White Mythologies
t If so-called ' so-called poststructuralism ' is the product of a single historical moment, then that moment is probably not May 1968 but rather the Algerian War of Independence  no doubt itself both a symptom and a product.
In this respect it is significant that Sartre, Althusser, Derrida and Lyotard, among others, were all either born in Algeria or personally involved with the events of the war.
But let us begin instead with Hlne Cixous's remarkable account of what it was like to grow up as an Algerian French Jewish girl at that time:
I learned everything from this first spectacle: I saw how the white (French), superior, plutocratic, civilized world founded its power on the repression of populations who had suddenly become ' invisible ', like proletarians, immigrant workers, minorities who are not the right ' colour '.
Women.
Invisible as humans.
But, of course, perceived as tools  dirty, stupid, lazy, underhanded, etc.
Thanks to some annihilating dialectical magic.
I saw that the great, noble, ' advanced ' countries established themselves by expelling what was' strange '; excluding it but not dismissing it; enslaving it.
A commonplace gesture of History: there have to be two races  the masters and the slaves.
Cixous has been criticized for lacking a politics and a theory of the social.
According to some criteria perhaps, but if so they would have to exclude from ' the political, considerations such as those described here.
Which is precisely the point: if there is a politics to what has become known as poststructuralism, then it is articulated in this passage which unnervingly weaves capitalist economic exploitation, racism, colonialism, sexism, together with, perhaps unexpectedly, ' History ' and the structure of the Hegelian dialectic.
A lot has been said already in the English-speaking world about poststructuralism and politics, much of it in the accusatory mode voiced from the opposing class-based verities of ' tradition ' or ' History '.
Such apparently secure grounds of objection amount to two narratives; their intriguing similarity brings out the extent to which poststructuralism challenges not just the politics and institutions of the right but also the politics and theoretical systems of the left.
Disturbing conventional assumptions about what constitutes' the political ', poststructuralism is correspondingly difficult to place itself.
In the passage just cited for example it is striking that Cixous includes the Hegelian dialectic in the forms of political oppression which she describes.
It is not a question of showing that such an allegation misinterprets or simplifies Hegel's texts.
Of course it does.
The problem involves rather the ways in which Hegel has been read, absorbed and adapted.
Nor is it just the Hegelian dialectic as such: Cixous includes' History ', and by implication therefore Marxism as well.
This can not simply be dismissed as another New Right invocation of the Gulag, for Cixous is arguing something much more specific: that Marxism, insofar as it inherits the system of the Hegelian dialectic, is also implicated in the link between the structures of knowledge and the forms of oppression of the last two hundred years: a phenomenon that has become known as Eurocentrism.
To this extent, Marxism's universalizing narrative of the unfolding of a rational system of world history is simply a negative form of the history of European imperialism: it was Hegel, after all, who declared that ' Africa has no history ', and it was Marx who, though critical of British imperialism, concluded that the British colonization of India was ultimately for the best because it brought India into the evolutionary narrative of Western history, thus creating the conditions for future class struggle there.
Such an arrogant and arrogating narrative means that the story of ' world history ' not only involves what Fredric Jameson describes as the wresting of freedom from the realm of necessity but always also the creation, subjection and final appropriation of Europe's ' others'.
This is why ' History ', which for Marxism promises liberation, for Cixous also entails another forgotten story of oppression:
Already I know all about the ' reality ' that supports History's progress: everything throughout the centuries depends on the distinction between the Selfsame, the ownself... and that which limits it: so now what menaces my-own-good... is the ' other '.
What is the ' Other '?
If it is truly the ' other ', there is nothing to say; it can not be theorized.
The ' other ' escapes me.
It is elsewhere, outside: absolutely other.
It doesn't settle down.
But in History, of course, what is called, other' is an alterity that does settle down, that falls into the dialectical circle.
It is the other in a hierarchically organized relationship in which the same is what rules, names, defines, and assigns' its' other.
With the dreadful simplicity that orders the movement Hegel erected as a system, society trots along before my eyes reproducing to perfection the mechanism of the death struggle: the reduction of a ' person, to a ' nobody ' to the position of ' other '  the inexorable plot of racism.
There has to be some ' other '  no master without a slave, no economico-political power without exploitation, no dominant class without cattle under the yoke, no ' Frenchmen ' without wogs, no Nazis without Jews, no property without exclusion  an exclusion that has its limits and is part of the dialectic.
(70C1)
Not that Hegel himself is responsible.
Rather the problem, Cixous argues, is that unfortunately Hegel wasn't inventing things.
The entire Hegelian machinery simply lays down the operation of a system already in place, already operating in everyday life.
