The Subject and the Speaking Subject

This paper schematizes with the aid of Venn diagrams possible relationships between the subject (the self, the agent, the person) and the speaking subject (the speaker, the linguistic subject) and comments on the character and sources of plausibility of each conception of the relationship. Key words: Identity, Subjectivity, Person, Speaker, Linguistic subject, Constitution of the subject, Sujet, Sujet parlant, le hors langage

Althusser, L 1971 'Ideology and Ideological State Apparatuses', in his Lenin and Philosophy and Other Essays. London: New Left Books

Bakhtin, M 1968 Rabelais and his World. Cambridge, Mass: M I T Press

Bakhtin, M 1981 The Dialogic Imagination. Austin: UNiversity of Texas Press

Barthes, R 1982 'Writers, Intellectuals and Teachers' in S Sontag ed A Barthes Reader pp 378 - 403. London: Jonathan Cape

Bateson, G 1972 Steps to an Ecology of Mind. London: Paladin Books

Benveniste, E 1966 'De la subjectivite dans le langage' in his Problemes de linguistique generale, pp 258 - 266 Paris: Gallimard

Bhaskar, R 1979 The Possibility of Naturalism Brighton: Harvester Press

Bickerton, D 1981 Roots of Language Ann Arbor, Michigan: Karoma Books

Bullowa, M ed 1979 Before Speech Cambridge: Cambridge University Press

Coulter, J 1983 Rethinking Cognitive Theory London: Macmillan

Feldman H Goldin-Meadow S and Gleitman L 1978 'Beyond Herodotus: the Creation of Language by Linguistically Deprived Deaf Children' in A Lock edAction Gesture and Symbol pp 351 - 414 London: Academic Press

Figes, O 1996 A People�s Tragedy London: Jonathan Cape

Freire, P 1972 The Pedagogy of the OppressedHarmondsworth: Penguin

Freud, S 1963 Civilisation and Its Discontents London: Hogarth Press

Giddens A 1979 Central Problems of Social TheoryLondon: Macmillan

Grice H 1975 'Logic and Conversation' in P Cole and J Morgan eds Syntax and Semantics, volume 3, Speech Acts pp 41 - 57 New York: Academic Press

Habermas J 1972 'Towards a Theory of Communicative Competence' in H Dreitzel ed Recent Sociology No. 2 pp 115 - 148 London: Collier-Macmillan

Habermas J 1979 'What is Universal Pragmatics?' in hisCommunication and the Evolution of Society pp 1 - 68 London: Heinemann

Heidegger, M 1962 Being and Time London: SCM Press

Hollis, M 1977 Models of Man Cambridge: Cambridge University Press

Lacan, J 1967 'Function and Field of Speech in Psychoanalysis' translated in A Wilden The Language of the Self Baltimore: Johns Hopkins Press

Lewis, D 1969 Convention Princeton NJ: Princeton University Press

Marcuse, H 1969 Eros and Civilisation London: Sphere Books

Pateman, T 1975 Language Truth and Politics Newton Poppleford, Devon: Stroud and Pateman (second ed 1980)

Pateman, T 1981 'Communicating with Computer Programs' in Language and Communication vol 1, no. 1, pp 3 - 12 (and also on this website)

Pateman, T 1987 Language in Mind and Language in Society Oxford: oxford University Press (andin part on this website)

Sartre, J P 1948 Qu'est que la litterature? Paris: Gallimard

Sartre, J P 1956 Being and Nothingness London: Methuen

Therborn G 1980 The Ideology of Power and the Power of Ideology London: New Left Books

Trevarthen, C 1979 'Communication and Co-operation in early infancy: a description of primary intersubjectivity' in M Bullowa ed Before Speech pp 321 - 347 Cambridge: Cambridge University Press

