Phoneme and Allophone

The nexus between

phonetics and phonology

Robert Mannell
Discriminative Elements (1)

z Trubetzkoy (1939) wrote:-

"It is the task of phonology to study which
differences in sound are related to
differences in meaning in a given language,
in which way the discriminative elements ...
are related to each other, and the rules
according to which they may be combined
into words and sentences."
Discriminative Elements (2)

z Linguistic units which cannot be substituted

for each other without a change in meaning
can be referred to as linguistically contrastive
or significant units.
z Such units may be phonological,
morphological, syntactic, semantic etc.
Discriminative Elements (3)

Logically, this takes the form:-

IF unit X in context A GIVES meaning 1

AND IF unit Y in context A GIVES meaning 2
THEN unit X AND unit Y belong to separate linguistic units

e.g.
IF sound [k] in context [ _ t] GIVES meaning cat
AND IF sound [m] in context [ _ t] GIVES meaning mat
THEN sound [k] and sound [m] belong to separate linguistic units
What is a phoneme? (1)

z Its not true to say that a phoneme is a sound,

or even that its a class of sounds
z Phonemes exist in human brains
z They are abstract cognitive (linguistic) entities
z They are conventions shared by a speech
community but vary, sometimes very
significantly, between speech communities
What is a phoneme? (2)

z You might well ask but surely they have

something to do with speech sounds?
z Well, yes, but indirectly
z When we speak we intend our listeners to
understand what words we have uttered.
z A word in the brain is represented as a
sequence of phonemes (or graphemes for the
written word)
What is a phoneme? (3)

z To communicate a sequence of words we must

utter a sequence of sounds (or write the words
or use sign language )
z A spoken word results from the production of a
sequence of vocal tract gestures
z These gestures result in a sequence of sounds.
z We interpret this sequence of sounds as a
sequence of phonemes
What is a phoneme? (4)

z When we learn a language we learn to

associate sequences of sounds in the
physical world with sequences of phonemes
(and therefore words) in the mental world.
z The same sound may belong to a different
phoneme in a different speech community or
even in a different phonetic context.
What is a phoneme? (5)

z OK doesnt that mean that a phoneme is

realised as a class (or set) of physical sounds?
z Yes, well almost as some sounds can, in
different contexts, belong to (or represent)
different phonemes
z This means that the phonemes are represented
in the physical world by potentially overlapping
sets of sounds
What is a phoneme? (6)

z The set of sounds (in the external acoustic

world) that represent a phoneme can be
referred to as a set of allophones.
z allo- indicates difference, alternation or
divergence (Macquarie Dictionary)
What is a phoneme? (7)

z An allophone is a sound that can represent a

particular phoneme.
z A phoneme can be said to have a huge
number of slightly different allophones.
z This is too complex to work with so we break
this down into a much smaller set of discrete
sounds that we can transcribe phonetically or
measure in some other way.
What is a phoneme? (8)

z One of the reasons for a phoneme having

different allophones is coarticulation.
That is, in different contexts the effect of
adjacent phonemes can affect a phonemes
physical realisation.
z Another reason is convention. A speech
community has an implicit (unconscious)
agreement that certain allophones be used in
certain contexts.
What is a phoneme? (9)

Classical phonology took a simple view of the

relationship between phonemes and allophones:-
z phonemes are contrastive, allophones are not
z an allophone belongs to a single phoneme
z allophones of the same phoneme are in
complementary distribution
z allophones are phonetically similar

