
‘You must know some Maori words’

‘Each’ is sometimes translated by ia, which appears to be a determiner. However, it is

equally often translated by repeated phrases, and sometimes by both, eg.
(2244) Pee·nei tonu ia raa, ia raa
like·proxI still each day each day
‘It was like this each/every day’
(2245) E kotahi aku aaporo i hoatu ai maa
num one plgenIsg apple T/A give part. intgen
teenaa, maa teenaa (maa teenaa) o ngaa
that intgen that intgen that gen the(pl)
tamariki
children
‘I gave each child an apple’

The number of repeated phrases is a matter of emphasis. Note that ‘every’ cannot be
distinguished from ‘each’ (or ‘all’, see below). ‘Other’ is rendered by (t)eetahi (atu), eg.
(2246) Teenaa pea ko eetahi atu o ngaa tamaiti
that perhaps top. some(pl) away gen the(pl) child
raa i kite i te taahae·tanga
dist T/A see DO the thief·nom
‘Other boys may have seen the theft’

However it should be noted that these forms are not the equivalent of the other in
English. An example like (2247) does not exclude the possibility of further cups of the set
existing:
(2247) Kua pakaru teetahi o ngaa kapu. Teetahi kei
T/A broken a(sp) gen the(pl) cup a(sp) at(pres)
roto i te kaapata. Teetahi (atu) kei runga
inside at the cupboard a(sp) away at(pres) top
i te teepu.
at the table
‘One of the cups is broken. Another is in the cupboard. Another is on the table’
 Maori 492

To capture this, it is necessary to use some form such as toenga ‘remainder’, or paunga
‘exhausting’, eg.
(2248) Kua pakaru teetahi o ngaa kapu. Ko te
T/A broken a(sp) gen the(pl) cup top. the
toe·nga kei roto i te kaapata.
remain·nom at(pres) inside at the cupboard

‘One of the cups is broken. The other is in the cupboard’

‘Any’ in questions is translated by he or (t)eetahi, eg.
(2249) He tamaiti raanei e moohio ana ki te
a child or T/A know T/A to the
whakautu?
respond
‘Does any child know the answer?’

The next group are frequently predicates in Maori, though they can also be attributive.
These include ‘both’, ‘few’ and ‘many’. ‘Both’ is rendered by rua ‘two’, eg.
(2250) Toko·rua ngaa tamaiti i kite i te taahae
pnum·2 the(pl) boy T/A see DO the thief
‘Both boys saw the thief

‘Few’ is rendered by numeral-based forms used predicatively, eg.

(2251) Taki·tahi noa ngaa taangata e moohio ana
gp·1 indeed the(pl) people T/A know T/A
ki te whakautu
to the answer
‘Few people know the answer’

Another possibility to replace takitahi in (2251) is rua·rua ‘2·dup’ or toru·toru ‘3·dup’.

‘Many’ is most often rendered by maha ‘much, many’ or tini ‘numerous’ used
predicatively, eg.
(2252) He/ka tini ngaa taangata e pai ana
cls/T/A numerous the(pl) people T/A good T/A
ki te kooura
to the crayfish
‘Many people like crayfish’

Maha cannot always occur in verbal predicates: he maha can replace he tini above, but
*ka maha was rejected. However, the ban is not absolute, as the following shows:
(2253) Kua maha kee hoki ngaa tau e noho ana
T/A many contr also the(pl) year T/A stay T/A
ia ki a au
IIIsg to pers Isg
‘He has been with me for very many years indeed’ (KH, 2)

They can also both occur as post-head attributives, eg.

 Morphology 493

(2254) Kua tae mai ngaa taangata tini noo

T/A arrive hither the(pl) people many actgen
Tuwharetoa
Tuwharetoa
‘The many people from Tuwharetoa have arrived’

Tini, but not maha, can also occur in pre-head position:

Negative quantifiers do not exist in Maori. Sentence negation or other lexical means
supply the need. ‘No’, ‘neither’, ‘none’, ‘not…any’ can be rendered as in the examples
below:
(2256) Kua pau ngaa miraka
T/A exhausted the(pl) milk
‘There’s no milk left’ (lit. The milk is finished’)
(2257) Kaahore he pukapuka i runga i te whata
neg a book at top at the shelf
‘There are no books on the shelf’
(2258) Kiihai ngaa tamaiti toko·rua nei i kite i
neg the(pl) child pnum·2 proxI T/A see DO
te taahae·tanga
the thief·nom
‘Neither boy saw the theft’
(2259) Tata kore ana he miraka i toe ana
near neg T/A a milk T/A remain T/A
‘There’s not much milk left’
(2260) Kaahore i maha/tini ngaa pukapuka i
neg T/A many/numerous the(pl) book at
runga i te whata
top at the shelf
‘There aren’t many books on the shelf’
 Maori 494

(2261) Kore rawa atu he miraka

neg intens away a milk
‘There isn’t any milk’

‘Every’ and ‘all’ are rendered by the quantifier katoa, eg.

(2262) I tika katoa ana paatai
T/A right all plgenIIIsg question
‘He got every answer right’,‘He got all the answers right’

‘Only’ is translated by anake, eg.

(2263) Ko ngaa tamariki pai anake ngaa mea ka
eq the(pl) children good only the(pl) thing T/A
whiwhi ki ngaa takoha
receive to the(pl) present
‘Only good children get presents’

2.1.6.6.1 Quantifier compounds

There are no quantifier compounds in Maori. Katoa can be used as a substantive, but
frequently full NPs are the equivalent of English quantifier compounds. The forms with
question words+raanei ‘or’ discussed in 2.1.2.1.14 are often equivalent A few examples
follow.
(2264) Ka pakipaki ngaa taangata katoa
T/A clap the(pl) people all
‘Everyone clapped’
(2265) Haere mai, haere mai, koutou katoa
move hither move hither IIpl all
‘Welcome, everyone’
(2266) Kore i kite·a ooku moowhitu i hea
neg T/A see·pass. plgenIsg glasses at where
raanei
or
‘I couldn’t find my glasses anywhere’

Note the form toko·maha ‘pnum·many’, eg.

(2267) He toko·maha ngaa taangata i mate i
cls pnum·many the(pl) people T/A dead from
tana mana maakutu
sggenIIIsg power magic
‘Very many people had died from his spells’ (KWh, 1)
 Morphology 495

2.1.6.6.2 Quantification by other means

The use of repeated phrases has already been mentioned with respect to ‘each’ above, as
has the use of determiners (see 2.1.6.6). Reduplication is not used in Maori to express
quantification except with the numbers rua and toru, see 2.1.6.5.

2.1.7 Adverbs
Adverbs are not a distinct part of speech in Maori, except for a small group of particles,
so there is no distinct adverb morphology. However, it should be noted that certain
classes of adverb agree with the verb in voice, see 1.2.1.3.1.1, 2.1.3.6.12.

2.1.7.1 Comparison of adverbs

The only means of rendering such concepts is through comparison of predicative
adjectives.

2.1.7.1.1 Equality
See 1.1.2.4.2.7, 1.9 and 2.1.4.4. Examples which in English involve adverbs are:
(2268) I rite tonu te horo o taau oma ki
T/A like indeed the fast gen sggenIIsg run to
taa Hone?
sggen John
‘Can you run as fast as John?’
(2269) I rite tonu taku pai ki te kaukau ki
T/A like indeed sggenIsg good to the swim to
toou
sggenIIsg
‘I can swim as well as you’

2.1.7.1.2 Comparison
See 1.1.2.4.2.7, 1.8 and 2.1.4.4. Examples which in English involve adverbs are:
(2270) I tere ake taana oma·nga i taau
T/A fast up sggenIIIsg run·nom compar sggenIIsg
‘He ran faster than you’
(2271) Pai atu ahau i a koe ki te kaukau
good away Isg compar pers IIsg to the swim ‘I can swim better than you’

2.1.7.1.3 Superlative
Just as there is no real superlative with adjectives, neither is there with adverbs (see
2.1.4.4).
 Maori 496

(2272a) Ko Hone te mea horo ki te oma

top. John the thing fast to the run
‘John runs fastest’, ‘John is the fast one at running’
(2272b) I horo atu a Hone ki te oma
T/A fast away pers John to the run
‘John runs fastest’, ‘John is fast at running’
(2272c) Ko Hone i horo ki te oma
top. John T/A fast to the run
‘John runs fastest’, ‘John runs fast’
(2273) Ko wai te mea pai ki te kaukau?
top. who the thing good to the swim
‘Who swims best?’, ‘Who swims well?’

2.1.7.1.4 Others
There are no other types of comparative with adverbs.

2.1.7.2 Degrees of quality

2.1.7.2.1 In large measure

Verbs can be modified directly by tino, which often has this type of sense, eg.
(2274) Kaahore ahau e tino moohio ki te koorero
neg Isg T/A very know to the speak
Maaori
Maori
‘I don’t speak Maori very well’

However, in most cases, an adjectival predicate is used, see 2.1.4.5.1, eg.

(2275) Tino tere te poti ki te haere
very fast the boat to the move
‘The boat goes very fast’
(2276) Tino pai tana tiaki i a Pare
very good sggenIIIsg care DO pers Pare
‘He looked after Pare very well’ (TR2, 57)

2.1.7.2.2 In superabundance
There is no way to distinguish this from ‘in large measure’. See 2.1.4.5.2.
(2277) I/E horo rawa tana oma
T/A fast intens sggenIIIsg run
‘He ran too fast’
 Morphology 497

2.1.7.2.3 In small measure

See 2.1.4.5.3. Aahua ‘somewhat’ is also used in instances corresponding to English
adverbs, eg.
(2278) I aahua aata haere ia
T/A somewhat slowly move IIIsg
‘He drove a bit slowly’

2.1.7.2.4 Others
There are no other modifications to adverbs.

2.1.8 Clitics
It is not clear what forms in Maori, if any, are to be counted as clitics. This is in part due
to the problem of defining a clitic: the forms which have been called clitics in the
literature are not a particularly homogeneous set. The closest definition seems to be that
clitics are elements with some of the properties of affixes and some of the properties of
words (see eg. Klavans, 1982, 3). In what follows, I review those elements of Maori
which fit this definition.
Most of the particles of Maori are bound forms which are not normally stressed
(although those prepositions which take pronouns without the personal article take the
phrase stress if they are accompanied by a singular pronoun). They thus form
phonological units with the rest of the phrase in which they occur. This is the basis for
Biggs’s observation that the phrase, phonologically definable, is the most significant unit
for the description of Maori grammar (Biggs, 1969, 17, 133). This is an affix-type
property. However, such forms can occur adjacent to forms of more than one class. Thus
verbal particles can be separated from the lexical verb by certain adverbs, eg.
(2279) Me aata haere koe oblig slow move IIsg
‘You’d better go carefully’

This freedom of occurrence is a word-like property. These two conflicting characteristics

are shared by all T/A/M particles, prepositions and articles. In the case of prepositions,
the following word can be either a determiner or a local noun; in the case of articles, there
are various intensifiers such as tino ‘very’ as well as various adverbial forms such as
deictics which can intervene between the determiner and head noun. However, the
presence or absence of an initial particle makes no difference to the placement of phrase
stress, so there seems to be no gain in treating such particles as clitics.
In negatives, however, where these particles are normally written as one word together
with a negative-sense verb, the stress pattern shifts, compare:
Ka ‘haere ‘Kaahore
T/A move neg
‘[He] went’ ‘No’
E ‘haere ana ‘Eehara
 Maori 498

T/A move T/A neg

‘[He] is going’ ‘No’

(The facts about stress are less clear with eehara: there appears to be some native speaker
variation.) Chung, 1970, 68 claims that these negative forms are still stressed on the base,
but this fails to take account of the fact that the rules for word stress apply only to final
phrases, and that different rules apply to non-final phrases (see Biggs, 1969, 133). Such
forms thus behave phonologically as single words. Furthermore, in these negative forms,
nothing can intervene between the T/A particle and the base. Thus there is reasonable
evidence for regarding these particles as affixes to the verbs, and not as clitics.
Various forms can attach to the article te ‘the’, eg. the deictic particles, and an open set
of possessive forms. Under these circumstances, there is a change in the phonological
form of the article: if the attached material begins with a consonant, the vowel of the
article is lengthened, eg.
teenei > te+nei ‘this’ >‘the’+‘proxI’

If the attached material begins with a vowel, the vowel of the article is dropped, eg.
taa Hone > te + aa Hone ‘John’s’ >’the’+‘gen John’
taa raaua > te+aa raaua their’ > ‘the’+‘gen IIIdl’

While te alone cannot be stressed, these complex forms can almost all be substantival,
and in such cases, if the complex form is a single word, the stress falls on the te thus:
‘teenei ‘this’

Thus there might be good reason to regard te as a clitic in these instances. Note that these
processes do not apply when heads of full NPs follow.
The forms of the singular pronouns which attach to possessive prepositions also sit on
the word-affix boundary. These forms are never stressed, unlike the dual and plural
pronouns in such constructions, eg.
maa ‘raaua ‘intgen IIIdl’
‘maaku ‘intgen Isg’

However, unlike the clitic pronouns in the Romance languages, for instance, these three
forms in Maori bear no formal relation to the independent singular pronouns which
follow other prepositions. The arguments for the clitic-like nature of these are less clear-
cut than for the particles, since they possess no obvious word-like qualities. However, it
was argued above that prepositions are probably clitics, and if these special singular
possessive pronouns were suffixes, a form of this kind would have a very strange
structure. If they are regarded as clitics, the situation is equally anomalous. However, it is
also far from clear that they are bases. Firstly, they are never stressed; secondly, bases in
Maori do not normally consist of a single mora; thirdly, they do not have the freedom of
occurrence of other bases in Maori. Nor does it seem possible to argue that the
prepositions in these forms are bases, since this would involve the claim that they are
 Morphology 499

bases if and only if they are attached to a singular pronoun. I do not see any way to
resolve the problems presented by these forms.
There are various other cases of phonological fusion in Maori, eg. with the
prepositions ki or i+a ‘pers’+singular pronoun, where the stress is moved from the
pronoun, compare
(2280a) ki a ‘raatou
to pers IIIpl
‘to them’
(2280b) ki ‘a ia
to pers IIIsg
‘to her’

where the personal article, normally unstressed, bears the stress. However, it is not at all
clear that such phenomena are to be accounted for in terms of cliticization, rather than as
arising from the operation of phrase stress rules, which seem to be applied fairly
superficially.

2.1.8.1 Types of clitic element

2.1.8.1.1 Personal pronouns

The only instances to be considered are the singular pronouns after ki and i mentioned in
2.1.8. See also 3.3.4.1 and 3.5.1.2.

2.1.8.1.2 Possessive pronouns

The only instance to be considered is the singular forms which attach to possessive
prepositions. See 2.1.8, and 2.1.2.4.3, 2.1.5.3.

2.1.8.1.3–4 Reflexive and reciprocal pronouns

There are none in Maori.

2.1.8.1.5 Auxiliary verbs

There are none in Maori, but note the discussion of verbal particles in 2.1.8. These
particles are extensively discussed and illustrated in 2.1.3.

2.1.8.1.6 Sentence particles

Sentence particles, as opposed to phrase-type particles, are not clitics in Maori.

2.1.8.1.7 Sentence connectives

These are not clitics in Maori.
 Maori 500

2.1.8.1.8 Anaphoric particles

Anaphoric particles are not clitics, except possibly (t)aua, which is a determiner. See
2.1.8 and 1.5.

2.1.8.1.9 Others
It was suggested in 2.1.8 that prepositions might also be clitics in Maori, as well as
articles. For detailed information on prepositions, see 2.1.1.4–6. For articles, see
1.2.5.2.4.

2.1.8.2 Positions occupied by clitics

2.1.8.2.1 Preverbal
Verb phrase-type markers occupy this position. However, since these particles are one
type of particle only, it is likely that this is not the most useful generalization about their
position. See 2.1.8.2.6.

2.1.8.2.2 Postυerbal
None of the candidates for clitichood in Maori occupy this position.

2.1.8.2.3 Sentence-final
None of the candidates for clitichood in Maori occupy this position, except by chance.

2.1.8.2.4 Sentence-initial
None of the candidates for clitichood in Maori occupy this position, except by chance.

2.1.8.2.5 Sentence-second position

None of the candidates for clitichood in Maori occupy this position.

2.1.8.2.6 Phrase-initial position

This is probably the most appropriate generalization for the position of many elements in
Maori which might be clitics. This covers all but the forms of possessive pronouns, which
always attach to specific prepositions (a feature which is un-clitic-like). Note, however,
that to preserve this generalization, it is necessary to regard articles as initial in NPs, and
prepositions as initial in prepositional phrases, not NPs.
 Morphology 501

2.1.8.3 Relative order of clitics

It is not possible to have more than one phrase-type particle in one phrase in Maori. Thus
ordering does not arise.

2.1.8.4 Combinations of clitics

No combinations occur with a phrase.

2.1.8.5 Expression of excluded combinations

Not applicable.

2.2 DERIVATIONAL MORPHOLOGY

The framework provided by the Questionnaire for the discussion of derivational

morphology presupposes that it is possible to determine the part of speech of the items
involved in derivational processes. As the discussion in 1.16 makes clear, for most bases
in Maori, it is not possible to determine the part of speech independent of the construction
in which the base occurs. Accordingly, the organization of the material below is a
compromise.
It has been observed by virtually all writers on Maori that the same word form can
function as several different parts of speech, without change of form. Consider the form
waiata in the following, for instance:
(2281) E waiata ana raatou
T/A sing T/A IIIpl
‘They are singing’
(2282) I tito·a te waiata nei e H.Melbourne
T/A compose·pass. the song proxI by H.Melbourne
‘H.Melbourne composed this song’
(2283) He pukapuka waiata teenei
cls book song this
‘This is a song book’
(2284) E haere waiata ana raatou
T/A move sing T/A IIIpl
‘They were walking along singing’

In some ways this resembles what in other languages is called conversion. However, the
name ‘conversion’ seems to imply two things: that a particular form can be identified as
originating as one part of speech, and that the part of speech of both ends of the
conversion process is determinable. Neither of these seems to be true of a considerable
proportion of those forms in Maori which can be used as nouns and as verbs, for instance.
What we appear to be faced with is a set of bases which are sometimes nouns and
 Maori 502

sometimes verbs, with appropriate difference in sense, without any overt derivational
process relating them. Many new items in the vocabulary immediately have this potential.
Krupa (1966) contains the only major study of derivational morphology in Maori. He
lists something approaching 500 forms which contain a base and an affix. There appears
to be little correlation between the parts of speech suggested by the glosses of either base
or affixed form, although Krupa points out that the majority of the derived forms with
prefixes are states or qualities (1966, 65). These prefixes appear to be relics of some
earlier word-formation process(es) which have little synchronic relevance in the
description of Maori. Some occur with fewer than five forms in Krupa’s lists. One or two,
however, occur in small sets of similar forms, and are thus transparent to native speakers
today, eg. puu- in
puu·whero ‘redd·ish’
puu·maa ‘whit·ish’

It is difficult to assess the current productivity of such forms. My consultant rejected

*puu·kikorangi ‘blue·ish’,

was uncertain about

?puu·aarani ‘orange·ish’

but thought
puu·koowhai ‘yellow·ish’

probably acceptable.
Krupa suggests that all 30 of the minor prefixes he finds (ie. those other than kai- and
whaka-) are allomorphs of one prefix. He also finds two unproductive suffixes (1966, 53–
4).
One of the difficulties of describing these sorts of phenomena in Maori today is that
many of the forms are not familiar to consultants, either because they were dialectally
restricted forms, or because they are no longer used. The vocabulary shrinkage is one of
the major signs of language decay in Maori. The Maori Language Commission (Te Taura
Whiri i te Reo Maori) has embarked on a process of word manufacture to provide
vocabulary necessary for Maori to function in today’s world. A study of the words they
have manufactured indicates that compounding and whaka- prefixing are the major
productive means of word-formation, along with the flexibility of forms to be used as
several parts of speech discussed above. In what follows I shall restrict myself to those
few forms which appear to be productive.

2.2.1 Formations producing nouns

There are two productive processes where the outcome is fairly clearly a noun. The first
is the -Canga nominalization, and the second the agentive prefix kai-.

-Canga nominalizations
 Morphology 503

It has already been suggested (1.16.3) that the possibility of forming -Canga
nominalizations can be used as a criterion for verbhood, ie. that all and only verbs in
Maori have nominalizations of this form. However, it must be noted that modifiers to
nominalizations get the nominal suffix added as an agreement phenomenon. A
considerable amount of information about these forms can be found in other sections.
Some information on the consonant is found in 2.1.3.1.1: the consonant is lexically
conditioned. The marking of the underlying subject is discussed in 1.10.6 and 2.1.1.3.3–
4, where the marking of other arguments is discussed. Clark, 1981 also contains useful
information. This process is basically semantically regular, although some
nominalizations have developed related senses other than the pure nominalization. It is
not an iterative process. The process applies to derived verbs as well as simple bases. The
nominalization is preceded by a determiner, as other nouns are. Since it has already been
illustrated in many other places, only a few further illustrations are given here.
(2285) E maumahara ana au ki te puta·nga
T/A remember T/A Isg to the appear·nom
ohorere·tanga mai o taua tangata i te
sudden·nom hither gen det aph man at the
kuuaha
door
‘I remember that man appearing suddenly in the doorway’
(2286) Ko te tikanga teenei o tana hakirara·tanga
eq the manner this gen sggenIIIsg insult·nom
i a ia
DO pers IIIsg
‘This is the manner of his insulting her’ (TA, 6)
(2287) Te oho·nga ake o Maaui i te awatea kua
the wake·nom up gen Maui at the dawn T/A
ngaro kee toona whaea
missing contr sggenIIsg mother
‘When Maui woke up at daybreak, his mother had gone’
(KM, 2)

Note that these nominalizations can still be followed by adverbial particles. Some forms
showing the variety of allomorphy follow:
suffix base gloss nom gloss
-nga
moe ‘sleep’ moenga ‘bed, place for sleeping,
marriage’
-anga
noho ‘sit, stay’ nohoanga (~ ‘place for sitting, chair’
nohanga)
-inga
kaa ‘take fire, burn, be kaainga ‘home, village, place where fire
lighted’ is lighted’
 Maori 504

-hanga
tangi ‘weep, cry, sound’ tangihanga ‘funeral, mourning’
-ranga
whakatupu ‘cause to grow’ whakatupuranga ‘generation’
-kanga
hopu ‘catch, seize’ hopukanga ‘catching, seizing’
-tanga
peewhea ‘like what’ peewheatanga ‘a plan, a means’
-
manga
tanu ‘bury, plant’ tanumanga ‘burial’

There are a number of forms where, although the base is commonly reduplicated, the
nominalization, like the passive, is not reduplicated, eg.
titiro ‘look at’ tirohanga ‘looking’

However, there are nominalizations of reduplicated forms, eg.

tirotiro ‘look about, investigate’ tirotirohanga ‘a sight’
huihui ‘gather, meet, assembly’ huihuinga ‘gathering’

Williams (1862, 45) generalizes the sense of this affixed form thus: “The noun thus
formed denotes the (a) circumstance, (b) time, (c) place or (d) purpose of the action
expressed by the verb”.
There are also a few cases where the nominalization ending -tanga is added to what
appear to be nouns rather than verbs, in which case they appear to create an abstract
noun, eg.
rangatira ‘chief’ rangatiratanga ‘evidence of breeding and greatness, sovereignty,
kingdom’
kaawana ‘governor’ kaawanatanga ‘Government’
kiingi ‘king’ kiingitanga ‘reign’

This does not appear to be a totally productive process. Thus the form
raakautanga < raakau ‘tree’

was not rejected outright by my consultant, but I was told “people don’t talk like that
nowadays".

Kai- agentives
 Morphology 505

Kai- can be added as a prefix to a transitive verb to create an ‘agent’ noun which is
always human, and not instrumental, eg.
mahi ‘work, do’ kaimahi ‘worker’
koorero ‘speak’ kaikoorero ‘speaker, story-teller’
ako ‘teach’ kaiako ‘teacher’

Kai can be added to transitives produced by whaka- derivation, eg.

whakawaa ‘judge, accuse’ kaiwhakawaa ‘judge’
whakahaere ‘run, organize’ kaiwhakahaere ‘organizer, foreman’

It can also be added to reduplicated bases, eg.

titiro ‘look at’ kaititiro ‘watchman’

This process is productive across the set of transitive verbs, and semantically regular.
Note also
kaitito waiata ‘song-writer’ > tito waiata ‘to song-write’

where the base is a compound verb formed by noun incorporation. The process of kai-
prefixation is not iterative. Some new formations from the Maori Language Commission
are (He Muka 3(1)):
kaiwawao ‘referee’
kaimaatai waa ‘time-keeper’
kaiwhakataetae ‘contestant’

Some verbal-type morphology remains if there are modifiers to these nouns. The
underlying DO can retain its marking with i, eg.
(2288) Ko Tama-te-hura te kai whakahau i ngaa
eq Tama-te-hura the ag command DO the(pl)
haka katoa
haka all
‘Tama-te-hura was the director of all the hakas’ (W, 200)

Note the fact that the kai here has not been written as a prefix to the following form.
Whether this is because it is in some sense prefixed to the entire phrase is unclear.
The only productive process creating nouns from nouns is reduplication, see 2.2.6.4.
Since adjectives in Maori are a sub-class of verbs, the only process for creating nouns
from adjectives is -Canga nominalization, discussed above. Nouns are not derivable by
affixation from other categories.
 Maori 506

2.2.2.1 Formations producing verbs

There are two productive processes in Maori which produce verbs, but it is not entirely
clear whether they are derivational rather than inflectional. The first is passive
suffixation, the second, whakaprefixation. The passive appears likely to be inflectional: it
is productive across all the possible input class with the exception of a handful of
irregular verbs which take no termination (eg. homai, hoatu ‘give’); it is semantically
regular, and it does not change the major class of the lexical item: derived passive forms
take the T/A markers which are verb-phrase markers as do underived verbs. Thus no
more will be said about the passive here (see 2.1.3.1.1).
Whaka- derivatives are less clear-cut Whaka- creates transitive verbs from
intransitives, including neuter verbs and adjectives. It thus appears to be fully productive
(see 2.1.3.1.3). However, it also appears that it can be added to nouns to make either an
intransitive or a transitive verb. This latter process does not seem to be as productive as
the first. While the first instance is class-maintaining (in the broad sense of class), the
second is class-changing. Whaka- derivatives are not particularly regular semantically—
many have become specialized in sense. Examples of this have been given in 2.1.3.1.3. It
is probably the case that whaka- formations are derivational (see also Williams, 1862,
43–4; Biggs, 1969, 83–4). The verbal formations have been discussed and illustrated in
2.1.3.1.3. The nominal construction is illustrated below:
tangata ‘man’ whakatangata ‘to change into a man’
koowhatu ‘stone’ whakakoowhatu ‘petrify, turn to stone’

eg.
(2289) Aananaa, kua whaka·tangata taua manu behold T/A cause·man det aph bird
‘Behold, that bird resumed human form’ (KM, 3)

Two recent creations by the Maori Language Commission are:

raa ‘day’ whakaraa ‘date (letters etc)’
koonae ‘small basket’ whakakoonae ‘file (papers etc)’

Whaka- is also sometimes added to forms which are neither nouns nor verbs, eg. in the
first example to an adverbial particle and in the second to a negator:
atu ‘away from speaker’ whakaatu ‘to show, point out, draw attention to’
kaahore ‘negative’ whakakaahore ‘to deny, resist’

Whaka- can be added to compounds, and to reduplicated forms eg.

tangata whenua ‘local person’
whakatangata whenua ‘to take on the attributes and duties of
a local person’
whiwhi ‘acquire, receive’
 Morphology 507

whakawhiwhi ‘to award, present’

It can also be added to kai- formations, eg.