Volosinov, V 1973 Marxism and the Philosophy of Language New York: Seminar Press

Winnicott D 1977 Playing and Reality London: Tavistock

Previously unpublished. This selectedworks website version published 2004

This essay had its origins in a pedagogic question: How might one approach the question of the relations between a person (the subject) and a speaker (the speaking subject) in the simplest possible way, in order to map the terrain and start a discussion? The answer I found was to make use of Venn diagrams, which are an easy way of illustrating relations of inclusion and exclusion. In the diagrams used below, 'S' stands forsubject; and SP for Speaking Subject. (See Endnote 1). Roughly speaking, I understand by subject what .is also called a person, a human individual, the human subject, a locus of subjectivity, that which experiences joy and suffering. It is something (essentially) embodied in space and time, finite and mortal. In contrast, speaking subjectis the linguistic subject, the symbolising subject, a locus of responsibility for truth, authority, and comprehensibility, as theorised notably in Habermas's universal pragmatics (Habermas 1976 and much else). It is at least arguable that the speaking subject is not essentialy embodied, finite and mortal - and this is one of the questions which is under discussion in this essay. By way of preliminary summary of the distinction one might say: human beings experience each other as subjects, and comrnurricate with each other as speaking subjects. The purpose of the essay is to explore what it means to think those subjectivities to be really different or actually no diffferent at all or, as a final possibility, partly the same and partly different.


Diagram 1 illustrates the supposition that subject and speaking subject (more generally, communicating or symbolising subject) are one and the same, occupying the same location in space and time. In time, one does not proceed the other nor survive it. In space, the speaking subject goes wherever the subject goes, and vice versa. A speaker just is (and must be) a person; and a person just is (must be) a speaker. Outside of language, there is no subject. If you like, il n' y a pas de hors langage - or, as Emile Benveniste said, "Est egoqui dit ego" (Benveniste 1966, p 26)

Genetically, to avoid any suggestion of a subject which is not a speaking subject, one has to think that the subject (as a person) develops in step with the speaking subject (the speaker), envisaged as put into place in social interaction. The subject as an individual person is thus entirely a matter of history - of culture, rather than nature � though rival theories try to offer us more than one way of describing that history. If we include an unconscious as part of the constituted subject then on this first view, we have to regard that unconscious as containing only symbolised material which has got there through repression. Repression, in turn, is envisaged as a two stage process: others repress me before I learn to repress myself.

This first view is really a modern way of expressing the position that a human being, rather than being a political animal, or a rational animal, is best thought of as a speaking animal � in other words, that human identity and individual identities are constituted through the possession of language or, more generally, conventionalised symbol systems. This position is strongly enough entrenched for some people to be as dismissive of the possibilities of teaching chimpanzees to talk as we all are when told of Lord Monboddo's eighteenth century theory that orangutans could talk, but refused to, in order to avoid being enslaved.

The position shown in Diagram 1 has proved overwhelmingly attractive to sociologists who, with whatever nuances, feel professionally obliged to insist on the social construction of subject identity through institutional and interactional shaping. To be guided by such a leading idea seems to make it impossible to attribute any active powers to the human mind as a generative and transformative mechanism. However thick the sociologist's verbal disguises, they can be peeled away to reveal Lockean doctrines of the mind as a tabula rasa and Pavlovian/Skinnerian theories of conditioning: in short, what Martin Hollis in Models of Man (Hollis 1977 ) called 'Plastic Man' theories. As an offender from the distant Dark Ages when we all believed in the latest ideas from New Left Books , I quote from an early work by Goran Therborn:

From what is known [ actually, from what is being assumed TP] about the ideological plasticity of human beings and their creative capacities [non-existent, it will soon be revealed TP] we should expect the given ideologies to be almost completely reproduced in societies whose internal conditions and relationships to the natural environment and to other societies remain exactly the same from one generation to the next ....A parental generation will always mould its children according to its own form of subjectivity ....the younger generation will face exactly the same affirmations and sanctions of the existing ideologies as the parental one (Therborn 1980, p 43 )

Such blithe social determinism is unable to accommodate to the achievements of contemporary cognitive science, since that has been paradigmatically an exploration of the ways in which the mind is not plastic, but rather structured and structuring in the way it goes beyond environmentally given information. Not only does the mind go beyond but, in some cases, it rigidly disregards what is given to it. The young child who obstinately insists that the plural of (English) sheepis sheeps despite overwhelming environmental evidence to the contrary, and even in the face of sanctions (endless correction from pedantic sociologist parents), is also making ridiculous all theories of the Therborn�determinist kind. This is more than inconvenient for social constructionists , and they have had to view cognitive science as espousing ideas which are wrong and, indeed, so wrong as to be pernicious. The philosophy of Wittgenstein has frequently been deployed in what has to be an all -out attack (see, for example, Coulter 1983 and, in contrast, my own cognitive - theoretic account of ideology in Pateman 1987 chapter 5: to be added to this website in 2005).