We will examine the last three points in more detail

later, but first we will examine phonemic analysis.
Phonemic Analysis

z Phonemic analysis uses a narrow transcription

of the speech of a language to determine what
are the phonemes (the their allophones) for
that language.
z The phonemic analysis relies on the
assumption that (a) the transcribed words have
different meanings and (b) the transcription
reliably captures the languages sound system.
Minimal pairs (1)

z Phonemes are the linguistically contrastive or

significant sounds (or sets of sounds) of a
language.
z Such a contrast is usually demonstrated by
the existence of minimal pairs or contrast in
identical environment (CIE).
z The search for minimal pairs is the most
important strategy in phonemic analysis.
Minimal pairs (2)

z Minimal pairs are pairs of words which vary

only by the identity of the segment1 at a single
location in the word (eg. [mt] and [kt]).
z If two segments contrast in identical
environment then they must belong to different
phonemes. That is, if we change one sound to
another and it changes the meaning then the
sounds belong to different phonemes.
(1) segment is another word for single speech sound
Minimal pairs (3)

z A paradigm of minimal phonological

contrasts is a set of words differing only by
one speech sound.
z In most languages it is rare to find a
paradigm that contrasts a complete class of
phonemes (eg. all vowels OR all consonants
OR all stops, etc.).
Minimal pairs (4)

e.g. English oral stops

z the English oral stop consonants could be defined by the
following set (paradigm) of minimally contrasting words:-
i) /pIn/ vs /bIn/ vs /tIn/ vs /dIn/ vs /kIn/
z Only // does not occur in this paradigm and at least one
minimal pair must be found with each of the other 5 stops to
prove conclusively that it is not a variant form of one of them.
ii) /n/ vs /pn/ vs /bn/ vs /tn/ vs /dn/
z Again, only five stops belong to this paradigm. A single minimal
pair contrasting // and /k/ is required now to fully demonstrate
the set of English stop consonants.
iii) /{In/ vs /k{In/
Contrast in Analogous Environment (1)

z Sometimes it is not possible to find a minimal

pair which would support the contrastiveness
of two phonemes and it is necessary to resort
to examples of contrast in analogous
environment (CAE).
z CAE is almost a minimal pair, however the
pair of words differs by more than just the
pair of sounds in question.
Contrast in Analogous Environment (2)

z Preferably, in CAE, the other points of

variation in the pair of words are as remote
as possible (i.e. not adjacent and preferably
not in the same syllable) from the pair of
sounds being tested.
z The further away the other contrast is, the
more unlikely it is to have any conditioning
effect on the selection of pair phones of
interest.
Contrast in Analogous Environment (3)

z eg. // vs // in English are usually supported

by examples of CAE pairs such as
"pressure" [pre] vs "treasure" [tre].
z The only true minimal pairs for these two
sounds in English involve at least one word
(often a proper noun) that has been borrowed
from another language (eg. "Confucian"
[knfjn] vs "confusion" [knfjn], and
"Aleutian" [ln] vs "allusion" [ln]).
Minimal pairs versus CAE

z Even one example of a minimal pair might be

considered good evidence that two sounds
are allophones of different phonemes.
z Minimal pairs are reliable evidence in
phonemic analysis.
z CAE is poorer evidence in phonemic analysis
and ideally requires other supporting
evidence (eg. similar patterns for similar
pairs of sounds confirmed to contrast by CIE)
Syntagmatic analysis (1)

z A syntagmatic analysis of a speech sound

identifies all of the locations or contexts within
the words of a particular language where the
sound can be found.
z Note that in the following examples (next
page), "#" is used to represent a word or
syllable boundary, "V" represents any vowel,
and "C" represents any other consonant.
Syntagmatic analysis (2)

For example, English [n] and [] :-

z a syntagm of the phone [n] in English could be
in the form:-
( #CnV..., #nV..., ...Vn#, ...VnC#, ...VnV..., etc.)
z whilst [] in English would be:-
(...V#, ...VC#, ...VV..., etc)
but would not include the word initial forms.
Syntagmatic analysis (3)

z For example, sequences of the type

"#CnV..." would include "snow" [sn],
"snort" [sno:t] and "snooker" [sn:k].
z In this case, the only consonant (for English)
that can occupy the initial "C" slot is the
phoneme /s/, and so the generalised pattern
could be rewritten as "#snV...".
Syntagmatic analysis (4)

