Steven Connor

A talk given at the Song, Stage and Screen conference, University of Winchester, 5 September 2010. [pdf version]

Chiasmus is defined as a grammatical figure consisting of two parallel clauses, in which the order of elements in the second clause inverts the order of the first: when the going gets tough, the tough get going: ask not what your country can do for you, but what you can do for your country. Chiasmus is so-named from the Greek χιαζω, meaning to mark with a sign of the letter χ, chi. The optic chiasm, located at the bottom of the brain immediately below the hypothalamus, is the place where the nerves from the eyes partially cross over each other, sending images from the right eye to the left hand side of the brain and vice-versa. In the International Phonetic Alphabet, χ is the symbol for what is known as the voiceless uvular fricative. This sound is made by vibrating the uvula, the little pendulous flap that hangs down from the soft palate. The related sign X is marks the voiceless velar fricative, a slightly thinner sound, made by vibrating the soft palate just in front of the uvula. Phonology distinguishes a number of voiceless fricatives formed at the back of the mouth: the voiceless epiglottal fricative; the voiceless pharyngeal fricative (Hebrew chet); the voiceless uvular fricative (French roche); the voiceless velar fricative (Spanish agua) and the voiceless palatal fricative (English hue).

My concern is not so much with the visible figure of crossing, as with the sound that it inscribes, though following the meanings and values of that sound will provide plenty of examples of the oscillatory operations of the chiasm. But there is an inaugural philosophical hiccup or totter from which this enquiry makes a start, the form of which seems to mark such a crossing or crisis in the voice, or the idea of the voice. Voice, says Aristotle in Book II.8 of the De anima, ‘is a kind of sound characteristic of what has soul in it; nothing that is without soul utters voice’. And yet, as he is quick to point out:

Not every sound…made by an animal is voice (even with the tongue we may merely make a sound which is not voice, or without the tongue as in coughing); what produces the impact must have soul in it and must be accompanied by an act of imagination, for voice is a sound with a meaning, and is not merely the result of any impact of the breath as in coughing

So voice is to soul (the animate) as soundless voice (noise) is to the inanimate. Like this:


But voice also contains the kind of sound which is characteristic of inanimate things. All voice has soul, but soul is not all there is of voice. The little deflection or decussation of the logic of identification, which we can picture like this, is the first of the chiasmus forms we will meet as we go.


Later commentators on the production of language have descried in it similar crossings or plaitings of contrasting elements. Notable among these is what might be called the amicable antinomy of the vowels and the consonants. Vowels, which for many are to be identified with voice as such, are produced by the modification of expired air in the larynx, to produce variations of tone and timbre. These sounds are then subject to further modification by the different parts of the mouth and nose – the glottis, tongue, teeth, palate and lips, whose work produces the consonants which differentiate the vowel sounds. There is a strong tradition within the various forms of mystical philology of seeing the vowels as the soul or essence of spoken language, with the consonants acting as their bodily frame or occasion. For such traditions, the absence of vowels in Hebrew script points to the uncapturability of speech in letters, so that the difference between vowel and consonant approximates to the difference between voice, the spirit that gives life, and writing, the letter that killeth.

But it is possible to read things the other way round (and you can probably guess that we are going to have to get used to this). The nineteenth-century physician James Yearsley, whose book Throat Ailments went through many editions after its first appearance in 1842, distinguished between voice, as the production of sound from the larynx, which animals as well as humans have the capacity to produce, and speech, the meaningful articulation of those sounds by the organs of the mouth, which only humans can produce. Yearsley develops a familiar comparison of the speech organs as a kind of musical instrument, with the lungs as the source of air, the different tones of the vocal cords corresponding to the holes of the instrument and the channels of the voice and nose corresponding to its pipes. But then he adds a level to the metaphor:

we cannot proceed further with the simile, if we confine ourselves to the organ alone. The power of articulation possessed by the tongue, palate, lips, and other parts under the direction of the mind, is not comparable to any parts of the organ itself, but rather with the hand and intellect of the performer on the instrument. (Yearsley 1853, 18)

We might recognise in this move the Aristotelian introduction of the element of soul, which is not only in the voice, but also acts upon it, in that act of imagination which produces meaning as opposed simply to sound. Yearsley’s particular interest was in diseases caused by inflammation of the tonsils and what he described as ‘elongation of the uvula’. He seems to have regarded these portions of the mouth as forming the gateway between voice and speech, and described the deleterious effects upon speech that enlargement of the tonsils and elongation of the uvula could bring about:

the clear resonance of the natural voice is changed for a harsh and disagreeable tone. This deterioration is known by the term thick speech; the voice cannot be understood at anything like the natural distance; either there is an unpleasant drawl, or the words are mumbled together in a confused manner; the variations and flexibility of the voice, which add so much to its expression, become altered to an unvarying monotony.(Yearsley 1853, 20)

The uvula marks the point where the throat emerges into the mouth, as the river into its delta. Here, the body of the voice might be regarded as entering into form, through articulation and differentiation.

The whirr of the fricative is a kind of insentience, as though it were blind, stalled, undirected force. The sound of the fricative is the sound of a turbulence, a miniature anticyclone of air turned back and round on itself. Of all phonations, the fricative appears most to detain the passage of the sound. And yet, unlike the other consonant sounds, which are restricted to the particular portions of the vocal apparatus that produces them, the teeth for the dentals, the lips for the labials and plosives, the fricative also has a kind of mobility and adaptability, which enables it to take up residence in other regions of the mouth – wherever the air can be forced through a narrow channel, for example in the sibilants, where the front of the tongue sends the air over the teeth, or, in Welsh, the side of the tongue against the molars, as in Llanelli.

Since sounds are produced by the coordination of movements and spaces, there is a tendency when attempting to read the meanings of sounds to focus on the position of the mouth or the quality of the movement of the vocal organs. Sounds often seem expressive, not because of any obvious or direct resemblance to sounds in the world, but on account of a kind of kinetic rhyming between the process of forming the sounds in the mouth and a corresponding process in the world. The mouth becomes a tactile cinema – a kinematograph, precisely – of the world. Socrates in Plato’s Cratylus is struck by the agitation of the tongue involved in the trilled letter ρ, rho, which may be regarded as the lingual equivalent of the velar or uvular fricative – especially at the beginning of the word, where it is sometimes marked with a diacritic to indicate a ‘rough breathing’:

Now the letter rho, as I was saying, appeared to the imposer of names an excellent instrument for the expression of motion; and he frequently uses the letter for this purpose: for example, in the actual words rein and roe he represents motion by rho; also in the words tromos (trembling), trachus (rugged); and again, in words such as krouein (strike), thrauein (crush), ereikein (bruise), thruptein (break), kermatixein (crumble), rumbein (whirl): of all these sorts of movements he generally finds an expression in the letter R, because, as I imagine, he had observed that the tongue was most agitated and least at rest in the pronunciation of this letter, which he therefore used in order to express motion.

