Data Pattern #1 – Minimal Pairs

Based on definition: Phomemes are contrastive.

When you have two or more sounds, found in identical 

environments, (transcribed identically except for one sound) that 

mean different things.

Example:

[bij] ‘bee’   [phij] ‘pea’/‘pee’

The sounds [b] and [ph] contrast in minimal pairs and thus belong 

to separate phonemes.

[suw] ‘sue’    [zuw]  ‘zoo’

The sounds [s] and [z] contrast in minimal pairs and thus belong to 
separate phonemes.

Data Pattern #2 –
Complementary Distribution, cont.

Some General Enviornments

 & Notations:

1. Word boundaries 

¡word initial   #_____

¡word final    ______#

2. Before certain sounds; before sound classes

¡___i, ____s, ____ N, ___ palatals, ___C, ___V

3. After certain sounds; after sound classes

¡i ___, s ____, N ____, palatals ___, C___, V___

4. Between certain sounds; sound classes

¡i____i, s____s, N___N, pal.___pal., C__C, V__V

Data Pattern #2 – 

Complementary Distribution, cont.

Example: Spanish

[dexo]    ‘finger’

[donde]  ‘where’ 

[naxa]  ‘nothing’

[pixe]   ‘3p sing. asks’

[pwexe]  ‘3p sing. can’ 

[pared]   ‘wall’

[r]= flap (symbol won’t show)

Also, [d] = dental

Note that phones are phonetically similar

¡Both dental (dental vs. interdental)

¡Both voiced

[d] occurs in:  elsewhere

  #___e

  #___o

  n___e

  e___#

[x] e___o  V___V

  a___a

  i___e

  e__e

 

The phones [d] and [x]  are found in 

complementary distribution and thus belong to

 the same phoneme. [d] and [x] are allophones of the same phoneme /d/.

/d/

[x]  V_V

[d] elsewhere

Data Pattern #3 – 

Analogous Environments

Based on definition: Phomemes are contrastive.

This pattern is when you find two sounds in 

nearly the same environments, but do not have 

data for a minimal pair.  The appearance of these 

two sounds does not seem to be conditioned by 

any specific phonetic environments, because 

they both occur in analogous environments.

Data Pattern #3 –
Analagous Environments, cont.

lExample:

Environment of [s]  Environment of 

#___a  #____a

i___a  i_____a

  i_____a

These phones are found in analogous 

environments and thus belong 

to different phonemes. These sounds are 

allophones of different 

phonemes.

Data Pattern #4 –
Free Variation

Based on definition: Phomemes are realized by allophones

This pattern is when you find two sounds in the same environments, and there does not seem to be a change in meaning. In other words, it appears that the phones can interchange with one another with no effect in meaning.  This is usually a result of phonetic distinction that you perceive that is not contrastive in this language.

Data Pattern #4 –
Free Variation, cont.

Example: English

Both of these phones are found in identical 

environments, and yet, unlike the analogous 

environments examples, alternation of these 

phones does NOT create differences in meaning.

Thus, both the [p] and the [p] (unreleased) 

are allophones of, and thus belong to, the 

same phoneme.

To summarize…

To show you have allophones of the same phoneme:

1.Complementary Distribution

¡phones occur in unique, separate, 

individual environments

¡environment of sounds conditions 

allophone

2.Free Variation

¡Phones occur in the exact same 

environments or near-identical 

environments but don’t change 

meaning

¡Seem to be used interchangeably, or 

substitute for one another,

are realized as allophones.

Phonemes

To show you have allophones of different phonemes:

1.Minimal Pairs –

¡sounds occur in exactly the same 

environments – in an IDENTICAL set 

- AND changes meaning of words

¡

2.Analogous Environments

¡Phones occur in overlapping 

environments, and in near-identical 

environments, but word meanings are

 different

●

Phonemes are contrastive.

Hardman’s Definition of the Phoneme

Negative Definition /Contrastive Definition

Phonemes act to keep words separate.

