What’s the Difference?

A phoneme is the smallest bit of language that makes a difference. Making a difference is just what phonemes do. They make one word different from another. In English, for example, the difference between the phonemes /p/ and /b/ makes “pin” different from “bin”, “pat” different from “bat”, “tap” different from “tab”, and so on and so on. A phoneme doesn’t mean anything in itself—/p/ doesn’t mean anything until it becomes part of “pin” or “pat” or “tap” or whatever. Its job is not to mean anything, but to make a difference between words that do mean something so you can tell them apart. But different languages have different differences, and native speakers of one language may find it difficult to hear the differences that make a difference in another language.

Each language has its own set of phonemes. My dialect of English has twenty-four consonants and nine vowels, as well as a set of pitches and intonation patterns. As I noted in my previous post, English has a “p” phoneme, as in the words “pin” and “spin”; but the “p” in “pin” in most English dialects is pronounced with a little puff of air, and the “p” in “spin” doesn’t have that puff of air. We can say that the “p” in “pin” is aspirated and the “p” in “spin” is unaspirated. We can write the aspirated “p” in square brackets with a little superscript “h”, [p^h], and the unaspirated “p” with a little superscript zero, [p^o]. The two sounds are different, but in English it never makes a difference that they are different. It makes a difference that “p” and “b” are different, because that difference makes the difference between “pin” and “bin”, but there are no words in English that are different because one uses [p^h] and the other uses [p^o].

In Classical Greek, however, the kind of Greek that Socrates spoke, this difference did make a difference. The Classical Greek alphabet actually has two different letters: “phi” (φ) for the aspirated /p^h/ and “pi” (p) for the unaspirated /p^o/. Those two phonemes make the difference, for example, between the word φάγος (“a glutton”), which we can transcribe /p^hágos/, and the word πάγος (“a crag”) which we can transcribe /p^oágos/. In Classical Greek aspirated “p” and unaspirated “p” are different phonemes, /p^h/ and /p^o/, but in English, aspirated “p” and unaspirated “p” are two allophones of a single phoneme, /p/. Classical Greek has two “p” phonemes, but English has just one.

English /p/ is not the only English phoneme with aspirated and unaspirated allophones. English /t/ and /k/ also have aspirated and unaspirated allophones, [t^o] and [t^h], [k^o] and [k^h]. (You might be tempted to think that [t^h] is like the “th” in “this” or “thigh”, but it’s not, it’s a real “t” sound with aspiration.) Just as /p/ is aspirated at the beginning of a word and unaspirated in the middle, /t/ and /k/ are aspirated at the beginning of a word and unaspirated in the middle, as we can see from the puff test with the words ‘top’ and ‘stop’ and the words ‘kin’ and ‘skin’. And once again, in each case classical Greek has two different phonemes where English has two allophones of one phoneme. In Greek, the letter tau (τ) represents unaspirated ‘t’ and the letter theta (θ) represents aspirated ‘t’, while kappa (κ) represents unaspirated ‘k’ and chi (χ) represents aspirated ‘k’. In Greek, the word τίς, which is the interrogative pronoun meaning ‘who?’, contrasts with the word θίς, which means ‘sea shore’: τίς is pronounced [t^ois] and θίς is pronounced [t^his]; the word Κρονος ([k^oronos]), the name of the father of Zeus, contrasts with the word χρονος ([k^hronos), which means ‘time’. The English pronunciation of these letters doesn’t match up with the Classical Greek pronunciation; most of the time we pronounce theta as the ‘th’ sound in ‘thermos’, but in Classical Greek the word ‘θερμός’ would have started with our aspirated ‘t’; it would have sounded like [t^hermos].

