Phones, Phonemes, and Allophones
Chapter Objectives
By the end of this chapter, you will be able to:
- Distinguish a phone, a phoneme, and an allophone.
- Explain contrast and describe complementary and conditioned distribution at an introductory level.
- Distinguish a minimal pair from a near-minimal pair.
- Interpret slashes / / and square brackets [ ] as two different analytical claims, not two ways of writing the same thing.
- Distinguish a citation form from a contextual realization.
- Analyze a small, self-contained dataset and justify whether a contrastive or allophonic interpretation is better supported.
- Avoid treating a spelling difference as proof that two sounds contrast.
An Observable Problem
Say the English words pin and spin slowly, paying attention only to the first consonant sound in each — not the spelling, the sound. If you hold the back of your hand close to your lips as you say each word, you may notice a small puff of air after the p in pin that is largely absent after the p in spin. These are two physically different sounds. Yet most English speakers do not perceive them as different "sounds" of English at all — both simply count as the consonant written p.
Now compare pin and bin. Here the difference between the first sounds is immediately obvious to any English speaker, and swapping one for the other changes the word's meaning. Something is clearly different about these two cases, even though in both cases two physically distinct sounds are involved. This chapter gives you the analytical tools to explain exactly what that difference is.
- Before reading further, write one sentence predicting why pin/spin might feel like "the same sound" while pin/bin feels like "different sounds," even though both pairs involve a physical difference in pronunciation.
Phones: Physical Realizations
Purpose: establish the most basic unit of phonetic observation, independent of meaning or category.
A phone is a single speech sound as it is actually produced and observed, described in purely physical and perceptual terms — articulation, acoustics, or audible quality. A phone is not yet claimed to belong to any particular category in a particular language's sound system. It is simply "a sound that occurred."
Phonetic transcription of a phone is written in square brackets [ ]. The aspirated p in pin and the unaspirated p in spin are two different phones:
- [pʰɪn]pin — aspirated [pʰ], a release of audible air after the closure
- [spɪn]spin — unaspirated [p], little or no audible release of air
Notice that describing these as two phones makes no claim yet about whether English treats them as "the same sound" or "different sounds" in any deeper sense. That claim belongs to the next concept.
A phone is a description of what a speaker did and what a listener could, in principle, perceive. It says nothing on its own about meaning, contrast, or mental category. Two phones can be physically quite different and still turn out to function as versions of "the same sound" in a given language — which is exactly what happens with English [pʰ] and [p].
Phonemes: Contrastive Categories
Purpose: identify the abstract sound categories a language's speakers treat as functionally distinct.
A phoneme is an abstract sound category in a particular language that contrasts with other categories — meaning that swapping one phoneme for another can change a word's meaning. Phonemes are not physical events; they are analytical categories that a language's sound system is built from.
Phonemic transcription is written in slashes / /. Compare:
- /pɪn/ vs. /bɪn/pin vs. bin — swapping /p/ for /b/ changes the word, so /p/ and /b/ are separate phonemes of English
This is the test for contrast: if substituting one sound for another, in an otherwise identical word, produces a different word with a different meaning, the two sounds represent different phonemes in that language.
Contrast is always evaluated within one language or variety. A distinction that is contrastive in one language may not be contrastive in another, even if both languages contain the same physical phones. Aspiration is exactly this kind of case: in English, [pʰ] and [p] never distinguish separate words, so they belong to a single phoneme, /p/. In other languages, such as Thai or Hindi, an aspirated stop and an unaspirated stop at the same place of articulation can distinguish separate words — meaning aspiration is contrastive, and the two phones belong to two different phonemes in those languages.
Reading the Delimiters as Claims
Purpose: treat the choice of bracket as a meaningful analytical decision, not a typographic preference.
Square brackets and slashes are not interchangeable. Each delimiter makes a different claim about what the transcription represents:
| Delimiter | What it claims | Example |
|---|---|---|
| [ ] | A phonetic transcription: an observed or stipulated physical realization, at whatever level of detail the task requires. | [pʰɪn] |
| / / | A phonemic transcription: the contrastive category a form belongs to, abstracting away predictable phonetic detail. | /pɪn/ |
You will sometimes see the same word transcribed two different ways depending on the delimiter, and both can be correct at once because they are answering different questions:
- /pɪn/phonemic: which contrastive categories make up this word
- [pʰɪn]phonetic: how a speaker actually realized the first phoneme in this context
An angle-bracketed form such as ⟨pin⟩, introduced in Chapter 2, represents the spelling, which is a third, separate kind of information. Do not place ordinary orthography inside phonetic or phonemic delimiters.
