Burmese

A Comprehensive Grammar

Burmese: A Comprehensive Grammar is a complete reference guide to mod-

ern Burmese grammar. It presents a fresh and thorough description of the
language, concentrating on the real patterns of use in modern Burmese.
The volume is organized to promote a thorough understanding of Burmese
grammar. It offers a stimulating analysis of the complexities of the language,
with clear explanations. Throughout, the emphasis is on Burmese as used by
present-day native speakers.
Features include:

• detailed treatment of the common grammatical structures and parts

of speech
• particular attention to areas of confusion and difficulty
• all examples given in Burmese script, IPA phonetic transcription, and
English
• glossary of linguistic terminology
The Grammar is the ideal reference source for intermediate to advanced
learners and users of Burmese and will remain the standard reference work
for years to come.

Mathias Jenny is Senior Lecturer/Researcher in the Department of Compara-

tive Linguistics at the University of Zurich, Switzerland, with a research focus
on the languages of Southeast Asia. He teaches General Linguistics and begin-
ners’ Burmese.

San San Hnin Tun is Maître de Conférences at Inalco/Lacito-CNRS. She

teaches Burmese and researches the Burmese language, in particular discourse
structure and functions of spoken contemporary Burmese using a corpus lin-
guistic approach. She is the author of the recent Routledge publication Col-
loquial Burmese.
Routledge Comprehensive Grammars

Comprehensive Grammars are available for the following languages:

Bengali
Burmese
Cantonese
Chinese
Catalan
Danish
Dutch
Greek
Indonesian
Japanese
Kazakh
Modern Welsh
Modern Written Arabic
Slovene
Swedish
Turkish
Ukrainian
Burmese
A Comprehensive Grammar

Mathias Jenny and

San San Hnin Tun
First published 2016
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2016 Mathias Jenny and San San Hnin Tun
The right of Mathias Jenny and San San Hnin Tun to be identified as authors of
this work has been asserted by them in accordance with sections 77 and 78 of
the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or
utilised 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.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and explanation
without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Jenny, Mathias, and Hnin Tun, San San, authors.
Burmese : a comprehensive grammar / Mathias Jenny and San San Hnin Tun.
pages cm
Includes index.
1. Burmese language--Grammar. I. Mathias Jenny and San San Hnin Tun, authors.
II. Title.
PL3931.J46 2016
495’.85--dc23
2015026471
ISBN: 978-0-415-73568-1 (hbk)
ISBN: 978-0-415-73569-8 (pbk)
ISBN: 978-1-315-65119-4 (ebk)

Typeset in Sabon and Gill Sans

by Bookbright Media
 Contents

Abbreviations xvii
Acknowledgments xxi

1 Introduction 1
1.1 Historical and social context 1
1.2 Myanmar and Burma: Literary and Colloquial
Burmese 2
1.3 Burmese today 3
1.3.1 The language and its speakers 3
1.3.2 Burmese globalized: Loanwords, code
mixing, and slang 4
1.4 Structure of Burmese: An overview 6
1.4.1 Sounds 6
1.4.2 Grammatical features 8
1.4.3 Role of context: Linguistic and extralinguistic 9
1.5 About this grammar 10

2 Sound system and script 13

2.1 Sound system 14
2.1.1 Consonants 14
Plain consonants 14
Consonant clusters 17
2.1.2 Vowels 18
Plain vowels 18
Vowels in closed syllables 19
2.1.3 Tones, voice quality, and prosody 21
Tones 21
v
 Stress and prosody 22
Contents
2.1.4 Syllable structure 23
2.1.5 Sandhi 24
Voicing 24
Assimilation 29
2.1.6 Regional and sociolectal variants 29
2.2 Script 31
2.2.1 Consonants 31
2.2.2 Ligatures – initial and medial 33
2.2.3 Vowels – free and bound 35
2.2.4 Rhymes 37
Vowels and final consonants 37
Representation of tones 39
2.2.5 Special symbols 40
2.2.6 Numerals 41
2.2.7 Spacing and punctuation 41
2.2.8 Finding words in dictionaries and spelling
words 42
Alphabetical order 42
Spelling out words 43
2.2.9 Chat script 43
2.2.10 Other methods of transcription 45

3 Parts of speech 49
3.1 Nominals 49
3.1.1 Nouns 50
3.1.2 Pronouns 52
3.1.3 Measure words 52
3.1.4 Classifiers 52
3.2 Verbs 53
3.2.1 Main verbs 53
3.2.2 Auxiliaries 54
3.2.3 Types of verbs 54
Intransitive verbs 54
Transitive verbs 55
Property verbs (adjectives) 55
3.3 Adverbs 56
3.3.1 Phrasal adverbs 56
3.3.2 Clausal adverbs 57
3.4 Markers and particles 57
vi
 3.4.1 Grammatical markers 57
Contents
Phrasal markers 57
Clausal markers 58
3.4.2 Pragmatic particles 59
Phrasal particles 59
Clausal particles 59

4 Lexicon 60
4.1 Pronouns 60
4.1.1 First person pronouns 61
4.1.2 Second person pronouns 62
4.1.3 Third person pronouns 62
4.1.4 Other pronouns 63
4.2 Kinship terms and personal names 65
4.2.1 Kinship terms 65
4.2.2 Personal names 68
4.3 Numerals 69
4.3.1 Cardinal numerals 69
4.3.2 Ordinal numerals 71
4.3.3 Fractions of whole numbers 72
4.3.4 Multipliers 72
4.3.5 Percentages and distributives 72
4.4 Classifiers and measure words 73
4.4.1 Classifiers 73
4.4.2 Measure words 76
4.5 Time and date 77
4.5.1 Units of time 77
4.5.2 Adverbs of time 79
4.5.3 Days, months, and years 80
4.6 Meteorological and natural phenomena 83
4.7 Feeling terms 85
4.8 Color terms 86
4.9 Specific vocabulary for different activities 87
4.10 Elaborate expressions 89
4.11 Ideophones 91
4.12 Loanwords 91
4.12.1 Pali/Sanskrit 92
4.12.2 English 92
4.12.3 Others 94
vii
 5 Word structure 96
Contents
5.1 Derivational affixes 96
5.1.1 The nominal prefix ʔə- 97
5.1.2 Nominal suffixes 100
Abstract nouns 100
People 102
Something to V 105
Modifying 108
Gender 110
5.1.3 Derivation by tone change 111
5.2 Compounding 112
5.2.1 Nouns 112
Modifying compounds 112
Additive compounds 116
Euphonic compounds 117
Exocentric compounds 117
5.2.2 Verbs 118
Verb-verb, not analyzable 118
Verb-verb, modifying 118
Verb-verb, additive 119
Verb-verb, euphonic 119
Noun-verb 119
5.3 Reduplication 121
5.3.1 Nominal reduplication 122
5.3.2 Measure words and classifiers 122
5.3.3 Verbal reduplication 123
Preverbal: adverbial 123
Postnominal: attributive 123
Special reduplication patterns 123
5.3.4 Adverbial reduplication 126

6 Phrase structure 127

6.1 Noun phrase 127
6.1.1 Plural markers 127
-twe/dwe ‘general plural’ 128
-tó/dó ‘associative plural’ 129
6.1.2 Classifier phrase 131
6.1.3 Quantifiers 133
-tàin/dàin ‘each, every’ 133
viii
 tə-N-lòun ‘the whole N’ 134
Contents
-səlòun/ -zəlòun, -lòun ‘all of’ 135
-pàun/bàun ‘sum, combined number, total’ 135
ʔəmyà.zòun ‘(at/the) most’, ʔənɛ̀.zòun ‘(at/
the) least’ 136
ʔəmyà.zú ‘most of, majority’, ʔənɛ̀.zú
‘fewest of, minority’ 136
təʨʰó ‘some’ 137
ʔà.lòun ‘all’ 138
tə-X-təle ‘some, a few’ 139
məká ‘more than, not as few as’ 139
6.1.4 Enumeration 140
Exhaustive lists 140
Representative lists 142
Non-specific parallel lists 143
6.1.5 Demonstratives 143
6.1.6 Interrogatives and indefinites 145
6.1.7 təʨʰà ‘other’, ‘another’ 146
6.1.8 Attributive modifiers 149
Verbal attributives 149
Nominal attributives 152
Adverbial attributives 153
6.1.9 Possessives 154
6.1.10 Postnominal markers 156
Case markers 156
Location markers 168
Derived postpositions 181
6.2 Verb phrase 185
6.2.1 Preverbal auxiliaries 185
la ‘come’ and θwà ‘go’ 186
pyan ‘return’ 186
tʰaʔ ‘pile up, impose, put on top’ 187
sʰɛʔ ‘connect, continue’ 188
laiʔ ‘follow, go around and’ 189
ɕauʔ ‘walk, go around and’ 190
sá ‘begin’ 190
ʨo/ʨo.tin ‘do in advance’ 191
ʨí ‘look’ 192
ku/ku.ɲi ‘help’ 193
wàin ‘surround’ 194
ix
 pàun ‘combine, add’ and kʰwɛ̀ ‘split, take
Contents
apart, separate’ 195
pè ‘give’ 196
6.2.2 Postverbal auxiliaries 197
-ʨʰin/ʥin ‘want to’ 197
-se/ze ‘let’ 198
-pʰù/bu ‘have had the experience, ever’ 199
-yá ‘get, must’ 200
-laiʔ ‘follow, just, for good, without
further consideration’ 201
-la ‘come’ 203
-θwà/ðwà ‘go’ 204
-ne ‘stay’ 205
-tʰà ‘keep’ 206
-pè ‘give’ 207
-θín/ðín ‘proper, suitable, appropriate,
should, ought’ 208
-yɛ̀ ‘dare’ 208
-taʔ/daʔ ‘able, skilled, know how’ 209
-nain ‘capable’ 209
-ʔà ‘free’ 210
-kʰàin ‘order’ 211
-pyan-bi ‘do again’ 212
yá ‘can, may’ 213
pì ‘finished’ 214
koun ‘all’ 215
6.2.3 Negation 216
6.2.4 Plural subject 217
6.2.5 Displacement in space and time 219
6.2.6 Importance, politeness, emphasis 220
6.2.7 Aspectual markers 221
-θè/ðè ‘yet’ 221
-ʔòun ‘further, more, again’ 222
-tɔ́/dɔ́ ‘change of situation’ 224
6.2.8 Status markers 225
-tɛ/dɛ ‘non-future, realis’ 225
-mɛ ‘future, irrealis’ 227
-pi/bi ‘new situation’ 229
-pʰù/bù ‘negation’ 231
-nɛ́ ‘prohibitive’ 232
x
 6.3 Adverbial phrase 233
Contents
6.3.1 V-V reduplication 234
6.3.2 N-lo, V-θə.lo/ðə.lo ‘as, like’ 236
6.3.3 N-nɛ́, V-ta/da-nɛ́ ‘with’ and N-mɛ́,
mə-V-pʰɛ̀/bɛ̀ ‘without’ 237
6.3.4 ʔə-V-ʨì/ʥì ‘very’ 239
6.3.5 -kʰənɛ̀/gənɛ̀ ‘abruptly’ 240