This sound, and some of the associations with motion that Socrates discerns, survive in English spellings like rhythm, rheumatic (incorporating rheum, which originally means a flowing of mucus), catarrh and diarrhoea. The Platonic hypothesis with regard to the rough sound of rho has inaugurated a durable and seemingly unextirpatable conviction, and especially among French writers, about the dynamic qualities of the r sound (Genette 1995, 41, 47, 71, 124-5)

As sounds move forward in the mouth, they tend both to become more distinct, that is, to involve less abrasive friction and reduced turbulence, producing the sense of diction, punctuality, division and precision, and to become higher in pitch. Higher-pitched sounds tend to be associated with distance and particular positions in space, partly because they provide a lot of sonic information of that kind (you can always pick out the sound of the ring tone or the xylophone amid the symphony). High-pitched sounds not only seem to come from the front of the mouth, they also seem to put us in front of things, in that we tend to face the source of sounds. Low-pitched sounds, by contrast, seem less located, more circumambient. The partial hearing loss that affects all human beings as they grow older returns them to something like the blurred hearing of the infant, because it tends to erase the consonants that are the least redundant portions of speech, that is, the portions of speech that give the most information by narrowing things down: that is why abbreviations use consonants rather than vowels – txtg is much more easily readable than ei as texting, and partially solved crossword clues with consonants in them are easier to fill out than when there are only vowels. When I suffered some sudden hearing loss last year, I had the odd experience of having to try to reconcile the experience of looking at people speaking to me, whose voices nevertheless sounded as though they had their backs turned to me.

There used to be a fashion for calling these throaty sounds ‘guttural’, a word that has markedly negative associations for English speakers, mostly because these sounds do not feature very much in modern English. English used to be as rich in guttural sounds as German and Dutch, and the presence of gh- forms in words like light, knight, brought, and daughter testify mutely to the once-active employment of the pharynx and uvula in their earlier pronunciations. English no longer has many guttural sounds, though such sounds are often heard by speakers of non-standard dialects or non-English speakers.The particular historical accident which encouraged the slow recession of such sounds (though it was, as it were, a retreat forwards, into phonic frontality), was the Norman Conquest, which accelerated the spread of Romance forms, in which all these gutural sounds first tended to be softened, into a simple aspirate h, and then to vanish altogether. This furnishes the essential meaning (though it arises from history, which is to say accident) of the guttural or of sounds that come from the throat, namely that they are early.

The vocal apparatus of the English speaker thereby becomes not only a mimetic theatre of physical actions and forces, but also an orographic chronotope, a knitting or weaving together of places and times in the mouth. Sounds that are made early, that is, at the initiation of the sound-making process, before the intervention of the tongue and its work of articulation, seem primitive, even (literally) prelinguistic. They seem to come from deeper and further back in time, both the time of the particular utterance, and the time of utterance in general.

Since they are produced at the back of the mouth, guttural sounds are also hidden from view – it is much easier to demonstrate the production of dentals, plosives and aspirates to the deaf or to speakers of other languages. The fact that guttural sounds seem sonically hard to place means that they can come to signify the cryptic or concealed as such – not just a hidden sound, but the sound of secret or mysterious intent. Kenelm Digby wrote of a deaf and dumb Spanish prince who had been taught to lip-read, which Digby represents as the seeing of sound. Digby is particularly impressed by his capacity to make out the sounds of Welsh and Hebrew, which appear much less visible than those of Spanish. Curiously, Digby notes, the man became mute only in the dark:

He could discern in another, whether he spoke shril or low: and he would repeat after any body any hard word whatever. Which the Prince tryed often; not only in English, but by making some Welchmen that served his Highness speak words of their language: Which he so perfectly ecchoed, that I confess I wonder’d more at that, then at all the rest. And his Master himself would acknowledg, that the rules of his art reach’d not to produce that effect with any certainty: and therefore concluded, this in him must spring from other rules he had framed to himself, out of his own attentive observation; which, the advantage that nature had justly given him in the sharpness of his other senses, to supply the want of this, endow’d him with an ability and sagacity to do, beyond any other man that had his hearing. He express’d it (surely) in a high measure, by his so exact imitation of the Welch pronunciation: for, that tongue (like the Hebrew) employs much the guttural Letters; and the motions of that part which frames them cannot beseen or judg’d by the eye, otherwise then by the effect they may happily make by consent in the other parts of the mouth, exposed to view. For, the knowledg he had of what they said sprung, from his observing the motions they made: so that he could converse currently in the light, though they he talked with whisper’d never so softly: and I have seen him, at the distance of a large chambers breadth, say words after one, that I, standing close by the speaker could not hear a syllable of. But, if he were in the dark, or if one turned his face out of his sight, he was capable of nothing one said. (Digby 1669, I.321)

The associations of guttural fricatives can be powerfully positive. The sound appears notably in the Hebrew word ruach, which means spirit, or wind of inspiration, and appears in the second verse of Genesis: ‘and the spirit of God moved on the face of the waters’. Michel Serres evokes the sound to suggest the animal delight in the breath in running and exertion

The first utterance of Genesis, at the dawn of the world, above the hubbub, says God, ruagh, a hoarse, alliterative breath, on the soft palate, at the back of the throat, before language, in front of the root of the tongue, where the gasping intake of breath acknowledges the divine; ruagh, breath, breathing, wind, breeze of the spirit (Serres 2008, 314)

But, for the most part, the associations of the gutturals and especially the guttural fricatives, are negative. Some commentators have pointed to the fact that English words beginning with gr-, seem frequently to have negative or unpleasant associations. Three clusters of association have been suggested: 1) unpleasantness (grim, grisly, gritty, gross, grotty, gruesome, greasy, gristly) 2) complaint (grumble, groan, grieve, gripe, grimace, grizzle, Scottish greet) and 3) relating to undesirable friction or rubbing, or its products (grind, groove, grate, grout, grub, gruel, gravel, grave, grit) (Bernard and Delbridge 1980, 151). One might easily see all of these as combining into the expression of what might be called negative friction – being rubbed up the wrong way. The complaint words with initial gr- might equally be thought of as a kind of articulate speech that has been broken or deformed by violent or painful pressure. Guttural sounds are routinely said by English speakers to sound ‘harsh’, a word that seems to imitate some of the asperity suggested by the sound and experienced in making it.

