Babel, page 4
Saussure did allow for what he thought of as an inconsequential exception. Onomatopoeias, he realised, resemble the sounds they signify, or the sounds made by the things they signify. Take ‘meow’, cats’ main contribution to the human lexicon. English has turned this into a verb, meow (or mew or miaow); in Vietnamese, MèO is the noun for ‘cat’, while MEO and MéO are verbs, the former referring to the workaday ‘meow’ of a cat giving orders, the latter a poignant rendering of its cry when scared or in pain.
Not all onomatopoeias are so alike across different languages. Farm animals in particular would seem to be giving their daily concert from different sheet music in different places. The rooster, for instance, cries cock-a-doodle-doo in English, COCORICO in French, KIKERIKI in German, GAGGALAGÚ in Icelandic, KKOKKIO KKOKKO in Korean, wōwōwō in Mandarin and Iʼííʼąóó in Navajo; in Vietnamese it sings ò ó O. There are two reasons for this wide variety. Animal cries contain sounds that our speech organs can’t faithfully reproduce. Moreover, languages are constrained in two ways: they have to choose from their particular sets of sounds and obey their particular word-forming rules (though onomatopoeias cut them more slack than most other words do). The differences between onomatopoeias from one language to the next enabled Saussure and later linguists to maintain that this special category was not much of a problem for their sweeping claim that ‘the sign is arbitrary’. If we, as English speakers, can’t tell that ò ó O represents the crowing of a rooster in Vietnamese, this suggests that the word can still be regarded as arbitrary.
Not that Saussure knew much about Vietnamese, Korean or other Asian languages. Western linguists have a somewhat shameful tradition of generalising about language on the basis of a smallish sample: English, French, German, Latin, Greek and perhaps a sprinkling of Arabic (mostly clichés) and Chinese (frequently misconceptions). Even though several Asian languages, dead and alive, have been studied in Europe for centuries, this specialist knowledge somehow failed to properly feed into general linguistic theory. As late as the 1960s and 1970s, Noam Chomsky and his followers tried to work out the universal grammar of human language on the basis of just a few specimens. Or rather, one: English.
This parochial perspective has changed. Today, many native speakers of Asian and African languages are questioning the long-held views of Western linguistics, which are also being challenged by scholars who have travelled to places that are as poor in modern conveniences as they are rich in linguistic treasure. One of the things that African and Asian linguists as well as linguistic field-workers have called into question is Saussure’s assertion that language is essentially arbitrary. Especially in the languages of sub-Saharan Africa and East and Southeast Asia, they’ve documented extensive use of what has come to be known as sound symbolism. These languages have words, numbering in their hundreds or even thousands, whose sounds map on to certain meanings: ideophones. This is a category of which onomatopoeias are merely a subgroup; we’ll meet other subgroups shortly.
The joy of ideophones
Two East Asian languages have been found to be particularly rich in this respect: Korean and Vietnamese. If I were more talented a language learner, I might now be able to draw on my first-hand experience with Vietnamese, but alas, no such luck. I did suspect that the words for ‘cat’, ‘cow’ and ‘goat’ (MEO, Bò and Dê), might be onomatopoeias, and the colloquial word for ‘sneeze’ left no room for doubt: HẮT Xì. But I seem to have missed all the sound symbolism that isn’t onomatopoeia.
Korean, the protagonist of this chapter, has literally thousands of ideophones. Indeed, in books about Korean, they are mentioned as one of its defining characteristics. Native-speaking linguists have long been aware of this, and have coined two terms for these ideophones: ŭISŏNGŏ are the ones that imitate sounds (onomatopoeias), while ŭIT’AEŏ convey visual, tactile or mental sensations.
The main point of using ideophones, in Korean as in other languages, is to make stories more lifelike. These are words that stand out, because there is something expressive and pictorial about them. They often display some sort of rhyme or even full repetition of syllables. In speech, the specialness of ideophones is frequently reflected in the performance: they may be set off by minute silences or accompanied by gestures; the volume, speed or pitch may be different from the surrounding words.