z A syntagmatic analysis can provide information

about different restricted distributions of two
allophones of the same phoneme (always found in
different locations).
z A syntagmatic analysis might help to strengthen a
case based on a CAE analysis. For example, we
might find CAE for [p, b] and we find that all other
oral stop pairs (e.g. [t, d] and [k, g]) have minimal
pairs in the same syllable location. So this would
strengthen the case for [p, b] being separate
phonemes.
Complementary Distribution (1)

z A phoneme may be realised by more than

one speech sound and the selection of each
variant is usually conditioned by the phonetic
environment of the phoneme.
z This is known as mutually exclusive or
complementary distribution (CD)
Complementary Distribution (2)

z The CD of two phonemes means that the two

phonemes can never be found in the same
environment (ie. the same environment in the
senses of position in the word and the
identity of adjacent phonemes).
z If two sounds are phonetically similar and
they are in CD then they can be assumed to
be allophones of the same phoneme.
Complementary Distribution (3)

z eg. in many languages voiced and voiceless stops

with the same place of articulation do not contrast
linguistically but are rather two phonetic realisations
of a single phoneme (ie. /p/=[p,b], /t/=[t,d], and
/k/=[k,]).
z Whether the voiced or voiceless allophone is chosen
depends upon syntagmatic distribution (e.g. where in
the word) or phonetic context (e.g. whether the
adjacent sounds are voiced or voiceless)
Complementary Distribution (4)

z In some Australian Aboriginal languages word medial

oral stops are voiced if both adjacent phonemes are
voiced (e.g. between two vowels) and are voiceless if
at least one of the adjacent sounds is voiceless.
z For initial stops the patterns varies from language to
language and even between dialects within a single
language. In some dialects of a language the
voiceless allophone is preferred, in others the voiced
allophone is preferred, and in others the choice of
allophone is a matter of individual choice.
Contrastive Distribution

z Contrastive distribution is the opposite of

complementary distribution.
z Sounds in contrastive distribution can occur
in the same location and when exchanged
change the meaning of the word.
z Sounds in contrastive distribution belong to
different phonemes.
Free Variation

z Occasionally speakers of a language are free

to choose whether they use one or another of
two possible allophones. The choice may be
word-specific (often a sign of language change
in progress). The choice may be pragmatic
(discourse context) or sociolinguistic
(e.g. Some French speakers choose to use the
alveolar trill [r] when in the village and the more
prestigious uvular trill [] when in Paris.)
Phonetic similarity (1)

z Allophones must be phonetically similar to

each other.
z In analysis, this means you can assume that
highly dissimilar sounds are separate
phonemes (even if they are in complementary
distribution).
z For this reason no attempt is made to find
minimal pairs which contrast vowels with
consonants. BUT
Phonetic similarity (2)

z Even the distinction between vowel and

consonants isnt totally unproblematic.
z For example, a sequence in one language
which is perceived as a diphthong might in
another language be perceived as a semi-
vowel (approximant) and a vowel
e.g. /ja/ ~ /ia/ OR /aj/ ~ /aI/
z What happens when the speakers of one of
these languages learns the other language?
What is the interlanguage phonology?
Phonetic similarity (3)

z Exactly what can be considered phonetically

similar may vary somewhat from language to
language and so the notion of phonetic
similarity can seem to be quite unclear at times.
z Sounds can be phonetically similar from both
articulatory and auditory points of view. One
finds pairs of sounds that vary greatly in their
place of articulation but are sufficiently similar
auditorily to be considered phonetically similar.
Phonetic similarity (4)

z According to Hockett (1942), "...if a and b are

members of one phoneme, they share one or
more features".
z Phonetic similarity is therefore based on the
notion of shared features.
z Such judgments of similarity will vary from
language to language and there are no
universal criteria for similarity.
Phonetic similarity (5)

z Example 1. Glottal [h] and palatal [] are

voiceless fricatives which are distant in terms
of places of articulation, but they share
features and are sufficiently similar auditorily
(both weak sounding, voiceless, non-tongue-
tip, non-labial, fricatives) to be allophones of
a single phoneme in some languages such
as Japanese.
Phonetic similarity (6)