Error: Treating slashes and brackets as stylistic alternatives, for example writing /pʰɪn/ when the aspiration detail is meant to mark a contrast.
Correction: Predictable, non-contrastive detail like English aspiration belongs inside square brackets, not slashes: [pʰɪn] phonetically, /pɪn/ phonemically. Including ʰ inside slashes would incorrectly imply that aspiration is contrastive in English.
Allophones and Distribution
Purpose: explain how a single phoneme can correspond to more than one phone, and how those phones are distributed.
An allophone is one of the phonetic realizations of a phoneme. English /p/ has at least two allophones relevant to this chapter: [pʰ] and [p]. Whether a phoneme's allophones are predictable, and where each one occurs, is described using the concept of distribution.
Complementary distribution
Two phones are in complementary distribution when each occurs only in environments where the other does not — their distributions never overlap. English aspiration follows a complementary pattern at an introductory level of description: [pʰ] occurs at the start of a stressed syllable, and [p] occurs after [s] in the same syllable.
- [pʰɪn], [pʰaɪ]pin, pie — [pʰ] at the start of a stressed syllable
- [spɪn], [spaɪ]spin, spy — [p] after [s]
Because the environments do not overlap, and because no pair of words is distinguished only by choosing [pʰ] instead of [p] in the same environment, both phones are treated as allophones of one phoneme, /p/, rather than as separate phonemes.
Conditioned variation, more generally
Complementary distribution is one common pattern, but allophonic variation can also be described more broadly as conditioned: the choice of allophone is predictable from the phonetic context, even in cases that are easier to state as a rule than as a strict either/or split. At this introductory stage, it is enough to recognize that an allophone's appearance is governed by context rather than by free choice or by meaning.
Claims about distribution should be checked against a reasonably full set of examples, not a single pair of words. A pattern that holds for pin and spin should also be checked against further data, such as pat/spat or pool/spool, before it is treated as a general rule. The dataset analysis later in this chapter gives you practice doing exactly this.
Minimal and Near-Minimal Pairs
Purpose: use word pairs as evidence for or against a contrast claim.
A minimal pair is a pair of words that differ in meaning and in exactly one sound, in the same position, with everything else identical. Minimal pairs are strong evidence that the differing sound represents two separate phonemes.
- /pɪn/ — /bɪn/pin — bin: differ only in the first consonant
- /pɪn/ — /pɪt/pin — pit: differ only in the final consonant
A near-minimal pair differs in more than one way, but the additional differences are independently controlled or judged not to affect the relevant comparison. Near-minimal pairs are useful when a true minimal pair is unavailable, but they must be labeled as near-minimal rather than presented as equivalent to a true minimal pair.
- /ʃɪp/ — /tʃɪp/ship — chip: a near-minimal pair some analysts use for /ʃ/ vs. /tʃ/, since the difference involves more than a single segment
Error: Concluding that two sounds are different phonemes because the words containing them are spelled differently.
Correction: Spelling is not phonetic or phonemic evidence. Phase and faze are spelled differently but are pronounced identically in many varieties of English, /feɪz/. Always test contrast using sound, not letters.
Citation Form Versus Contextual Realization
Purpose: distinguish an idealized, careful pronunciation from how a form is actually realized in connected speech.
A citation form is how a word is pronounced in isolation, carefully and deliberately — the form most likely to appear in a dictionary entry. A contextual realization is how that same word is actually produced within a phrase or sentence, where neighboring sounds, speech rate, and stress can all influence the outcome.
- [hæv]have — citation form, spoken in isolation
- [əv]have — a common contextual realization in fast, casual speech, as in "should have gone," informally written should've
Citation forms are a useful, controlled starting point for analysis, but a transcription of connected speech should not be assumed to match the citation form automatically. Chapter 23 returns to this relationship in much greater detail when you study coarticulation and connected-speech processes; for now, it is enough to recognize that the two are not the same thing and that a transcription task must specify which one is intended.
Guided Observation
Before working with the dataset below, complete this short self-observation task. No audio recording is required — you are listening to your own speech.
Say each pair of words aloud at a normal, careful pace. For each pair, note whether you can feel or hear a difference in the first consonant, and describe that difference in your own words before reading any further.
