Darian Leader

Since psychoanalysis is a practice based on speech, it is surprising that so little has been written on the voice as an analytic concept. Lacan made this one of the objects of psychoanalysis, yet commentaries have not been bountiful, and they have, for the most part, tended to confuse the voice with the dimension of sound. What I hope to show in this lecture is how the voice should not be equated directly with acoustic phenomena, and how, if we can find it in the field of sound, it is still a distinct concept and hence not identifiable with sound as such.

In Freuds work, the question of the voice is more or less posed in terms of the superego. After some early remarks on the role of auditory traces in the construction of phantasies, his interest turned to how the superego is established during childhood, and he argued that this involves the internalisation of the parental voice. Verbal residues, he tells us in The Ego and the Id, are derived from auditory perceptions, and these will eventually constitute the superego, defined as a residue of the punitive agency of childhood. Its internal voice will both admonish us for our trespasses and encourage us in the pursuit of impossible tasks, while the ego is left to suffer the consequences of these contradictory imperatives.

These themes were taken up by many of the early Freudians, and for most of the 1920s studies of the superego focused on the theme of guilt. Since the superego was assumed to be the voice of conscience, investigating the feeling of guilt and its absence promised to offer new material on the genesis of the psychical agency itself. While producing some fascinating material, these studies moved away from Freuds concern with the auditory side of the superego. Nonetheless, the voice was still there, and we find a marvellous example in a 1924 paper by Theodor Reik, Psychoanalysis of the Unconscious Sense of Guilt.

Reik reports a conversation with his son which must have delighted anyone familiar with psychoanalytic ideas. The boy describes what he calls an inner voice which says to him You mustnt play with your gambi (p.439). When asked to define this inner voice, he says Its a feeling in yourself and the voice of someone else. This vignette is of great interest. For the Freudian audience it must have been a perfect confirmation of the hypothesis that the superego is a voice, and that this voice is the agency of prohibition. And beyond this, it situates the voice as being both an inner phenomenon and an outer one : it is, as he says, a feeling in himself and the voice of someone else. It is thus both inside and outside.

In the subsequent literature, very little was to be written by the Post-Freudians on these aspects of the voice of the superego. There was a great deal on the prohibitive aspects of this psychical agency, and a great deal on questions of dating and the idea of archaic, preoedipal superego nuclei, but not much on the auditory dimension. The notable exceptions were papers by Otto Isakower and Robert Fliess, and Lacan was certainly familiar with the work of both of these authors. Fliess summarised much of his research in the 1956 book Erogeneity and Libido, where he examined both the content and the form of auditory introjections. Isakower's research was more specific, centering round the experience of falling asleep (Isakower 1939). By exploring the linguistic phenomena of this process, he came up with a remarkable model of superego formation.

Isakower had linked the auditory residues present while falling asleep with what he saw as the nucleus of the superego, and he needed a way to formulate how the initially external voice of the parent could be internalised. He found his model of this process in a curious marine phenomenon. The crustacean Palaemon, a tiny shrimp-like creature, was known to periodically insert grains of sand into a minute aperture known as the otocyst (or statocyst). These external objects were known as otoliths, and the Viennese physiologist Alois Kreidl had shown in a devilish experiment that their basic function was to regulate the organisms orientation. Placing some specimens in seawater where the salt had been replaced with iron filings, he waited for them to insert these foreign bodies into their otocysts and then placed a large electromagnet above the tank, playing havoc with their balance.

Isakower saw this experiment as giving a model of superego formation. The otoliths could be equated with fragments of the parental voice, incorporated by the subject, and, as the introduction of Kreidls magnet showed, they would still be predicated on an external agency. This analogy complicated the simplistic model of internalisation, since it showed how what was inside was also outside. As Isakower commented, the child has to build up his speech from linguistic material which is presented to him ready made. But this very fact sets in motion the process of developing an observing and criticizing institution (Isakower 1939, p.345), since ready-madeness means a link to the subjectivity ascribed to the parent. This led Isakower to a study of the linguistic aspects of this archaic incorporation of the Others voice.