7 Clause and sentence structure 242

7.1 The makeup of clauses 242
7.1.1 Clauses with verbal predicates 242
7.1.2 Clauses with non-verbal predicates 243
Identity: “zero copula”, pʰyiʔ ‘to be’,
mə-houʔ-pʰù ‘it is not’ 243
Existential and possessive: ɕí ‘exist, there
is, have’ and pa ‘be included’ 246
Locative: ne ‘be at, stay, remain’ 248
7.2 Clause types 249
7.2.1 Independent clauses 249
Status markers 250
Constituent order 251
Non-finite independent clauses 253
7.2.2 Dependent clauses 255
Complement clauses 256
Nominalized clauses 256
[CLAUSE-FIN, NP] ló 258
[CLAUSE, XP]-hmàn 260
Relative clauses 261
Adverbial clauses 264
Negative 265
Conditional 265
Purposive 267
Causal 270
Sequential 271
Temporal 275
Concessive 281
Clauses with postpositions 284
7.3 Transitivity 286
7.3.1 Intransitive 286
7.3.2 Transitive 288
xi
 7.3.3 Ditransitive 290
Contents
7.4 Valency changing 291
7.4.1 Passive 291
7.4.2 Causative 293
Lexical 293
Periphrastic 293
V-se/ze 293
V-kʰàin 294
pè-V 295
[CLAUSE]-ʔaun louʔ 296
7.4.3 Benefactive 297
7.5 Participation 299
7.5.1 Solitary 299
7.5.2 Reciprocal 301
7.5.3 Associative 302
7.5.4 Distributive 306
7.6 Combining sentences 306

8 Form and functions of utterances 308

8.1 Statements 308
8.1.1 Plain statements 308
General statements 308
Emphatic statements 311
8.1.2 Reported speech 313
Self-report or general report 314
Second person report 314
Third person report 315
8.2 Questions 317
8.2.1 Polar questions 317
8.2.2 Content questions 325
Pronominal interrogatives 326
Interrogative modifiers 328
Interrogative quantifiers 329
Temporal interrogatives 331
Locative interrogatives 332
Other interrogatives 334
Indefinites 336
8.2.3 Alternative questions 339
8.2.4 Tag questions 340
8.2.5 Elliptical questions 340
xii
 8.3 Requests 341
Contents
8.3.1 Direct imperatives 342
Softening a command 342
Intensifying a command 345
8.3.2 Prohibitive 348
8.3.3 Adhortatives 351
8.3.4 Indirect imperatives/optatives 352

9 Negation 356
9.1 Negating verbs 356
9.1.1 Standard negation 356
Basic negation 356
Negation with modifier 358
Negation with incorporated nouns 359
Negation in colloquial language 360
Negation in dependent clauses 360
Negated complement and relative clauses 361
9.1.2 Aspectual/temporal negation 362
9.1.3 Emphatic negation 364
Intensifying adverbial 364
Negated quantifier/indefinite 365
9.2 Negating non-verbal elements 367
9.2.1 Negating nouns and adverbials 367
9.2.2 Negating clauses 367
9.3 Nominal and clausal negation 368
9.3.1 Nominal negation 368
9.3.2 Clausal negation 369
9.4 Indefinite negation 370

10 Expression of tense, aspect, and modality 371

10.1 Tense 371
10.1.1 Non-future/realis 371
10.1.2 Future/irrealis 373
10.2 Aspect 377
10.2.1 New state of affairs 377
10.2.2 Old state of affairs 380
Past and present reference 380
Future reference 382
10.2.3 Ongoing event 382
xiii
 10.2.4 Displacement in time and space 385
Contents
Activity directed towards the center of interest 385
Activity done at another place 386
Activity removed from present 387
10.2.5 Change of state and direction 387
la ‘come’ 387
θwà/ðwà ‘go’ 389
la and θwà/ðwà compared 390
10.2.6 Experiential 391
10.2.7 Completive 392
10.2.8 Repetitive 394
10.2.9 Continuous 399
10.3 Aspect and manner 400
10.3.1 Without further ado: laiʔ 401
10.3.2 For the time being: tʰà 403
10.3.3 Have time to do: ʔà 405
10.3.5 Do rarely: kʰɛ̀/gɛ̀ 405
10.4 Modality and more 406
10.4.1 Ability 406
yá ‘get’ 406
taʔ/daʔ ‘able’ 408
nain ‘win, capable’ 408
10.4.2 Possibility 410
nain ‘win, capable’ 411
V-ʨʰin/ʥin V-mɛ ‘maybe’ 411
V kàun V-mɛ ‘possibly’ 412
lauʔ ‘as much as, probably’ 413
-léin.mɛ, -mɛ ‘will probably, certainly’ 414
V-mɛ tʰin-dɛ ‘will V, I think’ 415
10.4.3 Actuality 416
10.4.4 Necessity and obligation 416
10.4.5 Desiderative 419
10.4.6 Appropriateness 420
10.4.7 Do unintentionally 422
10.4.8 Daring 423
10.4.9 Trial 426
10.4.10 Empathy 426
10.5 Easy and difficult to do 427
10.6 Liking 428
xiv
11 Intensifiers and comparison 432
Contents
11.1 Intensifiers 432
11.1.1 Intensifying and softening adverbs 433
Intensifying 433
Too much 438
Softening 439
11.1.2 Negation of intensifiers 440
11.2 Comparison 442
11.2.1 Comparative 443
Marked on standard 443
Marked on verb 444
Marked on standard and property 446
Comparison with transitive predicates 447
Quantified comparison 448
11.2.2 Similative and equative 449
Similative 449
Equative 450
11.2.3 Superlative 452
Superlative with intransitive verbs 452
Superlative with transitive verbs 453
11.2.3 Correlative comparative 454

12 Pragmatics and language use 456

12.1 Minimal utterances 457
12.2 Incomplete utterances 458
12.3 Extended utterances 461
12.4 Making assumptions 462
12.5 Subjectivity 465
12.5.1 Properties 465
12.5.2 Quantities 465
12.6 Soliciting agreement 466
12.7 Sentence-final discourse markers 467
12.7.1 Speaker’s attitude 467
12.7.2 Politeness 472
12.7.3 Emphatic utterances 472
12.7.4 Negative emphatic 474
12.8 Information structure 475
12.8.1 Topic/comment 476
12.8.2 Topic and focus markers 476
xv
 12.8.3 Foregrounding 477
Contents
12.8.4 Exclusive 477
12.8.5 Additive 478
12.8.6 Contrastive 478
12.8.7 Reference tracking 480
12.9 Pragmatic case marking 481
12.9.1 Differential subject marking 481
12.9.2 Differential object marking 483
12.10 Social interaction 485
12.10.1 Politeness conventions 485
12.10.2 Greetings 486
12.10.3 Interacting with Buddhist monks 488

Glossary of grammatical terms 491

Index 501

xvi
 Abbreviations

1, 2, 3 1st, 2nd, 3rd pronoun for 1st, 2nd, or 3rd person

person
ADH adhortative imperative directed to the first per-
son plural (‘let’s V’)
ASS.PL associative plural marking the referent and his group
(rather than several referents)
CLF classifier unit marker used when counting
people and objects (concrete and
abstract)
COMPL complementizer morpheme that links a complement
to its verb (‘that’)
CONTR contrastive marker used to indicate a contrast,
either of referents or situations
CQ content question marker used at the end of a
wh-question
CSP consent seeking sentence-final particle indicating
particle that the speaker seeks consent (‘all
right?’)
DEP dependent marks the preceding element as
dependent on the following; inter-
pretations vary from possessive to
relative, among others
DIM diminutive marker frequently added to nominal
and other elements to indicate small
size or amount; often without any
concrete meaning
DISPL displacement indicates that an activity takes/took
place at another place or speech sit-
uation, or removed from the speech
situation xvii
 DISTR distributive distributive marker indicating that
Abbreviations
a certain amount is assigned to each
referent (‘per’, ‘each’)
EMPATH empathy particle added to a noun expressing
empathy for the referent (‘poor X’)
EMPH emphatic particle added for general emphasis
ENUM enumerative marks the amount assigned to each
referent (‘x each’)
EUPH euphonic word added for euphonic reasons
without obvious other meaning or
function
EXCESS excessive intensifier indicating that a prop-
erty or activity is present or done
excessively
EXCL exclusive marker added to a word or phrase
to mark that it is an exclusive choice
out of a number of possibilities
(‘nothing but’, ‘only’)
EXCLAM exclamation indicates surprise about a situation
F feminine noun/pronoun with female referent
FAM familiar informal, familiar form of a pro-
noun or particle, used in intimate
and casual contexts
FOREGR foregrounding marker added to a word or phrase
to foreground or emphasize it
FORM formal formal form of a pronoun, used in
official and public contexts
FUT future marker added to a verb to mark it
as referring to a future or expected
event
IMPORT importance particle/suffix added to an element
to mark the utterance as important,
frequently marking politeness
INSIST insisting particle added to an utterance to
mark an insisting note
INSTR instrumental instrumental suffix added to noun
phrases (‘with’, ‘by’)
INTENS intensifier intensifying particle added to add
weight to an utterance
INTER interrogative general interrogative element, part
of content question words (‘what’,
‘who’, ‘how’, etc.)
xviii
LOC location location marker added mostly to
Abbreviations
personal referents when used with
locative case marker (‘by’, ‘from’)
M masculine noun/pronoun with male referent
NEG negation negation marker; both preverbal
and postverbal
NFUT non-future marks a situation as past or present,
or actually happening
NML nominal marks a word as a noun formally,
with several possible interpreta-
tions, including adverbial
NP noun phrase a noun with its modifying and spec-
ifying elements, including attribu-
tive expressions, case markers, and
adpositions
NSIT new situation marks an event as one that is the
case after a complete or relevant
change of situation (‘(by) now’,
‘already’)
OBJ object marks a noun phrase as direct or
indirect object of a transitive verb
PAST past indicates past time reference
PL plural indicates a general plural, several
instances of the referent to which it
is attached
POL politeness marks an utterance as polite
POSS possessive marks the referent as possessor
PQ polar question marker attached to yes/no questions
PROG progressive marks an event as ongoing
PROH prohibitive denotes a negated imperative (‘don’t
V’)
QUOT quotation particle indicating that the pre-
ceding is a quote, reported by the
speaker
REL relative marks the preceding verbal clause as
relative or attributive
REP reportative marks the preceding word or clause
as reported speech
RESTR restrictive marks the preceding as only possi-
ble referent, amount, or time (‘only
when/if’, ‘not more than’, ‘only’,
‘just’)
xix
 ROY royal suffix added to noun phrases to
Abbreviations
indicate royal referents; today also
used with non-royal referents of
high standing
SBJ subject marks a noun phrase as subject of a
transitive or intransitive verb, or of
a non-verbal clause
SEQ sequential marks events as sequential (one fol-
lowing the other); also frequently
used with simultaneous events or as
general linker of verbs
SPEC speculative indicates that an event is not certain
to occur; often used like a future
marker
SUB subordinator marks a general subordinate rela-
tionship of a verb to the next
TOP topic marks a referent as topical, in the
center of interest
VP verb phrase a verb or verbal predicate with all
its modifying elements, includ-
ing auxiliaries, aspect and number
markers, and status markers