(2290) Kua whaka·kai·ako·tia ia
T/A cause·ag·teach·pass. IIIsg
‘He has been trained as a teacher’

It cannot be added to nominalizations:

(2291) *Kua whaka·noho·nga teenei tangata
T/A cause·sit·nom this man
‘That man turned into a chair’

(Some paraphrase with huri ‘turn’ is required to encode this.) Whakaprefixation is not an
iterative process.
Whaka- formations are an important means of creating new vocabulary items. Thus in
a recent list of words for the modern office prepared by the Maori Language
Commission, nearly half the verbs listed are whaka- formations. Not all of these are new,
but amongst the forms which appear to be new in addition to the forms whakaraa and
whakakoonae given above are:
waatea ‘free’ whakawaatea ‘to resign’
pou ‘post’ whakapou ‘to stamp (with a rubber stamp)’
weto ‘be extinguished’ whakaweto ‘to turn off (machines etc)’
kaa ‘burn’ whakakaa to turn on (machines etc)’

It will be seen that these new formations are both on bases which are verbs and on bases
which are nouns.
The only other means of creating verbs from verbs is reduplication, see 2.2.6.4.

2.2.3.1 Formations producing adjectives

Since adjectives are a sub-class of verbs, the processes involving them are dealt with in
the preceding section (See 1.16.3). However, it is not clear that new forms can be added
freely to this class.

2.2.4.1 Formations producing adverbs

There are basically no derivational means of producing adverbs in Maori. The adverbial
particles are monomorphemic. Forms which function as lexical adverbs are marked as
such by position, not form, although it should be noted that if they modify passive verbs
or nominalizations, they take the passive termination -tia or the nominalizing suffix -
tanga as an agreement phenomenon. (See eg. 2.1.3.1.3 and 2.2.1.) There is a group of
temporal adverbs such as inanahi ‘yesterday’ and aapoopoo ‘tomorrow’ which are fairly
transparently formed from prepositional phrases, but this is not productive. In some
cases, at least, such forms are interruptable by other adverbs, eg. inaianei ‘today, now’
 Maori 508

inaia tonu nei ‘right now’ (cp. Williams’s Dictionary entry for this form). While they are
commonly written as one word, one of my consultants sometimes writes them as two
words. These forms are only marginally to be considered single words.

2.2.5 Formations producing any other category

There do not appear to be any other clearly productive means of word-formation except
those with numerals, listed in 2.1.6.

2.2.6 Complex prepositions

2.2.6.1.1 Two prepositions

It is not possible to string prepositions together in Maori. However, in a few cases where
a time adverb has been formed from an original prepositional phrase, notably aapoopoo
> a poo poo ‘at(fut) night night’, another preposition is sometimes used before it, eg.
(2292) Hei aapoopoo,… at(fut) tomorrow
‘Tomorrow,…’

However, I have not found any similar cases with adverbs other than aapoopoo.

2.2.6.1.2 Nominal formations

The preposition complexes of the form ‘preposition+local noun+preposition’ have
already been discussed in 2.1.5.5. Because they potentially have substitutable items in all
slots, there is no reason to consider these as complex words rather than as syntactic
constructions, eg.
i runga (i)
kei raro (o)

The second prepositional phrase is often separated from the local noun by another
sentence constituent (see 1.13.2.5).
If haaunga ‘excluding’ is a nominalization from hau ‘excess’ it would be a relevant
formation, but this derivation is in considerable doubt The process is not productive.

2.2.6.1.3 Verbal formations

The form ahakoa (>aha koa ‘what indeed’) can be used as a preposition (see 2.1.1.4.24).
There is no evidence that this is productive.

2.2.6.1.4 Adjectival formations

There are none.
 Morphology 509

2.2.6.1.5 Other types

The bi-morphemic forms naa/noo/maa/moo should perhaps be mentioned. These consist
of m-, n- and aa, oo. The prepositions aa and oo also occur free. The prefixes (which
mark a realis/irrealis type of distinction) do not clearly recur with other forms, although
Clark (1976, 115) suggests that m- recurs in me ‘oblig’. This is not productive.

2.2.6.2 Simple derived prepositions

These do not occur in Maori.

2.2.6.3 Compound morphology

Firstly, it is necessary to address the question of whether Maori has compounds. Hohepa
(1967, 16) explicitly denies this: “There is no justification for treating modified—
modifier constructs as compounds; what are regarded as compounds in English do not
become isomorphic units when translated into Maori”. He does not provide any
arguments for that position, and I believe there is at least one important argument against
it. Adjectival modifiers in Maori cannot be strung together in one phrase (see 1.2.5.2.1).
However, constructions containing two elements after the head do occur , eg. (from
Hohepa, 1967, 25–6) te whare heihei nui ‘the large hen-house’. In a construction like
this, it is clear on semantic grounds that nui ‘big’ modifies whare heihei ‘hen-house’, not
heihei ‘hen’. Thus if heihei were a modifier of the same type as nui, a construction of this
type should be ungrammatical. It is possible to get two modifiers with one head in some
instances when those same modifiers with another head would require head repetition, eg.
(2293) he kopa·maaroo mangu
a wallet·hard black
‘a black briefcase’
(2294) he koowhatu maaroo, he mangu hoki
a stone hard a black also
‘a hard black stone’

(My consultant was uncertain whether waimangu (lit ‘water black’) ‘ink’ could be
modified by ‘incompatible’ colour adjectives, eg.
(2295) ?te wai·mangu whero
the water·black red
‘the red ink’

and suggested waiwhero (lit. ‘water red’) would be more appropriate. This judgement
may have been influenced by unfamiliarity with waimangu for ‘ink’; mangumangu was
given as the known alternative.) However, even without this last argument, there seem to
be good reasons for analysing some head-modifier constructions as compounds.
However, it must be pointed out that it is possible for a modifying nominal element to
be marked for plural in the case of those few nouns which show plurality. Thus the
following forms are found in the new words from the Maori Language Commission:
(2296) Mana·tuu Waahine
 Maori 510

power-stand women
‘Ministry of Women’s Affairs’

but
(2297) tohe mana wahine
argue power woman
‘feminist’

Once again, there are problems in the analysis of compounds in terms of part-of-speech,
because the part of speech of the components is not always clear.
Compounding is an extremely important means of word formation in Maori. Indeed it
is the principal productive means, and in the current political climate, when borrowing
with phonetic adaptation is an unacceptable means of word manufacture, it accounts for
an extremely high proportion of new nouns in particular. Thus from a list of 108 nouns
for the modern office from the Maori Language Commission, 73 are compounds,
although from a comparable sized list of verbs, there are fewer than 5 compounds.

2.2.6.3.1 Compound nouns

These have the order head+modifier, eg.
(2298) ipu·para
container-waste
‘rubbish bin’

While there must be some uncertainty in some cases about the contributing parts of
speech, the following combinations seem to be found:
Noun-Noun:
(2299) raarangi taonga
row treasure
‘catalogue’
(2300) aupuru teepu
cushion table
‘desk pad’
(2301) roro·hiko
brain·electricity
‘computer’
(2302) pouaka koonae
box basket
‘filing cabinet’

Noun-Verb:
These are much more difficult to find. However, the following possibly constitute
examples:
(2303) paanui whakamaarama
 Morphology 511

notice explain
‘pamphlet’
(2304) maar ama· taka
month·revolve
‘calendar’

Noun-Adjective:
(2305) kopa·maaroo
wallet·hard
‘briefcase’
(2306) wai·mangu
water·black
‘ink’
(2307) koopae piingore
round house flexible
‘floppy disk’

Verb-Noun:
This is a surprisingly large category:
(2308) whaka·koi pene
cause-sharp pen
‘pencil sharpener’
(2309) poro·pepa
cut·paper
‘guillotine’
(2310) wero·pepa
pierce-paper
‘hole punch’
(2311) huuhi maapere
cover finger
‘finger cone’

Note that in all these cases, the second element is a semantic DO for the verb element
This is not purely a new phenomenon (and thus is not a formation process borrowed from
English): pre-European contact examples include paretai (pare ‘turn aside’, tai ‘sea, tide,
wave’) ‘washboard of a canoe’, and toromoka (toro ‘thrust endways’, moka ‘end’) ‘bone
pin for fastening a cloak’. Note that all the examples found, both ancient and modern, are
instruments/implements. However, in the compound waka·rere·rangi ‘canoe·fly·sky’
‘aeroplane’, the modifying element appears to be a verb-noun combination where these
restrictions are not in force. However, rererangi does not have independent existence.
Other:
There are in addition many formations where there is uncertainty as to the parts of
speech involved. Consider
(2312) here·rapa
 Maori 512

tie·rubber
‘rubber band’
(2313) wai whakamaa
water whiten / shame
‘correcting fluid’

Here can be a verb ‘to tie’ or a noun ‘a tie, string, etc to tie with’. Rapa is borrowed from
English, and is presumably a noun. Either the nominal or the verbal sense of whakamaa
could be semantically appropriate in the second example. Such forms are in principle not
classifiable.
There are no clear-cut nominal compounds involving adverbial elements, but the
Maori Language Commission formations that follow must at least be considered as cases
in this category:
(2314) paepae reta mai/atu
dish letter hither/away
‘in/out-tray’
(2315) nama mai/atu
account hither/away
‘accounts receivable/payable’

There is a further type of formation involving the linking element -aa- which is being
used increasingly at present There are older forms like
(2316) waiata-aa-ringa
song-link-hand
‘action song’

cf.
(2317) waiata poi
song poi
‘poi song’, song to accompany a poi dance’
(2318) hui-aa-tau
meeting-link-year
‘annual meeting’

where the -aa- element may indicate a looser semantic relationship between the parts of
the formation than straight compounding.
However, when I asked about the formation
(2319) ?waiata ringa
song hand
‘hand song’

hoping to establish a contrast, I was told that there was no such word, and that there was
no way to tell what it would mean. There are quite a number of formations of this
structure in the new words proposed by the Maori Language Commission, eg.
 Morphology 513

(2320) tau-aa-ira
number-link-spot
‘decimals’
(2321) kaupapa-aa-rua
base-link-two
‘binary number’
(2322) utu-aa-tau
price-link-year
‘salary’
(2323) utu-aa-haaora
price-link-hour
‘wages’

It is not clear whether this aa is a preposition, but note the lack of articles with the
following element, normally found only with local nouns following prepositions. Aa is
the form of a future temporal locative preposition, and also the marked possessive
preposition in Maori (see 1.10). It is difficult to see that the -aa- in the above forms is
related to either.
It is possible in Maori to have compounds with either a compound head or a
compound modifying element, although the latter is more frequently encountered than the
former, eg.
(2324) wai pane kuini
water head queen
‘stamp dampener’

where
(2325) pane kuini
head queen
‘postage stamp’

is a compound modifier. Similarly

(2326) puurere hopu reo
machine catch voice
‘dictaphone’
(2327) puurere hopu tangi
machine catch sound
‘tape-recorder’
(2328) waka·rere·rangi
canoe·fly·sky
‘aeroplane’
(2329) kopa whaka·wiri tiitoki
bag cause-juice titoki
‘bag for squeezing oil from titoki’
 Maori 514

where the last is an older example parallel to the new formations. As an example with a
compound base, consider
(2330) maaramataka teepu
calendar table
‘desk calendar’

(see above (2304) under Noun-Verb combinations for the structure of ‘calendar’).
Compounds can also contain derived forms: there are examples of whaka- and kai-
derivatives, reduplications and nominalizations in compounds, eg.
(2331) puurere whaka·ahua machine cause-likeness
‘photocopier’

(see also ‘pencil sharpener’ (2308) under Verb-Noun combinations above)

(2332) papa pato·pato
board tap·dup
‘keyboard’
(2333) rau·rau tuhi
leaf·dup write
‘writing pad’
(2334) hui·nga ao
gather·nom world
‘universal set’ (mathematics)
(2335) utu tatau·ranga
pay number·nom
‘accountancy fees’

An example of a kai- formation from a compound base is

(2336) kai·tohe mana wahine
ag·argue power woman
‘feminist’

(cf. the base ‘feminist’ (2297) above). It is also possible to derive whakaforms from
compound bases, eg.
(2337) whaka·tangata whenua
cause-person land
‘to become a local person’ (RR, 12)

which must be analysed as whaka[tangata whenua].

2.2.6.3.2 Compound verbs

The process of object incorporation (see 2.1.3.6.12) creates compound verbs entirely
productively, and some of these compound verbs are used as elements in nominal
compounds discussed above. However, compounding is not a major means of creating
 Morphology 515

new verbs. In the Maori Language Commission lists, the only verb compounds use the
link element aa discussed above in noun compounds:
(2338) paanui-aa-waha
inform-link-mouth
‘to dictate’
(2339) whaka·maaori-aa-waha
cause·Maori-link-mouth
‘to interpret into Maori’
(2340) whaka·maaori-aa-tuhi
cause·Maori-link-write
‘translate into Maori’

One interesting usage from the mathematical terminology must be mentioned:

(2341) Ine·a-aa-mita·tia te roa o te ruuma measure·pass.-link-metre·pass. the long gen the room
‘Measure the length of the room in metres’

Here, when the verb is passive, the modifier in the construction takes the passive suffix in
agreement. This is characteristic of lexical adverbs, but not of prepositional phrases.
Compound verbs created by noun incorporation are not subject to whaka prefixation,
despite the fact that they are grammatically intransitive. Thus from hii ika ‘to fish catch’,
it is not possible to form *whakahii ika ‘to cause to fish catch’, eg.
(2342) *Ka whaka·hii ika a Pani i a Hata
T/A cause·catch fish pers Pani DO pers Hata
‘Pani made Hata go fishing’

However, it is possible to make kai- formations, eg.

(2343) kai·hoko pukapuka
ag·sell book
‘book seller’

This process is normally restricted to transitive verbs.

When nominalizations of such constructions occur, only the verbal element is
nominalized, eg.
(2344) I tana ruku·hanga kooura(·*tanga), ka
at sggenIIIsg dive·nom crayfish(·nom) T/A
whiwhi·a ia
entangle·pass. IIIsg
‘When he was diving for crayfish, he got entangled’

2.2.6.3.3 Other types of compound

There do not appear to be other types of compound in Maori. This is largely a reflection
of the fact that there are no other major parts of speech.
 Maori 516

2.2.6.4 Reduplication
This is one of the major types of word formation in Maori, but it is a process where there
are very many questions still to be answered, both in formal terms and in semantic terms.
Harlow (1991) has recently made some illuminating proposals with respect to the formal
description of reduplication, and what follows is indebted to his work.
Firstly, most bases in Maori have either two or three morae. (For the term ‘mora’ as
applied to Maori, see 3.2.3.3 and Bauer, 1981b.) The reduplication patterns for bases of
these shapes have to be specified separately. (Biggs (1969, 107) discusses reduplication
in terms of ‘partial’ and ‘complete’, but this fails to take account of base shape. The
partial/complete dichotomy goes back at least as far as Williams (1862, 25, 46), and is
also used by Krupa (1966, 47). In Bauer 1981b, an analysis in terms of morae was
proposed where ‘initial’ and ‘final’ reduplication was discussed. However, this, too falls
short of accounting for all the observed patterns.)
For two mora stems, partial-initial and complete reduplication appear to account for
most forms, although there are occasional forms in Williams’s Dictionary which must be
analysed as partial-final forms. These three formal patterns are illustrated here with
“marking divisions into morae:
nuˆi ‘big’ nuˆnuˆi ‘big (distributive)’
huˆi ‘meet, gather’ huˆiˆhuˆi ‘gather together’
aˆnga ‘face, turn to’ aˆngaˆnga ‘respect’

Of these, reduplication of both morae is by far the commonest

For three mora stems, Harlow allows for four patterns of reduplication. This analysis
depends on the consonant dissimilation rule which is Harlow’s chief interest in this paper.
Briefly, he suggests that in certain word positions, the second of two identical consonants
can be dropped from a Maori word. Harlow’s patterns are:
1. reduplication of first mora: 1123, eg.
hoˆiˆa ‘annoyed’ hoˆhoˆiˆa ‘annoyance’
2. reduplication of first and second morae: 11223, eg.
maˆneˆi ‘reach out to’ maˆaˆneˆneˆi ‘reach out to’
(<*maˆmaˆneˆneˆi)
3. reduplication of first two morae: 12123, eg.
haˆwaˆi ‘edible fungus’ haˆwaˆhaˆwaˆi ‘compost heap’
4. reduplication of all three morae: 112323, eg.
aˆkaˆhu ‘shallow’ aˆaˆkaˆhuˆkaˆhu ‘indistinct’

Of these, pattern 4 appears to be the only one likely to be productive.

From the semantic point of view, there are few generalizations apparent amongst the
lexicalized reduplicated forms. Thus with two mora stems, partial reduplication is found
with at least the following meanings:
(i) distributive plural, eg.
roˆa ‘tall’ roˆroˆa ‘tall (distributive)’
(ii) diminished intensity, eg.
 Morphology 517

paˆngo ‘black’ paˆpaˆngo ‘blackish, dark’

(iii) no change, eg.
haˆe ‘slit, lacerate, tear, cut’ haˆhaˆe ‘id.’
(iv) frequentative, eg.
haˆka ‘dance’ haˆhaˆka ‘dance on’
(v) other, eg
haˆke ‘unseemly, unbecoming’ haˆhaˆke ‘naked, bare’

With two mora stems, complete reduplications can be found with all meanings but the
first above:
(ii) diminished intensity, eg.
weˆra ‘hot’ weˆraˆweˆra ‘somewhat hot’
(iii) no change, eg.
aˆnu ‘cold’ aˆnuˆaˆnu ‘id.’
(iv) frequentative, eg.
hiˆko ‘flash’ hiˆkoˆhiˆko ‘flash repeatedly, twinkle’
(v) other, eg.
aˆni ‘resounding, echoing’ aˆniˆaˆni ‘disparage, belittle’

There are some stems where stem and both patterns of reduplication exist without a
determinable change in sense, eg.
haˆe, haˆhaˆe, haˆeˆhaˆe ‘lacerate’
Similarly, with three mora stems, the different patterns of reduplication are not
associated with different senses. Because there is relatively little data for pattern 2, it is
not listed below. (Harlow found in Williams’ Dictionary the following numbers of forms:
pattern 1:27; pattern 2:9; pattern 3:33; pattern 4:21.)
Pattern 1:
(i) transitive, eg.
paˆkaˆru ‘broken’ paˆaˆkaˆru ‘break in pieces’
(ii) nominalization, eg.
taˆkaˆi ‘wrap up’ taˆaˆkaˆi ‘bandage
(noun)’
(iii) no change, eg.
moˆhuˆa ‘yellowhead’ moˆmoˆhuˆa ‘id.’
(iv) other, eg.
maˆoˆa ‘cooked, ripe’ maˆmaˆoˆa ‘steam’

Pattern 3:
(i) intransitive, eg.
hoˆroˆi ‘wash’ hoˆroˆhoˆroˆi ‘wash, shower (intr.)’
(ii) adjective, eg.
taˆweˆe ‘noise’ taˆweˆtaˆweˆe ‘noisy’
 Maori 518

(iii) no change, eg.

paˆkeˆe ‘creak’ paˆkeˆpaˆkeˆe ‘id.’
(iv) frequentative, eg.
*haˆpaˆi lift up’ haˆpaˆhaˆpaˆi ‘id. (freq)’
(v) other, eg.
*whaˆngaˆi ‘feed’ whaˆngaˆwhaˆngaˆi ‘charm’

Pattern 4:
(i) diminished intensity, eg.
maˆraˆra ‘scattered’ maˆaˆraˆaˆraˆra ‘rather scattered’
(ii) no change, eg.
?maˆtoˆru ‘benumbed’ maˆaˆtoˆruˆtoˆru ‘benumbed’
(iii) frequentative, eg.
*paˆtaˆi ‘ask’ paˆaˆtaˆiˆtaˆi ‘to ask frequently’
(iv) other, eg.
*whaˆngaˆi ‘feed’ whaˆaˆngaˆiˆngaˆi ‘food to send visitors on their way’

(The small number of forms for Pattern 4 reflects its productivity: many such forms are
not lexicalized, and thus not in the Dictionary. However, this implies productivity with a
regular semantic outcome.)
Thus an approach to reduplication through study of those reduplications recorded in
dictionaries does not lead to any useful generalizations.
The only approach likely to shed any light on these forms is their use in the field of
coinages. A study of the reduplicated forms in the Maori Language Commission lists
reveals the following: all reduplications in this material have one form: for two mora
bases, complete reduplication; for three mora bases, Harlow’s pattern 4 (the only pattern
where all morae are reduplicated, so presumably also complete reduplication).
Semantically, all seem to be frequentatives. However, there are two grammatical classes:
those where the output is principally verbal, and those where the output is principally
nominal (but note that all verbal forms are potentially nominal). Examples of both types
are:
Nominal outcome:
meke ‘punch’ mekemeke ‘boxing’
hiki ‘lift’ hikihiki ‘weightlifting’
kawe ‘carry’ kawekawe ‘freight and cartage’
ata ‘reflection’ ataata ‘video’
wehe ‘share’ wehewehe ‘division’
*tatai ‘calculate’ taataitai ‘calculator’

Verbal outcome:
*pato ‘knock’ patopato ‘to type’
 Morphology 519

The other examples with verbal outcomes are not new formations, but are listed in
Williams, so this pattern is apparently less productive. Thus it would appear that the only
fully productive pattern of reduplication in modern Maori is the use of complete
reduplication to form a related noun. However, this may just be the result of using word
lists: I believe that the use of complete reduplication to form a verb with frequentative
sense is equally productive.
 3
Phonology

3.1 SEGMENTAL PHONOLOGICAL UNITS

The segmental descriptions below are based on two sources: specially elicited data
including minimal pairs and other crucial items from two consultants (a female Te
Aupouri speaker and a male Ngapuhi speaker), and the four cassettes Nga Ingoa o
Aotearoa recorded by Hugh Young (henceforth abbreviated NGIOA). The latter is a
recording of Maori place-names from the North Island spoken by Maori speakers from
each local area. They were collected to serve as a model for broadcasters. The speakers
were chosen by the local Maori communities. The majority of the speakers on the tapes
are male. There are over 100 different speakers. Unfortunately, no information was
collected on the age of the speakers, nor on their linguistic background. However, it is
clear that none of them are young speakers, and many are amongst the older members of
their community.

3.1.1 Distinctive segments

Ten consonant phonemes can be distinguished:
p t k
m n ŋ
f h
r
w

However, some of these segments have variable realizations across dialects, see 3.1.2.
Minimal pairs are found in 3.2.6.1.
There are five distinct positions for vowels, phonemically
i u
e o
a

It is often claimed that Maori also has five long vowels, corresponding to the five short
vowels, ie. a ten vowel system. However, if these long vowels are analysed as sequences
of two short vowels, as Biggs has proposed (see eg. Biggs, 1966, 11) then vowel length is
not phonemic. The appropriate analysis is still at issue, and is discussed further in 3.1.2.2.
 Phonology 521

Minimal pairs between long vowels are found in 3.2.6.1. Minimal pairs with short vowels
are:
taki ‘recite’
teki ‘scrape’
tiki ‘fetch’
toki ‘axe’
tuki ‘beat’

Some minimal pairs between short and long vowels are discussed in 3.1.2.2.1.

3.1.2.1 Non-syllabics

3.1.2.1.1 Plosiυes and affricates

There are three segments in this category. The basic articulations are
/p/ a voiceless labio-labial plosive
/t/ a voiceless lamino-prealveolar or apico-dental+laminoprealveolar plosive
/k/ a voiceless dorso-velar plosive

The degree of aspiration of these stops is extremely variable. Biggs reports (1961, 9)
slight aspiration for all, with aspiration increasing with loudness and stress, and more
aspiration before front vowels than back. On the basis of the data I have used, the first
and last of these observations require qualification. While the oldest-sounding speakers
from rural areas still have very little aspiration as the norm, many speakers on the NGIOA
recordings have very variable aspiration, and several have as much as in English.
Different speakers appear to have one of the stops more aspirated than the others, but
whether it will be /p/, /t/ or /k/ is not predictable. For almost none of these speakers is
there a clear link between degree of aspiration and frontness of the following vowel.
Indeed in some cases, the reverse of what Biggs found appears to hold: for some, back
vowels are associated with more aspiration than front vowels. I have no doubt that the
variable aspiration is a result of English contact. Early borrowings into English such as
piripiri > biddybid(dy) ‘a sticky plant burr’, Otaakou > Otago (a locality) seem to
indicate fairly clearly a lack of aspiration in earlier Maori. The difference between
Biggs’s observations 30 years ago and mine can be attributed at least in part to the years
that separate our observations.
At the onset of a stressed syllable, affrication is possible, rather than aspiration. With
/p/ and /t/, this is commonest in the environment of a high front vowel, whereas with /k/,
affrication appears commonest before /a/. With /t/, affrication is also heard before a
devoiced final /i/ or /u/, which is unstressed. (Biggs (1969, 9) comments that /t/ may be
palatalized in this environment) Thus the following pronunciations can be heard:
piu ‘swing’ [pç(i)u]
iti ‘small’ [itsi] [itçi]
 Maori 522

karanga ‘call’

/t/ also shows English influence in its place of articulation in the speech of a few speakers
on the tapes (in all cases, there were other features as well which suggest semi-speaker
status). Hohepa (1967, 5) claims that /t/ is “interdental through alveolar”. I did not
identify any interdental articulations of /t/ in the data I used.
/k/ shows a good deal of variation in articulation. It is usually advanced velar before
front vowels, and retracted velar before back vowels. In isolated tokens in Western
dialect speakers throughout the North Island, and with greater regularity in the Bay of
Islands and particularly Whangaroa areas of Northland, /k/ is realized as a velar fricative
primarily but not exclusively before back vowels, and in a few cases as a palatal fricative
before front vowels. Thus the following pronunciations of place-names can be heard:
Kaaea Kerikeri

Hopekako Rimariki

Such fricative pronunciations were generally intervocalic in areas other than Northland.
The low back vowels are much more likely to be accompanied by a fricative /k/ than is
/u/.