Now, what does all of this have to do with the phenomena associated with falling asleep and waking up, these momentary states known as hypnogogic and hypnopompic? These transitional states had received a certain attention from nineteenth and early twentieth century psychiatry (Heynick 1993). The verbal productions associated with these states were thought to provide clues to linguistic functioning and its degeneration. Examples range from the abruptness of the phrase Young trouts to the Joycean Or squawns of medication allow me to ungather or the sequential No, Id rather have spaghetti. Many of these phrases are incomplete, calling either for syntactic completion or embedding in a sequence of further sentences. In this respect, they might remind us of the interrupted messages of psychosis that so interested Lacan.

Isakower's study of hypnagogic states was intriguing. They displayed, he noted, an almost exaggeratedly elaborate grammatical and syntactic structure. The speech flows along in complex phrases, with strongly accentuated sentences of animated and changing form, but it loses its clarity more and more as it proceeds, and at length there remains only an impression of lively and complicated periods without any verbal elements which can be clearly grasped...until at last the periods gradually pass over into a scarcely articulated murmur (Isakower 1939, p.347). They seem to indicate that semantic content is becoming more and more absent.

Linking these phenomena to the structure of the superego, Isakower comments that What we see here is not so much content that is characteristic of the superego but almost exclusively the tone and shape of a well-organised grammatical structure, which is the feature which we believe should be ascribed to the superego. If these elements are linked to the state of falling asleep, waking up is frequently marked by the emergence of the sensation of an interpellation, the linguistic auditory phenomena reaching the sleeper like a call with a superego tinge, sometimes threatening, sometimes criticisizing - words for which the dreamer, as he wakes up, feels an inexplicable respect, although they are very often a quite unintelligible jargon (Ibid. p.348). This is a brilliant observation. Not only has Isakower drawn attention to the peculiar auditory form of hypnagogic language and linked it to the superego, he has noticed the sleepers odd relation to it, one of respect. It is as if these strange words or phrases exert a gravitational pull on the subject, even if their meaning is totally unintelligible. We'll come back to this tension between syntax and semantics later on.

Now, most of the later psychoanalytic work around the voice focused around its content, as we noted, until, from the seventies onwards, interest shifted to the dimension of sound as such, separated from meaning. The voice was now identified with pure sound, acoustic, sonorous and sometimes melodic presence. In Lacanian circles, it was often understood as the residue of the signifying operation by which the infant was caught up in the networks of language. The infant's cry would be interpreted by the caregiver and a meaning established: the code, as Lacan said, was that of the Other. Recent studies, indeed, have shown how the attribution of intentionality to infants is a prerequisite not only for their own intentional actions but for their very capacity for intentionality (Trevarthen 1977). The voice would be the remainder of these processes through which acoustic productions are given meaning. Logically, then, the voice would always be beyond meaning. It would be that part of the cry which wasn't absorbed in the network of meanings.

This interpretation suited people nicely. It gave the voice a kind of mystique, and it seemed to be at home in both analytic and cultural studies theory. From the analytic side, its background could be found in Winnicott, who had famously included sounds in his list of the transitional objects of children, alongside their bits of blanket and stuffed toys (Winnicott 1953). Winnicott had been alert to the function of such objects, and this allowed him to move beyond classical questions of their symbolism: it was now their use that defined their meaning, and so acoustic phenomena could count as objects just as much as material objects like bits of blanket could. It followed that our affective relation to music was predicated on transitional phenomena and their significance in childhood.

Developments in infant studies seemed to confirm all this. Whereas it had once been thought that the auditory apparatus was only fully functional some time after birth, it became clear that it is operative in a basic way from around five months after conception (Deliege and Sloboda 1996). Sounds were found to both increase fetal activities and at times to inhibit them. There was also clear evidence that sounds heard in utero could be singled out by the infant after birth. Even better was the moment when infancy researchers had the idea not only to see if such sounds were favoured over other sounds by the baby, but to add the variable of the mothers own affective relation to the sounds. Indeed, it turned out that music prefered by the mother was prefered by the child.