xx
 Acknowledgments

This grammar could not have been written without previous ground-
breaking work in particular by John Okell and Denise Bernot, among
others.
A number of people helped in bringing this grammar together, either by
discussing specific points with the authors and giving constructive feed-
back, or by providing examples and grammaticity judgements. Among
the many people in Myanmar who assisted in writing this grammar and
deserve special mention are: U Khin Aye, U Saw Tun and Patrick McCor-
mick for many stimulating discussions, as well as Than Hla Oo and Wai
Phyoe Aung, who helped in collecting and typing recorded material and
were always available when MJ had any questions. They also provided a
number of useful Burmese examples, and Wai Phyoe Aung typed many
of the examples in Burmese script.
The home institutions of the two authors, namely INALCO/Lacito-
CNRS in Paris (San San Hnin Tun) and the Department of Comparative
Linguistics at the University of Zurich (Mathias Jenny) not only allowed
the authors to spend a considerable amount of their research time to
compile this grammar over the last few years, but also granted travel ex-
penses for weeks of intensive cooperative work in Paris and Zurich and
made available the necessary infrastructure.

xxi
This page intentionally left blank
 Chapter 1

Introduction

This chapter gives a short general overview of the Burmese language in its
geographical, historical and social–political setting. This chapter covers the
major extra-linguistic facts important to the understanding of the language.
It also introduces the reader to the basic features of the language, which will
facilitate the understanding of the detailed descriptions and explanations
given in the main chapters of the book.

1.1 Historical and social context

Burmese belongs to the Sino-Tibetan language family, together with Ti-

betan and Chinese varieties, besides a number of smaller and lesser-known
languages in India, Myanmar, Thailand, and China. With approaching 50
million first and second language speakers, mostly in Myanmar but also
in neighboring countries and overseas, Burmese is the largest non-Chinese
language of the family. Being the only official language of government and
media in Myanmar, Burmese is spoken by the majority of the population of
this country either as a first or second language.

The written record of Burmese goes back to the twelfth century, when it
replaced Mon as the literary language in the Burmese kingdom of Bagan in
the central plains of Myanmar, on the bank of the Irrawaddy River. The
first securely dated document in Burmese appears on the four-language in-
scription written by Prince Rājākumār in the early twelfth century CE and
found at the Myazedi Pagoda near Bagan. The four faces of the stone pillar
are inscribed with parallel texts in Pali, Pyu, Mon, and Burmese. Pali, an In-
dian language related to Sanskrit, is the language of Theravāda Buddhism,
practiced in Śrī Lanka, Myanmar, Thailand, Laos, and Cambodia, as well
as parts of southwestern China (Yunnan). Pyu was the language of a people
who dominated Central and Upper Myanmar in the first millennium CE. 1
 1 Pyu is known from a number of short inscriptions. The language, written
Introduction in its own form of an Indic script, is poorly understood but belongs to the
same language family as Burmese. By the time of the Myazedi inscription,
Pyu probably had already been reduced to a purely ceremonial function.
Mon, an Austroasiatic language still spoken today in Lower Myanmar and
Thailand, seems to have been the main language of state and literature in
early Bagan. It is found on numerous stone inscriptions and on temple walls
in Bagan, but there is no evidence that it was widely spoken by the people,
who probably were mostly Burmese.

Old Burmese shows a number of differences with the modern language, es-
pecially in the orthography, though much less in lexicon and grammar. This
suggests that the pronunciation of Burmese changed markedly over the past
900 years, while the grammar remained relatively stable. Already the earliest
documents of Burmese show a large number of loanwords, mostly from Pali
and Sanskrit, the classical languages of Indian culture and religion.

With the replacement of Mon by Burmese later in the twelfth and thir-
teenth centuries, Burmese became the main official language of the expand-
ing Burmese territory. Through the expansion of Burmese hegemony, the
language came into contact with numerous other languages, both related
idioms and members of different language families. Some of these minor
ethnic languages disappeared as their speakers were absorbed into the body
of Burmese, while others survived but were affected to different degrees by
Burmese lexicon and grammar. Almost all ethnic languages spoken today
in Myanmar, of which there are over one hundred, have large numbers of
Burmese loans in their vocabulary, and calques of Burmese expressions are
easily found.

Burmese, on the other hand, has been rather resistant to influence from lan-
guages of the people lower on the social hierarchy than the Burmese, though
local varieties show features of their neighboring idioms especially in pro-
nunciation. In some cases, like for example in Lower Myanmar, the local
Burmese variety shows some syntactic influence from neighboring Mon, but
hardly any loanwords from this language.

1.2 Myanmar and Burma: Literary and Colloquial Burmese

The Burmese language comes in two distinct varieties, colloquial and liter-
ary. The former is used in everyday conversation, while the latter is mainly
used in writing and in formal contexts, such as public speeches and state
media broadcasts. Novels are typically written in the literary style, but di-
2 alogues of the protagonists appear in colloquial style so that the text is a
mixture of both, though with a clear and complementary distribution of Burmese today
roles. In recent years, the colloquial style has spread to domains previously
covered by the literary style, such as novels, journal and newspaper articles,
and media broadcasts, especially in non-government and foreign channels
such as the BBC and VoA.

In some text genres, such as magazine articles and advertisements, the styles
are frequently mixed.

Literary and Colloquial Burmese differ mainly in the forms of grammatical

function words and to some extent lexicon. Grammatical function words,
such as subject and object markers, are used more consistently in the liter-
ary than in the colloquial style. The literary language prefers longer words,
making use of more euphonic compounds. The present grammar is based on
the colloquial language, with literary variants added where they differ from
the described colloquial forms.

The competing names of the country, namely Burma vs. Myanmar, are ac-
tually an example of the two styles. The form myanma or myəma belongs
originally to the literary language, while bəma is the colloquial form. The
English name Burma represents the colloquial form, with the sequence ur
(sometimes ir) imitating the Burmese sound ə. In 1989 the Burmese govern-
ment changed the English form of the country’s name from Burma to Myan-
mar, together with a number of other place names throughout the coun-
try, like Rangoon to Yangon, Moulmein to Mawlamyaing, and Irrawaddy
to Ayeyarwady. This change aimed at using more indigenous, rather than
colonial terminology. In the case of the country name, another reason given
for the change was that Myanmar was not associated with a specific ethnic
group, but included all inhabitants of the country. This overt step towards
unifying the multi-ethnic state was defeated by the official use of Myan-
mar also for the Burmese language and culture, linking it with the majority
group. In this grammar, we use Myanmar to refer to the country, while we
keep the adjective Burmese for the people and the language.

1.3 Burmese today

1.3.1 The language and its speakers

Burmese is the main language of communication in Myanmar, used by eth-

nic Burmese as well as members of other ethnic groups among themselves
if their native languages are not mutually intelligible. Besides Burmese, a
number of other languages are recognized at the local level. These include
Mon, Shan, Karen, Jinghpaw, and others. With Burmese as the only means 3
 1 of formal education, Burmese is also learned by the vast majority of other
Introduction ethnic groups and serves as the lingua franca in the whole country.

With large numbers of speakers of Burmese as a second language in a vast

and geographically, as well as culturally, diversified area, it is no surprise
that Burmese is not a single, unified language. Some local dialects diverge
from the standard language to the extent that mutual communication is
not readily possible. This is the case with the archaic dialects of Arakan
(Rakhaing), which are, with some differences, also spoken in parts of east-
ern Bangladesh, where the language is known as Marma. Tavoyan (Dawei)
in southern Myanmar and Intha in the Shan State are other varieties not
mutually intelligible with Standard Burmese.

Other local variants of Burmese are closer to the standard language, and
communication is not usually a problem, in spite of differences, especially
in the pronunciation and syntax. In many areas, the dental fricative/affricate/
stop /θ/ is pronounced like the alveolar /t/, and in most, but not all, places
aspirated /sʰ/ merges with plain /s/. The mergers are not complete, though,
as the sounds show different voicing behavior. Other mergers apparently
found especially in Lower Myanmar and in the Yangon area, are the pro-
nunciations of the checked syllables /ɛʔ/ and /aʔ/, which are frequently
pronounced variously as /æʔ/ or /aʔ/, depending on the speaker and the
phonetic environment. On the level of syntax, there is some regional vari-
ation, especially regarding the placement of the preverbal negator mə- in
predicates consisting of more than one verb, as well as the divergent use
of auxiliaries. Again, these and other occurring syntactic variations do not
impede communication in general.

1.3.2 Burmese globalized: Loanwords, code mixing, and slang

Burmese never was closed to foreign influence, and in the increasingly glob-
alized world, these influences become even stronger. Technology introduced
by Westerners before the end of the colonial period mostly uses English
loans, such as kà ‘car’, tɛlipʰòun (or short pʰòun) ‘telephone’, and rediyo ‘ra-
dio’. More recent introductions are given Burmese names, like youʔ-myin-
θan-ʥà for ‘television’, literally ‘image-see-sound-hear’. The latest round
of technological introductions use English terms again, like bwidiyo ‘video’,
kunpyuta ‘computer’, and ʔintanɛʔ ‘Internet’.