In my view. a good cognitivist theory of the human mind allows us to show and understand why all cultural structures � languages, ideologies and art � are liable to unintended endogenous changes, independently of changes in their social context. All things change(Heraclitus: Panta rei) because the human mind is a generative and transformative mechanism.The mind isitself an environment with its own ways of dealing with the information it is obliged to handle, and the results may be unwelcome to those who wish that it was easier to mould. Social systems often devote considerable energies as, for example, in organised resistance to language change to slowing or stopping processes which cannot be stopped and which cannot be stopped because the dynamics of endogenous cognitive change exceed human capacities for reflexive monitoring and control. In contemporary social theory, interest in the idea that there are limits to rational control is obviously a motif in New Right thinking, where it figures (among other ways) as an argument against all forms of planification. But the motif also figures in thinking on the left, and both Roy Bhaskar and Anthony Giddens have concerned themselves with limits to possible reflexivity. Bhaskar makes use of the idea of the necessity of an unreflected background, which he adapts from Michael Polanyi's work (see Bhaskar 1979) and Giddens' concern with reflexive monitoring and control appears significantly as early as his Centrol Problems in Social Theory (Giddens 1979).


In Diagram 2, the binary opposite supposition is made that subject and speaking subject occupy disjoint spaces and times as if they have nothing in common and are, by origin and destination, in some way strangers to each other. Far from being identical, persons (as subjects) and speakers (as communicating partners), are different orders of reality. Is this a bizarre supposition? In at least two cases, the supposition clearly isn't bizarre, though it may be false or unhelpful as a characterisation.

(1) A first case which can be thought through in terms of the separation of subject and speaking subject is provided by written texts. Here, a reader assumes or is led to suppose the existence of an Author of the Text, but need not equate this Author of the Text with some individual person or personality. Most of what drops through one's letterbox has no obvious psychologically real and individual author at all. Gas bills and junk mail aren't like that. They do not identify or define an individual psyche, and if there exists a subject in or behind the gas bill, it can only be a collective or institutional one. But collective subjects are psychologically, as well as ontologically, hard to conceptualise. They can only be conceptualised at all if subject and speaking subject are separated. (The mistake of thinking that behind it all there must be an individual � albeit a mistake to which we are all in some sense driven � provides Kafka with the grit out of which he constructs the nightmare world of The Trial)

Characters in a novel are also speaking subjects, but evidently not really existing psychological individuals. And the author of a novel can also be thought of as something implied by a text, or the reading of a text, and thought of as in some way responsible for the text without it seeming necessary to suppose an identity between the Author�of�the�Novel and an individual person(ality) who was born, given a name, and who � unlike the characters of fiction � will die. On his deathbed, Flaubert raged that Madame Bovary would outlive him, which was true, but she does not outlive Flaubert�as�author�of�Madame Bovary. A contemporary way of putting this is to say that the author we call 'Flaubert' constructs himself in the act of writingMadame Bovary. If you like, Il n' y a pas de Flaubert hors Madame Bovary(2)

Now consider a second kind of case. In interaction with a computer, one may be led to suppose computer programs to be speaking subjects, conforming to the kind of rules of conversational cooperation theorised (for example) by H. P. Grice (1975). But this is not to say that one regards such programs as real psychological subjects � persons in other words � even though one may believe that one day they may become just that. The hope has been raised in numerous books and anathematised in others. The point is that a chess-playing computer program can make a perfectly worthy opponent, and be the object of admiration and irritation, but without being in any sense a person. It is a speaking (symbolising] subject, displaying a high level of rationality, but even if we choose to call it "Gary Kasparov" it does not belong to the same order of reality as does the real Gary Kasparov. Infuriatingly, there is never even a grin that you could at least wish to wipe off its face.