So English has three phonemes /p/, /t/, and /k/, represented by three letters, ‘p’, ‘t’, and ‘k’, and each of these has aspirated and unaspirated allophones: [p^h] and [p^o], [t^o] and [t^h], and [k^o] and [k^h]. But Classical Greek has six different phonemes, each represented by a different letter: π /p^o/ and φ /p^h/, τ /t^o/ and θ /t^h/, and κ /k^o/ and χ /k^h/. Here’s a diagram to show this difference between these phonemes in English and Classical Greek:

Greek             /p^o/    /t^o/      /k^o/

/p^h/    /t^h/     /k^h/

English          /p/      /t/       /k/

All of these sounds belong to a group of sounds called stops, because they are made when the airflow is totally stopped and then suddenly released. These stops are characterized by the position of the stoppage of the airflow: the “p” sounds are called bilabial stops, because the airflow is stopped by the lips; the “t” sounds are alveolar stops, because the airflow is stopped by the tongue against the alveolar ridge behind the teeth; and the “k” sounds are called velar stops, because the airflow is stopped by the velum. If you pronounce these in order, “p”, “t”, “k”, you can feel that they are progressively deeper in your mouth. One of the entertaining aspects of learning about phonology is getting to know your mouth better.

As we have seen, stops can be aspirated or unaspirated; in English the aspirated and unaspirated stops are allophones of a phoneme, but in Classical Greek they are different phonemes. In addition, stops can be voiced, if the vocal cords produce a kind of hum, or unvoiced, without the hum. The English stops /p, t, k/ are unvoiced, and the corresponding voiced stops are /b, d, g/. So the set of English stops looks like this:

                                    Voiced            Unvoiced

            Bilabial           b                      p

            Alveolar         d                      t

            Velar               g                      k

Classical Greek also has three voiced stops, “b” (beta, β), “d” (delta, δ), and “g” (gamma, γ). As we have seen, the Greek unvoiced stops can be either aspirated (φ, θ, χ) or unaspirated (π, τ, κ), but the Greek voiced stops don’t divide into aspirated and unaspirated. English has six stops, and Classical Greek has nine. Some other languages, including Hindi and Urdu, do make a phonemic distinction between aspirated and unaspirated voiced stops; these are usually written /b^h/, /d^h/, and /g^h/. Hindi and Urdu have five stop positions: bilabial, dental (rather than alveolar), retroflex, palatal, and velar, depending on where in the mouth the obstruction occurs, and all of these can be either aspirated or unaspirated, for a total of twenty stops, where English has six.

There is no such thing as a phoneme in isolation from the language it belongs to. There is no such thing as the phoneme /p/ in the abstract. We can say that a phoneme has no positive existence, it exists only in relation to the other sounds in its sound system.

According to the great Swiss linguist Ferdinand de Saussure (1857–1913), “In language there are only differences. Even more important: a difference generally implies positive terms between which the difference is set up; but in language there are only differences without positive terms.” This is an important point, but sometimes it has been misunderstood. Saussure doesn’t mean that sounds don’t exist on their own, he means that a sound in language doesn’t exist on its own. If some bird, for instance, makes a sound, it’s a sound, but it’s not a sound in a language, even if it’s similar to the sound of some phoneme in some language.

It’s reasonable to ask if Saussure’s principle extends beyond language. To some extent it certainly does. A little piece of wood in a particular shape isn’t a pawn until it’s part of a chess set. In fact it doesn’t even have to look like a standard pawn, so long as the two players agree that whatever it is counts as a pawn. If you lose a pawn from your set, you and your opponent can agree to use a pebble instead. What counts is the position of the piece within the system. A pebble has a positive existence as a pebble, but as a pawn it has no positive existence, just an existence defined by its place in the system of pieces.

A more challenging question is our own personal existence. Are we more like a pebble or more like a pawn? It would be comforting to feel that your self has some kind of positive existence, some kind of irreducible essence. I am I, just as much as a pebble is a pebble. Of course there are superficial aspects of my character which are influenced by other people or by what has happened in my life, but deep down at the core, there is some kind of irreducible, essential, individual, positive self. But some people would argue that there is no positive self, that the self exists only in relation to other selves. (I discuss these questions in my book, “Narrative Structures and the Language of the Self”.)