3.1.2.1.2 Fricatives
Maori has two consonants which are fricative in their basic realizations, /f/ and /h/. /f/ is
represented orthographically by wh, and as its phonetic realization is extremely variable,
the orthographic form is often used in the later sections of Chapter 3 as a cover for all
these realizations.
/h/ is a fricative produced at the point of articulation of the following vowel sound. It
thus ranges from palatal to retracted velar. It sometimes has a zero realization, and
between identical vowels may be realized as little more than the re-articulation of the
vowel, ie. with virtually no audible friction. Speakers from the Taranaki-Whanganui area
consistently realize orthographic h as [?] in all environments. Thus the place-name
Hikurangi is either depending on area.
The realization of orthographic wh varies sometimes by dialect, sometimes by speaker,
and sometimes varies with a single speaker from one token to another. The most
widespread realization in the North Island is [f] (hence the choice of phonemic symbol),
ie. a labiodental fricative. However, it has a lenis articulation in comparison with English.
It has variable lip-rounding. Other articulations are: voiceless labio-labial-velar fricative
[M], which seems to be in free variation with [f] for quite a number of speakers; [?w]
which is the realization in all environments for the Taranaki-Whanganui region; strongly
labialized [h] probably with back of tongue raising (cf. Harlow, 1979, 126); and [hw] or
[wh] (the latter usually intervocalic before nonrounded vowels) in the far north of
Northland. Whether there is really a difference between Harlow’s "[h] with secondary
rounding (and possibly high-back tongue position)" and Biggs’s (1961, 9) "voiceless and
strongly aspirated w" is an interesting phonetic problem which I do not feel competent to
 Phonology 523

resolve. However, it is clear that in some of the realizations from the Te Aupouri area,
there was an audible period of aspiration without lip-rounding following non-rounded
vowels, but following rounded vowels, only aspiration with lip-rounding was heard, ie.
awhi ’embrace’ was realized as while owha ’turn over’ was realized as
.
The question of the pre-European realization of wh is a vexed one. Many believe that
the realization as [f] is a post-European contact phenomenon. The fact that the
orthography has wh suggests that the informants who assisted Prof. Lee in devising the
orthography did not have [f], but it cannot be excluded that other dialects had [f] even at
that time, given the occurrence of [f] in cognates in other Polynesian languages. What
does seem fairly certain is that the spread of [f] to become the most widespread
realization is due in part to English influence, and largely at the expense of [M].

3.1.2.1.3 Nasals
Maori has three nasals:

/m/ is a voiced labio-labial nasal

/n/ is a voiced lamino-alveolar nasal
/ŋ/is a voiced dorso-velar nasal.

Biggs (1961, 9) describes /n/ as alveo-palatal, but in the data I used, there does not appear
to be any evidence for this more retracted position. LIke many other features of Maori
pronunciation, this may well have changed under the influence of English. Indeed some
second-language speakers can be heard producing dental or prealveolar /n/ as a hyper-
correction. One speaker on the NGIOA tapes produced a few palatal variants following
/i/.
/ŋ/ is retracted velar before low back vowels (Biggs, (1961,9) says ’back vowels’,
which may or may not include /u/, given its normal realizations), velar before /u/, and
advanced velar before front vowels. Biggs notes a palatalized variant before /e/ but this
did not occur in my data (there were very few tokens of /ŋe/ in the data). It should
perhaps be noted that three different speakers on the tapes NGIOA each produced one
token of [ŋg] for ng (all from the Auckland-North Auckland region) which points to the
pervasive influence of English.
In some areas in the Bay of Plenty, ng is realized as [n], although none of those who
recorded the place names for NGIOA showed this merger. Whether place names are
exempt from the merger, or whether the merger is regarded as informal, or whether these
were considered pronunciations for local people only, and thus not suitable as models for
broadcasters are tantalizing questions which I am unable to answer.

3.1.2.1.4 Liquids
Maori has only one liquid, a voiced lamino- or apico-alveolar tap. There is a considerable
area of contact between the sides of the tongue and the gums in the articulation of this
sound, which is the realization of orthographic r. Both Biggs (1961, 9) and Hohepa
(1967, 5) comment that it may be a short trill, but none of the tokens in my data involved
 Maori 524

more than a single contact. Intervocalically, the contact time is usually very brief.
Initially, it is sometimes a little longer. It may at times be followed by slight affrication,
especially before fronted realizations of/u/. Biggs notes that some speakers in a restricted
area of the East Coast of the North Island vary freely between a central and lateral
release, but this does not now seem to be as locally restricted as he suggests, since it can
also be heard from a variety of speakers from other areas of the country.

3.1.2.1.5 Glides/Semiυowels
Maori has one distinctive semi-vowel, namely /w/. This is a voiced labio-labial semi-
vowel with back of tongue raising and moderate lip-rounding. Biggs (1961,10) and
Hohepa (1967, 5) describe it respectively as having “a little friction” and “little friction”.
In most instances in my data, there is no audible friction: there is little difference except
in onset from a back realization of /u/.
It should perhaps also be noted that in vowel sequences in rapid speech, /u/ or /o/ can
be realized as a [w] glide, especially before /e/, and /i/ as a [j] glide. This is particularly
common following /p/, where the following were noted, for instance:
pio ‘extinguished’ [pjo] or [pço]
piu ‘swing’ or [pçu]

3.1.2.2 Syllabics
All the narrow transcriptions below use the notation of the International Phonetic
Alphabet revised to 1989.

3.1.2.2.1 Vowels
All the distinctive vowels of Maori are plain. However, their analysis presents an
interesting phonological problem. The traditional orthography uses five vowels, a, e, i, o,
u, but there are pairs of words not distinguished in the traditional orthography which are
minimal pairs in the spoken language, eg.
anuhe ‘caterpillar’

anuhe ‘sickly’

wheke ‘creak’

wheke ‘octopus’

One approach to the analysis is to establish two series of vowels for Maori, one long and
one short, thus analysing Maori as having 10 distinctive vowel phonemes:
 Phonology 525

This analysis is reflected in the orthographic practice of marking long vowels with
macrons. The fact that in some cases (most notably /a/ vs. /a /) the quality of the long
vowel is not identical to the quality of the short vowel is easily accommodated in this
analysis.
The second approach to the analysis of the Maori vowels takes as its starting point the
fact that all possible pairs of non-identical short vowels occur in Maori (see 3.2.3.3), and
so do all five pairs of identical short vowels, at least across morpheme boundaries, as in
the following forms: whakaara ‘raise’, haereere ‘stroll about’, iriiri ‘baptise’, orooro
‘rub back and forth’, uruuru ‘blade of weapon’. Forms like these have a variety of
pronunciations which depend on matters like formality, speed, emphasis, so that all of the
following can be heard:

and similarly for the other forms. The syllable break of the second pronunciation may be
no more than a new input of energy. The last of these pronunciations (the norm in casual
fast speech) is the same as the realization of the long vowel, even for those vowels where
the short and long vowels differ in quality. In the face of this sort of evidence, Biggs has
analysed (1961 and subsequent work) all long vowels in Maori as clusters of identical
short vowels. He has advocated ‘double vowel’ orthography to reflect this. On this
analysis, the minimal pairs above are phonologically

and thus Maori has only five distinctive vowels, all short, namely

(It should perhaps be noted in passing that each of the orthographies has disadvantages in
certain areas: when sequences of long+short identical vowel arise, as in whakaāhua
‘portrait’ from ‘form’, the traditional macron orthography retains the
etymological information, but does not represent the phonetic reality, (see
also 3.2.3.3); the double vowel orthography, whakaahua better represents the phonetic
reality, but loses the etymological information. The Maori Language Commission, which
advocates macrons, has, however, just decreed that forms like should be
written whakaahua (He Muka 5(2), 7). In general, when there is a clear morphological
boundary, the macron orthography writes vowel sequences, as whakaaro ‘decide’. Thus
macrons are used for long vowels only when there are no perceived morpheme
boundaries, but that is not consistently followed, as is shown by the passive of hanga
‘build’: hanga+-a ‘build+passive’ is written in macron orthography.)
There are considerable problems in deciding between these two analyses of the Maori
vowel system. At least the following points are relevant considerations:
(i) Forms with variable vowel length:
 Maori 526

There are a number of particles which vary between having a short or a long vowel by
context, among them ka ‘T/A’ and te ‘the’. Ka is [ka ] if the rest of the phrase contains
two morae; otherwise it is (see Biggs, 1969, 28). Te, on the other hand, is often
pronounced long in the compound determiners, eg. teenei ‘this’ [tε nεi] but te whare nei
‘the house here’
(ii) Native speaker variation:
There are many instances where native speakers are uncertain about vowel length, and
others where both short and long forms are regularly heard, eg. pohutukawa ‘a native
tree’: and hangi ‘earth oven’: [ha ŋi] and [ha
ŋi ]. On NGIOA, there are quite a number of instances where forms marked with long
vowels in the (macron) orthography of the accompanying text, presumably on
etymological grounds, are nevertheless pronounced short. (I did not note any instances of
the reverse, ie. short vowels pronounced long.)
(iii) Phonotactics:
All combinations of two non-identical short vowels occur in monomorphemic forms;
combinations of two identical vowels occur across morpheme boundaries, and thus are
part of the phonology of Maori whatever the analysis of the vowel system. Phonetically,
longer strings of identical vowels do not occur, although, phonologically, strings of three
can occur (or long V1+short V1 occurs). Comparable length strings of non-identical
vowels occur: aae ‘yes’, kaainga ‘home’, moounu ‘bait’. However, the combinatorial
possibilities for long vowels present a somewhat different picture. There are
combinations of non-identical long vowels (never in the same syllable, as far as I can
determine), such as kaaii ‘type of tree’, keeaa ‘false’, tootooaa ‘disrespectful’, including
some which arise through prefixation, such as pooaanini ‘giddy’, tuuaaoma ‘gait, pace of
travel’. However, there are fewer than twenty such forms in Williams’s Dictionary if
compounds and derivatives are excluded. All of these pairs of long vowels are in separate
syllables, unlike comparable length strings of non-identical vowels, such as raaua ‘IIIdl’
which in rapid casual speech is frequently one syllable (although in slower or more
formal speech it has two syllables). In contrast, I can find no examples of two adjacent
identical long vowels within a word, even across morpheme boundaries. Note that there
are no restrictions on long vowels in adjacent syllables with consonantal onset:
pooheehee ‘think mistakenly’, aapoopoo ‘tomorrow’, ngaawhaa ‘hot pool’, paateetee
‘creak’ are just a few of many examples, although whether any are monomorphemic is an
interesting question not readily resolved. Thus the clustering properties of long and short
vowels are different, and are more readily accounted for by the ‘double vowel’ analysis,
which claims that phonetically no more than two of the same vowel can cooccur.
(iv) Distributional patterns:
In terms of frequency of occurrence in text, four of the long vowels (/i e o u /) are
very rare. In a phoneme frequency count of 5000 words of running text (1000 words from
each of five different texts), the five rarest phonemes on raw scores were /i / /u / /e / /w/
and /o /. (Note that this presumes the 10-vowel analysis; and also that the texts counted
were written in macron orthography, so that across morpheme boundaries, two short
vowels occurred, although phonetically those two vowels might optionally be realized as
a long vowel.) In comparison, the five short vowels were all in the seven most frequent
 Phonology 527

phonemes in this count—the top order on raw scores was /a i t e k o u/ in descending

order.
Two words accounted for almost all the textual occurrences of /i /: ‘say’ and tīmata
‘start’. One accounted for more than half the occurrences of /u /: ‘stand’. Almost all
the occurrences of /e / were in compound determiners such as tēnei ‘this’ , tētahi ‘a(sp)’.
A very considerable proportion of /o / tokens were in possessive forms, eg. m ‘intgen’,
koutou ‘plgenIIpl’. Note that bases in Maori must have at least two morae, so that *ki
and *tu are impossible content words. It could thus be claimed that, in an extremely high
proportion of cases, the occurrence of these four long vowels is predictable.
Some support for the analysis of these long vowels as sequences of two short vowels
can be derived from a comparison of the frequencies of these long vowels and pairs of
non-identical vowels. Two approaches to this were taken. Firstly, a frequency count of 2-
vowel sequences was made, and the frequencies of the four long vowels in question (but
not /a /) fell within the range of frequencies found for 2-vowel sequences. Secondly, a
statistical calculation was made to see whether the statistical probability of occurrence for
each of these ViVi sequences given the frequency of Vi was of the same order as the
actual frequency of occurrence of the long vowels. In the case of the four in question (but
again not for /a /) the frequency figures were such as to suggest that this is a statistically
probable analysis.
The case of /a / seems to me interesting. Even given the very high frequency of short
/a/, /a / has too high a frequency to be likely accounted for as /aa/. It has a far higher
frequency than any other twovowel sequence. Also, the difference between its quality
(see below) and the quality of short /a/ is greater than the difference between any other
short-long pair. There may well be a case to be made for the claim that Maori has a 6-
vowel system, /a e i o u a /. I hope to develop this argument elsewhere at a later date.
Thus the issue of the proper analysis of the Maori vowels cannot be resolved here. It
will have to await further examination of the phonetic and distributional evidence. In
what follows, I will refer to long vowels where necessary, but leave open the question of
whether they are phonemic, or realizations of two adjacent identical short vowels, or
some combination of these positions.
Bearing this in mind, a consideration of the phonetic realizations of the Maori vowels
can now be made. Hohepa (1967, 5) articulates a commonly held belief about the vowels
of Maori when he says “Apart from slight centering towards mid-central position in
interconsonantal unstressed position, each vowel phoneme has a narrow allophonic
range”. He is at some pains to explain this (1967, 6): “The lack of appreciable allophonic
range for each vowel may be due to the high frequency of vowel clustering”, suggesting
that under these circumstances, Maori may “minimize ambiguity and maximize contrasts
with the aid of a tight allophonic range for each vowel phoneme”. These comments do
not accord well with the data I have analysed. Extremely careful, highly formal speech
from older speakers perhaps approximates to this description, but casual speech certainly
does not, and neither does all formal speech. The tapes NGIOA must be taken to represent
a fairly formal register: speakers were recording a list of names as models of
pronunciation. Of the 100 speakers on those tapes, all showed considerable variation in
each of the vowels without any apparent conditioning factors, and between speakers, the
variations in quality were considerable. Nor were these differences attributable to major
dialect differences, although age and sex undoubtedly play a part in the variations
 Maori 528

observed. Biggs (1961, 10) comments that “/e/… ranges higher than the lower allophones
of /i/, in unstressed position...”, and there is considerable overlap in my data between
realizations of /a/ and /o/. In general terms, the following seem to me the critical
parameters for the auditory distinctions between the vowels: /i/ is always more spread
than /e/; /u/ is always rounded, in contrast to the spreading of /i/, though /u/ can have
realizations which are as front as the most retracted /i/ tokens. There does not appear to
be any overlap between /o/ and /u/. There is considerable overlap between /o/ and /a/ in
terms of tongue position, but rounding is crucial for their distinction. In consonantal
environments rounded /a/ only occurs in environments from which /o/ is excluded (ie.
following /w/), and rounding of /a/ is optional there. However, the diphthongs /ou/ and
/au/ are extremely dose for some speakers, because both these environments involve
rounding. There does not appear to be overlap between /e/and /a/, since /e/ is higher and
fronter than the most advanced realizations of /a/.

Realizations of /i/
/i / is realized most commonly as ie. not as front or as spread as cardinal 1. It does not
show a great deal of variation, either in relation to consonants or adjacent vowels.
/i/ shows considerable variation. Word initially (and for many speakers, morpheme
initially) and word finally some speakers have a quality only slightly more centralized
than the quality for /i /. However, interconsonantally, and in vowel clusters, most

realizations for almost all speakers are lower, and less spread, ie [I] or Thus the

following pronunciations are characteristic: whakairi ‘to hang up’ whakairo

‘carve’ Interconsonantally, the highest variants of /i/ occur in the
environment of stops, and the lowest in the environment of /f/ and /r/. Some younger
speakers produce qualities of the variety, as is common in NZ English. Vowel
clusters lead to considerable variation in this short vowel. The following represent
common realizations in combination with other vowels:

In /ai/, /i/ is realized as or (For some speakers, this is the

environment where /i/ comes closest to cardinal 1)
In /ei/, /i/ is realized as [i] (For speakers with high realizations of /e/ in
this environment, the diphthong shows little movement)
In /oi/, /i/ is realized as [I] or
In /ui/, /i/ is realized as which given the usual realization of /u/ as
means that the main movement in this cluster is lip movement
In /ia/, realizations of /i/ are all centralized, but vary considerably from

about to

In /ie/, /i/ is realized as [I]

In /io/ and /iu/, /i/ is realized as [I] or [j]
 Phonology 529

Following long /a /, many speakers realize /i/ as as in kaainga

‘home’
In the middle of three-vowel clusters, in general there is a gesture
towards the ‘canonical’ quality, but most realizations are very centralized;
in /uia/, the /i/ may be a glide.

Realizations of /e/
In my data, /e / is cardinal 3 in quality, ie [ε ]. However, (thanks to Chris Lane for
drawing this to my attention) higher realizations (approaching cardinal 2) can be heard in
forms like teenei ‘this’, peenei like this’. It shows very little variation with preceding
consonants, and is rare in combination with other vowels. There is considerable variation
in the height of short /e/ from speaker to speaker, although individual speakers are
relatively constant with interconsonantal and final tokens. However, /e/ shows
considerable allophony in vowel clusters, and the /e/ of the article te ‘the’ is usually
considerably centralized: Those with least variation are the older males, who
produce consistent cardinal 3 realizations. The highest values come with greatest
frequency from younger female speakers, some of whom are fairly consistently close to
cardinal 2, which may be an influence from NZ English, which has very high realizations
of /e/. Many speakers show considerable fluctuation between cardinal 3 and cardinal 2.
The following represent common realizations in combination with other vowels:

In /ae/ /e/is higher and retracted:

In /ie/ /e/ is raised: however, if this diphthong is unstressed, it is

centralized for many but not all speakers: compare hie ‘shout’ and
rangimarie ‘quiet, peaceful’ The non-centralized
pronunciation is commoner among semi-speakers
In /oe/ /e/ is normally cardinal 3:
In /ue/ /e/ is raised almost to cardinal 2:
In /ei/ and /eu/, /e/ is considerably raised, again almost to cardinal 2:

hei ‘wear round neck’ heu ‘separate’

In /ea/ and /eo/, /e/ is most usually near cardinal 3

Realizations of /a/
The qualities of short /a/ and long /a / in general differ considerably, although there are
areas where they overlap. Short /a/ is treated first
The canonical realization of short /a/ is a slightly retracted low central unrounded

vowel, ie . Following /w/ for many speakers, it retracts and rounds to ,

especially in the environment of another velar, eg. in waka ‘canoe’, ngawe ‘yelp’, the
bold a has this realization for almost all speakers, but in awa ‘river’, only the second a
 Maori 530

has this realization for the majority of speakers. For some speakers wh also produces
rounding, but this depends to some extent on the realization of wh. Thus whare ‘house’

can be heard realized as or However, some speakers produce the

rounded realization after [f]. For many speakers, velars lead to greater retraction,
sometimes in preceding /a/, and sometimes in following /a/ as well. Quite a number of

speakers show fronting following /r/, eg. rangi ‘sky’ In combination with other
vowels, there is also considerable allophonic variation, although it varies from speaker to
speaker, and probably also with formality. Some common realizations are given below.

In /ai/, /a/ may be considerably raised, to or even

In /ae/, the basic quality is usual:

In /ao/, a retracted realization is normal:

In /au/, there is considerable variation. Older speakers, especially
males, have variants close to the canonical quality, eg. . However, it

is increasingly common to hear or for this diphthong. It

appears to be change which is dependent on a change in /ou/ (see below),
and to be led by female speakers. As females constitute a significant
proportion of the teachers of Maori, it is being taught to L2 speakers, and
is commonly used by both male and female L2 speakers. It appears to
have some prestige, at least in communities where semi-speakers abound.
Relatively few of the oldest-sounding male speakers on NGIOA had this
raised and rounded variant, but many—perhaps most—fluent younger
speakers use it.
In /ia/ /a/is considerably raised:
In /ea/ and /oa/, the canonical quality occurs: and
In /ua/ some speakers centralize, but this varies with the canonical
quality even within one speaker: or
Long /a / is characteristically further back and lower than short

/a/: . In the environment of velars, it can be retracted virtually to cardinal 5 quality.

Word finally, it is also common to hear retracted varieties. In those environments where
short /a/ is rounded and retracted, and long /a / rounded, there is considerable overlap in
quality, eg. Wawaa ‘locality’ [WDWD ]. Many speakers have fronter varieties,

especially following /n/ and /r/, and these may be as front as short /a/, ie. . There is an
increasing tendency to front as far as NZ English /a/, ie [a]. The article ngaa ‘the(pl)’ in
context is usually pronounced with the retracted quality of long /a/ in careful speech, but
seldom has the length the macron/double vowel suggests, ie [ŋa ] is a very formal
pronunciation, and [ŋa] is normal. In casual rapid speech it is often centralized to .
 Phonology 531

Long /a / is the only long vowel which occurs with sufficient regularity in vowel strings
for any generalizations to be worthwhile. In such strings, long /a/ has its most retracted

qualities before /o/, eg. raaoa ‘be choked’ . It is considerably fronted before /e/

and /u/, eg. aae ‘yes’ raaua ‘IIIdl’ . Its quality does not seem to be greatly
affected by preceding vowels,

Realizations of /o/

The basic realization of long /o / is a slightly raised cardinal 6, ie. with consequent
slight increase in lip-rounding. For some speakers, short /o/ has the same quality, but for
others it is a much more variable vowel, having near cardinal 6 quality in initial position
and following velars and sometimes /h/, but with considerable fronting or centralization

after other consonants, ie or or even Many speakers show considerable

variation in short /o/ without any evident conditioning. In combination with other vowels,
/o/ shows considerable variation.

In /ao/, realizations are often centralized, with noticeably less rounding

than the canonical realization, ie

In /eo/, realizations are close to cardinal 6, ie

In /io/, there is a great deal of variation, from the canonical quality, to

almost cardinal 7, and sometimes centralized varieties, eg. or

or
In /uo/,the quality of /o/ is usually about cardinal 6, but /u/ tends to be

much further back than is usual, ie

In /oi/and /oe/, /o/ is usually somewhat centralized, eg
In /oa/ /o/ is often centralized, ie
In /ou/, /o/ is very central, and unrounded or even spread, ie or
slightly retracted and/or lowered varieties of

Long /o / also shows some variation in combination with a following high vowel, ie in
/o i/ and /o u/, where it approaches cardinal 7 in quality.

Realizations of /u/
This is probably the most variable vowel in quality in Maori, with realizations ranging
almost from cardinal 8 (although these very retracted realizations are heard only from the
oldest male speakers, and even there are few and far between) to extremely front rounded
qualities, and there is also considerable variety in tongue height Some of this variation is
due to the influence of NZ English, where front qualities are the norm except before /l/. I
 Maori 532

have heard older speakers making fun of semi-speakers with extremely front rounded
qualities, but they are nevertheless not confined to semi-speakers nowadays. Long /u / is
normally with fronter variants after /t n r/, and for some speakers, slightly retracted
qualities after /ŋ/. Short /u/ is so variable that it is difficult to choose any one quality as
the norm. The most retracted variants tend to be word-initial, in the vicinity of but if
a /t/ follows, much fronter varieties occur. After /t / or / r/ the quality is often in the

vicinity of . Short /u/ also shows considerable variation in combination with other
vowels:

In /au/,it is often very front, eg. or although this depends to

some extent on the realization of /a/

In /eu/, the normal realization is

In /iu/, realizations are very variable, ranging from to [Y]
In /ou/, /u/ is usually but realizations of /o/ vary, and cause some
variation in /u/
Before other vowels, much more retracted qualities are normal: with
occurring in the combinations /ua ue ue/, and /uo/ showing

realizations ranging from to

In longer strings, /u/ also shows a great deal of variation from speaker to speaker, so that
it is not clear that valid generalizations can be made. However, the general tendency
seems to be for /u/ to become increasingly front, especially in the speech of younger
females, where it is particularly noticeable in the environment of /t/ and /r/.

3.1.2.2.2 Others
No other sounds function as syllabics in Maori.

3.1.2.3 Sounds in loanwords

The combinations wo, wu, who, whu occur only in English loanwords, and were clearly
felt by one of my consultants to be awkward. However, none of the individual sounds
specified as the Maori phonemic inventory above occur only in loanwords. Many totally
unassimilated English words can be heard in Maori discourse today, and it is probably
fair to say the the entire phoneme inventory of NZ English can occur in Maori in what are
clearly loanwords for the speaker. The use of unassimilated loanwords is frowned upon
by some, including some who frequently use them.
 Phonology 533

3.1.2.4 Restrictions in word classes

There are some interesting gaps in the sounds occurring in certain grammatical word
classes, but I suspect that they are accidental rather than systemic. The most obvious ones
are nevertheless noted here. No verb particles contain /o/, and the only consonants used in
these particles are /k t m n/. Because the determiners include possessive forms, they
contain a wider range of sounds, but /p/ and /w/ do not occur. No preposition contains /u/,
or any of the following consonants : /p t ŋ f w r h/. Passive terminations do not use /p/ or
/w/ (see 2.1.3.1.1). The absence of /p/ and /w/ from all these categories may indeed be a
systemic pattern, and may be linked to the restrictions on labials discussed in various
sections in 3.2 and 3.4 below.