I once compared human interaction and human � computer interaction at a time when one interacted with computers through primitive printouts spewed from cumbersome mechanical devices (Pateman 1981, on this website now as "Communicating with Computer programs"). In retrospect, it still seems to me that the centrally interesting question is the extent to which normal (Gricean, Habermasian) conversational presuppositions remain in force when human subjects interact with computer programs and, in contrast, to what extent they are suspended and allow, for example, for play in the Batesonian and Winnicottian sense: Something which does not mean what it would mean were it not play (Bateson 1972 Winnicott 1977 ). The ambivalent character of our relations to computer programs arises, at least partly, from the perception that they behave in important respects like speaking subjectswithout being in any way themselves subjects.

In sum, Diagram 2 may prove helpful in conceptualising some issues which have to do with the relations between subjects and speaking subjects. It also can provide a point of entry into the question of how subject and speaking subject ever get to link up together � a question which the theorisation of Diagram 1, in assuming their identity, effectively precluded.


Diagram 3 presents the speaking subject as part of the subject, entirely contained within the subject's space and time. They are interdependent but there is more to the subject than the speaking subject. It can, for example, possess a Nature which is not the deposit of Culture. And there are very good reasons for thinking that to be the case.

One could regard the speaking subject as emergent in time within the subject, which precedes it. A subject exists from birth, and exists always�already oriented towards the world of other subjects. The subject develops as a communicating (speaking) subject just insofar as it encounters another subject oriented towards it: without this there will be no linguistic development. Social interaction is a necessary condition of linguistic development. But from that it does not follow that linguistic development can only occur when there exists an always�already available conventionalised language. Work on creolisation and spontaneous signing by the deaf children of hearing parents shows any such claim to be empirically false: there can be language development in the absence of a ready�made language (Bickerton 1981; Feldman, Goldin�Meadow and Gleitman 1978 ). In other words, linguistic development does not have to be targetted on an existing language, and does not have to involve the acquisition (internalisation) of a target language. If somelinguistic development does not have to be targetted on an existing language, then it is possible that none is: even where there is a rich ready-made language in the child's environment, it may not be the case that the child's language development is adequately theorised by treating it as targetted. It may have its own internal dynamic, in which language states develop and succeed each other in a much more complex way than is envisaged by theorists of target�language development. (You do not learn to drive a car by targetting the behaviour of competent drivers. Most of what they are doing is unobservable or not easily observable. There are, I want to say, a succession of internally-generated driving-states which may beenabled by an instructor but not straightforwardly taught)

Remarkably, the very possibility of such actually�occurring situations as creolisation and spontaneous deaf�signing has often enough been denied on Wittgenstein�inspired theoretical grounds. In my Language in Mind and Language in Society,(Pateman 1987, especially chapter 6 now on this website as "Wittgenstein(ians) and Chomsky(ans)" ) I tried � and I hope you will forgive my lack of taste � to show how the facts of language development can be used to dispose of philosophical theories which tell us how things have to be only to be outsmarted by real human beings.

Human subjects can make themselves understood by means of invented signs, not shared by convention (in the sense of Lewis 1969). Likewise, even where there is a conventionalised language the evidence of developmental linguistics is that language is not so much 'internalised' (a word which is always indicative of a bad theory) as reinvented by every subject and changed in the process. Just because when you say English 'sheep' I also say 'sheep' does not mean that I have internalised 'sheep' . (That would be the fallacypost hoc ergo propter hoc ). The hand proof of that is that before I get to say 'sheep' I have, developmentally, replied to your 'sheep' with my 'sheeps'. I do not so muchinternalise the correct plural, as reinvent it as the least worst hypothesis to account for your odd persistence in the aberrant form 'sheep' as plural of 'sheep'. As for what our language ought to be, I have always, as it were, had my own ideas on the question, and children might be defined as those subjects who have their own ideas about everything. They suffer from overactive imaginations. Primary schooling is all about knocking ideas out of their heads, not putting ideas into them.