Linda Waugh quotes a New Yorker cartoon, in which one tiger is saying to another ‘griping, greedy, grasping, grotesque, gruesome, grisly – do you know any other good grr words?’ (Waugh 1993, 77). The comic idea that a creature like the tiger, whose very name is an encoded growl, would find positives where we tend to find negatives, is embodied in Tony the Tiger, the emblematic burning-bright animal who declares that Frosties are ‘grrreat’, turning friction into energetic frisson. The gutturals often suggest the animal in the voice, the animal being the chiasmic principle that is both within and against the animate, in Aristotle’s sense. Underlying all the negative gr-s is almost certainly the growl, which may be regarded as a kind of subvocal or prevocal utterance. In an animal, the growl is also prior in time, but not just because it precedes language, but rather because it is a rumbling warning of a roar. Leibniz called the r the canine consonant, for obvious reasons (Genette 1995, 47) – the dog not only is, but says, rough (though the name canis was sometimes explained as a derivation from canere, to sing.) When a girl asserts her aggressiveness, she may accordingly be called or call herself a grrl.

There are many other animals that appear to be invoked – that is, called up by, or called into voice – in the guttural fricatives. There is the frog that famously gets caught in the throat, and also crows that caw and other birds that crow. Hawking is the action of noisily and energetically clearing the throat of phlegm. Sometimes the sounds of the throat combine animal and supernatural overtones. Inuit shamans would sometimes have themselves harpooned in the back like walruses and then, in trance, go round every igloo in the camp ‘uttering guttural sounds like a walrus and spitting water and blood’ (d’Anglure 1993, 86).

The nasality that is often associated with guttural fricatives gives it a porcine grunting or snuffling cast. We tend to imagine that a nasal sound is a sound that has been displaced into the nose, and there is at least one people, the Dogon people of Nigeria, for whom nasal inflection signifies that the matter of the voice is out of place, and, since it is driven up into the nose, cannot escape into the air, and so putrefies and dies. Thus, it has been said, for the Dogon people, the nasal voice is powerfully and menacingly associated with death (Calame-Griaule 1993, 25). In fact, a nasal inflection occurs, not when sound is driven up into the nose, but when its passage in that direction is blocked – what we hear in nasality is actually not the voice unnaturally shunted up into the nose but the inhibition of nasal resonance.

Often, death is embodied in other peoples, or in the lethal or deathly sounds that seem to us to issue unwitting from their mouths. Coleridge remarks, somewhat cryptically, in the course of an admiring passage of talk on the fineness of the two English th sounds – as displayed in a phrase like the ether – on the comparative lifelessness of German when it comes to the articulation of the word death: ‘How particularly fine the hard theta is in an English termination, as in that grand word – Death – for which the Germans gutturize a sound that puts you in mind of nothing but a loathsome toad’ (Coleridge 1917, 185). I must admit I find out hard to make out what might be going through the mind, not to say the mouth, of the sage of Highgate here. For one thing, Coleridge seems to hear a hard th at the end of English death, which, while possible, would be a surprise, and for another, he seems to hear something guttural in the German Tod, death. I have played a few times with the sound, and find that it is just about possible, if you snort it as though you were a tank commander, to impart to it a little interior rumble that seems to implicate the throat, and therefore to introduce into the voice the mortuary toadish visitant, thereby inducing the hallucination of a croak in the animal’s name.

The animal overtones of guttural sounds may also give them a reference to eating, making them the site of interchange between the animal act of eating (an act which largely debars speech, or restricts it to guttural grunts) and the definitionally human or even divine act of speaking. This interchange is implicated in one of the most obviously echoic words in English, hiccup, which is an endstopped form of the word hiccough. (One might reasonably infer that hiccup is a later, lexicalised version of the somewhat more diffuse-sounding hiccough, though the judgement of the OED is that hiccough in fact is a later spelling, that puts ‘back’ in a sound that had not previously been audible in the word. Such chiasmic hysteron proteron is often to be found in relation to the letter h(Connor 2007, and we will see something happening a little later in relation to the word ghost.) A hiccough is caused by a disturbance of the mechanism that normally synchronises, or rather separates, the functions of swallowing and breathing, the epiglottis moving by reflex to seal off the trachea when food is moving into the oesophagus. In hiccough, involuntary movements of the diaphragm suddenly close off the trachea as though food were passing. The sound is produced by a sudden involuntary intake of air which meets the obstructed trachea, causing the hic- sound, followed by the -cup or -cough aftershock of eructation. A hiccough may therefore be seen as a kind of chiasmic spasm, as swallowing and breathing threaten to cross over into each other’s domains. Sadly extinct now is the word yex, from Old English ȝeocsian, ȝiscian, meaning to sob or hiccup, which implicates the velar fricative in the sound of the hiccup.

The fricative may be said to be negatively associated with eating, in that it strikes the ear as an expectorant sound, that is relating to the magically powerful actions of spitting and vomit. The guttural is at work in this way in the eecchh! and urghhh! of digust. ‘Eating the Wind’, by ‘Claude Searsplainpockets’ of the Speculative Grammarian, the ‘premier scholarly journal featuring research in the neglected field of satirical linguistics’, is an account of the elaborate ‘gastro-pulmonic’ system of an imaginary tribe called the Xoŋry, which sets out their elaborately-particularised designations and phonic enactments of compound activities of eating, speaking spitting, snorting, devouring and vomiting, including words for ‘to eat while talking’, ‘to snort up the nose while talking’, ‘to projectile vomit, while talking’ (Searsplainpockets 2006).