z Example 2. In English, /h/ and // are in

complementary distribution. /h/ only ever occurs
at the beginning of a syllable (head, heart,
enhance, perhaps) whilst // only ever occurs at
the end of a syllable (sing, singer, finger).
z The differ in place, manner and voicing and in
tongue body, velum and laryngeal gestures.
They are so different that no one regards them
as allophones of the one phoneme.
Phonetic similarity (7)

z So, if phonetic similarity is so difficult to pin

down, then is it a useful tool in phonemic
analysis.
z Yes, it is. What we need to be aware of are
the common patterns of phonetic similarity
across languages. What follows are some of
the common (but in no case universal)
patterns
Phonetic similarity (8)

z i) two sounds differing only in voicing:

[pb] [td] [k] [] [] [sz] [] [x] etc...
z ii) two sounds differing in manner of
articulation only as oral stop vs fricative. The
sibilant or grooved fricatives [s,z,,] are
excluded from this category as they are quite
different auditorily from the other fricatives.
[p] [kx] [b] [] etc...
Phonetic similarity (9)

z iii) Any pairs of consonants close in place of

articulation and differing in no other
contrastive feature:
[s] [z] [n] [l] [l] [m], etc...
z iv) Any other pairs of consonants which are
close in articulation and differ by one other
feature but are nevertheless frequently
members of the same phoneme
[l] [c] [t] [d]
Phonetic similarity (10)

z In languages where voicing is non-

contrastive, two phones differing in voicing
and only slightly in place of articulation might
be considered similar e.g. [c] etc.
z Further, for the purposes of this type of
analysis, the place of articulation of the
apicodental fricatives [,] is considered to
be close enough to that of the alveolar stops
[t,d] to be considered phonetically similar.
Phonetic similarity (11)

z v) Any two vowels differing in only one

feature or articulated with adjacent tongue
positions
[ ] [i I] [: ] [i y] [ ]
Phonetic similarity (12)

z Although it is implied above that the notion of

"phonetic similarity" is in some way less
linguistically abstract (more phonetic?) than
the notion of complementary distribution, it is,
nevertheless, a quite abstract concept.
z The are no obvious and consistent acoustic,
auditory or articulatory criteria for phonetic
similarity.
Phonetic similarity (13)

z There are many examples of very similar phones

which are perceived by native speakers to belong to
separate phonemes.
z In English, for example, a word terminal voiceless stop
may be either released and aspirated or unreleased.
The homorganic (same place of articulation) voiced
stop may also be released or unreleased.
z Often the unreleased voiced and voiceless stops may
actually be identical in every way except that the
preceding vowel is lengthened before the
phonologically voiced stop.
Phonetic similarity (14)

z In terms of phonetic similarity, the two

unreleased stops may actually be identical
and yet be perceived by native speakers to
belong to different phonemes because of
their different effect on the preceding vowel.
z /kp/[kp] ... [kp ]
/kb/[kb] ... [kb ] ... [kp ]
(nb. " " means unreleased stop and " "
means partially lengthened vowel)
Phonetic similarity (15)

z Conversely, phones which are very dissimilar

(at least from certain perspectives) may be
felt by native speakers to belong to a single
phoneme.
z eg. Japanese(1)
/h/ [] before /u/ eg.[uku] "luck
/h/ [] before /i/ eg.[ito] "man
/h/ [h] before /e,a,o/ eg.[hana]
(1) Japanese in the mid 20th century. This pattern has undergone recent change.
Phonemic Pattern

z A pair of phones in complementary distribution may

sometimes be classified into separate phonemes on
the basis of phonemic pattern.
z In other words, is there a group of phonemes which
exhibit a similar pattern of distribution (eg. clustering
behaviour, morphology, etc.) to one of the phones
being examined?
z In the case of the pair [h], [] there are some
similarities in patterning between [h] and certain
fricatives, and between [] and the other nasal stops.
Phonological Space (1)