- top vs. stop
- kit vs. skit
- tar vs. star
Now ask a partner, or record and play back your own speech, to check whether the pattern you found for p in this chapter's main discussion also appears to hold for t and k. Write one sentence stating your prediction before moving to the dataset analysis.
Dataset Analysis Lab
Purpose: apply contrast, distribution, and minimal-pair reasoning to a small, self-contained dataset.
The forms below are verified phonetic transcriptions of English words, General American reference variety, spoken in careful citation style. Use only this dataset to answer the questions that follow — do not rely on outside knowledge of English spelling.
| Orthographic form | Phonetic transcription | Environment |
|---|---|---|
| ⟨top⟩ | [tʰɑp] | Stressed syllable onset |
| ⟨stop⟩ | [stɑp] | After [s] |
| ⟨kit⟩ | [kʰɪt] | Stressed syllable onset |
| ⟨skit⟩ | [skɪt] | After [s] |
| ⟨tar⟩ | [tʰɑɹ] | Stressed syllable onset |
| ⟨star⟩ | [stɑɹ] | After [s] |
| ⟨din⟩ | [dɪn] | Stressed syllable onset |
| ⟨tin⟩ | [tʰɪn] | Stressed syllable onset |
Source: forms checked against a standard American English pronouncing reference for General American citation pronunciation. The transcription level is broad-to-moderate phonetic detail, sufficient to show the contrast under analysis.
Step 1 — Identify the recurring phones
List every distinct phone that appears in the onset position of these eight words. You should find more than three, even though English spelling uses only the letters t, k, and d to represent them.
Step 2 — Test for complementary distribution
For [tʰ] and [t], state the environment in which each occurs. Do the two environments overlap anywhere in this dataset?
Step 3 — Test for contrast
Compare din [dɪn] and tin [tʰɪn]. Is this a minimal pair? Does it provide evidence that [d] and [tʰ] belong to separate phonemes, or to the same phoneme? Justify your answer using the contrast test introduced earlier in this chapter.
Step 4 — Write the analysis
In two to three sentences, state which phones in this dataset are allophones of a single phoneme and which represent separate phonemes. Support each claim with a specific form from the table, and use the correct delimiter, / / or [ ], for each claim you make.
Practice — Phonemic versus phonetic: Decide whether each transcription below should be written phonemically or phonetically as given, then correct any that use the wrong delimiter for the level of detail shown.
- /stɑp/
- /kʰɪt/
- [tɪn] (used to represent the predictable allophone after [s] in a word like stin, a nonsense form for this exercise)
- /dɪn/
Common Mistakes
Error: Treating "phoneme" and "letter" as the same kind of unit, for example assuming English has exactly 26 phonemes because it has 26 letters.
Correction: Phonemes are sound categories established through contrast, not orthographic symbols. The number of phonemes in a language's sound system is established through phonological analysis, not by counting letters, and it commonly does not match the number of letters used to write that language.
Error: Describing phonemic transcription as simply a less careful or less accurate version of phonetic transcription.
Correction: Phonemic transcription is not an imprecise phonetic transcription; it answers a different analytical question by deliberately omitting predictable detail that does not affect contrast. Both levels can be equally rigorous for their stated purpose.
Error: Assuming that a contrast established for one language or variety automatically holds in another.
Correction: Contrast must be reestablished for each language or variety under analysis. Aspiration is non-contrastive in English but contrastive in some other languages; neither pattern is more natural or more correct than the other.
Chapter Review
Part A — Definitions
1. In your own words, define phone, phoneme, and allophone, and state how the three terms relate to one another.
2. Explain the difference between what [ ] and / / each claim about a transcription.
3. Define minimal pair and give an original English example not used elsewhere in this chapter.
Part B — Applying the Contrast Test
1. Sip [sɪp] and zip [zɪp] differ only in the first sound. Apply the contrast test and state whether [s] and [z] are separate phonemes in English. Justify your answer.
2. Explain why spelling alone cannot answer Question 1.
Part C — Distribution
1. Define complementary distribution in your own words.
2. Using the dataset from this chapter's Dataset Analysis Lab, state the environments for [kʰ] and [k] and explain whether they are in complementary distribution.
Part D — Citation Form and Context
1. Explain the difference between a citation form and a contextual realization, using the have example from this chapter or an original example.
Repaso acumulativo / Cumulative Review
1. Using Chapter 1's distinction among observation, inference, and judgment, classify the following statement: "The speaker realized /t/ as [tʰ] because that word began a stressed syllable." Is this an observation, an inference, or a judgment? Explain your reasoning.