Such associations are memorably dramatised in a passage from Samuel Beckett’s Molloy, in which the title character describes his attitudes towards his names for his mother: ‘I called her Mag, when I had to call her something. And I called her Mag because for me, without my knowing why, the letter g abolished the syllable Ma, and as it were spat on it, better than any other letter would have done’ (Beckett 1973, 17). Here, the m sound seems to suggest acceptance and assimilation, the closing of the lips on what has come in from the outside: Roman Jakobson proposes that the nursing infant, whose mouth and lips are monopolised by the nipple, will automatically associate the contentment of the maternal breast with the nasalised sounds which are the only ones it can make in this circumstance (though, actually, a guttural gurgle is also possible during suckling and is also of course sometimes to be heard in eating). The g is, by contrast, expectorant, a convulsive casting up and out of what has been within. In English, ‘spitting’ is a labial matter, but Beckett’s original French, ‘crachait dessus’ is an altogether more disgusted and disgusting.

Sylvia Plath articulates an inverted form of this ‘phlegmatopoeia’, in Luke Thurston’s phrase (Thurston 2009, 133) in her violent father-loathing poem ‘Daddy’. At the centre of the poem is the sense of constricted speech, a speech that finds in the German of Plath’s father an obstructedness itself:

I never could talk to you.
The tongue stuck in my jaw.

It stuck in a barb wire snare.
Ich, ich, ich, ich,
I could hardly speak.
I thought every German was you.
And the language obscene (Plath 1972, 54)

The poem pushes itself repeatedly forward, and back, to this sticking point of baulked or blocked utterance, the alternative of tongues and feet suggesting a bizarre literalisation of putting a sock in speech, or putting your foot in your mouth – or, as the poem has it, ‘the boot in the face’ (Plath 1972, 55). The poem finds release from the glutinous ‘gobbledygoo’ (‘I’m through’ being its final words), if it does, in an utterance that by mimicking the asphyxiating chokehold, tries to empty itself of it, even though the release it finds is itself a kind of muffling or amputation: ‘So daddy, I’m finally through./The black telephone’s off at the root,/The voices just can’t worm through’ (Plath 1972, 56).

The otherworldly or even the diabolical has also been thought to be audible in such voices. Ventriloquism, in its archaic form of ‘speaking from the belly’, without the involvement of the tongue and lips, has often been regarded as a kind of impious or irregular speech, a speech that has been corporeally displaced, or even, according to one popular theory, speech produced by drawing in rather than pushing out the breath, which produces a characteristically growling, hollow kind of utterance (Connor 2000, 180). The particular kind of turbulence of the airflow in the fricatives is an accelerated form of this reversal, in which, rather than being pushed unidirectionally, from inner to outer, sound is detained and thickened in the throat.

The suggestion that the guttural fricatives are primitive, even prelinguistic, and the accidental fact that English (like other languages) tends to move such sounds from the back of the mouth to the front suggests easily that the language of the barbarian is transitional, between noise and voice, still turning the corner from the throat to the mouth. There are two kinds of archaism involved in the English apprehension of gutturals. The first associates them with the so-called Gaelic languages of the margins of Britain, where guttural fricatives survive – in Scots and Irish och and loch, for example, and with words of Germanic origin. The relative forwardness of Latin compared with the guttural backwardness of German is emblematised by the very words for language in the two languages -the seemingly liquid language and the the clogged Sprache. Derek Attridge has elaborated the logic that produces the contrasting associations of the sunny, Mediterranean letter c and the suspicious Germanic k that it often pairs, inviting us to think of the difference between Caesar and Kaiser or America and Amerika (Attridge 1984, 1124).

The second associates the gutturals with Hebrew, Arabic and other allegedly primal languages. Though Hebrew had the credit of being believed to be the language spoken by Adam, it was also regarded by others as an occult precursor of the daylight, classical clarity of Greek and Latin. The so-called Semitic languages could then be regarded as equivalent to the Germanic undercurrent of English, proximate yet distant at once. Lancelot Addison, who spent some time living in the Maghreb, or what was then known as the ‘Barbary coast’ of Northwest Africa, instructed his readers of the derivation of ‘Barbary’ from the way in which the inhabitants pronounced their language, their accents thereby coming to serve as their names, like the tiger, or the allegedly stuttering Hottentots:

For if we listen to the Moors language, Barbary seems to be descended from Barbar, which signifies an inarticulate murmur and grumbling noise without accent or harmony, for their speech is harsh, being very guttural: which is esteemed an argument of its Antiquity. And indeed it hath gain’d the vogue of no less antient a pedigree, than to be bred of the old Punic and Arabian. (Addison 1671, 75)

The word from which Barbary derives is the Arab barbara, to talk noisily and confusedly, a word that surprisingly appears to have no relation to the Greek barbaros, even though this too involves a mocking echo of indistinct or unintelligible speech. Greek and Arab, both of them speaking a language richly provided with gutturals, seem to have heard labials as primitive babble or hubbub (the latter derived perhaps from the Gaelic cry of aversion or contempt ub! ub!). For Addison, by contrast, it is the guttural which sounds barbaric.

There is a long history of enquiry into the sounds and shapes of letters, perhaps nowhere more developed than in the Jewish tradition of kabbala, which is based upon the conviction that the articulation of verbal sounds is powerfully creative, repeating, as it does, the primal act of creation springing from divine utterance. The Hebrew letter that corresponds to Greek χ chi is ח chet. One recent commentator draws a contrast between this letter and the hey, the simple aspirate in Hebrew, suggesting that ‘Whereas hey signifies wind, breath, chet signifies a barrier’ (Souzenelle 1993, 93; my translation). Souzenelle identifies the chet with the paradoxical barrier of sin, that both obstructs the passage to the divine and yet also marks the gateway to it – the serpent, the barrier, is called נחשׁ, nachas, a word which has the phonic obstacle at its heart (Souzenelle 1993, 96). Other contemporary interpreters emphasise the oscillatory function of the chet. Yitzchak Ginsburgh associates it with the principle of ‘life itself as pulsation, the secret of “run and return” ’ (Ginsberg 1992, 122). The shape of the letter is traditionally understood as formed from the two previous letters in the Hebrew alphabet, vav and zayin, joined by a thin bridge or roof. Ginsberg tells us that this enacts a kind of hovering energy, made up of ‘touching yet not touching’ (Ginsberg 1992, 122). The principal reference here is to Genesis 1.2, ruach Elohim…meraphehet ’al, which, as we have seen, has often been translated ‘The spirit of God moved on the face of waters’, but which means, literally, to ‘flutter’, ‘flap’ or ‘shake’ (Yong 2006, 191n8). But there is a more metaphorical understanding of this tremor enacted in the ruach, the word for spirit or life which ends with the chet, which is the brooding of an eagle as it hovers over the young in its nest. In Ginsberg’s interpretation

‘by “hovering” over created reality, G-d continues to sustain and nourish His Creation while simultaneously allowing each creature…the ability to grow and develop “independently.” The letter chet thus hints at the delicate balance between the revelation of G-d’s presence to us (the vav of the chet) and the concealment of His creative power from his Creation the zayin of the chet). (Ginsberg 1992. 122-3).