As evidence for the existence of a non�trivially specified prelinguistic subject, I give as an example here the work of the developmental psychologist Colwyn Trevarthen, who has done very careful studies of early interaction between infants and their mothers. These succeed in showing, for example, that the infant spontaneously coordinates with its mother's head bobbing activity. This activity can be charted through focusing on the movements of the tips of their noses. ( An example of his work can be found in Margaret Bullowa's book, Betore Speech (1979, pp 321-47)) . If an infant of a few weeks of age is confronted with a video playback of its own mother's previous head�bobbing communication, it becomes rapidly distressed, since the head�bobbing is no longer coordinated with and responsive to its own movements. The communicative link has been broken. By the tips of their noses shall ye know them.

In this perspective, we can go on to view the emergent speaking subject as the point of capture (Lacan's point de capiton by means of which the subject takes hold of the world at the same time as the world takes hold of the subject. In a critical perspective, for example. that of Althusser (1971) the speaking subject polices the subject; interpellated in a first instance from the exterior (Who me?), the subject is socially complete when it begins to police itself. In Diagram 3, the central location of the small circle SP could be read as metaphoric for the control centre (though I indicate below some weaknesses of this control centre perspective).

In other words, the speaking subject can be seen as the social side of a larger self which may include aspects which are non� or even anti�social. If civilisation splits individuals from themselves through the manner in which it implants itself in them, it is not then surprising that it has its discontents. Such is Freud's theory, and it may well be time for social theorists to re�read Freud'sCivilisation and Its Discontents, along with Marcuse'sEros and Civilisiation. Both give us a vision of human subjectivity as (essentially) both conflicted and generative of genuine but doomed attempts to resolve such conflicts. The weakness of what Lacan (1967) criticised as (American) Ego Psychology is that it does think that the centrally located SP of Diagram 4 can successfully function as a control centre over the S which environs and includes it. But this is to place the speaking subject (the ego) above and outside the subject, as what one might call a Kantian (transcendental) Subject. (Unfortunatety, Venn diagrams are strictly two�dimensional; one would need a holographic third dimension to represent SP transcendentally). A reading of Freud and Marcuse might lead us to conclude that the speaking subject never transcends its environing subjectivity. The speaking subject is always�already a natural subject, and as such neither omniscient or omnipotent. It has to shift as best it can with its own subjectivity. This notion leads us readily into discussion of Diagram 4.


Diagram 4 may look, from a logical point of view, as if it is trying to represent something logically impossible. For how could the whole ( the Subject ) be part of the part (the Speaking subject) ? Despite this, it does seem to me that theories exist which imply something like this as the relationship between the individual subject and speaking subject.

For example, it seems to me that if you think of the unconscious in a Lacanian fashion, then you are conceiving it � roughly speaking � as posterior in time to the speaking subject, repression operating only on what has already been symbolised. The unconscious has its origins in culture, not nature. But if , nonetheless, and I realise this is a debatable transition, you regard the unconscious as the site of some kind of personal truth, then you are led in effect towards regarding the unconscious as itself a subject living inside the speaking subject and interfering, as is the way of the unconscious, with the speaking subject's pursuit of its own goals. Freud theorized this interference as an everyday occurrence, and beginning with the study of slips of the tongue and other parapraxes eventually came to regard it as part of the condition of universal neurosis. The pretensions of the speaking subject to act as a control centre are constantly thwarted and subverted: the bored committee Chair comes into the room, sits down and declares the meeting closed. Everyone laughs, because nowadays we know that this is what is wished for. The Chair remains embarrassed, because his or her job is to play a social part, not express a personal desire. (There are other kinds of slip. I once attended a meeting, back in the days of the Shah, in support of the people of Iran; the opening speaker urged us repeatedly to support the people of Vietnam ( soutenir le peuple du Vietnam...). We knew immediately that there stood before us a party hack).