Politics and knowledge have worked according to the same Hegelian dialectic, with its' phallo-logocentric Aufhebung '  whether it be Marxism's History, Europe's colonial annexations and accompanying racism or orientalist scholarship, or, in a typical conflation of patriarchy and colonialism, Freud's characterization of femininity as the dark unexplored continent (' within his economy, she is the strangeness he likes to appropriate' [ 68 ]).
For even Freud, according to Cixous, has not helped in any project to separate history from the history of appropriation or that of phallocentrism.
The patriarchal structures of psychoanalytic theory have often been defended on the grounds that they only describe the current customs of a patriarchal society.
But this does not alter the fact that psychoanalysis therefore repeats the same masculine ' Empire of the Selfsame ', and that as soon as such descriptions become institutionalized  as a structure of knowledge, or as psychoanalytic practice  then they become agents of the system they describe.
The point is to change it.
But why this emphasis on Hegel?
The problem of the Hegelian model, particularly of a historicism which presupposes a governing structure of self-realization in all historical process, is by no means confined to post-war French Marxism, but the dominance of Hegelian Marxism from the thirties to the fifties does explain the particular context for the French poststructuralist assault.
Here it is not a question of suggesting that Hegel is somehow answerable for the excesses of capitalism or even socialism in the past two hundred years: rather what is at stake is the argument that the dominant force of opposition to capitalism, Marxism, as a body of knowledge itself remains complicit with, and even extends, the system to which it is opposed.
Hegel articulates a philosophical structure of the appropriation of the other as a form of knowledge which uncannily simulates the project of nineteenth-century imperialism; the construction of knowledges which all operate through forms of expropriation and incorporation of the other mimics it a conceptual level the geographical and economic absorption of the non-European world by the West.
Marxism's standing Hegel on his head may have reversed his idealism, but it did not change the mode of operation of a conceptual system which remains collusively Eurocentric.
It is thus entirely appropriate that Hegelian Marxism has become generally known as' Western Marxism '.
As Cixous suggests, the mode of knowledge as a politics of arrogation pivots at a theoretical level on the dialectic of the same and the other.
Such knowledge is always centred in a self even though it is outward looking, searching for power and control of what is other to it.
Anthropology has always provided the clearest symptomatic instance, as was foreseen by Rousseau from the outset.
History, with a capital H, similarly can not tolerate otherness or leave it outside its economy of inclusion.
The appropriation of the other as a form of knowledge within a totalizing system can thus be set alongside the history (if not the project) of European imperialism, and the constitution of the other as' other ' alongside racism and sexism.
The reaction against this structure has produced forms of politics that do not fit into traditional political categories.
Here the problem rests on the fact that for orthodox Marxism there can be only one ' other ', that of the working class, into which all other oppressed groups, so-called ' minorities', must in the last instance be subsumed.
Such a position is by no means confined to a Marxism of a Stalinist past.
In Making History, published in 1987, for example, Alex Callinicos argues that the so-called poststructuralist critique of the category of the subject can be avoided by shifting the subject out of the problematic realm of consciousness into a theory of human agency.
This provides something closer to historical Marxism, although it does mean that he quickly becomes involved in assumptions about rationality and intentionality, and has to propose a ' principle of Humanity ', that is a common human nature, to hold them all together.
Perhaps not unexpectedly, the ' Principle of Humanity ' also turns out to involve the assumption that class is the primary form of collective agency, because, we are told, it is more fundamental than any other interests or forms of social power.
Callinicos writes:
Feminists and black nationalists [ sic ] often complain that the concepts of Marxist class theory are ' gender-blind ' and ' race-blind '.
This is indeed true.
Agents' class position derives from their place in production relations, not their gender or supposed race.
But of itself this does not provide grounds for rejecting Marxism, since its chief theoretical claim is precisely to explain power-relations and forms of conflict such as those denoted by the terms' nation ', ' gender ' and ' race ' in terms of the forces and relations of production.
The mere existence of national, sexual and racial oppression does not refute historical materialism, but rather constitutes its explanandum.
The only interesting question is whether or not Marxism can actually explain these phenomena.
The only interesting question?
So as long as gender and race can be satisfactorily subordinated to class then Marxism does not need refuting, and history can be reasserted as the single narrative of the Third International.
Conversely, as Callinicos implicitly recognizes, the problem with contemporary politics for the left is that the dialectic of class depends on a historicist History and vice versa; any failure of the former necessarily also involves a waning of the latter.
Marxism's inability to deal with the political interventions of other oppositional groups has meant that its History can no longer claim to subsume all processes of change.