The rhythm of paradoxical oscillation is also emphasised in Lawrence Kushner’s reading of the letter, which makes more of the parallel between sound and image:

In the Torah the Chet… is written with a sharp jagged notch on its forehead. It is almost as if there were two separate letters barely joined together. They need each other to stand. But they wish they did not. So they barely touch. Chet is the agony of a soul torn apart from itself. The top of your throat and the bottom of your throat fighting against one another create the sound of the chet. (Kushner 1975, 27-8)

It is for this reason that the chet gives rise to many ‘strange and conflicted word-pairs’ (Kushner 1992, 28). Thus, sin, chait, suggests ‘A soul torn against itself because it is sure that it is pious’, while chasid, a pious one ‘is a soul convinced that it commits many sins’. Pain is chavel, sonically and graphically twinned with chayim, life, suggesting ‘Almost dying from birthwork. Bringing forth life’; while destruction is churban, but linked with chuppa, the marriage canopy which the letter is often said to resemble (Kushner 1992, 28).

Theories of sound symbolism, or sonorously performative language, often involve the crossing of image and sound that Genette has called ‘mimeographics’. There are two interpenetrating forms of this. In one, the shapes of the mouth perform what will become the tensile shapes of letters. According to Francis Mercurious Van Helmont, the tongue ‘naturally forms the Hebrew Letters in the Mouth’ (Van Helmont 1694, 109). The semaphoric phantasy of letters danced out by the mouth is mirrored by the choreographic suggestiveness of the shapes of the letters on the page, which ‘begins to teem with little mouths opening, or closing, with gulets, with upside-down noses, with pointed or curled-up tongues – more or less recognisable, always both open to the view and hidden away in the thicket of letters’ (Genette 1995, 57).

All of this is said under the sign of immense contemporary discredit and scepticism. For it arises from a linguistic tradition of sound symbolism or phonosemantic speculation that most linguistics, governed by the Saussurean principle of the arbitrariness of the sign, still regards as fanciful and voluntaristic. The technique of phonosemantics is both easy and easy to disprove. One simply assembles all the examples of the letters that one can and observes the semantic patterns that emerge. There are few people who have not at some time been struck by the fact that the short i sound often seems to appear in English words denoting littleness or minimality. Few of those who tabulate these associations take much account of exceptional words like big, immense, mighty and miteand small. Still, there is some striking evidence of a regular correlation across a number of languages between sounds of relatively higher pitch – short i, for example – and ideas of smallness. Edward Sapir (1929) told 500 subjects that the words mal and milwere words for different sizes of table in an unnamed foreign language, and asked them to decide which was which: an overwhelming majority (96% of the adults) assumed that mal would designate the larger table, and mil the smaller.

This example recurs frequently in the literature, no doubt in part because the results obtained seem so compelling. There are other studies which try to find similar correlations between sounds and spatial values. The values implied by and correlated by the i/a contrast may be tabulated as follows, to suggest a general set of regular homologies between the values of what are felt to be closed and open sounds:

i a
front (mouth) back (throat)
closed open
high low
light dark
bright dull
definite indefinite
exterior interior
distant proximate
located placeless
articulate inarticulate
form force
image sensation
word cry
human animal
protasis (tic) apodosis (tac)
filling space filled with space

The essential difference here seems to be between a conception of closed or definite space, and vaguer, more indefinite space, with the primary locale being the semi-imagined spaces of the mouth and throat. Sounds made in the front of the mouth seem to partake in the strong and precise familiarity that most people feel with the geography of the mouth. It is an awareness with which Nabokov plays in the opening words of Lolita, which featly foots a little phonetic dance from the elements of the infant siren’s name, and in the process also wittily dramatises the relative lowness of o, from the back of the mouth, and the lightness of i: ‘Lolita, light of my life, fire of my loins. My sin, my soul. Lo-lee-ta: the tip of the tongue taking a trip of three steps down the palate to tap, at three, on the teeth. Lo. Lee. Ta’ (Nabokov 1997, 9). It is a movement of advance, ascent (despite the fact that the steps come down to the front) and emergence, from loins to teeth, from desire to articulation.

Sounds from elsewhere in the vocal apparatus may be strongly felt, but felt just like that – as belonging to a vaguer, more undefined elsewhere. It is much more disturbing to have a portion of one’s mouth numbed by anaesthetic than one’s throat, since one’s throat is any case visually more numb, and as it were less clearly outlined. In a curious way, since I am able to visualise the different areas of my mouth so well, it is as though I were in some sense outside it, looking in. That perceptual vantage point is not available to me in thinking of the throat and the even more occult parts of the bronchial apparatus, from which few sensations are likely to be reported, unless I swallow something hot or sharp, or have an inflammation. I am in my throat without somehow quite feeling able to look in on it. My mouth is an inside that is perceivable from an outside – I seem at times to work it by strong or remote control. My throat is an inside that I am, that is only perceivable from inside.

This contrast between defined and undefined space, at an even greater level of abstraction, a distinction between matter and the immaterial. Although matter exists in three, progressively more attenuated states solid, liquid and gas, matter means for most of us matter in its solid state. The characteristic of such matter is to be dense, concentrated, and definite, which is to say, with clearly defined boundaries between a material form and its outside. Matter, that is to say, occupies space; a material thing is a thing that has a certain extension in space (a ‘body’), but an extension that goes only so far, such that the thing can be distinguished from the space around it and from other things in that space. The immaterial does not in so clearly apprehensible a way divide space; rather it tends to occupy space, spreading into and blending with it indivisibly. Hence the fundamental distinction enacted by these two kinds of sounds, and the physiological processes we feel and imagine (and imagine we feel), is that between matter in space and matter as space.