3.2 PHONOTACTICS

3.2.1 Positions for consonants

3.2.1.1 Wordfinal consonants

Maori does not admit wordfinal consonants. Wordfinal vowels are added to forms
adapted from English, eg.

However, it must be noted that it is increasingly common to hear phonologically

unadapted English words in otherwise Maori contexts. With lists of words in isolation, as
in the data for segmental analysis from my consultants and in the lists of placenames on
the tapes NGIOA, it was noticeable that many forms ended with [h] or [?], the latter often
with audible release (cf. Interjections 4.2).

3.2.1.2 Initial consonants

All consonant segments of Maori can occur word-initially. The canonical syllable of
Maori begins with a consonant.

3.2.2 Consonant clusters

Maori does not admit consonant clusters in any position. Consonant clusters in words
adapted from English are either broken up by the insertion of an epenthetic vowel, or are
subject to cluster simplification, eg.
kiriimi < Eng. cream teihana < Eng. station
parau < Eng. plough taake < Eng. tax
 Maori 534

3.2.3 Positions for vowels

3.2.3.1 Wordfinal vowels

Every Maori word ends in a vowel (but note the comment in 3.2.1.1), and all vowels can
occur wordfinally.

3.2.3.2 Wordinitial vowels

It is possible for all vowels of Maori to occur word initially. However, there are relatively
few ‘content’ words beginning with vowels in comparison with those beginning with
consonants, as a quick glance through a dictionary of Maori will confirm. Nevertheless,
many particles begin with vowels, so the frequency difference in texts is not as great
There appear to be no ‘content’ words (as opposed to ‘grammatical’ words) beginning
with ee, except English borrowings like eekara ‘eagle’, and there are very few beginning
with ii.

3.2.3.3 Sequences of (syllabic) vowels

All possible sequences of two short vowels occur in Maori, both in monomorphemic
words and across morpheme boundaries. However, the question of syllabification is not a
simple one in Maori. In formal speech, and in, for instance, the recitation of poetry, each
short vowel clearly forms the nucleus of a rhythmic unit, and each seems therefore to
have a syllabic value. I have proposed that the label ‘mora’ should be used for such units
(Bauer, 1981b). However, there appears to be a second layer of syllabic structure, in
which certain combinations of vowels are grouped into larger units, for which I suggested
(1981b) that the term syllable should be retained. Within a single morpheme, all pairs of
identical short vowels form single syllables. Catford (1977, 215) distinguishes three
states for “geminate” vowel sequences: “…the two vowels may fuse into a single long or
ultra-long vowel, or they may be separated by a momentary diminution of initiator
power, or by the insertion between them of a glottal stop…”. All three realizations of the
Maori long vowels can be heard. The first is the normal casual speech realization; the
second is heard in contexts like the recitation of poetry; the third was produced by one of
my consultants on many occasions when I enquired whether a particular vowel was long
or short, and may thus be only a somewhat artificial realization.
Whether within morphemes all pairs of non-identical vowels form diphthongs with
one syllabic peak is a question not readily resolved. Clark and Yallop (1990, 107) say
that syllabicity is critical in distinguishing diphthongs from vowel sequences, and add
(108–9) that the term diphthong “is warranted only if the vowel genuinely counts as a
single vowel in the language in question”. They go on to point out that it may be a matter
of interpretation, and that “different criteria may point in different directions”. I believe
that this is the case with Maori, and that the issue is probably unresolvable. Thus the rules
for the use of e in imperatives and vocatives, and the reduplication rules treat such vowel
pairs as sequences, as do the rules for metrical regularity in certain types of waiata (song)
(see Bauer, 1981b and references therein). However, the rules for stress treat them as
single units.
 Phonology 535

Catford’s definition also relies crucially on syllabicity (1977, 215): “a diphthong may
be defined as a sequence of two perceptually different vowel sounds within one and the
same syllable”. But these clusters in Maori may all be treated either as one syllable, or as
two morae, and it is not clear that the notion of syllable appealed to in these definitions
excludes the mora.
Examples of all the possible two-vowel combinations in single morphemes follow:
haa ‘breath’ hea ‘where’ hia ‘how many’
hae ‘lacerate’ hee ‘wrong’ hie ‘shout’
hai ‘hey’ hei ‘wear on neck’ hii ‘catch (fish)’
hao ‘net’ peo ‘slip’ pio ‘extinguished’
hau ‘wind’ heu ‘separate’ piu ‘swing’
hoa ‘friend’ hua ‘fruit’
hoe ‘paddle’ hue ‘gourd’
hoi ‘deaf’ hui ‘meet’
moo ‘for’ tuota ‘a type of charm’
hou ‘bind’ puu ‘blow’

Catford (1977, 215–7) makes certain other distinctions amongst types of diphthongs
which are useful in characterizing these non-identical sequences in Maori. He
distinguishes sequential diphthongs from glides. Sequential diphthongs are characterized
by
steady state Vi+rapid change from Vi to Vk+steady state Vk while glides are
characterized by continuous change from Vi to VK. Maori diphthongs are clearly
sequential with the exceptions of /iu/, which is apparently always a glide, and /io/ and
/uo/ which are sometimes glides.
Catford also discusses the ‘falling/rising’ dichotomy, for which I shall use his
alternative terminology ‘crescendo’ vs. ‘decrescendo’. A crescendo diphthong has a
dominant second element, a decrescendo diphthong has a dominant first element. Only
/iu/, the glide, is characteristically crescendo. However, while in some cases the other
sequences are auditorily decrescendo, in many cases it is not clear that either element is
dominant, and a characterization like ‘sostenuto’ (following the musical analogy) appears
a much more realistic characterization. Thus I believe a three-way, rather than a two-way
distinction is needed here for Maori. Diphthongs with /e/ for their first element appear
normally to be decrescendo rather than sostenuto, but all the others show variation. Biggs
(1961, 13–14) divides the Maori diphthongs into three groups according to which
element is prominent: his characterizations do not entirely coincide with mine, although
he agrees that /iu/ is always crescendo, and all the /eV/ diphthongs are in his list of
decrescendo diphthongs. A number of factors may account for the other differences: we
are in agreement that it is often difficult to decide; there is considerable native speaker
variation; and there is possibly some lexical variation: hui ‘meet’ is characterized by
Biggs as either crescendo or decrescendo (in my data I have sostenuto and crescendo
tokens), but other /ui/ forms such as nui ‘big’ appear to exclude crescendo
 Maori 536

pronunciations; /oa/ in moana ‘sea’ is crescendo according to Biggs, who endeavours to

generalize the environment phonemically, but other examples show variation in my
consultants, and Moana as a proper name is not always pronounced in the same way as
moana used as a common noun.
Across morpheme boundaries, such combinations clearly have two syllabic peaks. A
brief glottal stop is sometimes inserted in such combinations, especially in emphatic or
slow speech, and semispeakers seem to include a glottal stop as the norm. Examples
illustrating the set of combinations across morpheme boundaries follow. Note that there
are two examples, marked ?, where the bimorphemic analysis is in doubt. Both involve e.
I was unable to find reduplicated forms in these instances due to the paucity of lexical
items with initial e.
whakaara ‘raise’ akeake ‘poor land’
whakaeke ‘attack’ eneene ‘coax’
whakairi ‘elevate’ ikeike ‘high, lofty’
whakaora ‘heal’ oneone ‘earth, soil’
whakautu ‘fondle’ ueue ‘shake’
akiaki ‘urge on’ akoako ‘split’
emiemi ‘be assembled’ ?koehu ‘turbid (water)’
iriiri ‘baptise’ inoino ‘tattoo marks’
oriori ‘sing a lullaby’ orooro ‘rub back and forth’
uriuri ‘dark (colour)’ utouto ‘use vindictively’
apuapu ‘crammed’
?puehu ‘dust’
inuinu ‘drink frequently’
okuoku ‘few’
uruuru ‘blade of weapon’

With the prefix whaka-, there are a few exceptions to the rule that the morpheme
boundary is marked by a syllable boundary, eg. whakairo ‘carve’ is normally pronounced
with three syllables, not the expected four, and the stress is accordingly in a different
place: whakairo not |whaka|iro, although the latter pronunciation is sometimes used in
careful speech. Across morpheme boundaries, these vowel sequences, even when there is
no intervening glottal stop, differ phonetically from the same sequence in a single morph.
The allophony described in detail for vowel sequences above (see 3.1.2.2.1) does not
occur, so that the first vowel tends to have the quality dictated by the preceding
consonant, and the second vowel has typically word-initial quality,compare
heu ‘separate’ [hey]
ureure ‘fruit of kiekie’
 Phonology 537

kau ‘swim’ [key]

whakautu ‘fondle’

Thus it is necessary to recognize that any sequence of two non-identical vowels may
represent a diphthong or may not, depending on whether there is an intervening
morpheme boundary. Since not all semi-speakers recognize the same morpheme
boundaries as older speakers, there are variant pronunciations in the Maori community,
and this is plainly an area where change occurs through time, as whakairo ‘carve’
indicates. (This is an interesting case, since the productivity and recognizability of
whaka- might be expected to counteract the non-transparency of iro in this form.)
Longer strings of vowels also occur, and some are in forms which are
monomorphemic. although many are polymorphemic. However, by no means all the
possible three-vowel strings actually occur, as far as I have been able to ascertain, but the
gaps do not seem to be entirely predictable. Strings of three identical short vowels do not
occur: where such arise morphologically, they are reduced to two-vowel length
phonetically, thus *whakaaahua is pronounced as written in double-vowel orthography,
whakaahua.
It is not possible to make any generalizations about which longer strings are permitted,
and which not Many arise from consonant loss, eg.
tiiaiaka < tiiwaiwaka ‘fantail’

(where these forms and several others are synchronically attested dialectal variants, see
3.4.1.3). I simply give here some more extended examples to indicate the sort of range of
forms found. Where it is clear that there is a morpheme boundary, I indicate this with a
decimal point, but it should not be taken that a lack of decimal point necessarily indicates
monomorphemic status historically or synchronically, nor that there are no other
morpheme boundaries than those marked. The left column contains four-vowel strings,
the right column has longer strings:
aaio ‘peaceful’ aaeaea ‘panting’
auee ‘alas’ hau·kooeo·eo ‘chilled’
hau·aa ‘crippled’ kaaea·ea ‘bush hawk’
hau·aitu ‘starved’ kaiua·ua ‘strenuous’
huuai ‘cockle’ kooau·au ‘flute’
kaaeo ‘mollusc’ maa·ui·ui ‘weary’
kaaii ‘kind of tree’ mooio·io ‘feeble
kaiao ‘living’ pae·aarau ‘shipwrecked’
koeae ‘kind of fish’ paa·ua·ua ‘strong’
koo·ua ‘drizzle’ piioi·oi ‘bird’
mauii ‘L.H.’
 Maori 538

peeia ‘kiekie’ (plant)

Problems regarding the syllabification of such forms are discussed in 3.2.5.1.

3.2.4 Comparison of morpheme structure and word structure

The structure of lexical morphemes differs from the structure of words only in length.
The canonical shape of a lexical morph is two morae, although there are quite a number
with three morae. Four morae is relatively rare, and often arises through forms which
were historically more than one morpheme. Words, on the other hand, are theoretically
open in length.

3.2.5 Syllabification

3.2.5.1 The syllabification of medial clusters

Since there are no medial consonant clusters in Maori, this is a question of the
syllabification of vowel clusters, and the issue is not a straightforward one. As has
already been mentioned (3.2.3.3) in certain styles, and in relation to certain rules (all, I
think, morphological), every (C)V unit of Maori (where V is short) is treated as a
separate unit, which I have called a mora (Bauer, 1981b). The division of forms into
morae does not seem to pose any problems. Every morpheme boundary will
automatically coincide with a mora boundary. A few examples of moraic divisions of
forms listed in 3.2.3.3 are given as illustrations:
hoˆa ‘friend’
heˆe ‘wrong’
whaˆkaˆaˆra ‘raise’
haˆuˆaˆiˆtu ‘starved’
koˆoˆaˆuˆaˆu ‘flute’

However, while moraic divisions seem to form the basis of Maori rhythm, and can be
heard in, for instance, the formal recitation of poetry, so that the rhythm is mora-timed,
Maori also has another level of prosodic structure in which vowels are grouped together
into larger units which I have called syllables (Bauer, 1981b). It is the rules for
syllabification which are complex.
One rule of syllabification is uncontrovertible: a consonant can only initiate a syllable,
so that there is always a syllable boundary immediately before a consonant. In addition,
the following principles apply: (i) if there is a perceived morpheme boundary, a syllable
boundary will normally coincide with it. (‘Perceived’, however, is critical here, cf. the
comments on whakairo ‘carve’ in 3.2.3.3.) However, there are exceptions to this. (ii)
Unless there is a morpheme boundary (and sometimes even when there is a morpheme
boundary), long vowels form single syllables. (iii) In general terms, no more than two
non-identical vowels occur in one syllable, but there are instances in rapid casual speech
 Phonology 539

where three non-identical vowels, or even four in the double vowel analysis, auditorily
form one syllable. (iv) Certain combinations of long V+short V also form a single
syllable.
Both Biggs and Hohepa have endeavoured to specify the rules that govern division
into syllables. The two accounts are not identical, and neither accounts for all the
observed data. Hohepa’s can be taken as a refinement of Biggs’s for practical purposes.
However, I give both accounts, point out certain differences and difficulties, and discuss
some of the actual data and how it relates to these rules.
Biggs (1961, 8) says “…a syllable boundary occurs at every plus juncture [which
“marks open transition between vowels” and “often occurs at a point of possible pause”
(1961, 12)] …and before every consonant. In sequences of vowels any two identical
vowels which are not separated by any juncture are in the same syllable. A syllable
division occurs after every second non-identical vowel”. He also lists the long V+short V
combinations in his data: aai, aao, oou (1961, 13).
Hohepa’s account (1967, 9) is as follows: “…syllable boundaries are established by
marking the following boundaries concurrently, and in the event of two boundaries
conflicting, the first in the sequence takes precedence. Boundaries occur at word space
(…), before each consonant, after identical vowels in close transition, before identical
vowels preceded by another vowel in close transition, and after every second vowel in a
non-identical close transition sequence”.
Both these accounts invoke the notion of close vs. open transition, but neither provides
a definition which is clearly independent of the syllabification issue: vowels are in close
transition if their quality influences each other, ie. if they form diphthongs, ie. if they are
in the same syllable, and in open transition if they have the quality of non-adjacent
vowels, ie. if they are in different syllables, or possibly, different morphs.
Examples of the following types are uncontroversial:
paa.ta.ka ‘raised storehouse for food’
poo.hee.hee ‘think mistakenly’
too.too.aa ‘disrespectful’

(although Biggs’ rules do not cover sequences of two long vowels, which are always in
separate syllables)
kau.ae ‘jaw’

(note that some dialects have kauwae)

ai.tu.aa ‘bad luck, accident’

(Biggs’ rules do not determine the syllabification of this one).

However, there are quite a number of problematic forms. Consider
kaainga ‘home’

Morphologically, this is kaa·inga, a nominalization of kaa ‘burn’. Biggs’ rules for the
syllabification of this are unclear, although he lists aai as a possible single-syllable
 Maori 540

combination. Hohepa’s rules predict that the syllable boundary should coincide with the
morphological boundary. However, the syllabification of my consultants was clearly
kaai.nga

(although at least one speaker from NGIOA says kaa.inga.) In contrast, raaihe
‘enclosure’, which as far as I can determine is monomorphemic, was syllabified
raa.i.he

by my consultants.
But it is not always the case that the syllabification proceeds against the morphological
boundaries. Compare
kau·ehu ‘turbid’ (cf. ehu ‘turbid’)

syllabified kau.ehu
where the morphological and phonological rules coincide, and
ma·ue ‘shake’ (cf. ue ‘shake’)

syllabified
ma.ue

thus coinciding with the morphological boundary, rather than the “after every second
non-identical vowel” rule, which would give
*mau.e

Incidentally, the quality of the /a/ here was more like long /a / than short/a/, and it may
be the case that vowel lengthening occurred to ensure better correlation between
phonological and morphological boundaries: maa.ue is rule-conforming on both criteria.
A similar treatment was accorded the following interesting pair by one consultant:
takai ‘wrap’ takai·a ‘wrap·pass’
taka ‘heap up’ taka·ia ‘heap up·pass’

Asked for the passive of these verbs, both my consultants gave

ta.kai.a

as the syllabification for both (somewhat to their surprise, although neither tried to
change the syllabification when they realized the coincidence in form). However, when
asked to read a sentence with each in context, one of the two produced
ta.kai.a ‘wrap·pass’

but
ta.kaa.ia ‘heap up·pass’,
 Phonology 541

again apparently modifying the phonetic form to produce coincidence between

phonological and morphological boundaries. The same treatment was accorded the
following forms by both consultants:
paoi ‘beater’ cf. oi ‘agitate’ pa·oi paa.oi
taoi ‘restless, eager’ ta·oi taa.oi

If underlyingly short vowels are lengthened in phonologically predictable environments

to preserve coincidence between phonological and morphological form, this may be
linked to the alternation between long and short vowels in certain particles (see 3.1.2.2.1).
In the following forms, morphological boundaries are, however, ignored:
haaereere ‘dup of haere’ haae.ree.re cf haa·ere·ere

and it is not uncommon to hear the /e/ of the first syllable dropped, giving
haa.ree.ree

Similarly,
Puungaereere ‘name of stream’ Puu.ngae.ree.re
cf. Puungaere·ere

(I have not been able to determine definitively the morphemic makeup of this place-
name, but all consultants agree that the last section is a reduplication.) The morphological
boundaries of reduplication seem to be more readily ignored than others (they are breaks
not corresponding to morphemes, of course).
The second type of problem form involves sets where there are strings of three non-
identical vowels. Some of the members of these sets clearly have bi-morphemic form,
and the syllabification reflects the morphemic structure. However, there may be one or
more forms with the same string of vowels where there is no evidence I can find for the
morphemic division, but where the syllabification follows that for the bi-morphemic
form, rather than the ‘after every second non-identical vowel’ principle. Some such sets
are listed here:
karioi ‘loiter, linger’ <oi ‘creep, crawl’ kari.oi
pioi ‘shake’ <oi ‘shake’ pi.oi
hookioi ‘mythical bird’ ? hoo.ki.oi
mariao ‘ulcer, pimple, star’ <ao ‘world, bud’ ma.ri.ao
niao ‘gunwale, rim, edge’ ? ni.ao
wheoi ‘quaking’ <oi ‘shake’ whe.oi
heoi (~ he oi) ‘however’ <oti ‘finished’ he.oi
peoi ‘a shrub’ ? pe.oi
 Maori 542

In the following set, I was unable to establish a clear case for a morphemic break in any,
but the patterns of syllabification are not predicted by the ‘after every second non-
identical vowel’ rule:
kaweau ‘tuatara, long tailed cuckoo ka.we.au
peau ‘be turned away’ pe.au
wheau ‘be long (of time)’ whe.au

The rule-conforming syllabification ea.u seems unnatural. In a number of these sets, the
first vowel is /e/, and it will be recalled that these diphthongs are the group that is most
consistently decrescendo. It is possible that syllabification tends to avoid such syllables,
but I have been unable to find sufficient evidence to be sure that this is the case.
Nevertheless, despite all this evidence that morphological boundaries take precedence,
in general, over phonological criteria in deciding syllabification, it must be remembered
that there are well established cases where the reverse is true, eg.
whakaae ‘agree’ < aae ‘yes’ whaka·aae

It is probably significant that of the long V+short V clusters Biggs lists as being possibly
monosyllabic, the long vowels involved are /a:/ and /o /. I do not think his lists are
complete: I think all combinations with /a / can be monosyllabic: kaai.nga ‘home’, aae
‘yes’, kaao ‘no’, raau.a ‘IIIdl’ are all attested. This might be another piece of evidence
for long /a / having a special status. In my data, none of the oou combinations was
uncontroversially monosyllabic, although kooura ‘crayfish’ and moounu ‘bait’ had
realizations about which I was uncertain, ie. I found it difficult to decide whether my
consultants had said koo.u.ra or koou.ra.

3.2.5.2 Canonical syllable structure

The canonical mora has the form (C)V. The canonical syllable probably has the form
(C)V(V), ie. it is not clear whether it is possible to determine any order of comparative
naturalness between the forms (C)V, (C)V1V1 and (C)V1V2. It is also difficult to
determine whether the canonical syllable or mora has an initial consonant. In content
words, initial syllables/morae with consonants are far more frequent than initial
syllables/morae with vowels, though the latter cannot be called rare. In non-initial
positions in words, the difference in frequency is smaller.

3.2.6.1 Restrictions between initial consonants and following vowels

The fact that there are no native vocabulary items with the combinations wo, who, wu,
whu has already been mentioned. All other combinations are attested, eg.
paa ‘touch’ taa ‘dash’ kaa ‘burn
pee ‘crushed’ tee ‘squeeze fluid from’ kee ‘different’
pii ‘chick’ tii ‘cordyline’ kii ‘say’
 Phonology 543

poo ‘night’ too ‘pull’ koo ‘sing (birds)’

puu ‘flute’ tuu ‘stand’ kuu ‘coo’
whaa ‘four’ haa ‘breath’ raa day’
whee ‘caterpillar’ hee ‘wrong’ ree ‘see!’
whii ‘can, be able’ hii ‘catch (fish)’ rii ‘screen’
whoounu ‘phone’ hoo ‘pout’ roo ‘go’
whurutu ‘fruit’ huu ‘resound’ ruu ‘shake’
maa ‘white’ naa ‘satisfied’ ngaa ‘satisfied’
meke ‘punch’ nee ‘eh?’ ngee ‘noise’
miiharo ‘wonder at’ niikau ‘palm’ ngiha ‘burn’
moohio ‘know’ nooku ‘mine’ ngoo ‘grunt’
muu ‘murmur’ nuku ‘earth’ nguu ‘squid’
waa ‘time’
wee ‘squeal’
wiiwii ‘rushes’
wooro ‘wall’
wuuru ‘wool’

Biggs (1961,13ff) and Hohepa (1967, 5ff) also discuss these combinatorial possibilities.

3.2.6.2 Restrictions between wordfinal units and preceding vowels

Not applicable.

3.2.6.3 Restrictions between consonants in adjacent syllables

There do not appear to be any such restrictions, although I failed to find adjacent
syllables with the following segments initially: n-ng, wh-w. These appear to be gaps
rather than systematic restrictions. However, there were relatively few forms with non-
identical nasals in adjacent syllables, and many combinations of labial consonants had
relatively few exponents. For instance, I failed to find w-m, w-p, w-wh in adjacent morae,
though they occur in adjacent syllables across morpheme boundaries in the forms
waimate ‘dread’, waipuu ‘red’, waiwhara ‘ornament of heron feathers’. Thus there may
be a tendency to consonant dissimilation in adjacent syllables, although this is clearly not
an absolute rule, as reduplicated forms regularly break it. It is possible that such
restrictions hold much more closely for adjacent morae in monomorphemic forms. With
labials, the only non-identical pairs under these narrower conditions appear to be m-p (eg.
mapu ‘pant’) and p-w (eg. pewa ‘bow-shaped’).
 Maori 544

3.2.6.4 Vowel harmony

Maori does not exhibit vowel harmony. I found forms with all possible combinations
(treating long vowels as separate phonemes) except the following: i-oo, u-ee, ee-ii, ii-uu.
Again these appear to be accidental gaps, due to the very low incidence of the long
vowels ii, ee, oo and uu in the vocabulary as a whole. Nevertheless, in the phonological
adaptation of words from English, the insertion of epenthetic and wordfinal vowels does
show a type of vowel harmony in some cases, in that the vowel inserted is often the same
as an adjacent vowel, eg. karaka < clock, tiriti < treaty, poroka < block, pereti < plate,
wuuru < wool. However, not all instances follow this rule, eg. kuki < cook, paramu <
plum (not *parama), pine < pin. Some, but not all, of the exceptions involve the use of i
next to h, as pointed out by Kearns (1990), eg. penihana < pension, pihikete < biscuit.
Thus vowel harmony may be a tendency, but is only one of a number of principles used
in arriving at a euphonious adaptation.

3.2.6.5 Consonant harmony

Maori does not have consonant harmony. If anything, it has a tendency to consonant
disharmony, see 3.2.6.3 and Harlow (1991).

3.2.6.6 Other restrictions on combinations

None have been found.

3.2.6.7 Phonotactic patterns and word classes

The particles of Maori allow different forms from lexical classes. No lexical stem can
consist of a single mora, but some particles have this form, eg. i, ki, me. The minimum
form for a lexical base is two morae, although such a base may consist of a long vowel
alone, eg. oo, ‘fit’, uu ‘breast’.

3.3 SUPRASEGMENTALS

3.3.1 Distinctive degrees of length

3.3.1.1 Length in vowels

Each vowel in Maori has two lengths, and the distinction is phonemic under one analysis
(see 3.1.2.2.1). Under the analysis which treats long vowels as a sequence of two short
vowels, sequences of three identical vowels can occur phonologically across morpheme
boundaries, eg. whaka·aahua ‘photograph’, puu·uruuru ‘abdomen’. However, these are
not phonetically longer than a long vowel.
 Phonology 545

3.3.1.2 Length in other syllabics

No other syllabics are found in Maori.

3.3.1.3–7 Length in consonants

Consonants in Maori do not have distinctive degrees of length, although in certain
emphatic types of pronunciation (see 3.3.2.2) some consonants are appreciably
lengthened.

3.3.2.1 The role of stress

Content words in Maori have a stress pattern which is predictable by rule if the
morphological make-up of the form is known. In addition stress plays an important
demarcative role in the phrase in Maori.