Because of these theatrics, it would be reasonable to assume that ideophones are limited to story-telling. But while they are indeed a valued literary resource in many cultures, they serve other purposes too, especially that of making a speaker more convincing and trustworthy. Ideophones reflect physical sensations and states of mind, and someone who makes a believable job of describing those must have experienced the story in person, or so the reasoning goes. Liars shun such details; when asked for an alibi, they will say they were ‘at a party’ and not become specific until prompted, whereas actual party-goers are likely to volunteer details (about people, feelings and incidents). Therefore, details bestow trust.
Korean talisman. Today, the Hangeul script runs left to right, but in previous centuries, top-to-bottom in right-to-left columns was preferred, as it was in Chinese and Japanese.
God’s wallop
In Western languages and literatures, there is a whiff of childishness about onomatopoeia and what we consider excessively expressive terms. The German-British linguist Max Müller wrote in 1862 that they ‘are the playthings, not the tools of language’. And in 1910, French anthropologist Lucien Lévy-Bruhl adduced ideophones as evidence that in ‘inferior societies’, the natives have an ‘irresistible tendency’ to ‘imitate all one perceives’.
It’s important to realise that the use or avoidance of such words is a cultural preference, not a universal. As Westerners, we would be taken aback if instead of ‘God smote the Philistines’ we were to read that he ‘whacked’ or ‘walloped’ them. Western missionaries translating the Bible into non-Western languages avoid such expressive terms, but in many languages this is a stylistic choice that weakens rather than strengthens the impact of the text. In East and Southeast Asia and Africa, very far from being childish, the effective use of ideophones is a mark of eloquence and literary sophistication.
Let’s have a look at some real-life specimens from Korean (and a listen, if you like). Overleaf, in the columns headed ‘Hangeul’ and ‘Romanisation’, you can see the words in the Korean Hangeul alphabet and in a Romanised version respectively.* The next column translates them into English. If you want to hear them, there are sound file links on the ‘BABEL’ page of my website (languagewriter.com).
Korean sets of ideophones
The pronunciation of Romanised Korean words often runs counter to English intuitions. For present purposes, the main thing that needs explaining is the spelling of four groups of consonants. These groups roughly correspond to the English sounds written as k, t, ch and p, each of which in Korean has three subtly different pronunciations. Let’s take the k-sounds as an example. One version, written as a simple k (or sometimes g) in Romanised Korean, requires the consonant to be spoken without exhaling, the same as in French and Spanish. This pronunciation is considered basic or neutral in Korean. In English, we hear this when the k is preceded by an s, e.g. skin. The second type of k-sound is called ‘aspirated’, meaning that it comes with a puff of air. This is the pronunciation that English and German speakers are most used to: the word kin is pronounced as /khin/. It’s written as k’ in Romanised Korean; in other contexts, kh is a more common notation. The third and final type of k is hard to describe, but there’s somehow more tension involved; it’s written as kk.
Thlv-thlv-thlv
Equipped with this knowledge and looking at the table once more, we may wonder: how do the sounds symbolise the meanings of these words? To answer that, we first have to find regular correlations between sounds and meanings.
The easiest regularities to detect are in the rows 1, 2 and 3. The first word in each of these – KAM-GAM, PING-BING, PANTCHAK – begins with a neutral consonant, whereas the second word begins with a tense consonant, and sometimes the second syllable does too: KKAM-KKAM, PPING-PPING and PPANTCHAK. This changes the meaning from neutral to intensive, e.g. from ‘dark’ to ‘pitch-dark’. Apparently the tense initial consonant conveys a more intense sensation. In rows 1 and 2, the third words in the list pull off a similar trick by substituting the aspirated variety of the consonant: K’AM-K’AM, P’ING-P’ING. This too intensifies the meaning, but with some sort of twist, adding the dimension of ‘spookiness’ in row 1 and ‘wideness’ in row 2. The fact that row 3 doesn’t have a third word is interesting in itself, because it shows that speakers know which of the potential forms are part of the Korean lexicon and which ones aren’t. If a Korean speaker were to say P’ANTCHAK, which is the form missing in row 3, it would be considered an inspired coinage: perfectly good Korean, but not conventional. It would be a bit like calling something ‘sensmashingsational’ in English: the word has a transparent meaning, only a pedantic stick-in-the-mud would call it ‘wrong’, yet it can’t be said to exist in any but the most ephemeral sense. In other words, ideophones are expressive, but they are not primarily a tool for individual expression.