Mobile communication, including Internet access, is available almost any-

where in the country and, with it, changing modes of communication. Until
a few years ago, Burmese language use in SMS and chat applications was
restricted to romanized forms due to the lack of Burmese fonts on mobile
4 devices. This has changed recently, with Burmese fonts being available on
most operating systems. Chat communication has consequently shifted to Burmese today
mostly Burmese script, but a few short forms using Roman script are still in
common use, especially among young people. These can be a mix of English
pronunciation, but with Burmese reading, as in 8p for ʔeiʔ-pi ‘I (am going
to) sleep now’ (the English numeral eight is commonly pronounced as ʔeiʔ
by Burmese), or just Roman letters as short forms of Burmese expressions,
as in apk for ʔè-ba kwa ‘that’s it, of course’, with p for Burmese orthographic
<pā> for pa/ba, the importance/politeness particle.

With the spread of English expressions also among people who do not nor-
mally speak the language, increasing mixing of codes can be observed. En-
glish words are frequently inserted in Burmese sentences, even if there is
a Burmese expression available. This is especially typical of modern ur-
ban speech, which spreads to other communities via media, such as radio
and television, which enjoy an increasing popularity with the availability
of private stations. These offer more entertaining programs than the state-
controlled channels and are an important factor in the spread of urban,
especially Yangonite and Mandalay, features also to rural areas. Examples
of code mixing found in natural language are the following:

discount မeပ iuငဘ ။

discount mə-pè-nain-bù.
discount NEG-give-capable-NEG
‘I can’t give you any discount.’

သ outside ရ တယ။
θu outside ɕa-dɛ.
3 outside seek-NFUT
‘He was looking for an extra job.’ or ‘He is looking for an extra job.’

A specialized jargon has developed in Burmese for computer and telecommu-

nication. While English loans are frequent in these domains, the Burmese el-
ements are often used in formulaic form, in many cases as calques of English
expressions. ‘She is online’ is translated as θu làin-bɔ-hma ɕí-dɛ, literally ‘she
is on the line’. ‘To download’ is called daunlouʔ sʰwɛ̀-dɛ ‘pull a download’,
and ‘to install a program’ is progrɛ(m) tin-dɛ ‘put on a program’. In most
cases, the noun is borrowed from English, while the verb is an indigenous
form. 5
 1 1.4 Structure of Burmese: An overview
Introduction

The following sections give a brief overview of the most important features
of Burmese. This is meant to facilitate the reading of the main chapters of
this book by introducing the reader to the major differences between Bur-
mese and English. It should be noted that the terminology used in this gram-
mar corresponds broadly to general descriptive linguistic usage, though in
a number of instances the choice of terminology may be controversial. It is
hoped that the numerous examples given to illustrate each point should be
sufficient to make the structure of Burmese presented here understandable,
though the reader may disagree with the choice of specific terminology in
some cases.

1.4.1 Sounds

Burmese has a number of sounds that are foreign to English and other Euro-
pean languages. All Burmese examples in this grammar are given in Burmese
script as well as in a broad phonemic transcription using the International
Phonetic Alphabet (IPA), with the exception that y is used for the palatal
semi-vowel instead of j. The following sounds need special attention for
speakers of Western languages.

Initial consonants
Burmese has a number of initial consonants not found in Western languages.
These are the velar nasal ŋ, which is pronounced like ng in English sing, but
occurs at the beginning of a word. The other initial consonant not usually
found in Western languages is the palatal nasal ɲ, which is pronounced like
the Spanish ñ or English ny in words like canyon. In Burmese it occurs at the
beginning of a syllable.

The palatal affricates ʨ, ʨʰ, and ʥ are similar, but not identical to English ch
and j, as they are pronounced more to the back of the mouth, with the back
of the tongue touching the hard palate. The same is true for the fricative ɕ,
which is pronounced further back than English sh.

The difference between aspirated and non-aspirated consonants is important

in Burmese, where, unlike English, it is used to distinguish words. While in
English the aspirated form occurs word-initially and the non-aspirated form
after s, as in port vs. sport, in Burmese both forms occur syllable-initially, as
in paʔ vs. pʰaʔ (‘go round’ vs. ‘read’). The aspiration is also found with initial
sʰ-, which is pronounced similarly to the medial sequence in English house-
6 hold, though the aspiration in Burmese is often lost in this case.
Aspirated, voiceless nasals, and liquids occur in Burmese at the beginning Structure of
of a syllable, namely hŋ-, hɲ-, hn-, hm, and hl-. In connected speech the aspi- Burmese: An
overview
ration is often barely heard, and the sounds are pronounced similar to their
voiced counterparts.

Final consonants
Burmese does not have final consonants in absolute final position, apart
from the glottal stop -ʔ. What is written here as final -n indicates a nasal
vowel. When followed by another syllable, these finals often appear as nasal
or stop at the same place of articulation as the following consonant. The
word for ‘spicy hot’ in isolation is saʔ, ‘cold’ is ʨʰã,̀ but in combination
with a following syllable they are pronounced with a final consonant, as in
sat-tɛ ‘it is hot’ or mə-sap-pʰù ‘it is not hot’, and ʨʰàn-dɛ ‘I am cold’ and mə-
ʨʰàm-bù ‘I am not cold’. The transcription used in this grammar does not
indicate these predictable changes, and the words are spelt as saʔ and ʨʰàn
throughout.

Nasal vowels
Burmese has a set of nasal vowels which are to be distinguished from oral
vowels. The nasal vowels found in Burmese are ã, ĩ, ũ and the diphthongs
ãi, ẽi, ãu, and õu. The nasalization is realized as a nasal consonant before a
following consonant in close juncture, with which it shares the place of ar-
ticulation. A sequence kãù -dɛ is pronounced as [kàun-dɛ] or [kàun-nɛ], while
kãù -bi is [kàum-bi] or [kàum-mi]. Nasalization in this grammar is always writ-
ten with a final -n, irrespective of the actual pronunciation, which is always
predictable from the context.

Tones
Burmese is generally described as having three (or four) tones. The exact na-
ture of the tones varies greatly among speakers and also shows considerable
regional variation. The tones are also pronounced differently in different
environments. The tone markers used in this grammar are a purely ortho-
graphic means to distinguish the tones, without attempting to indicate the
actual phonetic shape of each tone in each environment. The signs used here
are Ø for the neutral tone (a; mid, low level), the acute accent for the creaky
tone (á; high, creaky, short), and the gravis accent for the heavy tone (à;
high, falling, long).

In syllables ending in a glottal stop -ʔ, a remnant of old final stop conso-
nants, the tone is neutralized or merges with the acute (creaky) tone. No 7
 1 marker is used in this case, as the final glottal stop is sufficient to indicate
Introduction the value.

Voicing

In connected speech, consonants between vowels are frequently pronounced

as voiced in close juncture. This means that most consonants may appear
either as voiced or as voiceless variants, depending on the context.

1.4.2 Grammatical features

Word order

The grammatical structure of Burmese differs from English in many respects.

The most striking feature for speakers of a Western language is the consis-
tent word order with the verb at the end of a clause. Rather than saying ‘I
eat rice’, in Burmese, one says I rice eat. Similarly, instead of prepositions,
Burmese uses postpositions. ‘On the table’ in Burmese is therefore table on,
‘in the house’ is house in. Related to this word order structure is the position
of relative clauses before the head noun, as in ‘the books which I gave you’,
which in Burmese is rendered as I you gave which books. Clause linkers
(subordinators) occur at the end of a clause, so that ‘I didn’t come because
I didn’t have time’ in Burmese is I time not have because I not come.

Agglutinating structure

In Burmese, a large number of grammatical elements can be attached to a

lexical word, resulting in long strings which can be considered words. Each
element in the string has its own, more or less specific meaning. This is true
for noun phrases as well as for verb phrases, though the latter tend to be
more complex.

General possibility to drop noun phrases

Noun phrases with referents that are known or retrievable from the dis-
course context are frequently dropped in Burmese. Pronouns are therefore
less frequently used in Burmese than in English, as no overt subject or object
8 must be present.
Tense distinctions Structure of
Burmese: An
overview
Burmese makes a basic distinction only between future and non-future events,
the latter covering present and past situations. To specify a tense, temporal
adverbials are added.

Classifiers
Like many other languages of East and Southeast Asia, but unlike European
languages, Burmese makes use of classifiers whenever a noun is counted.
Instead of saying ‘three friends’, one says friend three person, and for ‘five
dogs’ the Burmese expression is dog five animal. Each noun is assigned a
classifier, which is not always semantically connected to it in the present-
day language, similar to gender assignment in languages like French and
German.

Adjective as verbs
No clear class of adjectives can be postulated in Burmese. What are adjec-
tives in English mostly function as verbs in Burmese. Instead of adjectives,
they are a class of stative verbs (or property verbs) in Burmese, with the
same syntactic possibilities as other verbs.

Adverbs as non-unified category

While nouns and verbs can be established as word classes in Burmese on
syntactic grounds, this is not true for adverbs. Adverbs show a vast range
of syntactic and morphological possibilities, some more noun-like, others
more verb-like. In many cases, verbs are used in adverbial function. When
speaking of adverbs, it therefore has to be kept in mind that this is not
a unified syntactic category in Burmese, but rather a category based on
semantics.

1.4.3 Role of context: Linguistic and extralinguistic

Burmese is a highly context dependent language. Knowing the semantic

meaning of the words of an utterance is often, perhaps in most cases, not
enough to understand the meaning of the utterance in a given context. Be-
sides frequently dropping nouns when they are felt by the speaker to be
sufficiently retrievable from the discourse context, many words are used in
meanings different from the corresponding lexical entry. The word θəmì has
the lexical meaning ‘daughter’, but in actual usage it frequently means ‘I’ or
‘you’, depending on the speech situation. 9
 1 Discourse particles are an important means in Burmese to structure the flow
Introduction of information and to add meanings related to the interlocutors. These par-
ticles have in some cases grammatical function, but frequently are used to
convey pragmatic notions, such as emphasis, speaker’s attitude, and others.
There is no easy way to translate these particles into English, and in many
cases they must remain untranslated. There is no other way to properly un-
derstand the pragmatic functioning of Burmese (and any other language)
than to see it in its cultural and social context and observe it in its actual ev-
eryday use. No written grammar can do full justice to this, and explanations
given here are to be seen as approximations at best.