So Diagram 4 can be made sense of, in at least one way, after all. Can it be made sense of in any other ways, or by extension of the way just indicated? The diagrams I have provided to structure this essay are obviously an invitation to metaphorical as well as logical thinking (and the set theoretic logic of exclusion and inclusion is already rich with metaphoric potential). So the central location of S in Diagram 4 could be taken to represent the soul or self of an individual. The speaking subject which does the work of social interaction surrounds this soul or self but need not express it: that is a condition of alienation, which may be social as well as, or rather than, psychological. So the region SP in Diagram 4 could be seen as the region in which roles are ocupied and social functions discharged without these, as it were, touching or interacting with some core or. at least, other self: this is the predicament of the bored committee chair. The soul of man is a far country,

Diagram 4 represents that thought visually once the single Venn diagram is visualised as located within a field in which it is surrounded by other Venn diagrams representing other subjects and speaking subjects who communicate across the SP region whilst containing an inviolate core of selfhood. I am not saying that I endorse this familiar enough picture; I am only trying to redraw it within the framework of the schematism I am developing. It is an important question whether the ' inviolate core of selfhood ' is simply a humanist myth, which loses its power to comfort the moment we read our histories of the Holocaust, or whether something like this idea constantly reappears as a theoretical necessity even in theories (social determinisms, for instance) which say they will have nothing to do with it.


Finally, in Diagram 5, I illustrate the situation in which there is an intersection, an interlacing, between the subject and the speaking subject, an intersection marked out in the shaded area. On this conception of the situation, subject and speaking subject occupy both common space and time and separate space and time. So there can be a subject which is outside the circuit of ordinary communication and a speaking subject which is no more than an idle chatterer, glib reproducer of ideological cliche, or lost soul floundering in empty rationalisation � such images are recurrent in Heidegger's Being and Time and Sartre's Being and Nothingness . That is to say, Diagram 5 shows how it is that one can theorise a speaking subject which, though oriented towards the world of others, does not speak or accent speech for itself, but is lost in idle discourse. This is the topic of chapter 3 in my Language Truth and Politics (1975), where I tried to show the social context and debilitating consequences of everyday inauthenticity. And a quarter of a century on I still find it unavoidable to have recourse to such essentialist language, or at least to the concepts it seeks to express. We have read Bakhtin, for example, more recently than Sartre, and his thought seems more in accord with the themes of postmodernism and poststructuralism. But in his notion of the (individual) accenting and reaccenting old speech there effectively reappears the notion of individual human freedom and creativity, conceived as the ability to remake and remark what has always�already been said. Like the child who has to reinvent in order to understand, the adult comes alive as a subject through reaccenting what has already been thought.

For Bakhtin, the paradigm of such reaccenting is irony, and we find the subject smiling, through the cracks of ironic reaccenting. The ironic mode ceases to be a historically conditioned literary trope, and becomes human subjectivity itself. If you ask where is the space of human freedom in Bakhtin's work, it is to be found in the capacity for irony, the capacity for taking the already�said, the preconstructed, the ready�to�hand, and reusing it with a gesture, a grimace, a grunt, a grain of the voice � indeed, anything which makes it one's own at the same time that it is always�already someone else's. This position is most readily lisible inMarxism and the Philosophy of Language (V'olosinov 1973) but can also be read out of the work, on carnival (Bakhtin 1968) as well as the works of literary theory (Bakhtin 1981).

But there are clearly instances of what one might call non�ironic subjectivity. Consider, for example, the role of personal doubt (the crisis of faith) in many versions of Protestantism. This is valued precisely because it allows the doubter, if the crisis is overcome, to move from a position in which he or she is merely the inheritor of a social (familial) faith, who has done nothing (as it were) to deserve it, to a position where the belief becomes a personal one, expressed and reaccented in a personal voice.

Every emancipatory practice such as psychoanalysis and Paolo Freire's pedagogy of the oppressed can be understood as an attempt to enlarge the area of intersection, the common space between subject and speaking subject. It is an attempt to enlarge the space of that full word (Lacan's parole pleine ) where I and Thou may succeed in cornmunicating each with the other free of domination (Haberrnas 1972 and much subsequently) This domination may be either by the past of an individual history, with which psychoanalysis as an emancipatory practice concerns itself, or domination by the present of oppression and exploitation, with which Paolo Freire was concerned.