The straightforward oppositional structure of capital and class does not necessarily work any more: if we think in terms of Hegel's master/slave dialectic, then rather than the working class being the obvious universal subject-victim, many others are also oppressed: particularly women, black people, and all other so-called ethnic and minority groups.
Any single individual may belong to several of these, but the forms of oppression, as of resistance or change, may not only overlap but may also differ or even conflict.
As soon as there is no longer a single master and no single slave, then the classic Hegelian reversal model on which Marxism depends and on which it bases its theory of revolution (literally, an overturning) is no longer adequate.
In fact it is arguable whether such dualistic conditions ever existed anyway: marginal groups which could not be assimilated into the category of the working class were merely relegated by Marx to the Lumpenproletariat.
Even the formulation of a dualistic class division, Laclau and Mouffe have argued persuasively, is itself nothing less than a nostalgic attempt to recreate for the nineteenth century the imagined simplicity of the conditions of the aristocracy/bourgeoisie conflict of the French Revolution which had originally inspired Hegel.
A similarly straightforward opposition also provides much of the attraction which has fuelled the recent growth of interest in the historical analysis of colonialism  in which you apparently have the simple binary of master and slave, colonizer and colonized.
With colonialism it's easy to tell the good guys from the bad guys, which makes it tempting to substitute the colonized for the lost working class.
Already in 1957 Roland Barthes was claiming that ' today it is the colonized peoples who assume the full ethical and political condition described by Marx as being that of the proletariat '.
But politics today are much more complex, much more difficult to disentangle.
The dialectical structure of oppositional politics no longer works for the micro-politics of the post-war period in the West.
This is the context of Foucault's critique of what he calls the sovereign model of power, of the idea that power has a single source in a master, king, or class  and can thus easily be reversed.
This shift from a conflictual dyadic political structure is not simply a question of historical change, of the recent appearance of ' minorities': after all the slave was already constituted simultaneously according to different groups (for example male or female), the Lumpenproletariat always had to be excluded.
The problem begins at a conceptual level with the initial division between master and slave as such, as if relations of power work according to the binary opposition of Hegel's fight to the death between two individuals.
This structure is not, as might at first be imagined, derived from a fantasy of power relations modelled on a medieval joust but from the phenomenological account of the constitution of knowledge that works according to the structure of a subject perceiving an object, a same/other dialectic in which the other is first constituted by the same through its negation as other before being incorporated within it.
No possibility of dialogue or exchange here.
As Cixous argues, nor can there be any place in this schema for the other as other, unless it becomes, like God, an absolute other, literally unknowable.
The difficulties which arise from this structure are familiar from the debates in feminism, where, woman, seems to be offered an alternative of either being the ' other ' as constituted by man, that is, conforming to the stereotypes of patriarchy, or, if she is to avoid this, of being an absolute ' other ' outside knowledge, necessarily confined to inarticulate expressions of mysticism or jouissance.
The only way to side-step these alternatives seems to be to reject the other altogether and become the same, that is, equal to men  but then with no difference from them.
Exactly the same double bind is encountered in any theorization of racial difference.
In his influential Le Mme et l'autre Vincent Descombes has described the entire history of twentieth-century philosophy in France as a succession of moves which attempt to get out of this Hegelian dialectic: the recent phenomenon of poststructuralism is part of a long philosophical story and distinguished only by what appears to be a certain success, or at least an avoidance of failure to the extent that it has at least managed to keep the game with Hegel in play.
The real difficulty has always been to find an alternative to the Hegelian dialectic  difficult because strictly speaking it is impossible, insofar as the operation of the dialectic already includes its negation.
You can not get out of Hegel by simply contradicting him, any more than you can get out of those other Hegelian systems, Marxism and psychoanalysis, by simply opposing them: for in both your opposition is likewise always recuperable, as the workings of ideology or psychic resistance.
Nor can you get away from Hegel by simply removing him, like the excision of Trotsky from the side of Lenin in certain official Soviet photographs.
This is the lesson of Althusser.
Althusser's historical interest derives from the fact that he represents the only orthodox Marxist theorist who has tried to get out of Hegel while remaining a Marxist  though for many Marxists he did sacrifice Marxism in the process, which only suggests how closely Marxism and Hegelianism are intertwined.
Althusser's theoretical interest, on the other hand, is that he demonstrates the impossibility of any attempt simply to exclude, excise or extirpate Hegel.
Other strategies are required.
II HISTORICISM AND IMPERIALISM
Metaphysics  the white mythology which reassembles and reflects the culture of the West: the white man takes his own mythology, Indo-European mythology, his own logos, that is, the mythos of his idiom, for the universal form of that he must still wish to call Reason.