1.5 About this grammar

The aim of this grammar is not to present the complete set of possibilities or
rules to build clauses, sentences, and discourse in Burmese, a task that would
be doomed to fail from the beginning. Rather, the intention is to explain
and illustrate the grammar of Burmese as a living and flexible language, a
system of communication used by some 50 million speakers in their every-
day life. Whatever rules or patterns are given in the following chapters can
be seen only as possible ways of expressing things and conveying certain
meanings in Burmese. In many or most cases there will be numerous other
ways to express the same meaning, and the English translations given are
only possible approximations, rather than ‘the one and only’ translation of
the respective Burmese expressions.

One problem in translating Burmese into a Western language is the dropping

of nouns and non-lexical use of lexemes, such as kinship terms instead of
pronouns, as well as the frequent lack of gender distinctions. In English in
such cases, a choice must be made, which is more often than not random
in the absence of more context. In many cases the reader will find that the
English translation is rather removed from the Burmese expression.

As the major aim of this grammar is to help the reader understand the struc-
ture of Burmese, that is, the interplay of form/structure and meaning/use of
expressions, the examples are given with glosses as well as English transla-
tions. In the glosses a ‘simple’ approach is followed, that is, each Burmese
word is consistently glossed by the same English word, irrespective of its
actual function in a given context. This function should be evident from the
translation in all cases, though it involves some degree of analytical work by
the reader. While this system leads to the glosses being very different from
the actual translation in many cases, it has the advantage that it can help
capture the structure of Burmese better than by using context-dependent
10 glosses. The choice of the glosses is usually not trivial or easy, as in many
cases there is no English word that covers all meanings or functions of a About this
grammar
Burmese lexeme. Whenever there is an apparent ‘basic’ meaning in a form,
this is chosen as the gloss. This is the case with verbs like pè, which is also
glossed as ‘give’ when it occurs as a benefactive or permissive marker, and
ne which is also glossed as ‘stay’ in its aspectual function to express con-
tinuous or progressive events. In other cases, the choice is more difficult,
as with the particle tɔ́/dɔ́, which is used to mark a contrast of some kind,
either between situations or between referents. In this case the grammati-
cal gloss CONTR for ‘contrastive’ is used. To keep the glosses concise, the
copula ‘to be’ as well as the infinitive marker ‘to’ are left out. It has to be
kept in mind that in some cases a Burmese word may be translated as a verb,
even though the gloss seems to suggest otherwise. Typical examples are kàun,
glossed as ‘good’, but translated as ‘be good’, and po, glossed as ‘in excess’,
but translated as ‘be in excess’. Standard glossing rules have been followed
throughout, linking one element in the text to one element in the glossing
line. All abbreviations used to gloss grammatical morphemes are given in
the list of abbreviations, together with short explanations of their use and
function.

Apart from the difficulty of glossing and translating, the presentation of

Burmese itself is a non-trivial issue. Different approaches are possible when
transcribing Burmese, which are based on and influence our understand-
ing of the language. The major point to be addressed here is what to take
as a ‘word’ in Burmese, a question that may seem easy enough to answer,
but poses a number of difficulties. The pieces of phrases can be written sepa-
rately, making Burmese look like a typical isolating language like Chinese or
Vietnamese, or one can choose to write them together, resulting in a rather
agglutinating language type more similar to Turkish and Japanese. In this
grammar the latter path is followed, based on phonological and syntactic
grounds, which suggest that Burmese indeed has agglutinating structures.
Whenever it is conducive to the understanding of the structure of the exam-
ples, the elements forming a ‘word’ are separated by dashes (-), while a full
stop (.) indicates that a ‘word’ is made up of two or more elements which
are not further analyzed. Depending on the point to be made, the same com-
pound word may in some examples be broken up into its constituent parts,
and in others remain unanalyzed.

The present grammar is laid out in a way that is believed to be useful for
different groups of readers. The first chapters (2–8) present Burmese struc-
tures, starting from the sound system and ending with sentence structures,
and links these structures to functions. This part answers questions like ‘how
is Burmese structured’ and ‘what does a given structure in Burmese mean’.
The following chapters (9–11) take a number of linguistic features and de-
scribe how these are expressed in Burmese. The question answered by these 11
 1 chapters is ‘how is feature X expressed in Burmese’. These are naturally
Introduction two different perspectives of the same phenomena, so it is unavoidable that
a number of statements will be found in both parts. The final chapter (12)
looks at how the Burmese language is actually used in different contexts and
to express meaning that goes beyond what is encoded by the semantics and
the syntax of an utterance. Linguistic terms as they are used in this grammar
are explained in a short glossary at the end of the volume.

The basis of this grammar is the contemporary colloquial language, as spo-

ken in Myanmar. Many examples are taken from natural language sources,
including recordings of conversations and narratives, radio and TV broad-
casts, newspapers and other written publications, as well as from Internet
chats and publications (such as BBC, VoA) and available corpora (such as
sealang.net and the individual research corpora of the authors). Additional
examples have been created to show specific points concisely. These ‘cre-
ated’ examples have been checked for grammaticity with a number of na-
tive speakers. The literary forms are added whenever there is a significant
difference to the colloquial style, usually with parallel examples to the ones
given for the colloquial language.

This grammar is the result of a long-time joint effort of the two authors,
who come from very different backgrounds and have very different ideas
on many questions. The aim from the outset was to produce a descriptive
account of Burmese which is accessible to the general-interest reader and
language learner, but also to linguistic typologists. This necessarily resulted
in a number of inevitable compromises, which, it is hoped, will not diminish
the usefulness and value of the present grammar.

12
 Chapter 2

Sound system and script

The sound system of Burmese has many interesting features, including a

number of uncommon sounds, such as voiceless nasals (hŋ, hm, etc.), an
aspirated sibilant (sʰ), and three tones. The aim of this chapter is to sys-
tematically describe the sound system and explain the use of the IPA-based
transcription method applied in this grammar. It will also give a detailed
account of Burmese pronunciation.

The transcription has been designed to meet linguistic standards without

overloading it with diacritic signs. The choice of symbols is based on linguis-
tic usage combined with ease of use for non-linguist readers. In some cases
we have chosen special IPA symbols over common Latin letters in order to
avoid possible confusion, and to retain the linguistic accurateness. This was
done for example with ʨ, ʨʰ, and ɕ for sounds that are represented by other
authors as c, ch and ʃ or sh. The use of c for the non-aspirated alveo-palatal
affricate sound may lead readers to pronounce this letter like in English,
that is, either as k or s. By choosing the IPA series there is no confusion
likely to arise, and the transcription could be kept symmetric. Aspiration is
always indicated by a superscript ʰ. This has the advantage that the unusual
aspirated sʰ sound is not mistaken for the English sh as in she.

The transcription is applied systematically so that readers who do not wish

to learn the Burmese script can still fully benefit from the grammar. Voicing
of intervocalic consonants in closely linked morphemes and words is an im-
portant feature of Burmese. Words that occur as voiced or voiceless are given
in both forms and the transcription given according to the actual pronunci-
ation in each occurrence. This system has been chosen over other methods
found in the literature (e.g. underlining of the voiceless consonant to indi-
cate that it is spoken as voiced, as done by John Okell’s Reference Grammar
(1969), as it is more readily comprehensible and easier to use. Other types
of sound change and assimilation in close juncture (Sandhi) are explained 13
 2 in this chapter, but not indicated in the examples, as they are less systematic
Sound system than voicing and, unlike voicing, never have distinctive function.
and script
This chapter also briefly describes the prosody (intonation) of phrases,
clauses, and sentences, and indicates some major regional and sociolectal
differences in pronunciation of individual sounds and sequences of sounds.
This is of interest not only to linguists, but also to language learners traveling
or working in Myanmar, who are exposed to real language use in different
regions of the country.

2.1 Sound system

The sounds of Burmese offer a number of difficulties for speakers of Western

languages. Apart from having a number of unfamiliar consonant sounds,
Burmese is also considered a tone language, where each syllable is pro-
nounced with a certain pitch or contour, though the tones of Burmese are
a composition of different voice-quality factors, rather than pure tones. In
the following sections the basics of the Burmese sound system will be pre-
sented in order to make the rest of the book easily accessible for speak-
ers of European languages. The transcription system used here to represent
the sounds of Burmese is adapted from the International Phonetic Alphabet
and aims to unambiguously represent the sounds of Burmese. The individ-
ual sounds and their pronunciations in different environments are explained
below, adducing English equivalents where possible.

2.1.1 Consonants

Burmese has 34 consonant sounds, all of which can occur in syllable-initial

position either alone or in consonant clusters, see 2.1.1.2.

2.1.1.1 Plain consonants

labial dental alveolar palatal velar glottal

stops p pʰ b θð t tʰ d ʨ ʨʰ ʥ k kʰ g ʔ

fricatives s sʰ z ɕ h

nasals m hm n hn ɲ hɲ ŋ hŋ

approximants w hw l hl r y

14
The stops can be divided into voiceless-inaspirate, voiceless-aspirate, and Sound system
voiced. The voiced stops and fricatives are partly or fully voiced, depending
on the context and, to some degree, on the speaker. In general, intervocalic
consonants are more voiced than in word-initial position. Word-initially, the
contrast between aspirated and unaspirated is much more prominent than
between voiced and voiceless.

The consonants of Burmese are pronounced as follows.

p as in pà ‘cheek’, is pronounced like the English p in sport, that is,

voiceless and without aspiration.
pʰ as in pʰà ‘frog’, is pronounced with aspiration, as in English pay.
b as in ba ‘what,’ is partly or fully voiced in most speakers’ pronun-
ciation and lies between English p and b. It is never aspirated. In
intervocalic position, b approaches [β] in casual speech.
θ as in θà ‘son’, is unlike English th in that it is mostly pronounced as
a dental stop, not as fricative or affricate. It is never aspirated. For
some speakers there is no clear distinction between θ and t.
ð as in ʨàun.ðà ‘student’, is the voiced counterpart of θ and is pro-
nounced as a dental stop, rather than as a fricative or affricate. It is
mainly found in medial position. For some speakers this sound falls
together with d.
t as in tà ‘forbid’ is voiceless and non-aspirated, as English t in stay. For
some speakers, the tongue is pushed back to post-alveolar position,
approaching a retroflex stop.
tʰ as in tʰà ‘keep, put, place’, is the aspirated counterpart of t and pro-
nounced like English t in table. As in t, the tongue is pushed back to
a post-alveolar position.
d as in da ‘this one’ is partly or fully voiced, like English d in day. In
casual speech, d in intervocalic position often approaches r [ɾ].
ʨ as ʨà ‘hear’, is an alveo-palatal or palatal affricate. The exact pro-
nunciation varies, frequently approaching ch in child, but without
aspiration, or j in joke, but voiceless.
ʨʰ as in təʨʰà ‘other’, is the aspirated counterpart of ʨ, similar to English
ch in child but the tongue pushed further back. The aspiration is
frequently realized as strengthening (or lengthening) of the fricative
part, rather than as pure aspiration.