z The greater the distance between a

phoneme and its nearest neighbours, the
greater the scope for allophonic variation.
z In other words, the larger the number of
redundant features (ie. features which when
changed will not create another phoneme)
the greater the number of allophones which
can actually occur.
Phonological Space (2)

z eg. English
/p/ = [-voice] [+bilabial] [+stop] [-nasal] [+/-aspirated]
(nb. + present, - absent, +/- optional)
z Changing the feature [-voice] to [+voice] will create /b/,
changing the feature [bilabial] may create /t,k/,
changing the feature [stop] may create /w,f/,
changing the feature [nasal] will create /m/.
z The only feature with complete freedom of movement
is aspiration, and variation of this feature does create
the main pair of allophones of this phoneme in English.
Phonological Space (3)

z eg. English
z /r/ [] alveolar approximant
z /r/ [R&] voiceless alveolar approximant (e.g. after /t/)
z /r/ [ ] retroflex approximant (West England)
z /r/ [] alveolar flap (Scottish) eg. [In]
z /r/ [] uvular fricative (Tyneside, UK)
z The possible varieties of /r/ seem to include variations
of manner, place and voicing. The only restrictions are
that its allophones may not overlap with those of /l/
and /w/.
Allophones & Narrow Transcription (1)

z It should now be obvious that broad

(phonemic) transcription is only possible
following phonemic analysis. We cant
transcribe the phonemes until we know what
they are.
z To what extent do narrow transcriptions
match the result of the identification of
allophones in phonemic analysis?
Allophones & Narrow Transcription (2)

z No narrow transcription captures all

allophones of each phoneme. As we have
already seen, each phoneme has an
extremely large number of slightly varying
allophones.
z In this course we greatly simplify narrow
transcription and ignore a lot of significant
Australian English allophones.
Allophones & Narrow Transcription (3)

z What might be a maximal narrow transcription

of a language (e.g. Australian English)?
z The maximal set of possible distinctions that
we can transcribe is limited by the availability
of IPA symbols, including the diacritics and by
our ability to hear (or otherwise measure)
these differences.
Allophones & Narrow Transcription (4)

z For example we can use, not only the IPA

vowel symbols but also the raising, lowering,
fronting and backing diacritics to indicate a
large variety of vowel qualities.
z We can also provide diacritics for vowel
nasalisation, breathy voice and creaky voice
as well as voiceless (for whispered vowels).
z For Arabic vowels we could add a velarised
diacritic (for velarised vowels).
Allophones & Narrow Transcription (5)

z Traditionally, phonemic analysis was limited

by the accuracy and degree of detail of a
narrow transcription. If a feature was missed
in a narrow transcription it wasnt part of the
phonemic analysis.
z Today we can add detail from acoustic and
physiological measurement when judging
allophonic variation.
Phoneme and Allophone

z This is a complex topic. You will not become

sufficiently familiar with it without practice in actual
phonemic analysis.
z You should review the unit web site. The web site
has some additional material not covered here
(especially The Premises of Phonemic Analysis).
z Read and try all of the phonemic analysis tutorial
exercises, including the complex ones. Solutions will
be available after the relevant tutorial.
Readings

z The unit web site:-

http://www.ling.mq.edu.au/speech/phonetics/phonology/phoneme/index.html
z Clark, J., Yallop, C., & Fletcher, J. (2007), An introduction to
phonetics and phonology, Blackwell (3rd edition).
References

Referred to in preparation of these notes, but not required reading

z Hockett, C.F. (1942) A System of Descriptive Phonology, Language,

18(1), 3-21
z Pike, K.L. (1947) Phonemics, U.Michigan
z Trubetzkoy, N.S. (1939) Grundzge der Phonologie. Travaux du
Cercle Linguistique de Prague 7, Reprinted 1958, Gttingen:
Vandenhoek & Ruprecht. Translated into English by C.A.M.Baltaxe
1969 as Principles of Phonology, Berkeley: University of California
Press.