There is only one more step of abstraction that it is possible to make, which, while we are in this stravaging mood. we should not be timid of making, for it may deliver to us an essential feature of the experience of language. The distinction between frontal and posterior sound values is finally a distinction between matter and immaterial, or between something and nothing. The mouth is a furnished room, with items disposed in an orderly way in its space; the throat, and all the other inner compartments with which one may feel it dimly and indistinctly gives on to is vacuity itself. The mouth is a place that can be selectively and sequentially lit up by the flashlight of the tongue; the throat is a dark place, a space therefore of indeterminate shape and extent. The throat is a space of nothingness. It is the place in the mouth in which the nothingness of pure gurgling, the pure agitation before articulation, can be heard as articulation. The sounds that issue from the throat, however, do not, insofar as they are sound, entirely belong to it. They function as sounds in a language, even though not belonging entirely to it, meaning that they do not belong entirely to the nocturnal realm of noise either. They are a place where darkness and light, though not yet fully distinguishable, alternate: they are an oscillating chiaroscuro, in which the distinctness and the indistinctness of light and dark chiasmically cross and recross.

Of all the vocal sounds, the guttural, and especially the fricative is the aptest to suggest the agitation of the mucus membrane and its associated substance, the phlegm, that is always present in the throat, is as it were the paradoxical embodiment of this nothingness, a glutinous sort of space-stuff, a stuff that clings to its space, that is never distinct enough from the space it occupies. The desire to clear the throat, which can become pathologically intense and unremitting, is a desire to move from the tangled thicket of the throat into the clearing of the mouth, and by getting ‘it’ up and out (‘better out than in’) to make it a something rather than a something-which-is-not-a-thing, and therefore something that can be fully expelled, and thereby rendered absolutely nothing.

The throat marks the point of transition at which the slimy, semi-solid quasi-nothingness or space-stuff of air is mantled in articulate speech, the point at which the animal becomes animate in becoming articulate. It is the point (which never exactly comes to a point, though) at which the nothing of pure, purblind guzzle or gargle becomes the something, or rather, the many different distinguishable things, of speech. Nothing becomes something. And yet, in thus becoming more definite, more articulate, speech also seems to become less material, less bodily. The glutinous, clinging ur-stuff of speech is sublimated in the crucible of the mouth into mobile sense and meaning. It is now, in Aristotle’s terms, ‘sound with a meaning, and is not merely the result of any impact of the breath as in coughing’. So, in becoming language, something (bodily matter) becomes nothing (meaning); but it is also the case that nothing (mere matter) becomes something (meaning). This transition has the classical cross-your-heart shape of the chiasmus.


It is for this reason that the guttural sound is not just the sound of the beginning of life and meaning, the rasp of the ruach that gives form to what is tohu-va-bohu, without form and void, dividing the light from the dark; for the throat is also that part of us from which issues the death-rattle, that, like the snore, can never be uttered wittingly, or with the full present-and-correct assent of the soul on the qui vive, but is nevertheless the sound we (?) make when we give up the ghost. Certain kinds of heavy metal bands, notably Napalm Death, but also Necrophagia, Master, Hellhammer and Cannibal Corpse (vocalist Chris Barnes), cultivated a growling, kind of vocal style, described somewhat diminishingly, though informatively in 2004 by Will York as the ‘Cookie Monster style’ in order to embody their preoccupations with ‘death, decay, and other such existential dilemmas’ (York 2004).

The association of the guttural with emptiness is assisted by many other associations. One of them, about which I have had a bit to say before this (Connor 2002), is the figure of the Gorgon, who may be regarded as something like the embodiment of gutturality, not least because she ends up decapitated, the blood gushing from her throat. But this fate seems to be anticipated in her name, deriving from the Sanskrit root garg, which according to Thalia Feldman, signifies ‘a gurgling, guttural sound, sometimes human, sometimes animal, perhaps closest to the grr of a growling beast’ (Feldman 1965, 487). The word garg, and the name gorgon to which it gives rise, means to make a noise which is neither a word or a name – a name that means nothing except that, that it means nothing. The apotropaic gorgoneion, or gorgon’s face, traditionally represented her with her tongue stuck straight out, as though to indicate the surrender of the functions of the mouth to those of the throat. Her legacy is to be found in the grotesque gargoyles which decorate churches and cathedrals throughout Europe, which are the facialisations of gurgling gutturality. The Gorgon or Medusa was interpreted by Freud as a figure of castration, as the threat of nothingness, though his focus, somewhat literalistically, was on the pubic tangles of her snaky hair (Freud 1981). But there is horrifying emptiness enough in the notion of a head devoted to the production of headless and voiceless utterance. Adapting the frequently-used Latin phrase vox et praeterea nihil, a voice and nothing more, Jane Harrison remarks that the Medusa ‘is a head and nothing more; her potency only begins when her body is severed, and that potency resides in the head; she is in a word a mask with a body later appended’ (Harrison 1991, 187). But it might just as well be said that this is a body facialised, an acephalic body that has not so much grown a head, as grown into one.

The horror of that which is there and not there takes a milder form in the idea of the ghost. Here, once again, gutturality cooperates in the compounding of something with nothing. Ernest Weekley remarks that the presence of the letter h in the word ghost makes the idea of the ghost even more ghostly (Weekley 1935, 4). In any case, the letter g has often been read and heard as connoting vacuity. Charles de Brosses wrote in his Traité de la formation méchanique des languages of 1765 that ‘gaping things are painted by the letter of the throat (g), such as gouffre (abyss), golfe (gulf). or even better by the aspirating letter, as in hiatus’ (quoted Genette 1995, 71). This sense of a being that is compact with nothingness, the sense it involves of something being there that isn’t, is well conveyed by a visible sound which cannot in fact be heard, though there seems nevertheless to be a hollow kind of resonance with words like ghoul and ghastly. These seem to be specifically Germanic ghosts, named with a word from the Germanic language that still hangs anachronistically around English. But, as Weekley points out, this ghostliness is itself a doubly phantom effect. Though giving the appearance of a haunting survival, the h in ghost is in fact a relatively new arrival, and one, what is more, that derives from the predilections of the eye as much as the ear. In Old English the word is spelled regularly as gaest, and in Middle English as gost or gast. Ghost first appears in Caxton, perhaps, the OED surmises, influenced by Flemish gheest. And yet, the ghost still haunts, for Geist seems at home, and ghost still wanders abroad; so, ‘I get your gist’, but ‘I haven’t the ghost of an idea’.