15
 2 ʥ as in ðəʥà ‘sugar’, is a voiced alveo-palatal or palatal affricate, sim-
Sound system ilar to English j in joke, but with the tongue pushed further back. In
casual speech, ʥ in intervocalic position sometimes approaches y.
and script

k as in kà ‘car’, is a voiceless velar stop, like English k in ski, that is,

without aspiration.
kʰ as in kʰà ‘bitter’, is the aspirated counterpart of k, pronounced like
English c in car. For some speakers, kʰ is slightly affricate.
g as in dəgà ‘door’, is partly or fully voiced, like English g in go. In some
high-frequency words, intervocalic g in close juncture is dropped,
resulting in syllables with vocalic onset without a glottal stop.
ʔ as in ʔà ‘be free’, marks the glottal onset heard in English fully
stressed words with initial vowel, like in old [ʔəuld]. In final position,
ʔ marks an abrupt end of the vowel.

s as in sà ‘eat’, is the voiceless alveolar fricative, similar to English s in

sun.
sʰ as in sʰà ‘salt’, is theoretically the aspirated counterpart of plain s,
but for most speakers there is no or hardly any distinction between
the two sounds in casual speech. In careful pronunciation, sʰ tends
to be pronounced as longer (but not quite geminate) s.
z as in ʔein.za ‘homework’, is voiced like z in English zoo.
ɕ as in ɕa ‘look for’, is pronounced similarly to English sh in shoe but
with the tongue pushed further back.
h as in ha ‘empty, vacant’ is pronounced like English h in house. In
intervocalic position, h is partly voiced, approaching [ɦ].
m as in ma ‘hard’, is pronounced like English m in me.
hm as in hma ‘order’, is the voiceless counterpart of m. It is pronounced
like a preaspirated m, where the first part of the consonant is voice-
less, similar to a whispered m. The aspiration (or voiceless part) is
not always clearly perceptible in connected speech.
n as in nà ‘nearby’, is pronounced like English n in near.
hn as in hna ‘nose’, is the preaspirated or voiceless counterpart of n,
similar to a whispered n. The aspiration (or voiceless part) is not
always clearly perceptible.

16
 ɲ as in ɲa ‘right (side)’, is similar to English ny in canyon, but pro- Sound system

nounced as such also at the beginning of a word. It approaches the

British English pronunciation of the initial consonant in new.
hɲ as in hɲa ‘be considerate’, is the voiceless or preaspirated counter-
part of ɲ. The aspiration (or voiceless part) is not always clearly
perceptible.
ŋ as in ŋà ‘fish, five’, is pronounced like English ng in song. It is pro-
nounced as such also at the beginning of a word and not followed
by a g sound as in some English words such as finger.
hŋ as in hŋà ‘hire, rent, borrow’, is the preaspirated or voiceless coun-
terpart of ŋ. The aspiration (or voiceless part) is not always clearly
perceptible.
w as in wa ‘cotton’, is pronounced like English w in we.
hw as in lɛʔ-hwé ‘box, punch’, is rare in Burmese. The pronunciation
of this phoneme varies between aspirated hw, like in some English
dialects wh in what, plain w, and pʰw, depending on the word and
speaker.
l as in la ‘come’, is pronounced like English l in like.
hl as in hlá ‘pretty, beautiful’, is the preaspirated or voiceless
counterpart of l. It is pronounced like a whispered l.
r as in rediyo ‘radio’ occurs mostly in English and Indic loanwords.
Its pronunciation is similar to the British English prevocalic r in
road, that is, as simple alveolar tap. This phoneme is not part of
the indigenous sound system of Burmese.
y as in ya ‘hundred’, is pronounced like English y in yard. In some
speakers’ pronunciation, y is fricativized and approaches [ʒ].

2.1.1.2 Consonant clusters

A limited number of consonant clusters is found in syllable-initial position.

The pronunciation of the clusters is generally the combination of the in-
dividual parts merged in one sound. The combinations ly and hly are rare
and mostly found in the formal language. Compared to other clusters, these
combinations are looser. In colloquial style, these clusters are commonly
pronounced as y- or ləy-/hləy-. There are some restrictions of clusters with
certain vowels, namely the sequences kwu kwo, kwɔ, which do not occur in
Burmese. 17
 2 labial dental alveolar palatal velar
Sound system
and script pw pʰw bw θw ðw tw tʰw dw ʨw ʨʰw ʥw kw kʰw gw

mw hmw nw hnw ŋw hŋw

sw sʰw zw ɕw

lw hlw yw

py pʰy by (ly hly)

my hmy

Examples of words containing consonant clusters are the following.

ပင pwín ‘be open’ ဖင pʰwín ‘to open’

သ θwà ‘go; tooth’ တ twà ‘crawl, creep’
eမ mwe ‘snake’ eင ŋwe ‘silver, money’
စ swɛ̀ ‘bear, hold’ ဆ sʰwɛ̀ ‘pull’
eဆ sʰwe ‘friend’ eရ ɕwe ‘gold’
လယ lwɛ ‘easy’ ရယ ywɛ ‘aim, intend’
ပ pyà ‘bee’ ဖ pʰyà ‘have a fever’
မ myà ‘many’ မ hmyà ‘to fish’

2.1.2 Vowels

Compared to the consonants, the vowel inventory of Burmese is rather sim-

ple. There are seven basic vowels, plus the neutral vowel ə and four diph-
thongs, which occur only in nasalized and stopped syllables. Vowel length
alone is not phonemic, but is taken by some as part of the tone system, which
will be described in section 2.1.3.

2.1.2.1 Plain vowels

Oral vowels in open syllables without a final glottal stop are close in pro-
nunciation to English vowels. The following table shows the plain vowels
according to their approximate place of articulation.

i u

e o

ɛ ə ɔ
18 a
The pronunciation of the plain vowels is as follows. Sound system

i as in θì ‘fruit’, is pronounced like English ee in see.

e as in θè ‘small’, is pronounced similar to English ay in say, but with-
out the [j] off-glide. It is a high, tense [e], similar to French é in été
‘summer’. Phonetically, it is often between [e] and [ɪ].
ɛ as in θɛ̀ ‘sand’, is similar to English e in bed. Phonetically, it is
between [e] and [ɛ].
u as in lu ‘person, human’, is similar to English oo in zoo.
o as in lo ‘want, need’, is similar to American English o in go, but with-
out the [w] off-glide. It is a high, tense [o], similar to French eau
‘water’.
ɔ as in lɔ̀ ‘hurry’, is similar to English aw in law.
a as in la ‘come’, is a low mid vowel, similar to English a in father.
ə as in ʔəko ‘elder brother’, is a neutral central vowel similar to English
a in ago. It occurs only in weak syllables which precede a full syllable.
This vowel never carries a tone or stress.

2.1.2.2 Vowels in closed syllables

There are two types of syllables that can be labeled closed syllables, namely
syllables containing a nasal vowel and syllables ending in a glottal stop ʔ. The
former are, in isolation, open nasal syllables, but the nasalization appears
as nasal consonant in close juncture, as will be described in section 2.1.5.2.
The transcription used here indicates nasality by a final -n, but it should be
kept in mind that this is not pronounced as a consonant in isolated words.
The quality of the vowels in closed syllables differs from open syllables, and
diphthongs occur only in closed syllables. Compared to the vowels in open
syllables, the high vowels i and u are lowered to [ɪ] and [ʊ] respectively.
The low vowel a is fronted and raised, approaching [æ], while ɛ in stopped
syllables is lowered and centralized. In some dialects, especially in Lower
Myanmar, aʔ and ɛʔ merge in [æʔ], while in other regions, especially Upper
Myanmar, the two are at least theoretically kept apart. The following vowels
and diphthongs occur in closed syllables.

in un iʔ uʔ

ɛʔ

an aʔ

ein oun eiʔ ouʔ

ain aun aiʔ auʔ 19

 2 The nasal vowels have no English counterparts. Their pronunciation is as
Sound system follows.
and script
in as in lìn ‘bright, light’, is pronounced as nasalized [ɪ].̃
un as in lùn ‘go beyond’, is the high back nasal [ʊ̃ ].
an as in làn ‘way, road’, is the low nasal [æ̃ ].
ein as in lèin ‘spread, apply (medicine, lotion, etc.)’, is slightly
diphthongized [ẽʲ] with nasalization throughout.
oun as in lòun ‘round’, is slightly diphthongized [õʷ] with nasalization
throughout.
ain as in làin ‘line’, is the diphthong [ãɪ]̃ with nasalization throughout.
aun as in làun ‘bet’, is diphthong [ãʊ̃ ] with nasalization throughout.

The vowels in stopped syllables are similar to English short vowels before
final stops, but without the stop actually pronounced. Their pronunciation
is as follows.

iʔ as in liʔ ‘slip away, sneak away’, is pronounced as [ɪ], similar to

English i in sit.
uʔ as in luʔ ‘be free’, is pronounced as [ʊʔ], similar to English u in put.
ɛʔ as in lɛʔ ‘hand’, is pronounced as [ɛʔ] or lowered to [æʔ], lower
between English a in bat and e in bet.
aʔ as in laʔ ‘fresh; medium (size)’, is pronounced as [aʔ] or raised to
[æʔ], merging with ɛʔ in some speakers’ pronunciation. It is between
English u in but and a in bat.
eiʔ as in leiʔ ‘turtle’, is slightly diphthongized [eʲʔ], similar to English a
in late.
ouʔ as in louʔ ‘do, make’, is slightly diphthongized [oʷʔ], similar to
English o in hope.
aiʔ as in laiʔ ‘follow’, is the diphthong [aɪʔ], similar to English i in like.
auʔ as in lauʔ ‘be enough, be equal, as much as’, is the diphthong [aʊʔ],
similar to English ou in about.