These studies should nuance the separation between the voice as a pure acoustic object and the voice considered in its relation to the Other. What they show, in fact, is how auditory phenomena are closely linked to this relation rather than situated in some space beyond it. Indeed, the more we scrutinise these studies, the more we see how they involve dialogue. Many recent developments in child psychology are based on the assumption that the apparently undifferentiated fetus and infant are in fact individuals right from the start. Without subscribing to the dubious baggage that goes with such theories, their merit lies in the emphasis on an archaic dialogue with the mother, which starts in utero. Whereas it was once assumed that such dialoguing only starts around the age of 16 months, Trevarthen and others have shown how it is operative much earlier ( Trevarthen 1974 ). The timing of exchanges between mother and infants shows interacting cycles of activity, where after making their contribution each participant retires slightly to allow the other to respond. The turn-taking that is required by any dialogue will, in turn, prepare the way for speech.

What such research shows is how even when the child speaks on its own, with no one else present, the formal features of dialogue are still present. Crib speech, the speech of babies as they are falling asleep, has no doubt existed as long as babies have, and it is amazing that it was only in the sixties that it began to receive any attention from linguists. The groundbreaking study was conducted by Ruth Weir in 1962, with her two-and-a-half year old son Anthony as subject, and it was published with an introduction by Roman Jakobson (Weir 1962). Weir placed a tape recorder next to little Antony's bed and then performed a linguistic analysis of the data recovered over a period of several months. Her results are exceedingly suggestive. What did she find?

Listening to the crib speech of her son, she noted first of all something that later studies confirmed: the remarkable frequency of imperatives in the child's speech. Lying alone in his bed, he would recite orders to himself, and this presence of another speech within his own was found throughout the recordings. Sometimes such imperatives could be taken for declaratives, as in phrases like Jump on yellow blanket or Make too much noise, and Weirs hesitation as to how best to classify these examples is instructive. It shows, we could argue, how what ends up as a declarative has its origin in an imperative: that is, in speech coming from the Other and addressed to the subject. All speech, perhaps, has this imperative root, and it is significant that linguists once wondered whether the imperative was the first mood of human speech.

As Weir studied his her sons evening soliloquies, she came to another remarkable conclusion: that these apparent monologues were not monologues at all but dialogues. Anthony produced what she calls a dialogue spoken by a single person (Weir p.146). It was as if Anthony was always in the process of addressing himself, and bedtime speech was a privileged moment for this. While his little stuffed toy Bobo was more or less ignored during the day and not particularly missed when left somewhere, at bedtime Bobo would become an interlocutor in crib speech, the addressee of numerous commands and calls. The fact that these dialogues would take place on the frontiers of sleep supports Isakower's argument, and implies that this is a privileged point for the internalisation of speech. Similar findings have been made by later researchers (Pickert 1981), but before trying to make sense of Weir's results, we could bring in another theme from child language studies.

If little Anthony was especially interested in having interlocutors with whom to speak, mothers spend a large percentage of their time dialoguing with their babies who are in no position to answer back directly. Cross-cultural studies have shown how around seventy percent of mothers' speech to babies consists of interrogative forms: Are you hungry?, Do you want a drink?, Are you too hot? etc. The puzzle here is less the frequency of these syntactic forms than the fact that they were not mirrored in the eventual speech of the babies themselves. There is no demonstrated correlation between the frequency of interrogative forms in maternal speech and in that of the children. In fact, the inverted word order characteristic of interrogative forms in some languages is hardly ever present in the early linguistic productions of native speakers (Elliot 1981).

This is surprising given what we know about imitative patterns, and it is interesting how researchers found that not only do children tend to imitate maternal speech, but they imitate more when the mother is imitating them. The relative frequency with which children imitate their mothers speech was correlated with the relative frequency with which mothers imitated their children's speech (Kuczaj 1983, p.6). The children, it turned out, were more likely to imitate maternal imitations than other speech acts. It shows how infants don't only learn imitating their mothers but learn, if you will, the process of imitation itself.