Something similar seems to have been at work a couple of hundred years later when, needing a name for the mysterious, unpredictable, airy stuff that nevertheless was not air given off when he heated coal, Jan Baptiste van Helmont invented the word ‘gas’, explaining that ‘by the Licence of a Paradox, for want of a name, I have called that vapour, Gas, being not far severed from the Chaos of the Ancients’ (Van Helmont 1664, 29). We should remember that, for the Flemish Van Helmont, gas and chaos would have sounded much more similar than they do in English today. The primary meaning of the Greek word chaos is a gulf or chasm, from the stem χα- to yawn, or gape – a sound made, according to Dwight Bolinger’s suggestion, when one’s jaw sags open and one looks ‘aghast’ (Bolinger 1940, 70-1) The ‘spiritus silvestris’, or ‘wild spirit’ for which Van Helmont sought a name was not much different from being a pure chaos, a tohu-va-bohu, a nothing, The wild spirit that it is, is a chiasmic crossbreeding of gaseous nothingness and what Helmont called ‘the thingliness of a Gas’ (Van Helmont 1664, 69).

Linguistics tends to assume the coherence, autonomy and continuity of a language, and for that reason has tended to treat individual languages as closed and self-sufficient systems. This makes it hard to account for the status of the guttural in English, from which it is largely absent. However, language is never entirely reducible to ‘a language’, for languages are constantly influencing and entering into one another. In the case of the gutturals, we have a class of sounds that have an impact on English precisely because they are felt not, or no longer to belong to it. In one sense, the phonology of German, for many reasons the reference language for English phantasms of gutturality, is separate from and therefore outside English. But, insofar as a kind of phonesthetic memory (or mnemonic phonology) seems to exert an influence on English, and to have an oblique but emphatic presence-in-absence within it, German is also inside English. So the guttural fricative both is and is not a sound in English. You will scarcely, by this time, expect me to pass up the chance to say that this chiasmic flicker between presence and absence is like an accelerated form of the guttural fricative itself.

The question at issue raised by the kind of ‘mimological reverie’ (Genette 1995, 309) which I have both documented, and have myself somewhat dementedly been indulging here is, as Genette puts it, ‘whether or not natural language respects and invests these imitative capacities in its functioning’ (Genette 1995, 323), whether it is a generative magic that language itself performs, or the effect of a kind of magical thinking that we perform upon language. But the further question this raises is whether it is possible to keep language ‘itself’ apart from the work of phonaesthetic phantasy. If language is not just what is abstractly given in structures and relations, but also what is historically made of those structures, or of what we profoundly seem to wish to be true of language, whether on rational or irrational grounds, then what Janis Nuckolls describes as ‘a kind of sound-symbolic creativity that is protracted through generations, below the threshold of awareness for most, yet assented to and thus engineered by entire communities of speakers’ must be regarded as something language is as well as something we do to it (Nuckolls 238)

I am not entirely opposed to the possibility that sounds begin to gather a kind of magical significance, though it comes about as all magical effects do, not by inherence but by a kind of supervenience. The number 13 has no power to bring luck or misfortune. But the attention paid to the number 13, by superstitious people, or by people who are determined not to surrender to superstition (like me), actually gives it a supervenient power of being fearfully attended to and having its effects scrutinised. This implies that when we examine the magic embedded or enacted in words, we are examining the aptitude to read those words in that way – the magical effect being an effect of magical according or attribution of magical power.

The difficulty here is that the two traditions are hard to separate in practice. In fact, there might be said to be a chiasmic oscillation between them, which is the same oscillation as is to be found in thinking about magical thinking. Magic does not exist in that there really is no omnipotence of thoughts. But the thought that there may be omnipotence of thoughts is a powerful and a dangerous thought (dangerous because it can itself become omnipotent). There is nothing to fear from magic, but plenty to fear from magical thinking. Rationalists like me have a hard time keeping our thinking about magical thinking free from magic.

This oscillation between the magical and the rationalist is mimicked in the oscillation between iconicist or mimological theories of word formation and conventionalist theories. For one does not need to give credence to the claims made by most iconicists that certain sounds have essential and archetypal meanings to accept that beliefs or hunches about such meanings might often put pressure on the formation and function of words. In such cases, the iconicity is not preformed but performed, not given but made. The fact that the performance depends upon a mistaken theory need not prevent the repeated error having a certain generative force. Language comes to consist of what is done to it, meaning that language is always inside and outside, behind and in front of itself.

The engine of this inside-outside alternation, this outering of language, is the fact of utterance, the fact that there is no language without the putting of language into play and at stake that is utterance. Voice and language envelop and exceed each other. Voice contains language, for there is nothing in language that cannot be given voice to or taken up into voice. But language also contains voice, since what we call a language is the sum total, or the imaginary horizon, of all possible occasions of speech. Voice is the soul of language, but, as Aristotle is at pains to assert, language is the soul of voice, what gives it its essential meaning (for without language and without meaning, the voice is merely noise). But voice is indeed a kind of noise within, or parasite upon, language. As Mladen Dolar observes, the voice is precisely that which is never audible to phonology – which is why phonology, from phone, may also be thought of as phonos, murder, the assassination of voice (Dolar 2006, 19). The voice is accent, accident, occasion, all of them signifying etymologically the way in which things fall out. We can gloss the word accent as ‘essential accident’. Everybody has an accent as everybody has a point of view; voice is the necessity of this contingency, the way it has to happen to have fallen out. Voice is the necessary disturber of language’s peace, that will not leave it one piece.