20
 2.1.3 Tones, voice quality, and prosody Sound system

2.1.3.1 Tones
Unlike tones in languages like Chinese, Thai, and Vietnamese, tones in Bur-
mese are less fixed in character and vary to a great extent within an utter-
ance. Burmese has three tones, which are characterized by pitch, contour,
length, and phonation type (creaky and breathy voice). No single factor ex-
haustively describes the tones, but it is the combination of these features
that makes the tones distinguishable, with different factors being more or
less prominent in different local varieties. Although pitch and contour, typi-
cal tonal features in other tone languages, are only two among the factors,
and are not dominant (or even relevant) in the standard language and Upper
Myanmar pronunciation, we use the term ‘tones’ here for convenience.

Tones in Burmese are mainly a lexical feature, that is, each lexical item has a
fixed tone, which distinguishes it from other words. In a few cases tone can
have grammatical function, for example in derivation processes, which are
not productive synchronically and can therefore be taken to be lexicalized.
The one productive grammatical process involving tone change is the creaky
tone indicating a dependence relationship, such as possessive on nouns and
attributive/relative on clauses.

We use the following notation to indicate tones. It should be kept in mind

that the accents used here are merely indicators of the ‘tone’ category and
do not represent the common IPA values assigned to them. The descriptive
factors of each tone are not necessarily present in all speakers’ pronunciation
or in all dialects, nor even in all words with the same tone spoken by the
same speaker.

No marker: a as in sa ‘written text’ is called here neutral tone. It is

pronounced with a low or mid pitch, usually of even
contour, sometimes rising. It has medium length and
normal (modal) voice quality.
Acute: á as in sá ‘begin, start’ is called creaky tone (high,
falling, relatively short, creaky voice quality)
Grave: à as in sà ‘eat’ is called heavy tone (high, falling,
relatively long, breathy voice quality)

In stopped syllables ending in a glottal stop /ʔ/, the tone is neutralized. It is

pronounced either with high or low pitch and some creakiness, and these
syllables are usually short. In some varieties, the stopped syllables are per-
ceived as similar or identical to the creaky tone. This is especially the case 21
 2 in varieties where nasalization is less prominent, such as southern dialects
Sound system of Mon and Kayin States. This leads to confusion and irregular spellings
in words like léin ‘roll, writhe’, which are sometimes erroneously spelled
and script

leiʔ.

2.1.3.2 Stress and prosody

Stress is not prominent in Burmese and is usually secondary to tone. What

superficially appears like stress in an utterance is frequently rather lexical
tone. Perceptionally the stress pattern in an expression like hnə-tʰaun ‘two
thousand’ is iambic, while the functionally parallel ŋà-daun ‘five thousand’
appears trochaic. This is a result of the tone and syllable pattern rather than
stress.

Stress usually falls on the main semantic element and may also be used for
emphasis to mark a word in an utterance as focal. Certain mostly grammati-
cal morphemes never receive full stress and are frequently weakened in con-
nected speech. These include, for example, the postverbal auxiliaries ne ‘stay;
progressive, continuous aspect’, ʨʰin/ʥin ‘want to; desiderative modal’, and
grammatical markers such as tɛ/dɛ ‘non-future verb marker’, mɛ ‘future verb
marker’, pʰù/bù ‘negative verb marker’, ko/go ‘object case marker’, hma ‘loca-
tive marker’, among others.

At least historically there is evidence that in compounds an iambic stress

pattern was common at some point in Burmese. This can be seen from the
fact that in many compounds the first element is weakened, with the vowel
appearing as schwa. This weakening is usually not represented in the orthog-
raphy, as in θəkʰò ‘thief’, spelled <θu kʰò>, tə-yauʔ ‘one person’, spelled <tiʔ
yauʔ>. Weakening also applies to loanwords, such as yəθé ‘hermit, recluse’
from Sanskrit ṛ́ṣi ‘sage’. There is no weakening of vowels to schwa in final
position.

On the clause and sentence level, the prosody is very much determined by
the semantics and pragmatics of the utterance. In general, the end of an
utterance is marked by a down-drift, irrespective of the lexical tone of the
final syllable, while an up-drift frequently marks an unfinished utterance
and often appears at the end of a subordinate clause. The down-drift at the
end of an utterance also appears in questions, both polar and content. For
emphasis or other pragmatic reasons, the final syllable of an utterance may
also receive a high pitch. This is also common in more formalized language,
22 such as news reports.
 2.1.4 Syllable structure Sound system

A word in Burmese can consist of one or more syllables. Burmese allows

only a limited number of syllable structures. Generally, every syllable be-
gins with a consonant. In loanwords beginning with a vowel, a glottal stop
ʔ is always present in the initial position. An exception to this rule is found
in some grammatical markers that undergo weakening in intervocalic posi-
tion and appear as pure vowels in some contexts. This will be illustrated in
section 2.1.5 on Sandhi rules. No syllable can have more than two initial
consonants, the second of which can be either y or w. In the transcription
applied here, an h before an initial consonant indicates a preaspirated or
voiceless consonant.

The syllable nucleus consists of a simple vowel or a diphthong, which carries

a tone and can be either oral or nasal. The only possible final consonant is
the glottal stop ʔ. No tone distinction is made in syllables ending in a glottal
stop.

Nasal vowels, transcribed here as -Vn, are, in isolation, pronounced as purely

nasalized vowels, while in close juncture, the nasal appears as nasal conso-
nant with the same place of articulation as the following consonant. The
vowel preceding the nasal may lose its nasality partly in close juncture. De-
tails of this assimilation process will be given in section 2.1.5 on Sandhi
rules.

The syllable pattern of Burmese is C(c)V(ʔ) + T, where c stands for a glide

and T represents a tone. In addition, there are weak syllables consisting of an
initial consonant and the reduced vowel ə, which never carries a tone. Weak
syllables always precede full syllables and are never stressed. Weak syllables
are orthographically often full syllables, suggesting that the weakening is a
rather recent phenomenon in Burmese. The following words illustrate the
syllable structure of Burmese.

CV
eလ le ‘wind’ လမ làn ‘way, road’
အ ʔà ‘be free’ အiu ʔo ‘old’
ဘ သ baða ‘language, religion’ ဘတ buda ‘railway station’

CcV
eပ pyè ‘run, flee’ eပ pyɔ̀ ‘speak, say’
ပ pwɛ̀ ‘festival, fair’ ပ pwà ‘grow, increase’
ဖင pʰwín ‘open’ က ʨwà ‘brag’ 23
 2 CVʔ
Sound system
and script
တက tɛʔ ‘go up, rise, တတ taʔ ‘skilled, know how to
ascend’ do’
လတ luʔ ‘come free, be လuပ louʔ ‘do, make’
free’
စ hniʔ ‘year; two’ iပ hneiʔ ‘press’

CcVʔ
မစ myiʔ ‘river’ ပတ pyaʔ ‘break off, be broken’
ထက tʰwɛʔ ‘go out, exit’ eမ က myauʔ ‘north’

Weak syllables
ကစ gəzà ‘play’ စက zəgà ‘(spoken)
language’
eခ င hnəkʰàun ‘nose’ လu hnəlòun ‘heart’
အပင ʔəpwín ‘blossom, အလuပ ʔəlouʔ ‘work’ (n.)
bloom’

2.1.5 Sandhi

Sandhi means a close juncture of words in compounds or phrases that in-

fluences the pronunciation of the individual parts. Sandhi rules are of two
types, voicing and assimilation. These rules are not automatic articulatory
processes that apply in all appropriate contexts in all speakers’ pronuncia-
tion, that is, there is some degree of variation among dialects and also within
a dialect as to the application of the Sandhi rules. Apart from the dialect or
sociolect, the speech setting and speed, as well as other factors, influence the
degree of voicing and assimilation.

2.1.5.1 Voicing
Consonants in intervocalic position tend to be voiced in close juncture, that
is, within a compound word or a phrase. While some frequent morphemes
are almost always voiced, others vary or are never voiced. The former in-
cludes the verb markers tɛ/dɛ ‘non-future’ and pʰù/bù ‘negative’, the latter
includes the secondary verbs pè ‘give; benefactive’, tʰà ‘keep, put, place;
do and leave the result’, kʰàin ‘order, command; causative’, among oth-
ers. The application or non-application of voicing may be triggered by the
24 grammatical status of a morpheme, that is, the more grammatical and de-
pendent a morpheme is, the more likely it is to be voiced in appropriate Sound system
contexts.

Voicing generally occurs whenever a voiceable consonant occurs between

two vowels, that is everywhere except the glottal stop. Voiceable consonants
are all consonants that have voiced counterparts, including the voiceless
nasals and liquids. Some high frequency items, especially the verb markers
tɛ/dɛ ‘non-future’ and pʰù/bù ‘negation’ are often also voiced after a glottal
stop.

The following degrees of voicing are observed, depending on the speaker and
speech situation, as well as the articulation speed. The third stage is found
normally only in grammatical morphemes, not in lexical words.

k, kʰ → [g] → [Ø]
t,tʰ → [d] → [ɾ]
p, pʰ → [b] → [β] (→ [Ø])
s, sʰ → [z]
θ → [ð]
hŋ → [ŋ]
hɲ → [ɲ]
hn → [n]
hm → [m]

The voicing and eventual deletion of the velar stop results in a phonemic op-
position of ʔ vs. Ø, though only in very restricted contexts. Compare the ca-
sual pronunciation of θú-go ‘him, her’ [θú-o] with θú ʔò ‘his pot’ [θú ʔò].

Voicing is found in the following contexts.

All elements following the main verb sà ‘eat’ in a verb phrase:

သမစ ခ ငeတ ဘ ။
θu mə-sà-ʥin-dɔ́-bù.
3 NEG-eat-want.to-CONTR-NEG
‘He doesn’t want to eat anymore.’

