03 Phonology

Sound categories and contrast

Phonology studies how a language organizes speech sounds into meaningful categories and patterns. Phonetics describes the physical properties of speech sounds; phonology asks which sound differences matter in a particular language and how sounds behave in context.

Phonemes and their variants

A phone is a speech sound, written in square brackets, as in [p]. A phoneme is an abstract sound category that can distinguish words, written between slashes, as in /p/. The phonetic realizations of a phoneme are its allophones.

A minimal pair can reveal whether two sounds are separate phonemes: the words must differ in meaning and in only one sound in the same position. English pat /pæt/ and bat /bæt/ form a minimal pair, showing that /p/ and /b/ are separate phonemes in English.

Allophones do not create a difference in word meaning. In many English varieties, /p/ is aspirated [pʰ] at the beginning of a stressed syllable, as in pin [pʰɪn], and unaspirated [p] after /s/, as in spin [spɪn]. Because these realizations occur in different environments, they are in complementary distribution. Their phonetic similarity and predictable distribution support analyzing them as allophones of /p/.

Complementary distribution is useful evidence, but does not by itself prove that two sounds are allophones: unrelated sounds may also happen to occur in different environments. Analysts also consider phonetic similarity, the pattern across the language, and whether a simple rule explains the data. If two variants can occur in the same environment without changing meaning, they may instead be in free variation.

Sound patterns and analysis

Phonological patterns describe how sounds are distributed or change in context. Common processes include:

• Aspiration: adding a puff of air, as with [pʰ] in pin.

• Assimilation: a sound becomes more like a neighboring sound in some feature, such as place of articulation.

• Deletion: a sound is omitted in a particular context.

• Insertion: a sound is added to ease pronunciation.

Analyzing sound patterns

A systematic analysis begins by listing the sounds in the data and their environments—the sounds or positions around them. Then look for minimal pairs, which provide evidence that sounds contrast as phonemes. If there is no contrast, check whether the sounds have complementary distributions and are phonetically similar.

Next, state a generalization or rule that accounts for where each sound occurs, and check that it fits all the data rather than just one example. For pin [pʰɪn] and spin [spɪn], a simple generalization is that /p/ is aspirated at the beginning of a stressed syllable, but not after /s/. One compact rule is:

The arrow means “is realized as.” This describes a common English pattern; exact details can vary by speaker and dialect.

Features and sound classes

A distinctive feature describes a property that helps characterize speech sounds. Features can refer to voicing, place or manner of articulation, and other properties. The sounds /p/ and /b/ are both bilabial stops, but differ in voicing, represented as $[\pm\text{voice}]$. In English, this difference can distinguish words, as in pat and bat.

Features also let linguists describe sound classes compactly. Sounds that share relevant features form a natural class. For instance, /p, t, k/ are voiceless stops in English. A sound rule that applies to this whole group can be expressed in terms of the features they share, rather than listing each sound separately.

Syllable structure

A syllable is a sound-pattern unit organized around a central, usually vowel-like segment. Its onset consists of the consonant or consonants before that segment; its nucleus is the central, most sonorous part, typically a vowel; and its coda consists of consonant or consonants after the nucleus. The rhyme comprises the nucleus and coda.

In the English word plant [plænt], the onset is [pl], the nucleus is [æ], and the coda is [nt]. The structure can be represented as $\sigma \rightarrow \text{onset} + \text{rhyme}$, with the rhyme dividing into nucleus and coda. A syllable may lack an onset or a coda, but in the usual analysis it must have a nucleus.

Phonotactic constraints are language-specific restrictions on the sound sequences permitted in each syllable position. English permits the onset cluster [pl] in plant, but not every imaginable sequence of consonants is a possible English onset. Syllable structure can therefore help explain why a sound occurs in one position but not another.

• Allophone