Ideophones should therefore not be confused with the expressive noises we make when reading a bedtime story to a child. After saying ‘And the snake slithered across the lawn’ we may go something like ‘THLV-THLV-THLV’ or ‘WZHV-WZHV-WZHV’ to paint a sound-picture of the slithering. Great fun of course, but it’s not how the typical ideophone works. Speakers of Korean and other languages don’t spend their days ad-libbing; for their ideophones, they mostly draw on a large set of vocabulary that can be looked up in dictionaries. For a snake slithering through grass, there’s probably a ready-made word.
Returning to the examples, we see that in rows 4, 5 and 6 the vowels play a significant role. The details are somewhat complex, but the long and the short of it is this: in ideophones, vowels known as ‘bright’ carry connotations such as small, affectionate, happy, flimsy and feminine, whereas ‘dark’ vowels are associated with concepts like big, heavy, clumsy, gloomy and masculine. (If you’re wondering if this has anything to do with yin and yang: yes, absolutely. Rather unusually, though, in Korean ideophones dark is associated with yin, bright with yang. We’ll see the consequences further on.) There are four bright vowels, and each one is paired with one or two dark vowels. For instance, since both the dark vowels ŭ (as in curl) and o (as in lot) have a (as in ah-ha) as their bright partner, the ‘dark word’ KKŭTTOK, with its heavy, gloomy undertones, has a more light-hearted counterpart in KKATTAK.
The two types of regularity discussed so far will sometimes team up. The word ping-bing appears in both row 2 and row 4. An even better example is the word PINGGŬL (similar in meaning to PING-BING), which has all the six forms that are logically possible: PINGGŬL, PPINGGŬL, P’INGGŬL, PAENGGŬL, PPAENGGŬL and P’AENGGŬL.
Finally, row 7 represents a less systematic group of Korean ideophones. Words ending in the same consonant often express similar sensations. Those ending in k tend to signify something abrupt, shrill or tight. Words ending in l often refer to smooth or flowing things, those in ng to round, hollow and open things (PING-BING!), those in t to small, fine things and pointed details, et cetera. Unlike the examples in row 1 to 6, however, the words under 7 do not form duos or other combinations, with different meanings based on their final consonant. They’re all loners.
Topsy-turvy symbolism
So yes, ideophones do display regular correlations between sound and meaning. But then, sound and meaning are sometimes correlated in English in a similar way: if you put un- or in- in front of an adjective, you (usually) negate it; if you add -s, you make it a plural. It would be silly to claim that those regularities are sound-symbolic; they aren’t. What is it about the Korean sounds then that does make them symbolic? Do they really convey meaning all by themselves?
To settle that question, we need to find out if people who don’t speak Korean have a hunch about these words. Of course, no one would be likely to guess the exact meaning of KAM-GAM, KKAM-KKAM and K’AM-K’AM, but it just might be possible for non-Koreans to intuit that the second and third words refer to a blacker darkness than the first.
They might – and they actually do, as Korean linguist Nahyun Kwon discovered in her PhD research. She had a number of Australians listen to quasi-Korean nonsense words displaying the sorts of regular patterns discussed above. Without knowing any Korean whatsoever, they correctly guessed (significantly more often than can be accounted for by mere chance) that KKAM-KKAM and K’AM-K’AM type words had a more intensive meaning than the KAM-GAM type, and not the other way round. So did Korean participants in the experiment – and remember these were nonsense words, so it wasn’t as if they knew them to begin with. The effect was not very strong in either group, but it was unmistakable.