Now, if we consider the frequency of such interrogative forms in maternal baby talk, it suggests that the baby, even though unable to speak, is being given a potential space within the mothers linguistic world. Even if they cannot reply with words, infants are being given the possibility of responding, which may take the form of gesture, cry and later words as the child grows older and learns the codes of the mother. But in its first months, the function of maternal questions must be to prepare a space for the subject, to offer a place within speech for the subject to be born. The interrogative forms of maternal speech not only create a context in which babbles become meaningful, but offer a space to the subject. This is by no means a given: think of those situations where the caregiver will not ask any question to a child, but, on the contrary, the child that they are hot, cold, hungry or thirsty. This leaves no place for the subject to emerge. The infant is simply an object for the Other here.

Question forms in maternal speech also bring to mind the phenomena discussed by Isakower. Remember how in the states of falling asleep and waking up, so many of the linguistic fragments seemed to require completion or elaboration. We've probably all experienced this on waking up: were either left with a verbal fragment that we can't make any sense of but that seems important, or we know that we've solved some mystery or puzzle during sleep but cant remember how we did it. Its like Bertrand Russell's famous proof of the existence of God: he knew he'd done it and threw his tobacco tin in the air to celebrate, yet tragically all he could recall later on was the image of throwing the tin. Unable to reconstruct the proof, its only legacy was this image and the feeling of certainty that he had solved something. An experience, perhaps, that we are all familiar with. Why this odd insistence, then, not only of incomplete, fragmentary bits of language but also the sense that they need to be completed or that they are important and concern us?

The answer requires us to extend the classical model of language. Despite its many vicissitudes over the twentieth century, linguistics has basically still remained faithful to Buhler's model of language as involving three functions: the referential, the emotive and the conative. Although many twists and nuances have been added - think, for example, of Jakobsons addition of the phatic, poetic and metalinguistic functions - the structure has really remained much the same. The referential treats the denoting and connoting aspects of language, how it relates to its objects; the emotive treats the speakers relation to their words (expressive of their attitude), and the conative treats the relation to the addressee (eg. questioning, ordering etc.) Now, this brings us to the crucial point: all of these perspectives on language explore the relation to the addressee but what they don't do is explore the experience of being addressed. And don't the experiences of the infant being spoken to and of the adult on the edge of sleep indicate exactly the contours of this function of language?

Being addressed is both something essential and something problematic for an infant. For two very simple reasons: firstly, the meaning of the adult's interpellation will initially be enigmatic and, crucially, the infant has no immediate defence against it. All other interactions with the adult can be the subject of some form of intervention. Intervening means showing ones subjectivity, and the most common form is to refuse what it seems the adult wants: the infant can refuse to eat, to drink, to go on the potty and so on. But it certainly cannot refuse to be addressed by the Other. Rather than seeing this as a trivial detail, we should not underestimate its importance as a central function of language and of the infants experience.

A parallel emerges here with the look, which is perhaps the only other form of presence which cannot be defended against. Infants can refuse to follow the direction of the adult's look, they can shut their eyes when they are supposed to be open, but they cannot block the fact of being looked at. And hence perhaps the reason children often have the fantasy that by closing their eyes they become invisible to others. The look and the fact of being addressed share this feature of being experiences imposed from the outside which concern the child directly, yet which cannot readily be defended against. And for that very reason, they can become invasive and threatening.

A further parallel might be drawn here. How does the child defend itself against the look of the Other? One solution, described by Lacan, involves the production of screens which function to distract the Other. Attention is drawn to some image or screen which the subject manipulates to keep the look way from themselves. There's a split, then, between the look of the Other and the screen offered up by the subject. Can't we also find a similar split at play in the field of sound? If the subject has to defend against the experience of being addressed, doesn't the production of sound - music, for example - have the same function as the screen in the field of vision, which could take the form, for example, of a painting? If we call the experience of being addressed by its Lacanian name - the voice - then we can posit a split between the voice and sound.

If the infant is initially unable to defend itself against the experience of being addressed through a refusal, what other possibilities are open to it? Perhaps there are more subtle, less evident forms of refusal at play here. One option might be for the child to act as if it were in fact being addressed not by the particular adult but by something else. Or, quite simply, to act as if one wasn't being addressed, a strategy well-known to children. Note that these options don't save the child from the fact of being addressed, but constitute forms of response. If someone refuses to hear, it means that they have heard very clearly what is expected of them. But there is also one more option, which I'll come back to later on by way of a conclusion.