Phonemes make words contrast.

A Phoneme is what it is because it is not something else.

Positive Definition/ Realization Definition

Phonemes are built up of a range of phones that trigger our perception.

This definition accounts for physical realization of phonemes.

Speech organs builds sounds which trigger phonemes.

Structural Definition

Phonemes build the structure of language and are part of the whole

 phonological structure of a language.

Humans have propensity for symmetry in language.

There is a linguistic tendency to make use of some phonetic 

possibilities to make contrasts and to use them consistently.

Analyzing Data to Determine Phonemes

There are four general patterns you will find in your data that help you to determine phonemic status.

With extensive data sets, there is a fifth pattern that also appears.

These patterns are built around the definitions of the phoneme itself.

Analyzing Data to Determine Phonemes

You analyze phonetic data to determine phonemic status.

Phonemic status means how native speakers perceive these 

sounds, and thus, how these sounds are organized in this language.

lDo native speakers “hear” a difference and use them 

contrastively? (allophones of different phonemes – these sounds 

belong to different phonemes)

lDo native speakers not “hear” a difference and thus do not use 

them contrastively? (allophones of the same phoneme – these 

sounds belong to the same phoneme)

So remember, 2 languages might make use of the same 

sounds [e] and [i], but organize and perceive those sounds 

differently.

Language A: /e/  /i/  language B:  /i/

      [e]           [i]  [e]

  [i]

A: 2 sounds belong to different phonemes, native speakers 

perceive them as different and they can be used to build 

differences in words, [e] and [i] belong to different 

phonemes, /e/ and /i/, respectively

B: 2 sounds belong to the same phoneme, native speakers 

do NOT perceive them as different, they may not be used to 

build differences in words, [e] and [i] belong to the same 

phoneme, /i/

That is your goal in phonemic analysis –

to determine the perception

and organization of sound

by a native speaker.

Phonemics

Phonemics, or Why Phonetic is so hard…

You do not hear physical sound directly.

If you did, phonetics would be easy.

Instead, you perceive all speech sounds through the sound system of your native language(s) and the languages you have studied.

What does

that mean?

You perceive speech sounds through structure.

When you deal with sound outside of the structure you are used to, it can become confusing and difficult to even perceive a sound.

When you hear human speech sounds, these sounds automatically “trigger” perceptual units in your brain/mind.

These units are abstractions and are used to organize and structure the “sounds” of your native language.

The phoneme is the basic unit of organization of sounds in language.

The phoneme is an abstract, structural and perceptual unit of speech.

To put another way, when

 someone utters a physical 

speech sound to you, that 

speech sound triggers a 

phoneme in your mind – you 

do not “hear” the phonetic 

distinctions directly. 

Instead, the sound triggers a perceptual unit and you

 perceive the sound as that unit.

When native speakers (without formal linguistics 

training) say they “hear no difference” between two 

sounds, it is probably because in their language 

those two sounds trigger the same perceptual unit.

Those two sounds belong to the same phoneme.

You will generally not “hear” a difference between two sounds that belong to the same phoneme.

You will generally “hear” a difference between two sounds who belong to different phonemes.

Phonemes are used to build words and contrast “sound unit” from “sound unit.”

Classic English Phoneme Example

Key Vocabulary in Phonemics

lPhoneme – abstract structural and perceptual unit /     /

lPhone – phonetic speech sound, unanalyzed according to 

phonemic status

  [ ]

lAllophone – after analyzing data, the phonetic speech 

sound that belongs to, and thus triggers, a phoneme

 [  ]  under /  /

Writing phonemes and the allophones that realize them:

/ / phoneme (choose one allophone as symbol)

  [ ] Allophone 1

  [  ] Allophone 2    etc.

Generally, when there is more than one allophone, each will occur in its own environment – if that  is the case, list environment, too

Analyzing Data to Determine Phonemes