Ah, the joys of surfing the web. Browsing through comments on language hat, I wound up at a new-to-me site, Postmodern Courtesan, and there discovered an interesting, although somewhat old, post about word choice and terms for female genitalia.

Most of the post concerned the connotations of different terms, and why the postmodern courtesan, who goes by the pseudonym Olympia, prefers cunt to the many other options, but she had two throw-away comments that caught my attention, commenting on both the phonology and the etymology.

I don't have time to tackle both this morning, so I'll start with the phonology. Olympia writes:

Yeah, I'll admit, the sound of the word cunt is harsh. Anything with a glottal stop is going to be. But I'm ok with that given the alternatives.

As several people noted in the comments, there isn't really a glottal stop here. A glottal stop is the sound that substitutes for [t] (in certain contexts) in some varieties of English (mostly, but not exclusively British). For Americans, the stereotype of Cockney English will be the most familiar. This is the sound in the middle of the word bottle (bo'le). Now it is true that some people glottalize final -t in words like cunt, but this is a rather subtle effect, the sort of thing that only a phonologist is likely to notice. I presume that Olympia was really thinking about the initial consonant, [k^h], which is an aspirated velar stop. Ignoring the aspiration, which is inserted by phonological rule rather than being an underlying phoneme, the difference between a velar and a glottal stop is in the place of articulation. [k] is formed on the velum (the soft palate). The glottis is in the throat.

Terminology, though, is only a minor point. What interests me is the assumption, widely shared, that there is a kind of inherent meaning to the sequence of sounds, that the harshness of the word comes from the sounds within it rather than its meaning. This effect is known as "sound symbolism." But whether, or to what extent, sound symbolism exists is by no means uncontroversial.

To start with, there is a long tradition in linguistics, going back to Ferdinand de Saussure, that insists on the arbitrariness of linguistic signs. If the relationship between sound and meaning is completely arbitrary, there is no reason to attribute any inherent meaning to phonemes. Many linguists since Saussure have denied the importance of sound symbolism and onomatopoeia, assigning them at best marginal roles in language.

On the other hand, others have come to the conclusion that there is something to sound symbolism. Otto Jespersen, for example, directly criticized Saussure's position. (See Language: Its Nature, Development, and Origin, 1922), and many studies since have claimed to find that certain sounds connote particular moods or meanings. The advertising world has bought into this line of argument. A great deal of money is spent to develop product names based on that supposition.

(In passing, I find it ironic that someone who calls herself a "postmodern" courtesan takes the sound-symbolism position, since postmodernism generally takes Saussure's arbitrariness of the sign as a foundational idea, but then many people use the term "postmodern" simply to mean hip or up-to-date.)

Now I'm not familiar enough with the literature on this subject to really express an opinion about the adequacy of the research, but I'm willing to accept that sound symbolism has some effect, and that effect is most obvious in invented words where prior meanings don't get in the way as much (although if they're formed from existing words, that may not be the case).

But in natural words, those meanings do interfere. For example, it's often claimed that low open vowels connote largeness while high, closed vowels connote smallness. But what about the words big and small, which have the opposite pattern?

Turning to the sound of cunt, is it really the [k^h] that makes it harsh? What about a word like caress? What about country, which even has the same first syllable? I suppose someone might argue that the later sounds somehow undo the initial harshness, but that leads us into somehow adding up emotional qualities like some sort of equation. The farther we go down that road, the more ridiculous it seems. To say that particular sound conveys a particular quality in the abstract, without considering the larger context seems problematic.

Even if we get a little more precise about the phonetic environment, I'm still dubious. We might speculate that the fact the word is a monosyllable also makes it pack more punch. It also starts and ends with stops. I'm not sure if the vowel matters, but let's assume it does. In many varieties of English it's a mid-central vowel (neither front nor back), but in some varieties, e.g., across Northern England, it will be a high back vowel. If we're to take the claims of sound symbolism literally, the connotations of the words will differ by dialect. Are cunts perceived as larger in Yorkshire because the word is pronounced with a back vowel there?

Even ignoring this problem, cut, and cup are both monosyllables starting with [k^h], ending with a stop, and containing the same vowel. Neither sounds particularly "harsh" to my ear, particularly not cup. And if you judge that cut is harsher than cut it's probably because you're being influenced by the words' meanings. I suspect that's what's happening with cunt too. To the extent that sound symbolism is real, the meanings of the words can easily override any vague emotional effect, making it impossible to judge what part of our reaction to the word comes from its meaning and what from its sound.