Lets continue the parallel with the look a bit further. How is the invasive dimension of the look of the Other dealt with? Why not invoke here the many games of peek-a-boo played by mothers and children, games which involve a rhythm of presence and absence. One might argue that the function of these games is to link the look of the Other to a structure, a ritual of presnece and absence where the cardinal feature is that the look isnt always there. A game is being used to socialise and tame a menacing presence. And isnt there something similar going on with the experience of being addressed? Doesnt the crib speech reported by Weir and others have exactly this function? Rather than following her interpretation of crib speech as an early language learning exercise, we coud see it as a kind of early incorporation procedure which goes towards alleviating anxiety. By generating a dialogue themself, doesnt the child manage to modulate the experience of being addressed? They have now become the organisor.

Just as games like peek-a-boo link the intrusive experience of being looked at to a rhythm and structure, so crib speech does the same with the experience of being addressed. It modulates the addressee function, and dont so many other childrens games continue this same task? So many of these games, after all, involve one player taking on a role that is different from the other players - being it for example - while the other players resist being assigned this role. Indeed, the strategies for not being designated it, such as ever more complex verbal constructions to block the possibilites of a predetermined outcome, become intrinsic parts of the game, or even, games in themselves ( Opie and Opie 1959 ). One could also evoke the games of dare you familiar from childhood where no one really wins and it is rather a question of just doing what someone else says or, significantly, trying to avoid doing this. Finally, one could think of the many door-knocking games in which a child will be elected or volunteer to go to knock on a door and then run off. The thread that runs through all of these examples is the different relations the subject has to the experience of being addressed. And in these games, this experience is played with, modulated, taken up into a structure. Being addressed becomes a variable in all of these games of interpellation.

Once we isolate both the linguistic function of interpellation and the experience of being addressed as its corollary, a whole range of phenomena become clearer. The linguistic peculiarities found at the borders of sleep involve, we can now see, a separation of this function. One has the experience of words or phrases that interpellate one even if the meaning is opaque, and the occasional sense of respect that Isakower drew attention to is a sign of subjective involvement. As we prepare to sleep, this function can be modulated as in crib speech, while during sleep itself it doesnt disturb us too much until it emerges again around the time of waking, and perhaps even plays a part in waking. This supports Freuds quite radical view of sleep as not a passive occurence but an active process : we dont fall asleep but ourselves sleep.

This linguistic function of interpellation is very simply what Lacan means when he refers to the voice. It is the experience of being addressed, isolated from any particular sensory modality and semantic field. As he says in the seminar on Angoisse, The voice is the voice qua imperative, in so far as it calls for obedience or conviction, that it situates itself not with respect to music but with respect to the word ( Seminar of 5-6-63 ). Once this meaning is clear, we can see why it was that in one of his first lists of objects, Lacan cites the familiar pregenital and phallic objects and then adds delusion ( Seminar of 20-5-59 ). By delusion he means the interpellative aspect of auditory hallucination, where the subject has the sense that someone or something is calling them, adressing them, persecuting them and so on.

Hallucination is indeed a case where the addressee function emerges in its pure form. We should distinguish here between those hallucinations which are the subject and those which are directed the subject : in the former, a voice may describe their actions continuously ( now hes getting dressed, now hes going to work.. ), while in the latter there are generally obscenities or sexual accusations, usually involving perjorative terms associated with women ( bitch!, cunt!.. ). Hallucinations which occur outside psychosis share this same feature, and the most famous example is perhaps Freuds experience described in On Aphasia, where he sees and hears the words simultaneously Now its all up with you at a moment when his life was in danger.

Lacans thesis about hallucination brings out the distinction between the voice as a linguistic function and the register of sound and acoustic materiality. A verbal hallucination, Lacan was always keen to point out, is not limited to any particular sensorium : it doesnt have to take on an acoustic form but can privilege any sensory modality. What remains constant, however, is the experience of being addressed, regardless of the channels through which this operates. Language itself, Lacan argues, contains this potentiality, and when the subject has no recourse to master signifiers to make sense of and respond to situations where the support of meaning is crucial, it returns as this potentiality in the form of murmurs, whispering, buzzing, verbal commenatries and so on. This is what Lacan calls the continuous current of the signifier ( Seminar 3, p.294 ).