3.3.2.2 The phonetic correlates of stress

The most constant correlates of primary word stress in Maori are pitch fall and length of
vocalic element (the first element in a diphthong). These features are often accompanied
by emphatic onset, which consists of a slight pause preceding the stressed syllable, and
the appropriate one of the following: aspiration or affrication of stops; preglottalization of
vowels; stronger friction for fricatives; longer contact for /r/; closer approximation for
/w/; lengthening for nasals. These features of stress may also be accompanied by
increased loudness.

3.3.2.3 Levels of stress

Hohepa (1967, 10) writes of secondary, tertiary and quaternary stress, but I suspect these
are theoretical constructs rather than auditorily discriminable levels of stress. I have been
unable to distinguish more levels than primary, secondary and unstressed syllables. The
following form shows all three levels, with no marking before unstressed syllables:
|
|wha.ka. maa.ra.ma ‘make clear, illuminate’

The realization of primary stress is given in 3.3.2.2. Secondary stress appears to be

principally a matter of length, although the length is shorter than for primary stress.
The question of whether secondary stress occurs in monomorphemic forms in Maori is
an interesting one. Hohepa (1967, 10) marks forms as follows (using ˘ for his tertiary
stress, but otherwise replacing his marking by the standard IPA conventions):
‘ma.|na ‘power’
‘ma.˘na.˘wa ‘heart’

but in equivalent forms in my data, I do not hear any evidence to suggest other than
 Maori 546

‘ma.˘na ‘power’
‘ma.˘na.˘wa ‘heart’

The final vowel of words in isolation is frequently devoiced, and thus the final syllable of
these words may be significantly less loud than the others, but loudness is by no means
always a concomitant of stress in Maori, and devoicing is optional, not obligatory, so that
I do not think that this can be used as a criterion for distinguishing the second and third
syllables of manawa, and since devoicing can also occur on the final syllable of mana, to
which Hohepa assigns secondary stress, there seems little reason to suppose devoicing is
correlated with stress. Rather, I think devoicing is restricted to unstressed syllables.

3.3.2.4 Constancy of position of stress

Stress does not constantly fall on a particular syllable specifiable by word position in
Maori.

3.3.2.5 Predictability of word stress

The position of primary word stress is usually claimed to be predictable in Maori in terms
of the phonotactic and morphemic structure of the word, although this view is challenged
by Schütz (1985). The rules given here apply only to monomorphemic forms. Stress in
complex forms is dealt with in 3.5. Biggs (1969, 132) and Hohepa (1967, 10) outline
rules for word stress, although their accounts differ slightly, due (as I have pointed out
elsewhere, see Bauer 1981b) to a dialectal difference. The rules for primary stress invoke
a hierarchy of syllable types:
(C)ViVi > (C) ViVk > (C)V

ie. syllables with long vowels (or geminate clusters) take precedence over syllables with
diphthongs, which in turn take precedence over syllables with single/short vowels. The
basic stress rule is this:

In a monomorphemic form, the syllable type highest on the hierarchy

receives the primary stress, and if more than one syllable of the same
hierarchic status occurs, the first will be stressed.

However, in certain dialects, a word-final syllable of the form (C) ViVk is lower on the
hierarchy than (C)V. Biggs adds that stress cannot occur more than four syllables from
the end, but I have not found any clear cases of monomorphemic words exceeding this
length. Some examples showing the application of these rules are:
‘koo.whai ‘kowhai’
ku.‘rii ‘dog’
‘aa.whi.na ‘help’
‘hae.re ‘move’
 Phonology 547

whe.‘nua ~ ‘whe.nua ‘land’

‘ta.nga.ta ‘man’
‘kau.wae ‘jaw’
‘haa.ngii ‘earth oven’

3.3.2.6 Exceptions to word stress rules

There do not appear to be monomorphemic Maori exceptions to these rules, but Schütz
(1985, 14) draws attention to English loanwords that do not conform. Some apparent
Maori exceptions are polymorphemic. Hugh Young (Introduction to NGIOA, Vol II, p.8)
comments that ‘some speakers “broke the rules’”. While there are undoubtedly a few
cases of this kind, they are not as widespread as some think. Because the phonetic
realization of stress in Maori is different from English, English speakers and Maori semi-
speakers do not always pay attention to the features salient from the Maori point of view,
but are influenced by loudness and relative pitch in their assessment of stress. Thus one
of my consultants said kohanga ‘nest’ with emphatic onset of /k/, length on /o/, but
increase in loudness and a jump in pitch on /ha/. In English terms, this sounded like stress
on /ha/ (against the rules, which predict stress on /ko/), but I believe that this would be a
misinterpretation of the phonetic data.
Phonologically adapted loanwords are subject to the stress rules of Maori, and in the
process of adaptation, English stress is not always matched, eg.

although both these adaptations have first syllable stress in Maori, in contrast to the
second-syllable stress of the English. The adaptation *mihiini which would have
preserved the English stress is not attested in standard texts or dictionaries, although
learners frequently produce it. Unadapted English loanwords retain their English stress.
Schütz (1985) uses the stresses on loanwords to argue that word stress is not predictable
in Maori. However, I think there are differences between older and newer loanwords,
with newer loans more likely to preserve English stress and contravene Maori patterns,
like konotaraka < contract, cited by Schütz. Some older speakers describe such
adaptations as “non-Maori”, claiming that in earlier days, the phonological adaptations
omitted phonological material from English, to preserve appropriate Maori length (as
haamanu < ammunition, perhaps). It may also be necessary to distinguish those forms
adapted by native speakers of Maori from those adapted by non-native speakers. These
are matters requiring further investigation.

3.3.3.1–9 Distinctive use of pitch

Maori does not have tones. Thus pitch is not used to distinguish lexical items or
morphologically different forms of the same lexeme in Maori. However, yes-no questions
can be distinguished from statements purely by pitch and intonation contour. Such
questions have a raised pitch throughout It is possible, though, to mark such forms as
questions by lexical means as well, and in the speech of younger semi-speakers, the
 Maori 548

lexical means are often used without the distinctive pitch raising which older speakers
use.

3.3.3.10 Interaction of tone and stress

As noted in 3.3.2.2 a small pitch fall is a normal concomitant of a stressed syllable in
Maori. (It may be a rise if the contour of the phrase as a whole is rising.) It is normally
clearly audible in slow styles, but in fast casual speech, I am often unsure whether it has
occurred. There does not appear to be any necessary relationship between the relative
pitch of stressed and unstressed syllables of words pronounced in isolation, nor between
primary and secondary stressed syllables.

3.3.3.11–12 Downdrift and updrift

These phenomena do not occur in Maori.

3.3.4.1 Syntactic function of intonation

To specify the intonation of Maori, it is necessary first to define the phrase, which is, as
Biggs has pointed out (eg. 1969, 17) the fundamental unit of Maori both in terms of
syntactic structure and in terms of suprasegmental phonology. The phrase can be defined
either in structural terms, as consisting of a phrase-type marker (T/A marker, determiner,
preposition)+optional pre-modifier+head+optional post-modifier(s), or it can be defined
in phonological terms as a chunk of speech potentially preceded and followed by a pause.
In slow and formal speech, these potential pauses are actualized. In rapid casual speech,
two syntactic phrases may be realized as one phonological phrase, although the pitch
and/or stress patterns still sometimes betray the two-phrase origin. If a pronominal
subject is not treated as a separate phrase, but is phonologically part of the preceding verb
phrase (quite a common occurrence, perhaps even the norm), it acts like a particle, and
does not take the stress (see also 3.5.1.2).
The normal intonation pattern for declaratives involves a risefall curve on each phrase,
with a lower fall and decreasing loudness sentence finally. The highest pitch point in each
phrase usually occurs on the syllable which receives the major phrase stress, although it
is not uncommon to hear it on the syllable preceding the major stress, especially, I think,
if the unstressed syllable is a determiner. The stressed syllable has a small pitch-fall, but
overall pitch may remain relatively high until the final syllable, which then has a rapid
fall. In some speakers, each phrase has a lower over-all pitch than the previous one, while
in others, fairly high pitch is maintained in all phrases before the sentence final phrase. I
have been unable to determine whether this is a matter of individual variation, or whether
there is dialectal variation in this. Examples are found in 3.3.4.2.
Imperatives start extra high, and have a fairly steady descending intonation contour if
they are one phrase long. If they are extended, the “tail” usually adopts the declarative
pattern, eg. (vertical lines mark phrase boundaries):
 Phonology 549

The intonation of questions has been dealt with in considerable detail in 1.1.1.2. The
salient points are recapitulated here. Yes-no questions have a higher pitch for their
entirety than statements, and this pitch may be the only formal mark of question status.
Neutral yes-no questions have a high-rise terminal, but others are more likely to have a
fall. Question-word questions have basically declarative patterns, but the question word
itself has a rise. However, there is a lot of variation in question intonation. In particular,
question-word questions may have rising contours on all phrases except the one
containing the question word, eg.

Sometimes even the phrase containing the question word has this pattern:

The high-rise terminal which is a notable feature of New Zealand English appears to be
on the increase in Maori too, especially among semi-speakers. In fact, English intonation
patterns as a whole have had an impact on many Maori speakers, making it difficult to
ascertain with any certainty the traditional patterns.

3.3.4.2 Non-contrastiυe intonation peak

It is not clear that sentences in Maori, as opposed to phrases, have an intonation peak.
There is for many speakers a general pattern of descending pitch height on the phrase
peaks from the predicate phrase to the sentence end, eg.
 Maori 550

However, there are so many variations on this basic pattern that it is difficult to be sure
that it does indeed represent the ‘norm’. In particular, some speakers maintain all pre-
final phrases at a constant level, but higher than the final phrase, and for others, the initial
phrases may have rising rather than falling contours, although I think this may be more
characteristic of female speakers than male speakers, eg.

The characterisation of the locus of the highest pitch point as the predicate phrase is not
without its problems. In canonical verbal sentences, there is no problem, and in most non-
verbal sentences, the same appears true, eg.

However, for this characterisation to be correct for examples like the actor-emphatic
(2350), a particular syntactic analysis is required, which is not supported by all scholars
(eg. Waite, 1990). (The analysis of certain ko-fronting constructions also interacts with
this, see Bauer 1991).
It certainly does not appear to be the case that an initial nonpredicate phrase
necessarily has the highest pitch. Such phrases are frequently low in pitch, eg.
 Phonology 551

However, sometimes such phrases are high, but still the pitch on the stressed syllable is
usually slightly lower than that on the predicate. Again, I think that high pre-predicate
phrases may be more characteristic of female than male speakers, but a major survey
would be required to confirm this.

3.3.4.3 Emphatic intonation

Emphatic intonation in Maori usually consists of a higher than normal jump up on the
stressed syllable of the predicate. Compare the neutral

and emphatic

However, other means besides intonation are frequently used to convey emphasis, such as
emphatic sentence structures, and sometimes the addition of emphatic particles.

3.3.4.4 Contrastive stress

Contrastive stress is always accompanied by emphatic intonation, as illustrated in 3.3.4.3.
Indeed, it is not clear that the two are distinct.
Basically, I think, it is not possible to have more than one contrastive stress in a simple
sentence in traditional Maori. However, under influence from English, such usages are
 Maori 552

quite widespread, and my consultant was unsure whether there were circumstances under
which more than one might be possible. See also 1.11.

3.3.4.5 Subtypes of intonation patterns

Given the variety of realizations that occur, it appears that subtypes are more closely
linked to sex, dialect and emotional factors than to syntactic or semantic effects.

3.3.4.6 Interaction of intonation and stress

As the illustrations above show, stress, which is normally marked by pitch-fall causes a
jump up which ‘disturbs’ the basic risefall curve of the intonation pattern. Emphatic
stress causes a higher jump which is then compensated for by a steeper fall. Note that if
the basic intonation pattern is a fall-rise, as in some questions, the stress is marked by a
step down, see the examples in 3.3.4.1.

3.3.4.7 Effects of intonation on segmental units

I am not aware of segmental units being affected by the placement of the intonation peak
or the type of contour, other than the coincidence of stress with intonation peak, and the
associated emphatic onset and lengthening of the vocalic segment. These are the effect of
stress, rather than intonation (see 3.3.2.2).

3.4 SEGMENTAL MORPHOPHONOLOGY

There is very little segmental morphophonology in Maori of synchronic relevance.

However, dialect differences provide evidence of a wealth of such relationships. In what
follows, some information is given of the sorts of cross-dialectal relationships which are
found. It has been beyond the scope of this book to research such matters in detail. The
information here should thus be taken as indicative only.

3.4.1.1 Assimilation
There appear to be no instances of consonant assimilation. Vowels, however, can occur
adjacent to each other, and the major instances of vowel allophony are caused by
assimilation. In the vast majority of cases, as far as I can determine, phoneme alternations
are not involved. However, there is some evidence that labialization causes phoneme
assimilations in some dialects, and possibly diachronically in single dialects. Consider the
following alternations:
houkeke ~ hookeke ‘obstinate’
hautaorua ~ hautoorua ‘dew’
koihoka ~ koohoka ‘impale’
 Phonology 553

It appears likely in these cases that the strongly labialized [o] has spread its labialization
to the adjacent segment, either following or preceding, which has then completely
assimilated to it. A similar process may be responsible for alternations like the following,
where the labial feature stems from [u], rather than from [o]:
hautoke ~ hootoke ‘winter’
kauhimu ~ kohimu ‘gossip’

(There are also many instances of variant forms with and without [w] in labial
environments. These are treated below in 3.4.4.1, but are further evidence of the
pervasive effect of labialization on Maori phonology.)

3.4.1.2 Dissimilatory processes

The fact that certain pairs of consonants in adjacent syllables are rare has already been
mentioned in 3.2.6.3. Harlow (1991) suggests that Maori has a rule which he calls
‘consonant dissimilation’, which involves the total loss of a consonant in a syllable
adjacent to another syllable of the same form. This rule is accordingly treated below
under deletions in 3.4.4.1.
The labial feature mentioned above under assimilations, and responsible for many
deletions (discussed below), also appears to be responsible for one phenomenon which is
probably dissimilation. It has already been mentioned in the section on phonotactics that
neither w nor wh in Maori can precede the rounded vowels [o] and [u], a restriction which
appears to be a reflection of a wider prohibition against sequences of labialized sounds. It
will be shown in 3.4.4.1 below that [w] is frequently subject to deletion when it follows
these same two vowels, but there are very few instances where wh is deleted under those
circumstances. However, in a large number of instances with appropriate input strings (ie.
…owh…, …uwh…), Williams’s Dictionary records a variant with [h] for wh, eg.
auwhi ~ auhi ‘sad, dejected’
huuwhaa ~ huuhaa ‘thigh’
koowhatu ~ koohatu ‘stone’

It thus seems possible that these instances of h—wh alternations serve to dissimilate these
labial environments. However, h—w h alternations are not restricted to rounded vowel
environments, and some of these alternations are attributable to a messy sound change (I
am indebted to Chris Lane for drawing this problem to my attention). A search of
Williams’s Dictionary revealed 25 alternations in nonrounded environments including
word-initially and 59 in rounded environments. Given that the vowels are not equivalent
in frequency, and that in texts the non-rounded vowels /a/ /e/ /i/ are considerably more
frequent than the rounded vowels, there would appear to be good reason to presume that
the rounded vowels do play a part in h ~ wh alternations. I have not found any relevant
data for this process amongst modern loan words. The mere fact that a few modern loan
words from English break the restriction against w and wh+[o] or [u] suggests that these
labial sequences are nowadays tolerated, although one of my consultants, when asked to
 Maori 554

read forms with w+/u/ or /o/ in several cases pronounced them with wh, and plainly
found it extremely counter-intuitive to produce /wo/ or /wu/.

3.4.1.3 Other alternations between segments

There are many other alternations between segments in Maori which are not readily
accounted for by the processes above. The only case where a phonemic alternation has
syntactic consequences is the handful of nouns for which plural is marked by an
antepenultimate long vowel where the singular has a short one (see 2.1.1.8.6), but it must
be borne in mind that this is not the only possible way to analyse this process. The
dialectal variation in the consonants in the passive suffix should also be mentioned here.
Apart from these, some of the observed alternations are clearly cases of systematic
dialectal variation, but many of them appear to be random. The major instances of
dialectal variation are dealt with first. (For further information on this topic, see Harlow
(1979), and Biggs (1971) and (1989).) The examples are ones listed in Williams.
Variation between [n] and [ŋ]:
Underlying [ŋ] usually becomes [n] in Ngati Awa and Tuhoe:
huuneieni ~ huungeingei ‘anger’
keno ~ kengo ‘night, underworld’

Variation between [k] and [ŋ]:

Underlying [ŋ] usually becomes [k] in Ngaitahu:
hango ~ hako ‘shovel’
hiitengi ~ hiiteki ‘hop’

Variation between [h] and [?]:

Underlying [h] usually becomes [?] in Taranaki-Whanganui. The glottal stop is not
recorded in the orthography. However, not all the instances recorded where h alternates
with Ø in the orthography are in fact instances of this alternation, since, for instance, the
negative kaaore which alternates with kaahore is by no means confined to the Taranaki-
Whanganui region. Some typical Taranaki-Whanganuivariants are:
‘buy’

‘move’

At least some of the variation between h and wh is dialectal, although I have suggested
above that in certain environments, it may be an instance of dissimilation. Such
alternations as
hea ~ whea ‘where’
haakure ~ whaakure ‘catch lice’

appear to have a regional basis, although further research into dialectal variation requires
to be done to establish precisely which regions use which form.
 Phonology 555

There are also a number of vowel alternations which have a regional basis. The most
important ones are variation between [ei] and [ai], and between [ou] and [au] after front
consonants, both of which occur as an East-West dialect split. However, the variation
between these strings seems to be restricted to particular forms. It also occurs in both
directions (examples from Biggs, 1989, 67):
West East
teina taina ‘younger sibling’ West innovates
hei hai ‘for’ East innovates
-tou -tau ‘pl. in pronouns’ East innovates
tuumou tuumau ‘permanent’ West innovates

There are also some instances of [u] becoming [i] in the East after front consonants:
tupu tipu ‘grow’

After these fairly well attested cases of dialectal variation, there are still large numbers of
forms recorded showing alternations where it is not at all clear whether dialectal
variation, diachronic variation, or even simple error is involved. Some show evidence of
patterning, such as those alternations which are due to the feature labialization, which
accounts for some deletions, dissimilations and assimilations. Firstly, [w] alternates with
[u] in eg.
maawewe ~ maaueue ‘flutter, tremble’

Secondly, [w] alternates with [o] in eg.

titihaoa ~ titihawa ‘shout of joy’

Neither of these alternations is particularly common.

In addition, many unsystematic variants occur, which are not productive as far as I can
determine. A fairly representative collection is given here, including those discussed
above, to present as complete a picture as possible, but I cannot guarantee
exhaustiveness. Plainly, it is an area of Maori phonology which would merit further
investigation. Variations between consonants
p~k parengo ~ karengo ‘sea-weed’
p~t hiato ~ hiapo ‘be gathered together’
p ~ wh wheeraa ~ peeraa ‘like that’ (probably Tuhoe variation)
t~p see above
t~k kaio ~ taio ‘lock of hair’
t~h kauati ~ kauahi ‘wood used to produce fire’
k ~ p, t see above
k~h kookako ~ hookako ‘N.Z. crow’
 Maori 556

k~ŋ kareko ~ karengo (~ parengo) ‘sea-weed’

k~m kanono ~ manono ‘a type of shrub’
k~r kokowhawha ~ korowhawha ‘anchovy’
m~ŋ kakaramuu ~ kakaranguu ‘a type of shrub’
m~k see above
m~h mokemoke ~ hokehoke ‘lonely’
n~ŋ huuneinei ~ huungeingei ‘anger’
n~r haamonemone ~ haamoremore ‘bare of branches’

ŋ ~ m, n, k see above
wh ~ p see above
wh ~ w wheekiki ~ weekiki ‘quarrel’
wh ~ h koowhatu ~ koohatu ‘stone’

(NB. there is sometimes a vowel change associated with this alternation, even when it is
not obligatory, eg.
hakori ~ whekori ‘be seen’
hoorite ~ whaarite ‘measure’)

h ~ wh, t, k, m see above

h~r huumene ~ ruumene ‘folded up’
r ~ k, n, h see above
w ~ wh see above

Variation seems particularly common in relation to flora and fauna. Some forms have
many variants, such as the following (not necessarily exhaustive) list for the N.Z. fantail:
hiiwaiwaka ~ piiwaiwaka ~ tiiwaiwaka ~ riiwaiwaka ~
hiiwakawaka ~ tiiwakawaka ~ tiirairaka ~ piirairaka ~ tiiaiaka…

In addition, it is possible for every consonant to alternate with Ø in at least some context,
see 3.4.4.1. Variations between vowels
a ~ aa anini ~ aanini ‘giddy’
a~e hamanga ~ hemanga ‘not full’
a~o aha ~ aho ‘open space’
a~u kaiwhata ~ kaiwhatu ‘a spell for well-being’
e ~ ee wheekiki ~ whekiki ‘quarrel’
e~a see above
 Phonology 557

e~i kemo ~ kimo ‘wink’

e~o hiiteki ~ hiitoki ‘hop’
e~Ø haere ~ hare ‘move’ (shorter form is poetic)
i ~ ii inanga ~ iinanga ‘whitebait’
i~u akitoo ~ akutoo ‘be lengthened’
i~e see above
i~o uhitea ~ ahotea ‘edible tuber’

i~ kaikama ~ kakama ‘brisk’

o ~ oo pohutukawa ~ poohutukawa ‘pohutukawa tree’

o ~ a, e, i,. see above
o~u hoata ~ huata ‘long spear’
o ~ au kauhimu ~ kohimu ‘gossip’
o ~ aw rauoa ~ rauawa ‘sideboards of canoe’
o~w titihaoa ~ titihawa ‘shout of joy’
u ~ uu kuaka ~ kuuaka ‘godwit’
u ~ a, o, i see above
tauweke ~ taweke ‘linger’
u~
u~w maaueue ~ maawewe ‘flutter, tremble’
aa ~ a see above
aa ~ ee hongaa ~ hongee ‘N.Z. crow’
aa ~ ii hurupaa ~ hurupii ‘fresh growth after cutting’
aa ~ oo kaamura ~ koomura ‘burn dry leaves over something’
aa ~ uu haarau ~ huurau ‘meddle with’
ee ~ e, aa see above
ee ~ ea putee ~ putea ‘bag’
ii ~ i, aa see above
ii ~ uu tiipaa ~ tuupaa ‘dried up’
oo ~ o, aa see above
oo ~ uu hootoke ~ huutoke ‘winter’
oo ~ au? hootoke ~ hautoki ‘winter’
oo ~ ao hautoorua ~ hautaorua ‘dew’
oo ~ ou hookeke ~ houkeke ‘obstinate’
 Maori 558

oo ~ oi koohoka ~ koihoka ‘impale’

uu ~ u, aa, oo, ii see above
uu ~ au? huutoki ~ hautoke ‘winter’

Many of these alternations also occur in vowel sequences, of course.

3.4.2 Metathesis
There are a handful of forms in Maori showing metathesis, but the process does not
appear to be productive. Some of these forms involve the metathesis of consonants in
adjacent syllables. The process often but not invariably involves a nasal consonant, and
often but not invariably arises in reduplicated forms, where the other order of syllables
may be present, thus presumably initiating the metathesis (as in ngahangaha ~
hangahanga, for instance, where the sequences ngaha and hanga are both present in each
form). In some cases, the process involves the metathesis of adjacent syllables, and in
some cases, it is impossible to be sure which of these processes is involved. There is also
one clear example involving the metathesis of vowels in adjacent syllables. I have no
examples of adjacent vowels metathesising, although this is perhaps the type of
metathesis that is most expected. I have not found any examples of modern loans
undergoing the process. The forms I have found are: Consonants in adjacent syllables:
engari ~ erangi ‘it is better, however’
honuhonu ~ nohunohu ‘nauseous’
kihukihu ~ hikuhiku ‘fringe’
ngaro ~ rango ‘fly’
raaitahiraa ~ raaitarihaa ‘day before yesterday’

Adjacent syllables:
koomuramura ~ kooramuramu ‘eat at odd times’
koongahungahu ~ koohungahunga ‘babe, infant’
ngakongako ~ kongakonga ‘crumbled to pieces’
†pepa ~ pape ‘slip in reciting karakia’
puutangitangi ~ puungitangita ‘prickle’

The following appears to involve a long-short vowel alternation as well:

wheato ~ aawheto ‘vegetable caterpillar’

Adjacent syllables/consonants from adjacent syllables:

koohumuhumu ~ koomuhumuhu ‘whisper, murmur’
koongarangara ~ koorangaranga ‘colic’
 Phonology 559

ngahangaha ~ hangahanga ‘frivolous’

ngaarehe ~ ngaahere ‘forest’

The following appears to involve [n]-[ŋ] alternation as well:

ngaawhaariki ~ whaanaariki ‘brimstone’

Vowels in adjacent syllables:

mahine ~ maheni ‘smooth, sleek (skin)’
†
pepa ~ pape ‘slip in reciting karakia’

†
This form can be analysed as involving either the metathesis of syllables or the
metathesis of vowels in adjacent syllables, and there is no evidence for preferring one to
the other.

3.4.3 Coalescence and split

Some of the examples involving the labial feature appear to show split. In the following,
[aw] is in variation with [o]:
aoake ~ aawake ‘2 days hence’
kaipaoe ~ kaipaawe ‘tramp’
rauoa ~ rauawa ‘sideboards of canoe’

The forms with [o] are clearly the source forms in at least the first two of these. There is
minimal auditory difference between these two variants. (Compare these with the
assimilations above.) It is difficult to determine whether this process is productive: I
suspect that if a new form met the input conditions, it could still apply.