Jacques Derrida
If poststructuralism has involved an attempted disruption or reworking of Hegelianism through the detection of its own fissures, it is not by any means unique in such an enterprise.
For, as Foucault argued in 1978, the work of the Frankfurt School could also be regarded as a reassessment of Hegelianism and the metaphysics of dialectical thought.
However, it is too simplistic a reaction to suggest that French poststructuralism can therefore be invalidated by judging it against the claims of a comparable endeavour in Germany, a procedure which can only operate by turning the former into a failed version of the latter, which obviously leaves open the possibility of exactly the reverse argument being made.
Though the two may have isolated similar problems, the political and intellectual context of their work was by no means the same.
The key Frankfurt School text in this regard is obviously Horkheimer and Adorno's Dialectic of Enlightenment of 1944.
The date, and the exiled place of composition of its authors, suggests tellingly that the situation with which it attempts to deal is the phenomenon of fascism which seemed to have stopped in its tracks the long march of the progress of reason, and its liberating enlightenment ideals, of which Marxism was the fullest political development.
Horkheimer and Adorno therefore pose the question: how has the dialectic deviated into fascism?
Why has History gone wrong?
Their answer, briefly, was that reason had always contained a measure of irrationality, which, despite its best intentions, had led to its involvement with tyranny and domination: ' Enlightenment is totalitarian '.
The very powers of rationality which enabled modern man to free himself from nature and control it had also become an instrumental device to dominate him.
If nature had been modelled by man into productive commodities, man's own subjectivity had also become reified into a self-identical instrument; man had become an empty and passive consumer.
The project, therefore, was to return to the enlightenment in the wake of fascism, to excise the forms of instrumental rationality that had produced this self-defeating and, ironically, irrational dialectical structure of domination, and to redefine reason and the forms of identity thinking that had defined the individual simply as an indistinguishable element in the collective.
In this way the autonomy and spontaneity of the individual subject that had been the original goal of enlightenment might be retrieved.
If for the Frankfurt School the problem to be dealt with was the relation of the phenomena of fascism, and particularly Auschwitz, to the ideals of the enlightenment and the progress of reason, for the French poststructuralists the historical perspective was similarly long.
But it comprised, rather, a history of the West in which fascism was itself merely a symptom, and included not only the history of European imperialism but also the defeats of the European colonial powers by Japan in World War II, the subsequent French (and American) defeat in South-East Asia, the war in Algeria, as well as the many other colonial wars of national liberation.
From this point of view the French have never regarded fascism as an aberration, concurring rather with Csaire and Fanon that it can be explained quite simply as European colonialism brought home to Europe by a country that had been deprived of its overseas empire after World War I. French poststructuralism, therefore, involves a critique of reason as a system of domination comparable to that of the Frankfurt School, but rather than setting up the possibility of a purged reason operating in an unblocked, ideal speech situation as a defence against tyranny and coercion in the manner of a Habermas, it reanalyses the operations of reason as such.
Here the focus is placed not so much upon the continued presence of irrationality, for irrationality after all is simply reason's own excluded but necessary negative other, but rather on the possibility of other logics being imbricated within reason which might serve to undo its own tendency to domination.
Here we have a major difference from the historical pessimism of Adorno's negative dialectics which in certain respects poststructuralism might appear to resemble.
Another such project was initiated by Adorno's contemporary Jean-Paul Sartre, whose attempt to define a new form of Hegelian Marxism via a reworked philosophy of consciousness in many ways more closely resembled that of the Frankfurt School.
In both cases revisionary work of this kind was embodied in a historical and political analysis of the relation of individual consciousness to society, from within the aegis of an Hegelian Marxism in some respects still impossibly bound to its own enlightenment heritage.
This meant that Adorno in particular tended to project science as something exterior and exclusively instrumental; he reacted against it, as well as' objective ' reason generally, by trying to retrieve the individual subject as the means to salvation.
In France, however, there also existed a very different tradition, that of the history of the sciences, a tradition in which Foucault has placed himself.
Having been overshadowed since the war by Sartrean Marxism, in the crises of the 1960s it emerged as the more influential of the two.
Foucault traces its history back to Husserl's Cartesian Meditations (1929) and the Crisis of the European Sciences (1936)  in which Husserl ' posed the question of the relations between the ' Western ' project of a universal deployment of reason, [ and ] the positivity of the sciences and the radicality of philosophy'.
This enabled the establishment of a critical position in relation to science which for Adorno remained so elusive.