In other words, the whole world starts to speak, not just the humans in it. What normally blocks out this noise is the presence of certain significations which have been established unconsciously by key signifiers. Take away these meanings and the buzzing starts. The idea here is that it is foreclosure which unleashes this continuous current, the infinity of minor paths rather than the highway of centrally established meaning that Lacan investigates so carefully in this seminar. As he puts it in Seminar 5, the paternal metaphor establishes the signification of the desire of the mother in the unconscious, but when foreclosure is present, the desire of the mother cannot be symbolised and hence the speech of the Other is not inscribed in the unconscious : it just speaks to the subject all the time, not necessarily ascribed to another subject but simply the field of perception itself ( Seminar 5 p.480 ).

The Other speaks to the subject in a way homologous with the early, primitive speech predicated on demand and hence everything is sonorised. When the subject is called on to respond in certain charged situations and the ego has no respondent in the signifier, then the totality of the signifier responds ( Seminar 3, p.307 ). On a more local level, the so-called elementary phenomena found in psychosis are an example of the pure interpellative function : the subject is addressed, but doesnt know too much more about it. Interestingly, these moments do not always have a menacing aspect.

Dont these ideas give us a clue to situating certain clinical features of transference, especially as find in psychosis? It is a well-known fact that what the subject searches for here is often a witness, and hence Lacans encouragement to the clinician to function as a sort of secretary. It might seem a favourable sign if an appeal to an addressee is constituted, and Schrebers dedication of his book to the attention of scientists is frequently cited as an example. Without wishing to suggest any kind of continuity across clinical structures, cant we see a shadow of this in little Anthonys addressing commands and calls to his stuffed toy, as if he were creating an interlocutor and hence taming, as it were, the interpellation function? When a patient says to the analyst I have to make you into a hearer, isnt it to distribute what it means to be a hearer oneself and to defend against it? And isnt making oneself addressed a response to the fact of being addressed?

Separating hallucination from particular sensory modalities was a consequence of Lacans model of the signifying chain. Language is a structure which operates and has its effects at all levels of sensory perception. Hence it follows that if one of the properties of the signifying chain is the addressee function, this can return in any modality. It could emerge through silence, for example, or through vision. There are plenty of situations where a silence gives us the sense of being interpellated, . and so it is a question not of equating the voice and sound but rather of finding the effects of the voice in the field of sound. For example, the way that the subject punctuates their speech, their rhythm, their verbal style and so on, brings out their modes of incorporating the addressee function. Ä The way that speech is punctuated always involves an implicit placing and supposition of the listeners presence, and, more generally, of the supposition of being addressed. Hence one persons speech may be organised to blot out any possibility of being addressed or, on the contrary, may invite it. We could note how the use of connectives always indexes the presence of this aspect of the Other : when children start to use terms like and and but, it points to the supposed presence or indeed, intrusion of another speaker. Wasnt Peirce right when he suggested that thinking always takes the form of a dialogue, whether we know it or not?

If we turn to vision,we can find further examples of this function. A film like The Ring tells the story of a video tape that strangely interpellates its viewers, sending them a lethal message. The visual field here calls the subject and they desparately and futilely try to avoid being addressed by it. Similarly, in the phenomenon of voodoo death often discussed by medics and those interested in the so-called mind-body problem, the little bit of symbolic matter, be it a doll or a bit of hair, has catastrophic effects on the person who finds it. The object here, although presented visually, consists of a concentration of the addressee function : it is pure interpellation, a vector designating the finder. Beyond the visual, it calls the subject. And hence its terrifying effects.