3.4.4.1 Deletion processes

The first groups of deletions considered arise like the cases of dissimilation, to prevent
sequences of labialized segments. Again, it is not possible to determine from the data
available to me whether the variation recorded is dialectal, diachronic, both, or neither.
Loss of [w] following [o]:
arowaa ~ aroaa ‘be comprehended’
koowaro ~ kooaro ‘upside down’
kowera ~ koera ‘broken (clouds)’
moowai ~ mooai ‘become moist, water’
mokowaa ~ mokoaa ‘space, interval’
raaowa ~ raaoa ‘be choked’
rowau ~ roau ‘motionless’
 Maori 560

rorowai ~ roroai ‘type of fish’

tooroowai ~ toorooai ‘weapon’

Note that the formulation of this alternation as loss of [w] is clearly justified
morphologically in some cases, such as moowai, but in most cases I do not have
sufficient information to be sure that the form with [w] represents the older or more
conservative form. Nor have I found any instances where the vowel following [w] is [i];
indeed, I have been able to find only one form with the required input string, namely rowi
‘South Is. kiwi’, and I have no evidence about the existence or otherwise of the variant
roi. (Note that [o] and [u] do not occur following [w], which is another facet of the same
restriction.)
Loss of [w] following [u]:
auwee ~ auee interjection of distress
haakuwai ~ haakuai ‘a mythical bird’
hauwarea ~ hauarea ‘insignificant’
hikuwaru ~ hikuaru ‘crooked’
kauwaeraro ~ kauaeraro ‘youngest-born child’
kuuwaha ~ kuaha ‘mouth’
kuware ~ kuare ‘ignorant’
rauwiri ~ rauiri ‘interlace with twigs’
ruwahine ~ ruahine ‘old woman’
tapuwae ~ tapuae ‘footprint’
tauwehe ~ tauehe ‘separate’

Notice that, although in most of these cases, [w] is followed by [a], there are also
examples where [i], and [e] follow. This suggests that the lack of examples with [wi]
following [o] is merely a gap. The same rider about the description as the loss of [w]
applies.
Rather than loss of [w] following [u], the following form shows loss of [u] preceding
[w]:
tauweke ~ taweke ‘linger’

This is not a common deletion.

I suspect that the loss of [w] is productive, and that loan words are subject to it, but
have not been able to find many relevant examples. However, Williams lists the
following alternants:
Noema ~ Nowema ‘November’

This is parallel to the forms above showing loss of [w] after [o].
 Phonology 561

Wh shares with [w] the restriction that [o] and [u] cannot follow it, which suggests that
wh shares the labialization feature with w, although some of its phonetic realizations do
not appear to be labialized. It is perhaps noteworthy that wh does not normally occur in
the sorts of deletion processes just illustrated for [w]. I have been able to find only one
instance:
towhari ~ toari ‘clear, bright’

There is also one form which has wh, w and Ø alternants:

huuware ~ huuwhare ~ huare ‘spittle’

From the information available it is not possible to ascertain which consonant here was
deleted. However, as noted in the section on dissimilation, the sequence of labial
segments can be dissimilated under these circumstances by the replacement of wh with h.
Harlow (1991) proposes a rule which he describes as consonant dissimilation, but
which involves the deletion of a consonant under certain circumstances. The rule
accounts for quite a large number of the observed morphophonemic variations between
long and short vowels in Maori. The rule Harlow proposes is that a mora can lose its
initial consonant if the mora is antepenultimate or earlier in a word and the preceding
mora has the same consonant. He proposes that this rule will account for the ‘irregular’
passives involving reduplication, eg.
pupuhi ‘shoot’ puˆuˆhia ‘pass. of pupuhi’
tatari ‘wait’ taˆaˆria ‘pass. of tatari’

He proposes that the underlying forms of these passives are regular, ie *puˆpuˆhia and
*taˆtaˆria respectively, and that the attested forms arise from the action of the consonant
“dissimilation” rule by deleting the bold consonant. Similarly, the rule operates in the
formation of reduplications from bases of three morae, eg.
koˆpuˆu ‘blistered’ koˆoˆpuˆuˆpuˆu ‘having blistered skin’
moˆhuˆa ‘yellowhead (bird)’ moˆoˆhuˆaˆhuˆa ‘id.’

The underlying forms proposed for these reduplications are koˆkoˆpuˆuˆpuˆu and
moˆmoˆhuˆaˆhuˆa, and the deletion of the consonant in bold gives rise to the attested
form. The rule proposed by Harlow thus accounts for quite a number of the systematic
instances where a long vowel apparently alternates with the corresponding short vowel.
(However, it probably does not account for the noun plurals (see 2.1.1.8.6) since forms
like tuaakana do not have the appropriate input form, thus casting doubt on this as the
appropriate derivation for those noun plurals that show an appropriate form.) In many
instances, the derivation of such complete reduplications requires the postulation of an
unattested base form. Consider the following data:
*paˆtaˆi paˆaˆtaˆi ‘ask’ paˆaˆtaˆiˆtaˆi ‘ask freq’
*moˆkaˆi moˆoˆkaˆi ‘pet’ moˆoˆkaˆiˆkaˆi ‘id’
 Maori 562

Harlow proposes that the underlying forms for the longer reduplications are
paˆpaˆtaˆiˆtaˆi and moˆmoˆkaˆiˆkaˆi, which derive regularly from the unattested three
morae bases given. The attested bases in the second column are themselves
reduplications of a different pattern but also show the action of the ‘dissimilation’ rule:
they derive from paˆpaˆtaˆi and moˆmoˆkaˆi respectively. Although I have not been able
to find any examples, I believe that this rule is productive, and that it potentially applies
to the reduplications of recent loan words.
The dissimilation rule Harlow proposes may also account for some other types of
variation found. He suggests that the variation between Ngaati and Ngaai in tribal names
may be another manifestation of this rule. The second variant occurs exclusively with
tribal names beginning with [t], although it is not obligatory under those tircumstances.
Some typical data:
Ngaati Porou Ngaai Tahu
Ngaati Toa Ngaai Te Rangi

Rather similar to the variation in tribal prefixes is the variation between whaka and whaa
as the causative prefix. However, this rule appears to operate optionally in both
directions: there are whaa forms before consonants other than [k], and [k] forms are not
obligatorily prefixed with whaa rather than whaka. The following four examples illustrate
the range of possibilities:
kii ‘full’ whakakii ‘fill’
kore ‘neg’ whaakorekore ~ ‘cause not to be’
whakakorekore
inu ‘drink’ whaainu ~ whakainu ‘give drink to’
kana ‘stare wildly’ whaakana ‘make grimaces’

There are also isolated variants like the following, which could be accounted for by such
a rule:
Autahi ~ Atutahi ‘the star Canopus’

However, there are also many instances of alternations which involve consonant loss
which do not seem to be accounted for by such a rule. A handful of representative
examples follows:
huuhi ~ uuhi ‘cover’
huumaarire ~ huumaarie ‘peaceful’
kaatahi ~ aatahi ‘then’
maoka ~ maonga ~ maoa ‘cooked’
hiiroki ~ hiioi ‘lean, thin’
heoti ~ heoi ‘so’
karapoti ~ karapoi ‘surround’
 Phonology 563

kekerenguu ~ keekereuu ‘black wood bug’

It is not at all clear how alternations of this sort are to be accounted for. In many
instances, it appears that sequences of three vowels in Maori arise through the deletion of
a consonant
There is one further deletion process which should be mentioned here, and that is the
deletion of the initial [t] from the determiner te in the compound determiners of Maori to
form the plural, as in
teenei ‘this’ eenei ‘these’

This process of plural formation applies throughout the set of compound determiners,
including possessive forms. The set of forms to which it applies is, however, closed.
Some dialects form the plural of some of these differently, using the plural determiner,
ngaa, eg. teetahi ‘indef spec’ most commonly has the plural eetahi, but in some areas the
plural ngaatahi is used. (The possible analysis whereby a singular morpheme {t-} is
added to the base in the singular cannot be sustained. The alternation of forms like teenei
whare ~ te whare nei (see 2.1.2.5.4) and parallel alternations in the possessives, see
1.10.1, and the existence of the plural ngaatahi are strong evidence for identifying the
first element as the determiner te.)

3.4.4.2 Insertion processes

There do not appear to be many clear cases of insertion in Maori, although the insertion
of epenthetic vowels in phonological adaptations makes an interesting study (see eg.
Kearns, 1990). However, the forms for the first person singular pronoun appear to show
consonant insertion:
au ~ ahau ~ awau ~ awahau ‘Isg’

The forms ahau and awau are apparently made up of the personal article a+au, with a
consonant inserted epenthetically. They cannot be preceded by the personal article.
However, it is not clear to me what the analysis for the last variant is. It looks as though
this process has applied twice, but I have no evidence that that is so.
A further insertion process can be seen in the forms of quite a number of temporal
adverbs. Consider the following forms:
inaaianei
noonaaianei ‘today’ cf aaianei ‘now’
onaaianei
inanahi ‘yesterday’
inakuanei ‘just now’ cf aakuanei ‘presently’
inakuaraa ‘a little while ago’ cf aakuaraa ‘a little while’
 Maori 564

The [n] following the initial preposition appears to be a case of the insertion of a
consonant to break up a long vowel sequence. The preposition forms a unit with the
following NP. The temporal adverbs appear to use [n] exclusively for this purpose.

3.4.5 Reduplication
The processes of reduplication have been discussed fully in the section on word
formation, since the interest of these formations in Maori lies not only in their
phonological form, but also in their meaning.
See 2.2.6.4.

3.4.6 Other processes

It will be clear from the sections above that the area of segmental variation in Maori is
under-researched. Much of the research which needs to be undertaken to throw light on
this area is a thorough dialect study. It has been impossible for me to undertake this. Thus
while I am not aware of other processes, it cannot be precluded that such a study might
reveal them.

3.5 SUPRASEGMENTAL MORPHOPHONOLOGY

3.5.1 Stress in morphologically complex forms

3.5.1.1 Complex words and compounds

The prefix whaka- and the passive suffix in Maori do not affect word stress, but other
prefixes usually take the major word stress in discourse, although toko, the numeral
prefix for people, shows some variation. Compare the stresses on the following forms:
‘hoko ‘barter’ ‘kai|hoko ‘salesperson’
‘tiaki ‘care for’ ‘kai|tiaki care-giver’
‘rima ‘five’ ‘toko|rima ‘five (people)’
‘whitu ‘seven’ ‘hoko|whitu ‘140’
‘rua ‘two’ ‘tua|rua ‘second’

but
|
‘hoki ‘return(intr)’ |whaka hoki ‘return (tr)’
|
‘rima ‘five’ |whaka rima ‘fifth part’
‘kimi’ ‘look for’ ‘kimihia ‘pass. of kimi’

In these last cases, if the normal rules for word stress applied, the stress would occur on
the affix, although there are some dialects which do not stress a final diphthong, so that
 Phonology 565

the stress for the passive kimihia would not alter. (Note, however, that some consultants
put primary stress on the passive suffix as well as the root, although all examples of this
were on longer forms, eg.
‘karanga ‘call’ ‘karanga’tia ‘pass. of karanga’

This probably has to do with the avoidance of long strings of unstressed syllables. Note
also that for some speakers, roots which end with a diphthong change their stress when
the passive suffix is added, conforming to word stress rules, eg.
‘horoi ‘wash’ ho’roia ‘pass. of horoi’)

There are one or two exceptions in the case of whaka, of which

the commonest is
wha’kairo ‘carve’

Iro does not appear to exist with a transparently related sense, although Williams lists the
whaka form under iro ‘maggot, vermin’. What appears to happen here is that the form no
longer has a transparent prefix, and thus is no longer stressed by the prefixing stress rule.
Partial reduplications appear to follow the stress rules for monomorphemic words:
‘nui ‘big’ ‘nunui ‘dup of nui’
‘kino ‘bad’ ‘kikino ‘dup of kino’

However, some uncertainty exists over this, because the forms most commonly heard
(papai ‘good’, nunui ‘big’) both have final diphthongs, and are thus subject to dialectally
variable stress, so that one hears nu’nui, which could be interpreted as the retention of
stress as on the simple base.
Two mora reduplications receive stress on both parts:
‘muhu’muhu ‘muttering’

Biggs (1969, 133) says that the first stress is heavier than the second, ie.
‘muhu|muhu

but in my data, that is not always clearly the case: there are instances which appear to
have stronger stress on the second element, and many cases where the stress appears to
me equal.
In the other common pattern of reduplication (Harlow’s Pattern 4, see 2.2.6.4), all
three units receive stress, but various patterns of primary and secondary stress are heard.
The commonest pattern seems to be
|
|pa tai|tai ‘to question repeatedly’
|
|ma hara|hara ‘remember’

but there are tokens where the initial syllable appears to receive primary stress in addition
to the primary stress on the second syllable, and other cases where the first syllable
 Maori 566

appears to receive primary stress and the two remaining elements receive equal secondary
stress.
Compounds in general appear to have primary stress on the second element (the
modifying element), with secondary stress on the head, eg.
|
kohaŋa ‘nest’
|
reo ‘language’
|
kohaŋa ‘reo ‘kohanga reo’
|
taŋata ‘man’
|
fenua ‘land’
|
taŋata ‘fenua ‘local person’

This coincides with the rules for phrase stress (see 3.3.4.1). However, with compounds in
isolation, one of my consultants varied between stress patterns like those above, and the
reverse
|
kohaŋa |reo

Compounds with the linking element -aa- take primary stress on the linking element:
|waiata ‘a |riŋa ‘action song’

or in less formal speech

|
|waia ta riŋa
With the verb-noun compound nouns (see 2.2.6.3), consultants varied as to which
element received primary stress:
‘poro ‘cut’
‘pepa ‘paper’
‘poro|pepa ~ |poro|pepa ‘guillotine’

3.5.1.2 Phrases
Phrases are the major unit in Maori phonology, see 3.3.4.1. Each phrase has one major
stress. Biggs (1969, 133) distinguishes stress placement in final and non-final phrases.
“In final phrases the phrase stress occurs on the last base in the phrase” according to the
rules for word stress given in 3.3.2.5. “In non-final phrases, however, the phrase stress
occurs on the last syllable in the phrase if that syllable contains more than one vowel,
otherwise on the second to last syllable”. (This last is presumably equivalent to stressing
the penultimate mora.) However, there is considerable deviation from these rules
observable in the Maori-speaking community. In particular, it is not uncommon to hear
non-final phrases stressed in the same way as final phrases.
The major exception to these phrase stress rules is phrases of the form
prep+pers+sg pers pron
 Phonology 567

eg.
i a ia
from pers IIIsg
‘from her’
ki a koe
to pers IIsg
‘to you’

With these phrases, whether non-final or final, the stress falls on the personal article, and
this article is pronounced with long vowel length and quality before ia and koe:

Note that these pronominal bases contain two morae, and so are phonologically possible
bases. The form ahau, however, behaves as expected, and with au, the two /a/’s form a
long vowel which carries the stress:

Thus these combinations of personal article and singular personal pronoun form a single
base in phonological terms. Other personal pronouns (which all have more morae) are
treated as bases like any other.

3.5.2 Tone in morphologically complex forms

Not applicable.

3.5.3 Interaction between intonation and segmental elements

Exclamatory sentences have not only a special intonation, but also other suprasegmental
features which are realized by certain changes in the articulation of segments. In
particular, drawl occurs, with noticeable lengthening of the segments, both consonants
and vowels. The specific effects on each consonant type are specified in 3.3.2.2.
 4
Ideophones and Interjections

4.1 IDEOPHONES

Maori makes some use of ideophones. However, there were many cases where my
consultant was in doubt as to whether particular forms were ideophones or not, notably
for bird names. The following list is fairly conservative, and contains only those of which
there is reasonable certainty. Other consultants might well have included others, and
excluded some of these.
Form Gloss
amuamu grumble
au howl (of dog)
aue to cry with distress
haruru dull, heavy sound
ihiihi ?kea make a hissing, rushing noise, as the wind N.Z. bird (from call)
keekee quack (duck)
kekee creak
keo screech (as bird)
?kiwi native flightless bird (from call)
kihikihi cicada
kiki(tara) cicada sound
kitaa chirp of cicada
kuuii bird call, call
ngawee ~ ngawii squeal, yelp, howl
ngawiingawii cry of puppy
ngetengete ~ ketekete palatal click for surprise
ngongoro snore
ngunguru rumble (as thunder)
paatiitii splash
pakuu loud sudden sound, pop
piipii young of birds, chick (from call)
 Ideophones and interjections 569

?tuuii native song bird (from call)

wawaa make a loud rumble or roar
whio whistle

4.2 INTERJECTIONS

It is not clear to what extent Maori makes use of interjections which do not conform to
the regular principles regarding the phonological structure of words. Perhaps the one
totally clear example is the use of a palatal click to express surprise or disapproval (the
latter use possibly derived from English). However, some interjections expressing disgust
appear to have vowels not normally part of the Maori system, and can end with a
voiceless glottal (sometimes as far forward as velar) fricative, eg.
[næh]~[nεh]

Sometimes the initial consonant in this interjection is a devoiced alveolar or dental stop,
neither of which normally occurs in Maori:

However, the vast majority of interjections in Maori conform to the normal phonology of
the language.
 5
Lexicon

One of the many signs of the decay of the Maori language is a general depletion in the
vocabulary known to those speakers who remain. Even those older speakers who have
lived most of their lives in predominantly Maori-speaking communities frequently find
themselves using English words, often not even phonologically adapted. Sometimes this
is for items not part of traditional Maori life, but such Anglicisms in many cases replace
the traditional Maori terms, with the subsequent loss of those traditional terms. This has
made it difficult to deal with the questions in this section. Williams’s Dictionary records
many terms not known to my consultant Thus I culled about 200 terms relating to kinship
from Williams, and of these, almost half were unknown to my consultant Some of this is
no doubt due to the recording of forms restricted in dialect, but much of it is due to
language impoverishment under pressure from English. In the sections below, I have in
large measure restricted myself to items that are clearly still in use, as only in these cases
is there adequate information available.

5.1 STRUCTURED SEMANTIC FIELDS

5.1.1 Kinship terminology

Maori kinship terminology reflects the traditional social units of Maori society, in
particular the whaanau, or extended family. However, many of the terms for distant
relations have disappeared from use today. Even with respect to some of the distinctions
in age and gender which were plainly keystones of traditional whaanau, modern speakers
confess to uncertainty as to the appropriate terminology.

5.1.1.1 By blood
Five generations are distinguished, two older and two younger than the point of
reference. Thus the basic terminology is:
tipuna ~ tupuna matua grandparent and earlier ancestor parent
tamaiti child
mokopuna grandchild and subsequent descendant

No distinctions for sex are made in the two outer generations, although terms such as kuia
‘old woman’ and koroua ‘old man’, which are not necessarily for relatives can be used if
sex distinctions are required. No distinctions are made within the older generations for
 Lexicon 571

paternal or maternal relationships, either, at least not to the knowledge of my consultants,

although certain glosses in Williams’s Dictionary for terms not known to my consultants
suggest that such distinctions may have had a place in kinship terminology formerly.
There are other terms for ancestors in general, such as puutake (from ‘base, root’), and
heinga ~ wheinga, used for instance as a term of respectful address to a very old person.
Within a tribal grouping, the elders are kaumatua, a term normally restricted to males,
although that restriction is changing for many speakers today. More general terms for
descendants are kooiwi ‘line of issue’ (also ‘bone’), normally rather formal, and uri,
which can be used for any blood relative, but is most commonly used for offspring.
Within the category parent, the term matua is used primarily for the male parent, and
the term matua wahine ‘female parent’ can be used as a parallel term, but this is rather
formal; more usual terms are whaea, or in some dialects kookaa. (The term whaereere
‘animal mother, dam’ can also be used jokingly or insultingly.) More informally, paapaa
is used for ‘father’ and there is an equivalent maamaa for ‘mother’. These seem to be
used as the normal terms of address for parents in modern times.
In the category ‘child’, distinctions can be made for sex. Some speakers appear to
reserve the term tamaiti for males, although the form tamariki, normally described as the
irregular plural of tamaiti, is used for groups of mixed sex by all speakers. It seems to me
far more common nowadays for tamaiti to be used regardless of sex. Tama is ‘son’, but is
sometimes used more widely; tamaahine is a somewhat formal equivalent for ‘daughter’,
but can also be used for non-relatives, ie. for ‘girl’ in general. This is also true of the
more usual term kootiro ‘girl, daughter’.
Position of birth in the family was traditionally of considerable importance, and there
are many words in Williams’s Dictionary reflecting this. Only two seem to be in common
use nowadays, mataamua for ‘first born’ and pootiki for ‘youngest’; the latter is also used
for any infant or young child. However, the terms whakamutunga and mutunga for the
youngest are still sometimes used.
The traditional terminology for siblings involved factors of both position in the
whanau and sex. Some of these distinctions seem to be in the process of disappearing. As
far as I can determine, the following were the traditional terms:
tuakana older sibling of the same sex
teina younger sibling of the same sex
tungaane brother of a female
tuahine sister of a male

However, there seems to be a tendency nowadays for the sex distinctions to be ignored:
speakers can be heard using tuakana for older siblings and teina for younger siblings,
regardless of sex. Wills (1960, 83) says “Modern Maori women please note” in relation
to the term tungaane and also the terms for in-laws, which suggests that these distinctions
were often ignored at that time.
Traditionally, all these terms applied to all members of the extended family. Thus
aunts were whaea and uncles matua; same sex cousins were tuakana or teina depending
on age (or in at least some cases on whether they belonged to the older or younger branch
of the family); mokopuna is used for any group of young children, regardless of the
 Maori 572

directness or otherwise of the relationship. However, nowadays the following terms are
common:
matua keekee ~ matakeekee uncle, aunt
iraamutu niece, nephew

These terms are probably on the increase, due to influence from English family systems. I
have heard Maori speakers regretting the fact that Maori doesn’t have a word for
‘cousin’.
The term kirimate ‘near relative of the deceased’ should perhaps also be mentioned, as
being of importance in the tangihanga or mourning ceremony.

5.1.1.2 By partial blood

No distinction is made between partial and full blood relationships, see 5.1.1.1.

5.1.1.3 By marriage
The commonest terms for relations by marriage are:
hoa husband, wife (also ‘friend’)
wahine wife (also ‘woman, female’)
taane husband (also ‘man, male’)
hungawai ~ hungarei parent-in-law (variants are dialectal)
~ hunarere ~ hunarei
~ huungoi
hunoonga ~ hunaonga step-child
~ hinaonga
taokete sibling-in-law of same sex
auwahine sister-in-law of male
autaane brother-in-law of female

For some speakers nowadays, taokete is used for all same-generation relations by
marriage. The distinctions in these last three terms are among those which Wills (1960,
83) draws the attention of “Modern Maori women” to.

5.1.1.4 By adoption
Maori society had no formal adoption procedure prior to the arrival of the Pakeha.
Fostering, however, was, and still is, common. See 5.1.1.5. Within modern religious
groups, the terminology of the whaanau is used for the religious family, see 5.1.1.1.
 Lexicon 573

5.1.1.5 By fostering
The usual terms for foster relations are:
matua whaangai foster parent
tamaiti whaangai foster child

Whaangai can be glossed as ‘feed, nourish, bring up’. Sometimes atawhai ‘show
kindness to’ is used instead of whaangai. However, it is entirely usual for foster relatives
to be called by the same terms as are used for full blood relatives.

5.1.1.6 By affiliation
This does not seem to be a relevant parameter for Maori kinship.

5.1.1.7 By tribe
Tribal relations are of considerable importance in Maori society. Beyond the whaanau, or
extended family, there are several other groupings of importance. The next in size is the
hapuu, or sub-tribe, followed by the iwi or tribe (although this last term is sometimes
used to encompass all tribes). Sometimes the term waka ‘canoe’ is used to designate
tribal groupings. My consultant deemed it equivalent to the hapuu. (The tribal groupings
can be traced back to the various canoes in which the Maori people arrived in New
Zealand.) The term mata(o)waka is also used for these groupings on formal occasions.
Reflecting these ‘nested’ groupings, are the two terms whanaunga ‘relative, blood
relation’; huaanga ‘relative, member of the Same hapuu’. Williams’ Dictionary records
many other words for tribal relations which were not known to my consultant, which is
no doubt a reflection of the fact that during this century, tribal membership has been of
decreasing importance, although the renaissance of the Maori people has seen a reversal
of this trend. Some of these old terms may yet see a revival.

5.1.2 Colour terminology

It is not easy to determine what the colour system for Maori was prior to the arrival of the
Pakeha. However, it appears possible that it was a three-term system. My consultant was
uncertain of the range of application of the available colour terms, and it is not clear that
the typical values given are traditional. Those colour terms which are clearly traditional
are
black pango/mangu range possibly any very dark colour; most typical value given as ‘night’
white maa range possibly any light colour; most typical values given as Christ, sand in
Northland
red whero the range is unclear, but probably extending to any reddish brown and also to
some orange colours and tan; most typical values given as blood,
pohutukawa flowers
 Maori 574

It is uncertain whether there was a traditional colour name for blue or green. The terms
used today are
blue kikorangi, kahurangi blue sky
green kaakaariki native green parrot

My consultant clearly found it hard to imagine that Maori lacked these terms
traditionally, but was uncertain of their traditional range, for instance. These coexist
beside the borrowings from English puruu, kiriini. A further term which, like these, refers
to a natural phenomenon with a particular colour, and which nowadays is used as a colour
term is
yellow koowhai a native tree with bright yellow flowers in Spring

There are a number of other colour terms in use in Maori today which are either derived
from the basic terms or borrowed from English; whether the derivatives are traditional,
and if so, what their values were is information I have been unable to retrieve. Today,
they are clearly used with values derived from English: Derived terms
grey puumaa
pink maawhero
Borrowings
brown paraaone
grey kerei
orange aarani
purple paapura

There are many who prefer to avoid such borrowings, and terms for some of these
colours such as para-karaka for ‘orange’ have been proposed as alternatives. These
follow the pattern of koowhai ‘yellow’: they are the names of some natural phenomenon
of an appropriate colour, in this case, the berry of the karaka tree. It is difficult to judge
how successfully they will supplant the borrowings.
The general word for colour appears to be tae, ‘stain’, originally the name for a bark
dye, apparently. A few other traditional terms related to colour terminology are
striped whakahekeheke
faded puumaahue
blaze, flame, blush, redden mura

Grey or white hair is matenga hina.