In the second part of some verbal compounds:

ဆင ခင sʰin-ʥin ‘consider’
စuစမ soun-zàn ‘investigate’
ခမ သ ʨʰàn-ða ‘rich, wealthy’ 25
 2 In the second element in reduplicated verbs with adverbial or attributive
Sound system function:
and script
eက င eက င eနeန ။ အiမ က က
kàun-gàun ne nɔ. ʔein ʨì-ʥì.
good-good stay CSP house big-big
‘Be good, will you?’ ‘a (really/rather) big house’

Numerals and classifiers in numeral phrases:

eခ eလ eက င င ဆယ
kʰwè lè-gaun ŋà-zɛ
dog four-CLF five-ten
‘for dogs’ ‘fifty’

Grammatical elements, such as plural and case markers, in noun phrases are
voiced:

eခ ကeလ eတကiu ဆရ တiuက

kʰwè-gəlè-dwe-go sʰəya-dó-gá
dog-DIM-PL-OBJ teacher-ASS.PL-SBJ
‘the little dogs (object)’ ‘the teacher and his group’

In the verbal element in noun-verb compounds that function as nouns:

ထမင ခ က tʰəmìn-ʥɛʔ ‘cook’ tʰəmìn ‘cooked rice’ and

ʨʰɛʔ ‘cook’
အiမeစ င ʔein-záun ‘guard’ ʔein ‘house’ and sáun
‘wait, watch over’
ဗiuလခ ပ bo-ʥouʔ ‘general’ bo ‘army, forces’ and
ʨʰouʔ ‘control’
လက lu-ʥì ‘adult, lu ‘person’ and ʨì ‘big’
senior
person’

No voicing occurs in the following contexts. Note that sporadic voicing may
occur also here. Generally the faster the speed of the speech, the more likely
voicing is to occur.

26
Main verbs and preverbal auxiliaries are never voiced: Sound system

ထမင စ tʰəmìn sà ‘eat rice’ tʰəmìn ‘cooked rice’, sà ‘eat’

ပနe ပ pyan pyɔ̀ ‘say again, pyan ‘return, do again’, pyɔ̀
reply’ ‘speak, say’

Reduplicated verbs with non-adverbial or attributive function are sometimes

voiced, but remain unvoiced:

ဘ ပe ပ e ပ ba-bɛ̀ pyɔ̀-pyɔ̀ ‘whatever one says’

ဘယeလ ကပစ စ bɛ-lauʔ-pʰɛ̀ sà-sà ‘however much one eats’

Some verbal compounds are not voiced:

ဆu ဖတ sʰòun-pʰyaʔ ‘decide’ sʰòun ‘end’ and pʰyaʔ ‘cut’

ရင ပ ɕin pyá ‘explain’ ɕin ‘untangle’ pya ‘show’
တင ပ tin pyá ‘report’ tin ‘put on’ pya ‘show’
သင က θin ʨà ‘teach’ θin ‘learn’ ca ‘hear’

Noun-verb collocations functioning as verbs:

ဝမ သ wùn-θa ‘happy’ wùn ‘stomach’ and θa

‘comfortable’
ဝနခ wun khan ‘admit’ wun ‘burden’ khan ‘accept,
receive’
eဈ ဆစ zè sʰiʔ ‘bargain’ ze ‘price’ sʰiʔ ‘chop off pieces’
လမ eပ က làn pyauʔ ‘be lost’ làn ‘street’ pyauʔ ‘lost’
သeဘ eက င ðəbɔ̀ kàun ‘kind’ ðəbɔ ‘nature’ kàun ‘good’

Reduplicated nouns are never voiced. These are mainly personal names and
kinship terms:

eဖeဖ pʰe-pʰe ‘daddy’

ကiuကiu ko-ko ‘big brother’
eက eက ʨɔ-ʨɔ ‘Kyaw Kyaw’

After the demonstratives and interrogatives di ‘this’, ho ‘that’, bɛ ‘which’,

there is no voicing. Exceptions are the high frequency expressions bɛ-ðu
‘who’, di-hma∼di-ma ‘here’, and bɛ-hma∼bɛ-ma ‘where’. 27
 2 ဒခက di kʰwɛʔ ‘this glass’
Sound system
and script
ဟiuကeလ ho kʰəlè ‘that child’
ဘယ စ bɛ hniʔ ‘which year’

The prefixes ʔə- ‘nominal’ and mə- ‘negative’ do not trigger voicing:

အe ပ ʔəpyɔ̀ ‘speech’
အ ပစ ʔəpyiʔ ‘fault’
မeက င mə-kàun ‘not good’
အeက င ʔəkaun ‘body, corpse, classifier for animals’ cf. tə-gaun
‘one (animal)’

Aspirated consonants are not voiced after weak syllables with the vowel ə.
This holds also for sʰ, though this phoneme otherwise usually merges with s
in most speakers’ pronunciation. Compare the two columns below.

တစeထ င tə-tʰaun ‘one thousand’ cf. င eထ င ŋà-daun ‘five

thousand’
စခ hnə-kʰa ‘two times’ cf. င ခ ŋà-ga ‘five times’

No voicing occurs normally in loanwords, especially recent ones from En-

glish, but also in some Indic loans.

ကနပ တ kunpyuta ‘computer’

ကuမဏ kounpəni ‘company’
eမ eတ က mɔtɔkà ‘motorcar’
ဧက ʔeká ‘acre (English); one (Pali)’
eလ က lɔ̀ká ‘world’
ဓမဒတ dəmá-dutá ‘missionary’

In some cases voicing is not restricted to the medial consonant, but spreads to
the initial as well. This is a kind of assimilation, described in the next section.
Voicing is not reflected in Burmese orthography, apart from informal new
ad-hoc spellings sometimes used in chats and text messages, such as လ ဘ la
bi ‘has come, is coming’ for correct လ ပ. In this grammar, voicing is always
indicated in the transcript and voiceable grammatical morphemes are given
in both forms, voiced and voiceless. Voicing is not indicated in the case of
nasals and liquids, as it seems to be less systematic with these and there
appears to be more individual variation, as in ʔein-dɛ̀-ma ∼ ʔein-dɛ̀-hma ‘in
28 the house’.
2.1.5.2 Assimilation Sound system

Assimilation may occur between two adjoining sounds in close juncture.

Most commonly, a nasal vowel or final glottal stop is assimilated to the
following consonant, that is, it takes the same place of articulation. In the
case of nasal vowels, the nasalization is realized as a nasal consonant, with
the vowel partly or completely denasalized. Frequently, a stop following a
nasal vowel is nasalized, especially in casual speech, as in kàun-dɛ ∼ kàun-nɛ
‘it’s good’ and mə-kàum-bù ∼ mə-kàum-mù ‘it’s not good’.

Final glottal stops are dropped before consonants or assimilated to them in

place and kind of articulation, resulting in long (but not quite geminate) con-
sonants, as in ʔeiʔ-mɛʔ [ʔéimːɛʔ] ‘dream’ and houʔ-kɛ́ [hóukːɛ́] ‘right’. The loss
of the final glottal stop is usually compensated by the creaky tone.

Another kind of assimilation is the spread of voicing to the initial conso-

nant if the word contains a voiced medial consonant. This is seen in words
like dəgɛ ‘really’, dəgà ‘door’, zəgà ‘(spoken) language’. The unvoiced forms
are represented in the orthography as <təkɛ>, <tan.kʰà>, and <səkà>, respec-
tively.

Being more sporadic than intervocalic voicing, assimilation other than voic-
ing is not represented in this grammar, neither in Burmese orthography nor
in the transcriptions.

2.1.6 Regional and sociolectal variants

With probably around 40 million first language speakers and another 10

to 15 million second language speakers, it is evident that Burmese com-
prises a number of local varieties, which differ mostly in pronunciation.
Standard pronunciation is generally taken to be the variety spoken in the
Mandalay area in Upper Myanmar (Anya), though the Burmese language
as spoken in Yangon, the former capital and still the economic center of
the country, is more widely used. Markedly different varieties are spoken
in the peripheral regions, such as Mon and Kayin States in the southeast,
Kayah and Shan States in the east, Kachin State in the north, and, to some
extent, the Irrawaddy Delta. No dialect survey of Burmese is available, apart
from Okell’s¹ description of four Burmese dialects, namely Yaw, Arakanese,
Intha, and Tavoyan, which are different languages rather than dialects of

¹Okell, John. 1989. The Yaw Dialect of Burmese. In Davidson, Jeremy (ed.) South East
Asian linguistics: essays in honour of Eugénie J A Henderson. London: School of Oriental
and African Studies, 199–217. Okell, John. 1995. Three Burmese dialects. In Bradley, David
(ed.) Papers in Southeast Asian Linguistics No. 13: Studies in Burmese languages. Canberra:
Pacific Linguistics, 1–138. 29
 2 the same language, except Yaw, which is very close to Standard Burmese.
Sound system The following observations are therefore very general and no regional pref-
and script
erences can be defined in most cases.

Variation exists among speakers in the pronunciation of the initial conso-

nants, especially sʰ, θ, and ð. For many speakers, sʰ all but merges with s, θ
with t, and ð with d. The merger is in no case complete, though, as the sounds
behave differently in some contexts, especially in close juncture. While s is
voiced also after weak syllables, sʰ does not undergo voicing in this context.
The voiced alveolar stop d is frequently pronounced as an alveolar tap [ɾ]
between vowels, which is never the case with ð.

Voicing and assimilation is more or less regular or prominent in different

regional varieties. Generally, Yangon and Upper Myanmar speakers apply
more Sandhi voicing and assimilation than speakers of the peripheral re-
gions, though there are also local and individual differences. Voicing is re-
stricted to plain initials in Arakanese and does not occur at all in Intha.

Nasal vowels are often less nasalized in the peripheral regions than in Stan-
dard Burmese, and final nasals may appear also in isolated syllables. This
may be due to influence from local languages that lack nasal vowels, such
as Shan and Mon.

A major difference between regional varieties is found in the production of

tones. While in Standard Burmese, as spoken in Yangon and Upper Myan-
mar, the creaky tone is realized as such, that is, with creaky voice but no
clear pitch or contour distinction, it is pronounced as relative high pitch in
southeastern varieties where no creakiness is present. It is possibly shorter
than the other two tones in Yangon and the southeast, but no such length
distinction is found in Upper Myanmar pronunciation.

Young urban speakers of Yangon especially seem to favor English-style pro-

nunciation not only in English loans, but also in indigenous words. Intervo-
calic d often approaches English (American) r, the vowel of weak syllables
is occasionally dropped, and y is pronounced close to [ʒ] in word-initial po-
sition. The dental stop is more fricativized, approaching English th, and the
alveopalatal sounds are more fronted, resembling English j, ch, and sh. In
loanwords, and occasionally in indigenous words, pʰ is pronounced as f, es-
pecially when representing English f. Similarly, English v is pronounced as
such, rather than the naturalized pronunciation bw, as in vidiyo for more tra-
ditional bwidiyo or bidiyo. This speech is especially frequent in popular music
30 and radio talk aimed at a young audience.