In his book on the Russian Revolution, .A People's Tragedy, which I happen to have as bedside reading, Orlando Figes (1996) describes an interesting precursor of Freire's pedagogy of the oppressed:

Dora Elkina recalls how she came to write the first Soviet primer. In 1919 she was sent to the Southern Front to teach the soldiers how to read and write. Having got hold of some old school textbooks, she wrote out the first sentence on the blackboard: 'Masha ate the kasha'. But the soldiers only laughed and heckled. Close to tears, she lit upon the idea of turning the lesson into a political discussion and explained to the soldiers why they could not go home to their Mashas, and why the country was short of kasha. Then she turned to the blackboard and wrote: 'We are not slaves, slaves we are not!'. It was a great success among the soldiers, for whom the idea of not being slaves had always been a vital aspect of the revolution. This simple expression of human dignity later became famous as the opening line of her reading book. It was used in primary schools throughout the 1920s and 1930s. For millions of Russians, many of them still alive, it was the first sentence they ever learned to read' (p 601)

It may seem a big step, but I would argue that the task facing the would�be creative writer is generally to develop from the stage of writing poems or stories which tell us blithely that 'Masha ate the kasha' towards writing, with both difficulty and relief, that 'We are not slaves, slaves we are not' . This is not simply or at all a question of commitment or engagement in what we take to be the Sartrean sense (Sartre 1948 ). It is to say more generally that the task of the writer is to find a voice, in which his or her subjectivity is engaged. To find one's voice as a writer just is to succeed in writing a full word, a word expressive of a human subjectivity liberated from both personal repression and ideological distortion. To find one's voice is to be able to call a spade a spade and Rolet a rascal: to narne and to express. Of course, the passage of tirne will show one's voice (if one is a painter, one's style: if a philosopher, one's vision) to be socially conditioned and expressive of the currents of one's time. How could it be otherwise? We are subjects in time and space, not transcendental egos. But as in psychoanalysis we seek, asymptotically, to free ourselves from the tyrannv of our past, so in writing, we must seek, asymptotically, to free ourselves from the tyranny of our times. This is why the writer is often contrasted (see, for example, Barthes 1982) (favourably) with the academic, the Prof., whose job it is to enrol under already�unfurled banners (Marx, Freud, Derrida, Baudrillard) and sustain their cause rather than subvert them. The job of the academic is always to be a conservative, and academicism and the Academy are not incidental aberrations but the heart of the role. The academic is like the linguistic prescriptivist who is always trying to turn back the tides of linguistic change. Academic radicalism is a contradiction in terms.


I said at the outset that I had pedagogic aims in mind in using Venn diagrams to introduce ideas about the subject and the speaking subject. But the diagrams leave out a lot, though they could be made to include more. Undoubtedly, as things to think with, these diagrams are too monologising. They limit us to isolated subjectivities. They do not represent the real presence of other subjects and speaking subjects or the omnipresence of the past sedimented as cultures and traditions. In a more advanced class with a less simple�minded teacher, one would need to inscribe the circles into a two dimensional field of other circles, representing other people, and into a three dimensional (holographic) field in which the past (and anticipated futures) could be figured. In this more advanced class, we might � if our powers of reflexive monitoring permit � try to reflect on the role of spatial metaphors in social theorising, for it is with such metaphors that we have worked. That discussion might generate a companion paper to this one which tried its hand at a time�based approach to the study of the subject within social theory (clock time, lived time, historical time, time's arrow, reversible time, temporal discontinuity ...)


1 'SP' (standing for Sujet parlant) is the trace of the fact that the original of this essay was a short talk given in French and using a piece of chalk at the Colloquium on Mikhail Bakhtin held at the Centre International Culturel, Cerisy la Salle, in July 1995. My daughter, Isabella Pateman, subsequently prepared the Venn diagrams on my computer at a time when this was no mean feat - at any rate, one quite beyond me.