Voice is therefore the friction of language, that rubs it against the grain, and fricatives are the emblem and occasion of the action of vocality as such. How are we to understand friction or frication? Friction encompasses a spectrum of actions. At one end, there is the caress, that conserves and curates that which it touches. The caress is a non-appropriative touch. It is also productive; in the words of Jean-Paul Sartre, ‘[t]he caress is not a simple stroking; it is a shaping. In caressing the Other, I cause her flesh to be born beneath my caress, under my fingers’ (Sartre 1984, 390). The caress is both a touching and a keeping intact; like the moving or brooding of ruach on the waters, it seems to imply the suspiration, or the insufflation, the tenuous, tremulous touch of breath alone. The caress illuminates and accentuates the form of what it touches, giving it its distinctness and integrity. At the other end of the spectrum from this lightest of all touches is the violent abrasion, that grinds, rips, pulverises, reducing unity to atomised multiplicity. The voice moves between these two extremes in its relations with itself. Sometimes, the voice seems to touch itself with the most erotically delicate of touches, a touch that withdraws from and in the process yields place to itself, a doubling that unifies, like an echo or an aura. But sometimes the voice also seems to turn on itself, shaking itself to pieces as though it held itself terrier-like in its own teeth. It is cooling zephyr or wrathful whirlwind. When voice delicately doubles itself, it seems to shape for itself an ideal and serene body, made out of nothing but sound. When voice rancorously grates, scrapes or abrades itself, it aims to mill itself down into a different kind of nothingness, the nothingness of the dross, the dissolute, of mere empty frication. More than any other sound, guttural frication performs voice’s self-relation, or language’s relation to itself via the scratchy intercession of voice. Voice is nothing more or less than this friction between the two extremes of frication, that make the voice something out of nothing, or nothing in the midst of something.


Addison, Lancelot (1671). West Barbary, or, A short narrative of the revolutions of the kingdoms of Fez and Morocco with an account of the present customs, sacred, civil, and domestick. Oxford: John Wilmot.

Attridge, Derek (1984). �Language as imitation: Jakobson, Joyce, and the Art of Onomatopoeia.�MLN, 99, 1116-1140.

Beckett, Samuel (1973). Molloy. Malone Dies, The Unnamable. London: John Calder.

Bernard, John and Arthur Delbridge (1980). Introduction to Linguistics: An Australian Perspective. Sydney: Prentice-Hall of Australia.

Bolinger, Dwight L. (1940). ‘Word Affinities.’ American Speech, 15, 62-73.

Calame-Griaule, Geneviève (1993). `La nasalité et la mort.’ In Pour une anthropologie des voix, ed. Nicole Revel and Diana Rey-Hulmann. Paris: L’Harmattan/Langues d’O, 23-33.

Connor, Steven (2007)‘Whsper Music.’ Online at
————–(2002) ‘Seeing Sound: The Displaying of Marsyas.’ Online at
————–(2000). Dumbstruck: A Cultural History of Ventriloquism. Oxford: Oxford University Press.

dAnglure, Bernard Saladin (1993) ‘The Shaman’s Share, or Inuit Sexual Communism in the Canadian Central Arctic.’ Trans. Jane Philibert. Anthropologica 35, 59-103.

Digby, Kenelm (1669). Of Bodies and of Mans Soul to Discover the Immortality of Reasonable Souls… 2 Vols. London for John Williams.

Dolar, Mladen (2006). A Voice and Nothing More. Cambridge, Mass. and London: MIT Press.

Feldman, Thalia (1965). ‘Gorgo and the Origins of Fear.’ Arion, 4, 484-94.

Freud, Sigmund (1981). ‘Medusa’s Head.’ The Standard Edition of the Complete Psychological Works. Vol. 18, Beyond the Pleasure Principle, Group Psychology and Other Works, 1920-22. Trans. James Strachey. London: Hogarth Press, pp. 273-4.

Genette, Gérard (1995). Mimologics. Trans Thaïs E. Morgan. Lincoln NB and London: University of Nebraska Press.

Ginsburgh, Yitzchak (1992). The Hebrew Letters: Channels of Creative Consciousness. Jerusalem: Linda Pinsky.

Harrison, Jane (1991). Prolegomena to the Study of Greek Religion. Princeton: Princeton University Press.

Kushner, Lawrence (1975). The Book of Letters: A Mystical Alef-bait. New York, Evanston, San Francisco and London.

Nabokov, Vladimir (1997). Lolita. New York: Vintage.

Nuckolls, Janis B. (1999). ‘The Case For Sound Symbolism.’ Annual Review of Anthropology, 28, 225-52.

Plath, Sylvia (1972). Ariel. London: Faber and Faber.

Serres, Michel (2008). The Five Senses: A Philosophy of Mingled Bodies (I). Trans. Margaret Sankey and Peter Cowley. London: Continuum.

Sapir, Edward (1929). ‘A Study in Phonetic Symbolism.’ Journal of Experimental Psychology, 12, 225-39 (1949) In Selected Writings of Edward Sapir in Language, Culture, Personality, ed. D.G. Mandelbaum. Berkeley: University of California Press., 61-72

Sartre, Jean-Paul (1984). Being and Nothingness: An Essay on Phenomenological Ontology. Trans. Hazel E. Barnes. London: Methuen.

Searsplainpockets, Claude (2006). ‘Eating the Wind: An Antropological Linguistic Study of the Xoŋry.’ Speculative Grammarian, 151. Online at Accessed 1 September 2010

Souzenelle, Annick de (1993). La Lettre chemin de vie: Le symbolisme des lettres hébraiques. Paris : Albin Michel.

Thurston, Luke (2009). ‘Outselves: Beckett, Bion and Beyond.’ Journal of Modern Literature, 32, 121-43.

Van Helmont, Jan Baptiste (1664). Van Helmont’s Works. Trans. John Chandler. London: L. Lloyd.

Van Helmont, Francis Mercurius Van (1694). The Spirit of Diseases, or, Diseases from the Spirit Laid Open in Some Observations Concerning Man and his Diseases… London: for Sarah Howkins..

Weekley, Ernest (1935). Something About Words. London: John Murray.

Yearsley, James (1853) Throat Ailments: More Especially The Enlarged Tonsil and Elongated Uvula, In Connexion With efects of Voice, Speech, Hearing, Deglutition, Respiration, and the Imperfect Development of Health, Strength, and Growth, in Young Persons. 5th edn. London: John Churchill. [first edition 1842]

Waugh, Linda R. (1993). Against Arbitrariness: Imitation and Motivation Revived, With Consequences for Textual Meaning.’ Diacritics, 23, 71-87.

Yong, Amos. ‘Ruach, the Primordial Chaos and the Breath of Life.’ In The Work of the Spirit: Pneumatology and Pentecostalism, ed. Michael Welker (Grand Rapids, MI and Cambridge: William B. Eerdmans Publishing Co.), 183-204

York, Will (2004). ‘Voices From Hell.’ San Francisco Bay Guardian, July. Online at Accessed 30 August 2010.