5.1.3 Body Parts

The body parts that follow come from a number of sources. Most have been checked with
my principal consultant; those unknown are marked with a † below. Many of these come
 Lexicon 575

from Williams’s Dictionary. The remainder come from a list of medical terms issued by
the Maori Language Commission, and may represent recent coinages to satisfy medical
requirements. A number of terms exist in both stem and reduplicated forms. This is
indicated by brackets round the reduplicated material.
body tinana
†
limb (pe)peke
head maahunga, maatenga, upoko (head of food animal: pero)
neck (back) kakii
throat korokoro
†
nape porongaaua, †porongaaue, †ue
shoulder pokohiwi, pokowhiwhi, pakihiwi
chest uma, rei
breast uu
waist hope
arm/hand ringa(ringa)
elbow tuke(tuke)
†
wrist kawititanga
leg/foot waewae
thigh kuuwhaa, huuhaa
†
hip humu, †himu, hope
knee turi
heel rekereke
ankle pona(pona)
joint in arm/leg pona, punga
back tuaraa
†
spine tara
buttocks tou (polite), nono (slang), kotore (blunt)
stomach, puku, kopu
abdomen
navel pito
armpit keekee
ribs rara
finger matihao, matimati (many variants are found)
toe matikara
 Maori 576

thumb/big toe koonui

index finger/toe kooroa
2
middle finger/toe maapere
ring finger/toe 4 manawa
little finger/toe koiti
†
knuckle mona
hair (head) makawe
hair (body) huruhuru
crown of head tumuaki
face kanohi
forehead rae
eyebrow kape
eye kanohi, karu, mata
ear taringa
cheek paparinga
nose ihu
lip ngutu
tooth niho
mouth waha, maangai
jaw, chin kau(w)ae
tongue arero
beard paihau, paahau
†
sole of foot kapu, tapuwae
†
palm of hand kapunga, †papanui
nail on finger/toe ma(t)ikuku
eyelid kamo
eyeball kanakana
nostril ponga-aa-ihu, puta-ihu
†
ear-lobe hoi, †pokopoko, †toke
penis ure
liver ate
heart ngaakau
 Lexicon 577

lungs pukapuka
kidney ate-whatakuhu
vein uaua
skin kiri, hioko
fat ngako
flesh kiko
blood toto
guts whekau
brain roro

5.1.4 Cooking terminology

Here, as elsewhere in the lexical area, it is difficult to determine the traditional
terminology and cooking methods, and distinguish them from modern practices. There is
also considerable dialect variation in the terminology used. Traditional foods included
seafood (fish, shellfish, etc) and birds; root crops such as kumara and taro, and some fern
roots; puha, taro leaves; berries.
The general term for ‘cook’ is whakamaoa, which is a causative of maoa, ‘cooked’.
This is opposed to mata ‘raw’. There appear to be three principal methods of cooking
traditionally, in an earth oven, or over a fire, or boiling by means of heated stones.
The large earth oven is basically a method of steaming food. The term haangii is used
for the process and the resultant food. For at least some speakers, poki is used for a small
earth oven, and possibly also for ‘cook in ashes’. The term haangii is used for ‘to steam’
using modern methods as well.
The term for ‘to cook in smoke over a fire’ is tunu, which extends to modern barbecue
cooking. Tunu is also used for ‘bake’, and tunutunu for ‘roast’. My consultant also
produced the term waamuu ‘to smoke’; I have not been able to determine how this differs
from tunu.
The term koohue ~ koohua is the term used for ‘to boil’. My consultant was unsure
what sort of vessels were traditionally used for this purpose.
Preparation of food included marinating, for which whakatuu ‘stand’ is used.
However, my consultant was not able to provide more information on food preparation.
Food preservation was important: tutu or korotutu is ‘preserve in fat’; pakipaki is
‘preserve by drying’. No other methods were known to my consultant

5.1.5 Other structured fields

I have no doubt that there were other structured fields in Maori, such as fishing and
hunting terminology. However, so many of the terms are lost to any one individual, and
there was so much dialectal variation, that only a major cross dialectal study involving all
the older speakers of Maori would reveal sufficient information. Such a study is well
beyond the scope of this work. A study of insect names in Northland by Wendy Pond
 Maori 578

suggests that these too are a structured domain. The following examples are indicative
(information generously supplied by Wendy Pond, personal communication):
ngaro: winged insects with stout bodies and supportive stances and zig-zag manners
of flight
ngaro paaera: humming hover flies
ngaro wiiwii: spider-hunting wasps, especially mason wasp which buzzes inside its
cell
ngaro huka: ‘sugar fly’ probably lesser house fly
ngaro rere rangi: the original term for the introduced honey bee which was observed
to fly higher than flies do
ngaro miere: introduced honey bee (from French?)
ngaro ngahere: introduced honey bees which went wild or biting flies
roo: stick insects which are unwinged and one species of praying mantis which is
winged when adult but not in its juvenile form roo rere: winged roo or adult praying
mantis

5.2 BASIC VOCABULARY

1. all katoa
2. and no single equivalent, see 1.3
3. animal (4-footed) kararehe
4. ashes pungarehu
5. at i/kei/hei/ko/a, (see 2.1.1.5.1, 2.1.1.6, 2.1.5.1)
6a. back (of body) tuaraa
6b. back (behind) muri
7. bad kino
8a. bark (of tree) hiako
8b. bark (of dog) auau, tautau (many dialectal variants)
9. because naa/noo/moo/taa te mea
10. belly puku, koopuu
11 big nui, rahi (now infrequent)
12. bird manu
13. bite ngau(ngau)
14. black mangu, pango
15. blood toto
16a. blow (vb.) pupuhi
16b. blow (n.) meke
17. bone wheua, kooiwi
 Lexicon 579

18. breast uu
19. breathe whakaaea
20a. burn (intr) kaa, toro (=ablaze)
20b. burn (tr) tahu (=set fire to), wera
21. child tamaiti
22. claw matikuku
23. cloud kapua
24. cold makariri, maatao, koopeke
25. come haere mai
26. count tatau
27. cut (vb.) tapahi, poro
28. day raa, rangi
29. die mate
30. dig keri
31. dirty paru(paru)
32. dog kurii, kiirehe
33. drink inu
34. dry maroke
35a. dull (not sharp) puhuki
35b. dull (of weather) kinokino, poouri
36. dust puehu
37. ear taringa
38a. earth (world) ao
38b. earth (soil) one(one), whenua
39. eat kai
40. egg hua (manu)
41. eye kanohi, mata
42a. fall (autumn) ngahuru
42b. fall (descend) taka
43. far taawhiti
44a. fat/grease hinu (=dripping), ngako (=fat)
44b. fat (overweight) momona
 Maori 580

45. father matua (tane), paapaa

46. fear mataku, wehi
47. feather huruhuru
48. few takitahi, ruarua, torutoru
49. fight whawhai
50. fire kapura, ahi
51. fish ika
52. five rima
53. float maanu, tere
54. flow rere, tere
55. flower pua, puaawai, putiputi (<Eng. pretty)
56a. fly (insect) ngaro, rango
56b. fly (vb.) rere
57. fog kohu
58. foot waewae
59. four whaa
60. freeze haupapa
61. fruit hua raakau, whurutu (Eng.)
62. full kii
63. give ho(o)mai, hoatu
64. good pai
65. grass otaota
66. green kaakaariki
67. guts whekau
68a. hair (of body) huru(huru)
68b. hair (of head) makawe
69. hand ringa
70. he ia
71. head upoko, matenga, mahunga, pero (of animals)
72. hear rongo, whakarongo
73. heart ngakau
74. heavy taimaha
 Lexicon 581

75. here nei (see 2.1.2.5)

76a. hit (give blow to) meke, patu
76b. hit (collide) paoro
77. hold/take mau (=carry), pupuri (=hold), tango (= take)
78. horn probably no equivalent; tusks of pig are niho
79. how no simple equivalent, see 1.1.1.2.2.1
80. hunt aru
81. husband taane, hoa
82. I au ~ ahau ~ awau
83. ice haupapa
84. if me, mehemea, ki te (see 1.1.2.4.2.5)
85. in prep roto prep (see 2.1.1.5.1, 2.1.1.5.3)
86. kill patu
87. knee turi
88. know moohio, maatau
89. lake roto
90. laugh kata
91. leaf rau
92. leftside maui(i)
93. leg waewae
94. lie (ie. be prone) takoto
95a. live (be alive) ora
95b. live (dwell) noho
96. liver ate
97. long roa
98. louse (of head) kutu
99. man/male taane
100. many maha, tini
101. meat/flesh kiko
102. moon marama
103. mother whaea, matua wahine, kookaa
104. mountain maunga, puke
 Maori 582

105. mouth waha, maangai

106a. name (n.) ingoa
106b. name (v.) tapa, whakaingoa, whakahua
107. narrow whaaiti, kuuiti
108. near tata
109. neck kakii
110. new ho(o)u
111. night poo
112. nose ihu
113. not kaahore, kaua, eehara (see 1.4)
114. old taawhito (not for people)
115. one (ko)tahi
116. other atu (see 2.1.6.6)
117. person tangata
118. play taakaro
119. pull kukume
120. push pei
121. rain ua
122. red whero
123. right/ correct tika
124. rightside matau
125. river awa
126. road ara, huarahi
127. root pu(u)take
128. rope taura
129. rotten pirau
130 round (circular) porotaka, porohita
131. rub muku, miri(miri)
132. salt maataitai, tote (Eng.)
133. sand onepuu
134. say koorero, kii, mea
135. scratch raku(raku)
136. sea moana, tai
 Lexicon 583

137. see kite

138. seed kaakano
139. sew tuitui
140. sharp koi
141 short poto
142. sing waiata
143. sit noho
144. skin kiri, hiako
145. sky rangi
146. sleep moe
147. small iti, nohinohi
148a. smell (tr. vb) hongi, rongo
148b. smell (n) haunga, kakara
149. smoke auahi, paoa
150. smooth maene
151. snake ngarara
152. snow hukapapa
153. some he, eetahi (see 1.2.5.2.4, 2.1.1.10.2)
154. spit (vb) tuha
155. split koara, wehe
156. squeeze kuteetee, korotee, nanati (=throttle)
157. stab/pierce wero
158. stand tuu
159. star whetuu
160a. stick (of wood) raakau, manga
160b. stick (adhere) raparapa, piri
161. stone koo(w)hatu, toka
162. straight tika
163. suck momi
164. sun raa
165. swell pupuhi
166. swim kau(kau)
 Maori 584

167. tail whiore (animal), hiku (fish, reptile), kurutou (bird, animal)
168. that teeraa, teenaa (see 2.1.2.5.2)
169. there raa, naa (see 2.1.2.5)
170. they raaua (dual), raatou (pl)
171. thick maatotoru, matatengi
172. thin rahirahi
173. think whakaaro
174. this teenei
175. thou koe
176. three toru
177. throw panga, titiri, opa
178. tie here
179. tongue arero
180. tooth niho
181. tree raakau
182. turn huri, tahuri (upside down)
183. two rua
184. vomit ruaki
185. walk haere(ere)
186. warm mahana
187. wash horoi
188. water wai
189. we taaua, taatou, maaua, maatou (see 2.1.2.1.10)
190. wet maakuu(kuu)
191. what aha
192. when hea
193. where hea
194. white maa
195. who wai
196. wide whaanui
197. wife wahine, hoa
198. wind (weather) hau
 Lexicon 585

199. wing paakau

200. wipe muku(muku)
201. with me, ki (see 2.1.1.4.3, 2.1.1.4.4)
202. woman wahine
203. woods (bush) ngahere
204. worm toke, noke
205. ye koorua, koutou
206. year tau
207. yellow koowhai
 TEXTUAL SOURCES OF DATA

Many of the texts used have appeared in several places. In most instances, the versions I
have used were reproduced by the (Anthropology and) Maori Department at Victoria
University of Wellington for the use of students. My page references are to these
reproductions. In some cases, the reproductions keep the page numbers of the originals.
Where I am aware of the existence of other reproductions, I have included that
information in the references below. The abbreviations used for referring to the sources in
the text of the book are on the left in bold.
H ‘Hinemoa’
(n.d.) Recorded by Rev. H.J.Fletcher, Whitcombe and Tombs, Christchurch
HHM He hokinga mahara
Hēmi Pōtatau (1991), Longman Paul, Auckland NZ
KH ‘Kupe raaua ko Hoturapa’
Himiona Kaamira. Refs to 1 below. From 4 below, p.16
KM ‘Te Kimi a Maaui’
Refs to 1 below. From 3 below, p.2
KP ‘Te Koorero moo Pootaka-Tawhiti’
Paora Temuera. In 6 below, pp.40–44
KW ‘He Koorero moo ngaa Waka’
Recorded by Richard Benton. Refs to 1 below. From 4 below, p.20
KWh ‘He Koorero moo Whakaari’
Refs to 1 below. From 3 below, p.14
P ‘Te Tipuna nei a Pou’
Refs to 1 below. From 3 below, p.5
PT Paipera Tapu
(1977) The Bible Society in New Zealand, Wellington
R ‘Rona’
In 6 below, pp.19–20
RR Te Reo Rangatira: a course in Maaori for sixth and seventh forms
T.S. Karetu (1974), A.R. Shearer, Government Printer, Wellington
TA ‘Te Aohuruhuru’
Refs to 2 below. Also in Grey, Sir George (1928) Ngaa Mahi a Ngaa Tupuna Thomas
Avery & Sons Ltd, New Plymouth
TP ‘Tutae-poroporo’
Wiremu Kauika. Refs to 2 below. From JPS, XIII, No. 50, 1904, pp. 89–98
TR 1/2 Te Rangatahi 1 & 2
 Textual sources 587

H.R.Waititi (1962, 1969), Government Printer, Wellington

TWh ‘Tuuwhakairiora’
Mohi Tuurei. Reproduced by Dept of Anthropology & Maori, Victoria University of
Wellington from JPS, Vol. XX, 1895
W ‘Wairangi’
Refs to 2 below. From JPS, 19, 197–205; and Te Whare Kura, 1961, p.3. Also in 6 below
WhWh ‘Whare Whakairo’
Katarina Mataira. Refs to 1 below. From 5 below, p.4

1 ‘Maori Literature: Te Whare Kura Series’. Mimeo. VUW Dept of Anthropology &
Maori
2 ‘Selections of Oral and Written Prose: Reprints’. Mimeo, VUW Dept of Anthropology
& Maori
3 He Pakiwaitara, Te Whare Kura, 2. 1960. Government Printer, Wellington
4 He Kohikohinga, Te Whare Kura, 4. 1962. Government Printer, Wellington
5 He Whakaraapopototanga, Te Whare Kura, 11. 1965. Government Printer, Wellington
6 Biggs, B., P. Hohepa & S. Mead (eds.) (1967) Selected Readings in Maori, A.H. &
A.W. Reed, Wellington
7 Te Taura Whiri i te Reo Māori (1990) Maori for the Office, Te Taura Whiri i te Reo
Māori, Wellington
8 He Muka. He pitopito kōrero nā te Taura Whiri i te Reo Māori. (The newsletter from
the Maori Language Commission.) Te Taura Whiri i te Reo Māori, Wellington
 BIBLIOGRAPHY

Bauer, W.A. (1981a) ‘Aspects of the Grammar of Maori’, unpublished PhD thesis, University of
Edinburgh
______ (1981b) ‘Hae.re vs. ha.e.re: a note’ Te Reo, 24, 31–36
______ (1982) ‘Relativization in Maori’ Studies in Language, 6, 305–342
______ (1983) ‘On the category ‘Direct Object’ in Maori’, unpublished paper presented to the
Linguistic Society of N.Z. conference, Auckland, 1983.
______ (1984) ‘Experience Verbs in Maori’ Te Reo, 26, 3–28
______ (1991) ‘Maori ko again’ Te Reo, 34, 3–14
Benton, R.A. (1979a) The Maori Language in the Nineteen Seventies, NZCER, Wellington, N.Z.
______ (1979b) ‘Who speaks Maori in New Zealand?’ SET: Research Information for Teachers,
No. 1, NZCER, Wellington, N.Z.
______ (1981) The Flight of the Amokura: Oceanic Languages and Formal Education in the South
Pacific, NZCER, Wellington, N.Z.
______ (1984) ‘Bilingual education and the survival of the Maori language’ Journal of the
Polynesian Society, 93, 247–266
Biggs, B. (1955) ‘The compound possessives in Maori’ Journal of the Polynesian Society, 64, 342–
347
______ (1960) ‘Morphology-syntax in a Polynesian language’ Journal of the Polynesian Society,
69, 376–379
______ (1961) ‘The structure of New Zealand Maaori’ Anthropological Linguistics, vol. 3, no. 3,
1–54
______ (1966) English-Maori Dictionary, A.H. & A.W.Reed, Wellington, N.Z.
______ (1968) ‘The Maori language past and present’ in Schwimmer, E. (ed.) (1968) The Maori
People in the Nineteensixties: a Symposium, Longman Paul, Auckland, N.Z.
______ (1969) Let’s Learn Maori: a Guide to the Study of the Maori Language, A.H. & A.W.
Reed, Wellington, N.Z., Revised edn. 1973
______ (1971) ‘The languages of Polynesia’ in Sebeok, T.A. (ed) (1971) Current Trends in
Linguistics Vol 8: Linguistics in Oceania, Mouton, The Hague
______ (1989) ‘Towards the study of Maori dialects’ in Harlow & Hooper (eds) (1989) Part I
Campbell, C. & M.C. Muntzel (1989) ‘The structural consequences of language death’ in Dorian,
N.C. (ed.) (1989)
Catford, J.C (1977) Fundamental Problems in Phonetics, Edinburgh University Press, Edinburgh
Chafe, W.L. (1976) ‘Givenness, contrastiveness, definiteness, subjects, topics, and point of view’ in
Li, C.N. (ed.) (1976) Subject and Topic, Academic Press, New York
Chapin, P.G. (1974) ‘Proto-Polynesian *ai’ Journal of the Polynesian Society, 83, 259–307
Chung, S. (1970) ‘Negative Verbs in Polynesian’, unpublished ms, Harvard University
______ (1973) ‘The syntax of nominalizations in Polynesian’ Oceanic Linguistics, 12, 641–686
______ (1978) Case Marking and Grammatical Relations in Polynesian, University of Texas Press,
Austin and London
Chung, S. & W.J.Seiter (1980) ‘The history of raising and relativization in Polynesian’ Language,
56, 622–638
Clark, J & C.Yallop (1990) An Introduction to Phonetics and Phonology, Basil Blackwell, Oxford
Clark, R. (1973) ‘Passive and surface subject in Maori’, unpublished paper presented at the LSA
Annual Meeting, San Diego, December 1973.
 Bibliography 589

______ (1976) Aspects of Proto-Polynesian Syntax, Te Reo Monograph, Linguistic Society of New
Zealand, Auckland, N.Z.
______ (1981) ‘Inside and outside Polynesian nominalizations’ in Hollyman & Pawley (eds)
(1981)
______ (1983) ‘Does Maori have parts of speech?’ unpublished paper presented to the Linguistic
Society of NZ conference, Auckland, 1983
Comrie, B. (1976) Aspect: an Introduction to the Study of Verbal Aspect and Related Problems,
Cambridge University Press, Cambridge
______ (1981) Language Universals and Linguistic Typology: Syntax and Morphology, Basil
Blackwell, Oxford
______ (1985) Tense, Cambridge University Press, Cambridge
Dahl, Ö. (1985) Tense and Aspect Systems, Basil Blackwell, Oxford
Dorian, N.C. (ed.) (1989) Investigating Obsolescence: Studies in Language Contraction and
Death: Studies in the Social and Cultural Foundations of Language vol. 7 Cambridge
University Press, Cambridge
Firbas, J. (1959) ‘Thoughts on the communicative function of the verb in English, German and
Czech’ Brno Studies in English, 1, 39–63
Foster, J. (1987) He Whakamarama: a New Course in Maori, Heinemann, Auckland, N.Z.
Hale, K. (1968) ‘Review of P.W. Hohepa (1967)’ Journal of the Polynesian Society, 77, 83–99
Harawira, K.T. (1950) Teach Yourself Maori, A.H. & A.W. Reed, Wellington, N.Z.
Harlow, R.B. (1979) ‘Regional variation in Maori’ N.Z. Journal of Archaeology, vol. 1, 123–138
______ (1982) ‘Discourse and “movement rules” in Maori’ unpublished paper presented to the 4th
NZ Linguistics Society Conference, 1982
______ (1986) ‘The Actor-Emphatic construction of the Eastern Polynesian languages’ in
Geraghty, P., L. Carrington, S.A. Wurm (eds) (1986) FOCAL 1: Papers from the Fourth
International Conference on Austronesian Linguistics. Pacific Linguistics, Series C No 93,
A.N.U., Canberra
______ (1989) ‘Ka: the Maori injunctive’ in Harlow & Hooper (eds) (1989) Part I
______ (1991) ‘Consonant dissimilation in Maori’ in Blust, R. (ed.) (1991) Currents in Pacific
Linguistics: Papers on Austronesian Languages and Ethnolinguistics in Honour of George W.
Grace. Pacific Linguistics C-117. A.N.U., Canberra
Harlow, R.B. & R. Hooper (eds) (1989) VICAL I: Oceanic Languages. Papers from the Fifth
International Conference on Austronesian Linguistics, Auckland, New Zealand, January 1988.
Linguistic Society of New Zealand, Auckland
Head, L. (1989) Making Maori Sentences, Longman Paul, Auckland, N.Z.
Hohepa, P.W. (1967) A profile-generative Grammar of Maori Indiana University Publications in
Anthropology and Linguistics, Memoir 20, IJAL, Waverley, Baltimore
______ (1969a) ‘Not in English and kore and eehara in Maori’ Te Reo, 12, 1–34
______ (1969b) ‘The accusative-to-ergative process in Polynesian languages’ Journal of the
Polynesian Society, 78, 295–329
______ (1970) ‘Embedding, deletion and gapping in Maori narrative texts’ in University of Hawaii
Working Papers in Linguistics, vol. 1, no. 2, 1–37
______ (1981) ‘A look at Maori narrative structure’ in Hollyman & Pawley (eds) (1981)
Hollyman, J. & A. Pawley (eds) (1981) Studies in Pacific Languages and Cultures in Honour of
Bruce Biggs Linguistic Society of New Zealand, Auckland, N.Z.
Hooper, R. (1982) ‘Neuter verbs and stative aspect in Polynesian’, unpublished MA thesis,
University of Auckland
______ (1984a) ‘An unusual sentence type: complements of verbs of completion in some
Polynesian languages’ Te Reo, 27, 3–28
______ (1984b) ‘Neuter verbs, stative aspect, and the expression of agency in Polynesian’ Journal
of the Polynesian Society, 93, 39–70
______ (1988) ‘Focus constructions in Tokelauan’ Te Reo, 31, 51–71
 Bibliography 590

Hopper, P.J. & S.A. Thompson (1980) Transitivity in grammar and discourse’ Language, 56, 251–
299
Johansen, J.P. (1948) Character and Structure of the Action in Maori, Danske Videnskabernes
Selskab, Historisk-filologiske Meddelelser, vol. 31, no. 5, Munksgaard, Copenhagen
Joos, M. (1968) The English Verb: Form and Meanings, University of Wisconsin Press, Madison,
Milwaukee & London
Kearns, K. (1990) ‘A note on the glottal fricative in Maori’ Te Reo, 33, 65–81
Klavans, J.L. (1982) Some Problems in a Theory of Clitics, Indiana University Linguistics Club,
Bloomington, Indiana
Krupa, V. (1964) ‘On the category of possession in Maori’ Bulletin of the School of Oriental and
African Studies, 27, 433–435
______ (1966) Morpheme and Word in Maori, Mouton, The Hague
______ (1968) The Maori Language, Nauka Publishing House, Moscow
______ (1981) ‘On word order in Maaori and other languages’ in Hollyman & Pawley (eds)(1981)
Lynch, J. (1973) ‘Verbal aspects of possession in Melanesian languages’ Oceanic Linguistics, 12,
69–102
Mark, A. (1970) ‘The use of ki and i in New Zealand Maaori’, unpublished paper, MIT
Maunsell, R. (1842) Grammar of the New Zealand Language, W.C. Wilson, Auckland, N.Z.
McKenzie, D.F. (1985) Oral Culture, Literacy & Print in Early New Zealand: the Treaty of
Waitangi, Victoria University Press with the Alexander Turnbull Library Endowment Trust,
Wellington, N.Z.
Mithun, M. (1984) ‘The evolution of noun incorporation’ Language, 60, 847–894
Mutu-Grigg, M. (1982) ‘The manner particles rawa, tonu, noa, kee and kau in Maori’, unpublished
MPhil thesis, University of Auckland
Mutu, M. (1989) ‘An overview of theoretical orientations to Polynesian syntax’ in Harlow &
Hooper (eds)(1989) Part II
Orsman, H.W. (ed.) (1989) Heinemann N.Z. Dictionary, Heinemann, Auckland, N.Z., 2nd edn.
Polinsky, M. (1991) ‘Maori he revisited’ Paper presented to the 6th International Conference on
Austronesian Linguistics, Honolulu, May 1991. Extended version to appear in Oceanic
Linguistics
Prince, E.F. (1978) ‘A comparison of WH-clefts and it-clefts in discourse’ Language, 54, 883–906
Reedy, T.M. (1979) ‘Complex sentence formation in Maori’, unpublished PhD thesis, University of
Hawaii
Sanders, G. (1990) ‘On the analysis and implications of Maori verb alternations’ Lingua, 80, 455–
502
(1991) ‘Levelling and reanalysis in the history of Polynesian passive formations’ Journal of the
Polynesian Society, 100, 71–90
Schütz, A.J. (1985) ‘Accent and accent units in Māori: the evidence from English borrowings’
Journal of the Polynesian Society, 94, 5–26
Smyth, P. (1939) Te Reo Maori: a Guide to the Study of the Maori Language, Whitcombe &
Tombs, Christchurch, N.Z. Page refs to the 6th edn., 1957
(1946) Maori Pronunciation and the Evolution of Written Maori, Whitcombe & Tombs,
Christchurch, N.Z.
Stowell, H.M. (1911) Maori-English Tutor and Vade Mecum Whitcombe & Tombs, Christchurch,
N.Z.
Waite, J. (1987) ‘Negatives in Maori: a lexical-functional approach’ Te Reo, 30, 79–100
(1989) ‘Tough- and pretty-movement in Maori’ Te Reo, 32, 61–94
(1990) ‘Another look at the Actor-Emphatic’ Journal of the Polynesian Society, 99, 395–413
Williams, H.W. (1917) A Dictionary of the Maori Language, Government Printer, Wellington, N.Z.
Refs to the 7th edn., 1971
 Bibliography 591