2. Using Chapter 3's delimiter conventions, correct any delimiter errors in this string: /pʰɪn/, intended as a careful phonetic transcription.
Chapter Performance Task
Performance Task
Building a Small Contrastive Dataset
Choose one consonant or vowel distinction in English that interests you (for example, the [l] sounds in light versus full, previewed further in Chapter 11). Working independently or with a partner:
- Collect at least four words illustrating the distinction, written in ordinary spelling.
- Transcribe each word phonetically, in square brackets, at a level of detail sufficient to show the distinction.
- State whether the distinction appears to be contrastive or allophonic in English, and justify your conclusion using the contrast test or a distribution argument.
- Identify at least one piece of evidence that would strengthen or challenge your conclusion.
Answer Key (click to expand)
Dataset Analysis Lab
Step 1: The onset phones in this dataset are [tʰ], [t], [kʰ], [k], and [d] — five distinct phones, even though spelling uses only three letters (t, k, d) to represent them.
Step 2: [tʰ] occurs at the start of a stressed syllable (top, tar, tin); [t] occurs after [s] (stop, star). The environments do not overlap in this dataset, so the two phones are in complementary distribution.
Step 3: Din [dɪn] and tin [tʰɪn] is a minimal pair: they differ in meaning and in exactly one sound in the same position, with the rest of the word identical. This is evidence that [d] and [tʰ] belong to separate phonemes, /d/ and /t/, since substituting one for the other changes the word.
Step 4 (model answer): [tʰ] and [t] are allophones of a single phoneme, /t/, because they are in complementary distribution and no minimal pair in the dataset distinguishes them in the same environment, as shown by [tʰɑp] versus [stɑp]. By contrast, [d] and [tʰ] represent separate phonemes, /d/ and /t/, because din /dɪn/ and tin /tʰɪn/ form a minimal pair. Other well-reasoned answers that correctly apply the contrast and distribution tests are acceptable.
Practice — Phonemic Versus Phonetic
1. /stɑp/ is correctly phonemic; no aspiration mark is needed or expected after [s]. 2. /kʰɪt/ incorrectly places the aspiration mark inside slashes; the phonemic form should be /kɪt/, with [kʰɪt] reserved for the phonetic level. 3. [tɪn] is correctly phonetic, representing the predictable unaspirated allophone after [s]. 4. /dɪn/ is correctly phonemic.
Chapter Review — Part A (model answers)
1. A phone is a physically observed speech sound, described without reference to category; a phoneme is an abstract, contrastive sound category in a specific language; an allophone is one of the phonetic realizations, or phones, that correspond to a given phoneme. 2. [ ] claims to represent a phonetic realization at a stated level of detail; / / claims to represent the contrastive phonemic category, abstracting away predictable detail. 3. Answers will vary; e.g., cat /kæt/ and bat /bæt/.
Chapter Review — Part B
1. Yes: sip and zip form a minimal pair differing only in voicing of the first consonant, and the substitution changes the word's meaning, so /s/ and /z/ are separate English phonemes. 2. Spelling reflects orthographic convention, not phonetic or phonemic evidence; many languages, including English, have letters that do not correspond one-to-one with sound categories, so contrast must be tested with pronunciation, not letters.
Chapter Review — Part C
1. Complementary distribution describes a relationship between two phones in which each occurs only in environments where the other does not occur, so their distributions never overlap. 2. [kʰ] occurs at the start of a stressed syllable (kit); [k] occurs after [s] (skit). The environments do not overlap in this dataset, so the two phones are in complementary distribution, parallel to the [tʰ]/[t] pattern.
Chapter Review — Part D
1. Answers will vary in wording but should state that a citation form is the careful, isolated pronunciation of a word, while a contextual realization reflects how the word is actually produced within connected speech, which may differ due to neighboring sounds, rate, or stress, as in have [hæv] versus its reduced realization [əv] in fast speech.
Cumulative Review
1. This is best classified as an inference: it goes beyond a directly observed fact (a puff of air was present) to propose an explanation connecting that fact to syllable structure and stress. 2. The string already uses correct delimiters for a phonetic transcription; no correction is required, since aspiration is appropriately shown inside square brackets, not slashes.
Looking Ahead
What You Can Now Do
You can now separate a physical speech sound from the abstract category it may belong to, test whether two sounds contrast, recognize predictable allophonic variation, and choose the correct delimiter for a transcription claim. These distinctions will be active in every remaining chapter of this book.