phonology english

Phonology

Phonemes

Phonology is the study of phonemes. How are those different from phones? While phones include all the speech sounds that are possible – or at least potentially useful – phonemes are the ones that a particular language actually employs. This means that phones (the subject of phonetics) are universal, but phonemes (the subject of phonology) are unique to the language you’re looking at.

Got all that straight? Don’t worry if it seems confusing; the distinction gets easier with practice. Notation can also help. Remember that phones are written between brackets, like [æ]? Phonemes are written between slashes, like this: /æ/. So if you see an IPA symbol between slashes, you’ll know that the sound is being considered not for its objective value, but for its function in a certain language.

But why bother looking at speech sounds within one language? Isn’t, say, a p sound always a p sound, no matter what languages use it?

Not always. Consider the city of Busan, South Korea. From the time it was first written using Roman letters (the ABCs we use in English) all the way up to the year 2000, the city’s name was spelled Pusan.

Then, in 2000, the P became a B. Signs, maps, and official documents bore witness to the change. Pusan was wiped from the face of the earth, and the new city of Busan was established in its place.

Except of course it was the same city. And its pronunciation in Korean was exactly the same. The English-speaker, once he learns this, will suspect he’s being lied to: if the pronunciation is unchanged, surely a change in spelling must be misrepresenting that pronunciation, right? Well, no – at least not any worse than before.

Because the first consonant sound in P/Busan – a phoneme that the Korean alphabet represents as ㅂ – is neither a P nor a B, at least not as English conceives those sounds. To an anglophone, B is a voiced bilabial stop, and P is an unvoiced bilabial stop, aspirated at the beginning of a syllable except in consonant clusters. That is to say, E. /b/ = [b] and /p/ = [p] or [p^h], depending on the environment (^h denotes that a phone is aspirated). Notice that the E. phoneme /p/ contains two phones, [p] and [p^h]? Each of those is called an allophone, a term we get by prefixing allo- (Greek for ‘different’) to phoneme (allophones are different versions of the same phoneme).

But K. /p/ (written in Korean as ㅂ) is not the same phoneme as E. /p/. Like E. /p/, K. /p/ has two allophones, but they are [p] and [b], not [p] and [p^h]! Whereas E. /p/ varies between [p] and [p^h] based on its position in a syllable and the presence or absence of other consonants, K. /p/ varies between [p] and [b] based on its position in an spoken phrase. If it’s the first sound, it’s going to sound like unvoiced [p]; but if there’s a voiced sound before it, it will inherit that sound’s voicing and be realized as [b].

The confusion arises because different features matter in these languages. English distinguishes between voiced and voiceless consonants but pays no mind to aspiration, which happens automatically where it’s supposed to. In contrast, Korean distinguishes between aspirated and unaspirated consonants but pays no mind to voicing, which happens automatically where it’s supposed to. No wonder Korean-speakers and English-speakers have such a hard time pronouncing each other’s languages! How many people will they ever meet who can explain this all-important difference to them?

So in the end, neither Pusan nor Busan is a wholly accurate representation of the city’s name. English-speakers will pronounce Pusan as [p^husɑn], which will sound wrong to Koreans because they have a whole other phoneme for [p^h] – /p^h/ (written ㅍ in Korean). As for Busan, anglophones will pronounce it [busɑn], which will sound right to Koreans only in certain environments, viz., in the middle of a spoken phrase. Phrase-initially, Koreans will expect [pusɑn], a sound English-speakers aren’t likely to make because it violates the rule of English phonology that aspirates any syllable-initial consonants not in a cluster. (Consonant clusters aren’t allowed in Korean.)

It seems the only way to be sure of pronouncing words correctly is to learn 1) the IPA and 2) the phonology of the language you’re speaking. This includes phonotactics, our next subject.

Phonotactics

Just because a language gives you a bunch of phonemes to play with doesn’t mean you can stuff them together in any combination. Each language has its own rules about what it will allow in a syllable. Collectively, these rules are studied as the phonotactics of a language. Before observing what English syllables allow, let’s learn the parts of a syllable.

What is a syllable, exactly? Basically, it’s a vowel with optional consonants. There are two main parts to a syllable: the onset and the rhyme. The onset is the beginning of the syllable; the rest of it is the rhyme. We can further divide the rhyme into a nucleus and a coda.

The letters in red show you what kind of phoneme can occur in each position. C stands for consonant, V for vowel. The parentheses around the Cs mean that the consonants in the onset and rhyme are optional; all that’s required for a syllable is a vowel in the nucleus.

Here’s a relatively simple syllable, dog:

In English, the onset and coda can each have more than one consonant, as in skunk:

Check out all the consonants in strengths, which is a single syllable in English:

Some speakers even pronounce a k between ŋ and θ. The coda can accommodate all four consonants!

The nucleus also has room for more than one vowel, though only one can be the peak vowel, the one pronounced with the most intensity. Less-intense vowels, called glides, can come before and/or after the peak vowel. An onglide comes before the peak vowel and belongs to the onset, while an offglide goes after the peak vowel and belongs to the rhyme:

The first G is an onglide. Because it belongs to the onset and not the nucleus, an onglide does “consonant work,” even though it’s phonetically identical to a vowel. Offglides, like the second G, are part of the nucleus, which admits no consonants. For this reason, glides are often called semivowels.