This terror factor that so often accompanies the isolated functioning of the voice is perhaps one of the reasons why Lacan called it the sadomasochistic object. This must have surprised his audience, since he had been trying to persuade them for years that there is no such thing as sadomasochism ( Seminar of 26-3-69 ). Just when his message might have been received, the new category of the sadomasochistic object pops up, and Lacan described the voices function in sadism and in masochism. The aim is either to impose the voice on the other or to subtract it. And clinically, we often see that perverse scenarios actually involve a staging of some encounter where the subject makes himself addressed or addresses the other in a pure form. Doesnt this interpretation also echo the emphasis Lacan put on anxiety rather than sexual practice as the central variable in perversion?

The parameters of the invocatory drive thus become clearer. If the drives object is the voice, what is it that the invocatory drive aims at? On one level, to make oneself heard, but on a more fundamental level, to make oneself addressed. This is also something that we should be alert to particularly as it functions in sadim and masochism. And in science. When Mersenne succeeded in measuring the velocity of sound in air, it was through measuring the time it took for his voice to return to him as an echo over a known distnace. In other words, through creating a scenario in which he made himself addressed. When Borelli and Viviani came up with finer calculations, it was through the use of a cannons boom rather than a human voice. Although the subjective dimension became eclipsed, there is still the presence of the drive at the origin of these scientific studies.

To conclude, lets return to the question we posed earlier about the possibilites of response to the experience of being addressed. We described two of them : to act as if one were being addressed by someone or something else, or to pretend that one simply wasnt being addressed. But why not hypothesise a third option, one which would consist in the rejection of this linguistic function in its entirety : language minus the voice, minus the experience of being addressed. Isnt this a clinical picture we find in some so-called autistic states?

And wouldnt certain clinical consequences follow? On an immediate level, it would provide an explanation for what most people who work with these subjects know - that is, dont try to address them directly. And secondly, that words can be experienced potentially as invasive. This means that it isnt semantics that is to blame, and furthermore, that it isnt through semantics that one will be able to make any progress. If interpellation is to play a part, it is through its modulation rather than through its direct exercise.

Alison Elliot, Child Language, Cambridge, Cambridge University Press, 1981.
RIrene Deliege and John Sloboda, Musical Beginnings, Oxford University Press, 1996.
:Robert Fliess, Erogeneity and Libido, New York, IUP, 1956.
Sigmund Freud, The Ego and the Id (1923 ), Standard Edition, Vol 19, pp.12-66, London, Hogarth, 1961.
Sigmund Freud, On Aphasia ( 1891 ), New York, IUP, 1953.
Frank Heynick, Language and its Disturbances in Dreams, New York, John Wiley, 1993.
Otto Isakower, On the Exceptional Position of the Auditory Sphere, International Journal of Psychoanalysis, Vol 20, 1939, pp.340-48.
Stan Kuczaj, Crib Speech and Language Play, New York, Springer Verlag, 1983.
Jacques Lacan, Le Seminaire Livre 3, Les Psychoses ( 1955-6 ), ed J-A Miller, Paris, Seuil, 1981.
Jacques Lacan, Le Seminaire Livre 5, Les Formations de LInconscient ( 1957-8 ), ed J-A Miller, Paris, Seuil, 1998.
Jacques Lacan, Seminar 10, LAngoisse, 1962-3, Unpublished.
Jacques Lacan, Ecrits, Paris, Le Seuil, 1966.
Jacques Lacan, Seminar 16, Dun Autre a lautre, 1968-1989, Unpublished.
Iona and Peter Opie, The Lore and Language of Schoolchildren, Oxford University Press, 1959.
Theodor Reik, Psychoanalysis of the Unconscious Sense of Guilt, International Journal of Psychoanalysis, Vol 5, 1924, pp.439-50.
Colwyn Trevarthen, Conversations with a Two-Month Old, New Scientist, 62, pp.230-3, 1974.Ä Colwyn Trevarthen, Descriptive Analyses of Infant Communicative Behaviour, in H.R. Schaffer ed, Studies in Mother-Infant Interaction, London, Academic Press, 1977.
Ruth Weir, Language in the Crib, Mouton, The Hague, 1962.
Donald Winnicott, Transitional Objects and Transitional Phenomena ( 1953 ), in Playing and Reality, London, Tavistock, 1971, pp.1-30.

Darian Leader is a psychoanalyst and writer working in London.