Kwon also had her English-speaking participants listen to pairs of Korean-style nonsense words, one with a dark vowel, the other with a bright one (as exemplified in rows 4, 5 and 6 of the table). Again, they had statistically significant intuitions, but here’s another surprise: this time, they were wrong somewhat more often than chance would predict. Weirder still, her Korean-speaking participants simply seemed at a complete loss when trying to interpret these nonsense words. These startling results begin to make sense if we assume that the Korean language somehow has its vowel symbolism the wrong way round. And well we might, because there is a well-documented tendency for humans worldwide to associate low, open vowels such as /ah/ with bigness – words like vast and large fit the bill, as does yang – and high, closed vowels such as /ee/ with smallness – think mini, teeny-weeny and wee as well as yin. Other research has shown that Koreans actually share this widespread intuition, but only when hearing words that are evidently not Korean. Words that appear Korean but aren’t, as in Kwon’s research, put them in a bind: their general intuition clashes with their specific knowledge of Korean, the result being that they can’t make up their minds. Just why the Korean language has reversed the widespread correlation is anybody’s guess; no explanation has proved satisfactory so far. Let’s chalk it up as a linguistic accident.
What is understood is why most of us associate /ah/ with big and /ee/ with small: when we say /ah/, our oral cavity (otherwise known as ‘mouth’) is big, when we say /ee/, it’s small. This is sound symbolism at its most basic: size of mouth reflects size of thing referred to. You may object that when you’re talking, you’re hardly aware of the size of your mouth, which is true. But try thinking of sound symbolism as a special manner of gesturing. Even though you hardly realise what you’re doing with your hands during a conversation, your manual gestures are highly meaningful and symbolic. Similarly, ‘oral gestures’ can be meaningful and symbolic. However, the oral gestures also do something that manual gestures don’t: they form the sounds that you utter. Oral gestures therefore, being both unconsciously expressive and the mechanism to produce our speech sounds, could easily be a mechanism that links meanings to sounds. Exactly this is thought to be the case for the sounds of many words in some languages, such as Korean and other Asian as well as African languages, and also for some words in languages elsewhere. More about those later.
Turning from vowels to consonants, we’ve seen that Korean ideophones beginning with tense consonants rather than neutral ones (pp or kk instead of p or k) convey a more emphatic meaning, and so do the aspirated ones (p’, k’). Those correlations, especially the former, make even more obvious and intuitive sense than a wide-mouthed /ah/ being suggestive of something big.
Korean handwriting, as it were.
This leaves us with the third type of Korean sound symbolism seen above, the one that manifests itself in the final consonant. Kwon didn’t examine it, but this type seems to reflect another general human intuition: words ending in k sound more abrupt, because the consonant itself consists in an abrupt, sharp burst of sound, unlike g or m. So does the t, but here the burst of sound is smaller, produced by only the tip of the tongue – hence the connotation of something small, fine or with pointed details. It’s also tempting to see the t as another oral gesture, presenting the tip of the tongue: a pointed part or detail that makes small, nimble movements.
There are other real-world phenomena that can help explain symbolic links between sound and meaning, some widely accepted, others more speculative. If a certain movement can be heard and therefore expressed in an onomatopoeia, say sand sliding down a slope, the same word can then also be applied to other movements that are similar yet silent: think of a raindrop sliding down a leaf. Another link can form when certain types of sounds often go together with conspicuous features: if time and again we notice that children, short people and small animals such as mice make softer, higher noises than sturdy adults and large beasts such as lions, we will come to associate loudness and pitch with size. This is also an (additional or alternative) explanation for the ‘ah-as-in-vast’ versus ‘ee-as-in-wee’ contrast: /ah/ reminds us of a roaring lion, /ee/ of a squeaking mouse. Yet another conceptual link may be found in duration, with longer sounds representing longer and therefore slower things and movements.
Taking it further
Once such links are established in our minds, we can base ever more abstract associations on them. If we think of a certain sound as ‘soft’, for instance /b/ as compared to /p/, we can easily extend this to a general sense of ‘pleasantness’ and from there to ‘sweetness’. If we feel that /ah/ represents ‘big’, and thus high on the scale of size, we may also think of it as high on other scales, for instance thickness, danger or bitterness. In this way, ideophones can become abstract to the point of having cognitive rather than sensory meaning.