Williams, W.L. & H.W. (1862) First Lessons in Maori, Whitcombe & Tombs, Auckland, N.Z.
Refs. to the 13th edn., 1965, revised by W.T. Ngata
Wills, W.H. (1960) Lessons in the Maori Language, Thomas Avery & Sons, New Plymouth, N.Z.
 INDEX

a as temporal preposition: 338–352, 488

-a-: 118, 521–522, 524, 576
A/O
(see also oo): 117, 197–198, 203–204, 205–207, 209–215, 278–282, 375, 376–377, 489, 491
actor-emphatic: 8, 10, 11, 14, 15, 16–17, 25, 31, 37, 38, 52, 55, 57, 62, 63, 73, 74, 75, 85, 87, 91,
135, 157, 160, 162, 165, 166, 176–180, 220, 223–229, 230, 233, 243, 252, 271, 277, 284, 292, 303,
398, 408– 409, 412, 424, 425, 426, 432, 434, 436, 443
addressee: 31, 32, 33, 35, 36, 143, 307, 467
adjective/adjectival predicate: 32, 80, 95–99, 107, 119, 137–138, 230, 233, 241, 259–260, 266, 275,
276, 278, 286, 287, 288, 289, 291, 409, 413, 430, 448, 450, 459, 480–484, 505, 515, 516, 518
adverb(ial): 231, 251, 398, 503–506, 516;
of manner: 48,49, 65, 92, 93, 95, 290–291, 418–419, 478– 479;
of place: 13, 14, 16, 18, 19, 25, 44, 45, 90, 219, 252;
of time: 3, 13, 16, 25, 44, 60–65, 89, 90, 92, 122, 219, 227, 338–352, 468, 477, 517, 574
adverbial clause: 13, 20, 59–74, 133, 160–162
adverbial particle: 92, 95, 114, 133, 512
agent
(see also passive agent): 276– 277, 413
aha: 5, 12
ahakoa: 57, 72, 304–305, 373
ai: 10, 11, 14, 67, 69, 72, 73, 89, 90, 92, 123, 224, 227, 229, 252, 253, 292, 339, 344, 418, 454,
496;
as habitual marker: 64, 445–446;
in purpose clauses: 65–66, 74;
in relative clauses: 50, 53–55, 56, 57, 58, 392;
in time clauses: 62, 63
ake
(see also directional particle): 189, 319–320, 321, 477–478, 488
ana: 60, 63, 416–419, 448
anake: 152, 156, 165–185, 186–188, 369, 375, 418, 503
anaphora
(see also (t)aua and reira): 124, 148–165
anaphoric determiner
(see also (t)aua): 110, 152–153, 158, 159
anaphoric pronoun: 152
animacy: 9, 242, 244, 247, 249, 271, 365
anoo: 152, 156, 164, 165–185, 186– 188, 305, 375
apposition: 42, 117–118, 153
article
(see also determiner): 109– 110, 254, 263, 307, 308–309, 506
atu
(see also directional particle): 24, 189, 192, 305, 329, 475–476
aspiration: 530
 Index 593

benefactive/beneficiary: 271–272, 282–283, 295

body part: 33, 585–587
borrowing: 289, 339, 354, 398, 542– 543, 544, 554, 557, 564, 571, 580, 584
breathy voice: 37, 38

Canga
(see also nominalization): 48, 49, 214–215, 253–254, 258, 278, 296, 397, 469, 499, 512–514;
Canga agreement: 48, 258
canonical transitive verb
(see also transitive verb/sentence): 8, 32, 55, 86, 214, 215, 223, 228, 230, 265, 267, 270–271,
279, 280, 456, 478
cataphoric reference: 154, 156
cause: 67, 291–294, 409–412
causer with neuter verb: 10, 25, 88, 90, 229, 276, 294, 413
classifying sentence
(see also he): 6, 136, 143–144, 287, 430, 481
clefting
(see also ko): 25, 220–221, 367
clitic: 52, 199, 202, 208, 216, 223, 256, 264, 367, 371, 506–510
cognate object: 266, 407
comitative
(see also me): 88,128, 130–132, 285–286, 410
communicative dynamism: 248
comparative/comparison: 96, 100, 114–115, 119, 189–192, 260, 273, 471, 484–485, 504
complementation: 39–49, 87
complementizer: 165, 247
compound: 59, 101, 116, 118, 147, 299, 408, 480, 511, 514, 516, 518–525, 574, 576
compound possessive
preposition see possessive
preposition
coordination: 23–4, 60, 61, 64,121– 139, 156–159, 233, 374, 451, 480, 484

default article: 110, 420

definite
(see also determiner and te): 113, 154, 355–361
definitive
(see also determiner): 263
deictic particle
(see also nei, naa, raa): 50, 53–55, 56, 57, 58, 81, 92, 114, 119, 152, 153, 219, 229, 257, 360,
362, 369–370, 381–388, 392, 393–395, 418
deletion: 42, 43, 55–56, 79, 80, 85, 134–139, 149–150, 156, 160, 162, 404, 409, 470;
of consonants: 526–528, 564, 567, 571–573
demonstrative
(see also deictic particle): 111–112, 151, 153–154, 156, 158, 159, 257, 353, 381–388
determiner: 57, 93, 106, 108, 109–112, 120, 137, 153, 155, 190, 198, 200, 202, 216, 233, 254–255,
263, 266, 274, 345, 353, 355–363, 380, 478, 495, 535, 536, 543, 559, 573
devoicing: 556
dialectal variation see variation
direct object: 8–10, 15, 17, 18, 19, 25, 26, 32, 45, 53, 54, 55, 57, 86, 89, 90, 91, 117, 135, 147, 156,
171– 173, 220, 225, 230, 232, 241–242, 245, 248, 267–271, 273, 396, 400, 402, 407, 478, 514, 521
 Index 594

DO clause: 13, 15, 17, 39, 40, 45, 424, 436

direct speech see (in)direct speech directional particle
(see also mai, atu, ake, iho): 30, 92, 418–419, 470–478, 485
distributiveness: 482–483
drawl: 216–217, 230, 231, 232, 577
dual: 368, 369, 371, 374, 375
durative: 61, 444

e
(see also passive agent); as T/A marker: 12, 14, 35, 72, 139, 142, 143, 224, 415–416, 419–420,
421, 422, 430;
in imperatives: 30–31;
in vocatives: 30, 306–307, 489;
with numerals: 31, 83–84, 496
e… .ana: 61, 66, 86, 92, 139, 252, 340, 414–419, 421, 426, 444, 447–448, 449, 456
Eastern dialect: 445, 565–566
Eastern Polynesian: 225
East-West dialect split
(see also Eastern dialect and Western dialect): 421, 426
echo question: 8, 9, 10, 13, 14, 15, 16, 19, 20, 21, 26–28
eehara: 29, 139, 143–145, 224
ellipsis: 12, 61, 74, 90
emphasis: 216–235;
emphatic particle: 92, 114, 185, 219, 277;
emphatic pronunciation: 37, 38, 230, 367, 555, 562
English, interference from: xxii, xxiii, 129, 131, 284, 289, 295, 343, 530, 532, 533, 538, 539, 541,
542, 560, 582;
borrowing from see borrowing
epenthesis: 544, 554, 573
equative
(see also ko): 5, 78–79, 115, 137, 143–144, 193–196, 221, 273, 431
ergative: 10, 34, 224, 413
exclamatory NP: 80
exclusive: 256, 368, 371
existence: 78
exophoric reference: 152, 154
experience verb: 7, 9, 32, 41, 53, 54, 86–87, 155, 214, 215, 259, 267, 269–271, 280, 282, 424, 447,
450, 456
extraposition: 225, 244, 246

finite see non-finite focus: 229;

in questions: 3, 17, 234–235
fronting: 6, 7, 8,11,14,16,19, 24, 25, 26, 87–88, 221, 224, 230, 237, 238, 251–253, 339
future: 58, 63, 68, 69, 75, 430–434

generic: 110–111, 144, 365, 445

genitive
(see also possessive): 52, 53, 56, 58, 108, 184, 225;
local genitive
(see also local noun): 54
goal
 Index 595

(see also ki): 88, 91,268–269, 271–272, 311–312, 315

habitual: 59, 63–65, 444–447

Hawaiian: 57
he as determiner: 110–111, 137, 260, 356–363, 408, 493, 499–501;
in classifying predicates: 6, 78–79, 80, 83, 274, 290, 430, 481–482, 493–494
heavy NP: 120, 241–247
hei: 58, 78, 109, 274–275, 279, 295–296, 297, 298, 430, 489, 493–494;
in nominalizations: 19, 49, 63, 280, 294–295, 469
human: 52, 58, 129, 149, 152, 155, 156, 291, 514

i: 103, 259, 265, 489, 492;

as cause marker
(see also causer): 276– 277, 291–293;
as comparative
marker: 70, 189–192, 273;
as DO marker: 86–87, 267–271;
as equative marker: 193, 273;
as locative preposition: 82, 93, 298, 309–338;
as source marker: 283, 312;
as T/A marker: 14, 139, 224, 414, 423, 425–428, 436, 442–444;
as temporal preposition: 60, 338–352;
in nominalizations: 296
iho
(see also directional particle): 189, 320, 477–478
i te see pseudo-verbal continuous
imperative: 30–37, 38, 47, 72, 85, 86, 87, 123–124, 142–143, 400, 432, 458, 463, 465, 467, 559–
560
inaa: 63, 69, 419, 458
inchoative: 414, 432, 440
inclusive: 256, 368, 371, 374
indefinite subject fronting: 7, 16, 25, 79, 89
indirect object: 10, 52, 53, 54, 56, 87, 155, 166, 168, 173, 250, 271–273, 402–403
(in)direct speech: 1, 44, 46, 47, 241–242, 271, 392
instrument: 88, 283–285, 521
intensifying adverb
(see also adverbial particle and adverbial of manner): 92, 102, 461, 485, 486, 487
interrogative/question: 2–28, 44, 46, 57, 87, 123, 270, 388–392, 560;
question-word: 5, 257, 388, 391;
time question: 10, 11, 16, 17, 25;
why-question: 12
intonation: xxiii, 38, 39, 50, 97, 119, 121, 123, 125, 217, 559–563;
in exclamations: 37, 38;
in imperatives: 30, 31, 32;
in questions: 1, 2, 3, 4, 5, 14, 25, 26, 27, 28, 32
intransitive verb/sentence: 7, 30, 32, 43, 44, 57, 58, 66, 74, 85–87, 95, 134, 214, 215, 223, 225,
226–227, 250, 260, 266, 278, 292, 313, 404, 407, 408, 409–410, 456, 467, 515, 524

ka: 61, 62, 63, 64, 66, 69, 139, 141, 192, 259, 260, 340, 414, 419–420, 422, 426, 433, 435, 441–
442, 450, 451, 495, 535
kaahore: 29, 139, 140–141, 146, 217, 429
 Index 596

kaatahi…ka: 62, 250, 440

kai-: 514–515, 516, 523, 524–525, 574
katoa: 112, 137, 300, 418, 499, 502–503
kaua: 35, 142–143, 461
kee: 129, 189, 406, 428
kei: 136, 341, 489;
as mood marker: 40, 458, 464, 465;
in negative imperatives: 37
kei te see pseudo-verbal continuous
ki: 489, 491;
as comparative marker: 189;
as DO marker: 86–87, 267–271;
as equative marker: 193, 273;
as goal marker: 311–312;
as IO marker: 87;
as instrumental marker: 283–285, 290;
as locative marker: 311;
as purpose marker: 294–295;
with adjectives: 97
ki te: 18, 39, 104, 247, 255, 278, 459, 460, 466, 469, 470;
in adverb clauses: 65, 66, 68, 73, 458;
in noun clauses: 41, 43–44, 47–48, 159
kia: 17, 32, 39, 62, 63, 70, 141, 143, 194, 247, 425, 458, 459–460, 464–465;
in noun clauses: 41–43, 47, 159, 466;
in purpose clauses: 65–66
kiihai: 139, 145–146
ko: 490;
as specifier in coordination: 23, 127–128, 374;
in equatives: 5, 57, 78–79, 83, 113, 221, 253, 274, 308;
for subject fronting: 6, 7, 8, 15, 16, 25, 79, 89, 91, 135, 165, 193, 220–221, 228, 233;
for topic switch: 127, 220, 236–241
kore: 72, 139, 141–142, 146, 147–148, 288, 460, 465
kua: 139, 141, 259, 260, 340, 405–406, 414, 428–429, 430, 432, 433, 437–442, 450, 487

labial(ization): 397, 543, 554, 563, 564, 566, 569–571

language death/decay
(see also semi-speaker): xx, 212, 511, 580
left dislocation: 222, 236
lento: 37, 38
light constituent: 241, 243, 248–250
listing: 109, 254–255, 355–356
loan word see borrowing
local noun: 52, 81, 93, 103–104, 106, 109, 155, 255, 257, 258, 260, 262– 263, 298, 310, 313, 355,
390, 492, 494, 517
location: 197, 204–205;
locative preposition/phrase: 82, 140, 144, 226, 309–352, 403, 414, 424, 426, 431, 435;
locative proform: 393–395
loudness: 217, 530, 556, 557

maa: 272, 277, 282–283, 284, 285, 291, 318, 319, 328, 334, 335, 490;
in hei nominalizations: 19, 49, 280
 Index 597

maha: 501–502
mai
(see also directional particle): 327, 349, 419, 474
manner see adverbial of manner
me: 490;
as comitative: 23, 127, 128, 129, 285–286, 305;
as equative particle: 194;
as T/A/M marker: 18, 40, 72, 224, 276, 398, 405–406, 458, 458, 460, 461;
as weak imperative: 34, 460;
for similitude: 309, 493
me te mea: 65
mea
see also pseudo-clefting: 372, 410–412
mehemea: 46, 67–69, 458
Melanesian: 204
modifier in NP: 21, 24, 167, 168, 188, 260, 265, 266, 288, 289, 313, 380, 518–519
moo: 272, 282–283, 295, 296–297, 302–303, 347, 490;
in hei nominalizations: 49, 279, 280
mood: 434, 458–466
mora: 30, 31, 33, 306–307, 452–453, 458, 482–483, 507, 525–528, 536, 544–545, 548, 554, 555,
571, 572, 575
mua: 317–318, 350, 352, 494
muri: 318–319, 329, 350, 494

naa: 277, 291, 293, 324, 491

narrative: 149–150, 236–237, 240, 355, 383–385, 396, 399, 416–417, 421, 441–442, 471, 472
negation/negative: 4, 15–16, 29, 31, 35, 38, 40, 41, 74, 75, 78, 123, 139–148, 250, 252, 285, 286,
287, 290, 292, 302, 305, 349, 421, 424, 427, 431, 436, 438, 460, 461, 502, 506–507
nei: 10, 54, 108, 202, 349, 384, 446
neuter verb: 7, 10, 11, 25, 33, 41, 66, 85–86, 134, 214, 215, 227, 229, 249–250, 259, 265, 266, 276,
278, 291, 409, 413–414, 430, 447, 450, 456, 515
neutral possessive: 199, 279, 280, 282, 376–378
ngaa: 110–111, 359
Ngaitahu: 565
Ngapuhi: 529
Ngati Awa: 565
Ngati Porou: 2, 445
noa
(see also adverbial particle): 189, 419
nominal expression, preference for: 65, 67, 70, 71–72
nominalization
(see also Canga, hei and plain stem): 19, 48, 60, 62, 63, 64, 67, 70, 73, 85, 87, 117, 132, 133,
160, 182–183, 190–192, 195, 196, 214–215, 253–254, 277, 278–281, 293, 312, 397, 434–435,
466, 467, 470, 488, 512–514, 523
non-finite: 14, 47, 58, 61, 66, 73, 277, 434, 456, 466–470
non-past: 62, 68
non-verbal sentence: 3, 5, 12, 22, 26, 58, 62, 71, 78–84, 94, 95, 104, 136, 140, 166, 180–182, 197,
221, 222, 233, 252, 253, 259, 261, 267, 274, 290, 293, 303, 304, 309, 311, 407, 420, 424, 426, 444,
449;
non-verbal predicate: 311, 315, 316, 318, 319, 321, 322, 324, 333, 339, 348, 561
noo: 60–61,199, 227, 298, 312, 315, 317, 324, 338–352, 491
 Index 598

noun clause: 13, 39–49, 74, 75, 159, 424, 431

noun phrase: 21, 22, 106–120, 137, 184, 190–192, 195–196, 233
number see numeral
numeral: 7, 31, 83–84, 85, 113, 261–262, 299–300, 368–369, 494–499

oo: 298, 299, 341, 491

object incorporation: 86, 91, 117, 266, 270, 356, 358, 362, 364–365, 408, 454, 478–479, 514, 524
oblique NP: 10–12, 16, 17, 19, 52, 53, 87, 89, 90, 91, 97, 98, 129, 135, 271, 396, 404
orthography: xx, xxiv–xxv, 1, 199, 531, 532, 534–535

partitive: 298–302
part of speech: 254–263, 510–511, 519
passive: 10, 15, 25, 32, 35, 41, 55, 59, 66, 74, 75, 90, 91, 117, 133, 134, 143, 157, 214, 215, 220,
224, 227, 230, 248–249, 266, 269, 278, 293, 396–407, 409, 412, 413, 430, 435, 443, 456, 458, 467,
469, 470, 478–479, 488, 524, 551;
passive agent: 10, 28, 32, 33, 35, 58, 87, 90, 91, 135, 149, 174–175, 228, 232, 243, 248–249,
260, 276, 293, 366, 396, 404–405, 489;
passive agreement: 92, 257–258, 478– 479, 504;
passive suffix: 34, 87, 143, 254, 257, 262, 306, 396–398, 410, 517, 543, 565, 571, 574–575
past (time): 60–62, 68, 228, 399, 425–428, 441
patient: 248, 268–269
pause: 222
pee-: 193, 395
perfective: 400, 442–444
personal article (a): 106, 109–110, 254, 256, 263, 274, 355, 506, 508, 573, 577
personal noun
(see also proper name): 262, 355
personal pronoun: 52, 56, 109, 127, 144, 150–152, 155, 159, 165–185, 186–188, 202, 256, 262,
279, 366–371, 392, 506, 507, 508, 559
pitch: 2, 3, 4, 26, 27, 28, 30, 38, 216, 217, 557, 559
plain stem: 73–74, 116, 192, 195, 215, 254, 281–282, 469
plural noun: 106, 254, 352–353, 354, 483–484, 519, 565, 572
possessive: xxv, 21, 40, 79, 107–108, 114, 117, 151, 156, 165, 184, 193, 197–216, 239, 256, 287,
288, 353, 375–381, 389, 536
possessive preposition
(see also maa, moo, naa, noo, oo): 82, 90, 140, 145, 155, 208–209, 223, 251, 278–282, 283,
299, 366, 367, 371, 376, 379–380, 424, 426, 518
possessive-relative clause: 8, 9, 10, 13, 14, 17, 18, 19, 25, 26, 45, 53, 54, 57, 70, 190–192, 195–
196, 220
post-verbal particle
(see also adverbial particle and ai): 345, 346, 418–419
predicate: 6, 78, 83, 98, 126, 199, 223–227, 230, 233, 236, 244–246, 251, 303, 304
preposition: 162, 255, 260–261, 264, 309–352, 434, 488–494, 506, 508, 517
prepositional complement: 40, 139
prepositional phrase: 22, 23, 88, 93–94, 98, 101, 103–105, 114, 119, 127, 133, 138, 184, 200, 216,
233, 261
prepositional predicate: 6, 12, 81–83, 85, 136, 140, 143–145, 275
present tense/time: 421–425
productivity: 59, 86, 118, 453, 479, 511, 512, 514, 515, 516, 519, 521, 524, 527, 528, 570, 571, 572
pronominalization: 156–165, 249, 400
pronoun see personal pronoun
 Index 599

proper name: 31, 33, 51, 106, 108, 109, 117, 127, 255, 256, 274, 306, 360–361
Proto Polynesian: 146, 392
pseudo-clefting: 25, 221–222, 231, 367, 412
pseudo-verbal continuous: 61, 76, 103, 106, 139, 141, 255, 261, 414, 421, 425–426, 428, 444, 447–
448, 449, 468

quantifier: 261, 499–503;

float: 112, 243, 262, 300–302
question see interrogative

raa
(see also deictic particle): 54
raanei: 3, 4, 23, 125, 126, 130, 372–373, 503
raising: 15, 35, 247, 250–253
raro: 323–325, 335, 494
rawa: 190, 192, 443
reciprocal: 186–188, 375, 379, 407, 412, 456, 470
reduplication: 259, 327, 353–354, 398, 452–453, 482–484, 503, 513, 514, 516, 523, 525–528, 546–
547, 551–552, 554, 568–569, 571–572, 574, 575–576
reflexive/reflexivization: 87, 152, 155, 165–185, 186–188, 271, 375, 379, 400–401, 412, 456, 470
reira: 152, 257, 393
relative clause/relativization: 12, 13, 19, 21, 50–59, 70, 71, 74–77, 87, 107, 115, 184, 209, 220,
252, 270, 286, 288, 348, 380, 392–393, 404, 406, 425, 428, 432, 436–437, 466
rhythm: 31, 388
right dislocation: 223
rite: 193–196
roto: 314–315, 332, 335, 351, 352, 494
route: 318, 319, 324, 328, 336
runga: 319–323, 494

Samoic: 146
semi-speaker: xxii, xxiii, 210, 531, 539, 542, 546, 547, 558
simultaneity: 62
source: 91, 283, 291, 312
stative: 32, 58, 85, 139, 257, 406, 413, 450, 456, 459
stative intransitive verb: 74, 169, 259, 313, 406
stress: 33, 38, 216, 218–219, 222, 229, 236, 506–507, 508, 530, 555–558, 559, 562–563, 574–576;
in phrases: 576–577;
strong stress: 217, 230, 231, 232, 367, 562, -563
subject: 127, 128, 134, 149, 168–171, 216, 221, 230, 245, 260, 308, 466, 470, 473;
in imperatives: 32;
in nominalizations: 48, 49, 67, 73, 117, 278–282;
in non-verbal sentences: 5, 6, 8, 78, 81, 182, 267;
in subordinate clauses: 13–15, 17, 41, 43, 44, 45, 53, 54, 58, 65, 69, 74, 159, 232, 238, 239, 251;
in verbal sentences: 7, 8, 14, 24, 85, 248, 266–267, 396, 404, 413;
subjectless sentences: 78, 84–85, 226
subject clause: 39, 40, 140, 224, 252, 253, 424
subject fronting
(see also ko and indefinite subject fronting): 26, 69, 79, 80–81, 89, 90, 142, 164, 220, 222, 236,
238
subject raising: 26, 35, 40, 140, 141, 142, 143, 144, 145, 163, 224, 252
 Index 600

subjunctive: 32, 62, 194

subordinate clause: 13–21, 31, 39–77, 141, 159–164, 232, 239, 251, 253, 480;
tense-marking in: 422, 424–425, 427–428, 431–432, 436–437, 448, 449
superlative: 96, 100, 115, 119, 485–486, 504
suprasegmental: 216, 559, 574
syllable: 536, 543, 544–553, 554, 557, 559, 564, 569

taha: 324, 325–332, 337

T/A(/M) marker: 30, 31, 39, 48, 49, 50, 63, 66, 76, 89, 139, 142, 224, 251, 254, 258, 261, 263, 339,
344, 346, 399, 405–406, 414–437, 458, 467, 506, 543;
absence of: 61, 142, 276, 339, 418–419, 422–423, 426, 448, 467, 487
Taranaki(-Whanganui): 225, 445, 531, 532, 565
(t)aua: 110–111, 152–153
te: 110–111, 356–360, 535
Te Aupouri: 146, 445, 529, 532
tee: 146
teeraa: 237–238
(t)eetahi: 110–111, 255, 256, 300, 356–363, 499–501
tense: 1, 5, 8, 10, 11, 12, 14, 25, 54, 73, 74–77, 80, 81, 82, 207–208, 310, 414–437
tempo: 38
temporal preposition
(see also adverbial of time): 338–352
tini: 501–502
toko-: 7, 83, 355, 496–497, 574
tonu: 419, 423, 453
topic: 222, 229, 236–241, 277
topic switch see ko
transport, means of: 211, 285
transitive verb/sentence: 11, 13, 25, 32, 37, 43, 74, 75, 86, 117, 134, 157, 226, 249, 267, 399, 407,
411, 456, 458, 467, 514, 515, 525
Tuhoe: 565
type: 289

universal: 254, 257

unmarked NP
(see also subject): 88, 193, 223, 228, 303, 414

variation: 30, 31, 32, 89, 122, 142, 211–214, 268, 307, 315, 376, 397, 483, 496, 506, 531, 533, 535,
537–542, 546, 547, 559, 576;
by age: 2, 3, 7, 8, 10, 34, 53, 54, 129, 145, 218, 295, 305, 398, 483, 530, 537, 538, 540, 542,
581, 582;
by sex: 2, 37, 122, 537, 539, 540, 542, 561, 562, 587;
by region, dialect, or tribe: xxiii, 2, 35, 69, 79, 82, 103, 145, 198, 212, 225, 310, 350, 370, 371,
397, 431, 444, 446, 498, 529, 530, 557, 563, 565, 566, 574, 580
verb: 12–13, 15, 16, 17, 18, 95, 133, 135–136, 231, 257–259, 480;
verb phrase: 23, 32
vocative: 31, 32, 35, 306, 467
vowel length: xx, xxiv-xxv, 121, 397, 398, 507, 529–530, 534–538, 551, 555, 571
vowel sequence: 534, 535–537, 538–542, 544–553, 573
 Index 601

waenganui: 332–334, 494

waho: 316–317, 329, 494
weak imperative see me
Western dialect: 531
whaka-: 25, 86, 258, 260–261, 409–410, 412, 491, 511, 515–516, 523, 524, 547, 572, 574
word order: 2, 3, 22, 25, 92, 101, 116, 120, 165, 168, 193, 242, 265;
in non-verbal sentences: 78, 79, 80–81, 97, 244;
in subordinate clauses: 49, 51, 56, 59, 63, 65, 66, 67, 69, 72;
in verbal sentences: 89–91, 94, 99, 104, 248, 271