There are two kinds of glides, labial and palatal. There are four glide phonemes in English: the labial onglide w, labial offglide ʊ, palatal onglide j, and palatal offglide ɪ. (To understand why the glides are called “labial” and “palatal,” think about what your lips and tongue are doing when you pronounce them.) We can observe two different glides in the word quail,

where w is an onglide and ɪ is an offglide. The peak vowel is e.

Whereas quail featured a labial onglide and palatal offglide, yowl reverses the order of the glides:

There, j is an onglide and ʊ an offglide. Note that there is no C in the onset, only G.

As we’ve seen, English can stuff a whole lot into its syllables: up to three consonants in the onset, two vowels in the nucleus, and four consonants in the coda. Most languages are not so accommodating.

Prosody

Now that the phonotactics of your language has decided what a syllable can sound like independently, the prosody of your language decides what syllables sound like in relation to each other. By letting you make certain syllables more prominent and tuck others into the background, turning the stream of syllables you speak from a flat plain to a chain of rolling hills, prosody gives you a useful way to distinguish words from each other.

If you speak English, you’re familiar with the difference between the verb record (to set down in writing) and the noun record (something that has been set down in writing). The words are obviously related; both have the same two parts, re- ‘again’ and cord- ‘heart’. When you meet them in print, the words look identical, so you have to use the surrounding sentence to tell which is meant. When you hear them spoken, however, the words sound noticeably different. The second syllable of the verb, and the first syllable of the noun, is stressed – spoken a bit more loudly and a bit higher in pitch than the other syllable. This difference is useful, allowing a word to split into two words that share the same meaning but have different grammatical functions.

Different languages have different ways of handling prosody:—

English: Accented syllables are stressed, slightly louder and higher pitched than unaccented syllables. Stress can be primary (quite strong) or secondary (less strong); long words tend to have one syllable with primary and one with secondary stress. Vowels in unstressed syllables often reduce to /ə/, the unrounded mid central vowel called “schwa.” Words from the Anglo-Saxon core are usually stressed on the first syllable, a fact that lent Old English words to alliterative poetry; words borrowed from French are often stressed on the last syllable, a fact that lends modern English words to iambic poetry.
Spanish: Like English in that accented syllables are stressed (louder and higher pitched) and that stress can be primary or secondary, but unlike English in that unstressed vowels don’t reduce to /ə/ but keep their quality. A simple rule tells readers which syllable to stress in any word. If a word breaks that rule for any reason, the stressed syllable is shown by an accent mark over the stressed vowel.
Chinese: Neither syllable in a word (most words have two) is accented above the other. However, each one-syllable morpheme has its own assigned tone. A Chinese tone is a pitch pattern: over the length of a syllable, depending on which morpheme the syllable represents, the speaker will either (1) keep the pitch steady, (2) start low and rise high, (3) start in the middle and drop low before rising high, or (4) start high and drop low. The same collection of phonemes, given different prosodic pitch patterns, can become several distinct morphemes.
Japanese: Accented syllables are higher in pitch but not in volume. “Accentless” words by default start low and go high from the second syllable, including the particle on the end of a noun; accented words drop low after the accented syllable and stay low, including the particle. Though syllable accent is not, well, stressed in Japanese-language classes for foreigners, knowing the accent of every word is as important in Japanese as it is in English. Besides accent, syllables borrowed from Chinese words that in that language had certain tones are, in tone-less Japanese, held long, as are some consonant-eroded syllables in native Japanese words. Long syllables are marked in writing by an extra vowel.
Korean: Accent is, at least in the modern language, unfixed and varies from speaker to speaker, sentence to sentence. Rules for deciding accent from a word’s first consonant have been suggested but not firmly established. Where it occurs, accent raises pitch but not volume. Until recently, syllable length was also used to distinguish words from each other (e.g., short 눈 ‘eye’ vs. long 눈 ‘snow’); but that distinction, though still recorded in good dictionaries, has been largely lost. Even before it used length, Korean prosody used tone: yesterday’s long syllables were Middle Korean’s rising-toned syllables, which started low and ended high in pitch.
Latin: Both accent and length distinguish syllables. Accented syllables are stressed (louder and higher), while long (tense) vowel phonemes are actually held out twice the length of short (lax) vowel phonemes. Accented syllables needn’t be long, and long syllables needn’t be accented. (In English, by comparison, tense vowels last only slightly longer than lax ones and are usually stressed.) In Old Latin, every word was stressed on the first syllable (even compound words, which explains why so many root vowels got weakened and raised), but by the Classical period a simple rule decided which syllable got the accent.
Ancient Greek: Like its younger cousin Latin, ancient Greek used both accent and length to distinguish syllables; unlike Latin, it raised only the pitch (not the volume) of accented syllables and also used tone: some long syllables have a high-to-low contour (the inverse of Middle Korean). As Greek evolved, tone and length distinctions died out, and accent became stress, raising volume in addition to pitch.