In larger terms, however, the questions posed in the French tradition were comparable to those of the Frankfurt School, particularly the interrogation of rationality in its claims to universality.
As Foucault puts it:
In the history of the sciences in France, as in German critical theory, it is a matter at bottom of examining a reason, the autonomy of whose structures carries with it a history of dogmatism and despotism  a reason, consequently, which can only have an effect of emancipation on condition that it manages to liberate itself from itself.
(54)
The final emancipatory gesture of enlightenment thought would thus be its own liberation from itself, so that it is no longer recognizable as reason.
But what was it that brought about the return of the question of the enlightenment to contemporary philosophical enquiry?
Foucault identifies three reasons: first of all the ever-increasing importance of technology, secondly the place of rationalism in the optimism attached to the notion of ' revolution '  as well as in the despotism that so often followed its realization  and thirdly:
the movement which, at the close of the colonial era, led it to be asked of the West what entitles its culture, its science, its social organization, and finally its rationality itself, to be able to claim universal validity: was this not a mirage associated with economic domination and political hegemony?
Two centuries later, the Enlightenment returns: but not at all as a way for the West to take cognizance of its present possibilities and of the liberties to which it can have access, but as a way of interrogating it on its limits and on the powers which it has abused.
Reason as despotic enlightenment.
(54)
Foucault's account is particularly useful insofar as it gives a good indication of the characteristic French, as opposed to German, emphasis on the relation of Marxism to enlightenment rationality and the questioning of enlightenment claims to the universality of its values.
The first element, the role of science, is common to both, though they approach the question from opposite perspectives.
The second, the role of enlightenment thinking in the subsequent history of European despotism, is the particular focus of interest for the German critical theorists, most memorably and most forcibly articulated in Benjamin's ' Theses on the Philosophy of History '.
It is the third element, however, which represents the special interest of the French, whether of the Sartrean or Foucauldian tradition: the relation of the enlightenment, its grand projects and universal truth-claims, to the history of European colonialism.
This need not necessarily involve a direct analysis of the effects of colonialism as such, but can also consist of a relentless anatomization of the collusive forms of European knowledge.
For Foucault this has comprised a vigorous critique of historicism, including Marxist historicism, and its relation to the operations of knowledge and power.
It is from this perspective that it becomes possible to understand the basis of the distrust of totalizing systems of knowledge which depend upon theory and concepts, so characteristic of Foucault or Lyotard, both of whom have been predominantly concerned with the attempt to isolate and foreground singularity as opposed to universality.
This quest for the singular, the contingent event which by definition refuses all conceptualization, can clearly be related to the project of constructing a form of knowledge that respects the other without absorbing it into the same.
It is in the work of Edward Said that we can find the problematic of historicist forms of knowledge linked most forcibly to the question of European imperialism.
He writes:
So far as Orientalism in particular and European knowledge of other societies in general have been concerned, historicism meant that one human history uniting humanity either culminated in or was observed from the vantage point of Europe, or the West...
What... has never taken place is an epistemological critique at the most fundamental level of the connection between the development of a historicism which has expanded and developed enough to include antithetical attitudes such as ideologies of Western imperialism and critiques of imperialism on the one hand, and on the other, the actual practice of imperialism by which the accumulation of territories and population, the control of economies, and the incorporation and homogenisation of histories are maintained.
This was the difficult project of his own book, Orientalism (1978), which we shall later be examining in detail.
For the moment let us focus on Said's subsequent point that if Orientalism and anthropology derive from historicism, this is by no means a thing of the past: of more recent sciences, Said singles out in particular that of world history as practised by Braudel, Wallerstein, Anderson and Wolf, which he contends is still derived from the enterprise of Orientalism and its colluding companion anthropology, and which has refused to encounter and to interrogate its own relationship as a discipline to European imperialism.
For Said, the problem amounts simply to historicism and, the universalising and self-validating that has been endemic to it':
the theories of accumulation on a world scale, or the capitalist world state, or lineages of absolutism depend (a) on the same displaced percipient and historicist observer who had been an Orientalist or colonial traveller three generations ago; (b) they depend also on a homogenising and incorporating world historical scheme that assimilated non-synchronous developments, histories, cultures, and peoples to it; and (c) they block and keep down latent epistemological critiques of the institutional, cultural and disciplinary instruments linking the incorporative practice of world history with partial knowledges like Orientalism on the one hand, and on the other, with continued ' Western ' hegemony of the non-European, peripheral world.
(22)
A new type of knowledge, Said contends, must be produced that can analyse plural objects as such rather than offering forms of integrated understanding that simply comprehend them within totalizing schemas.
Already across a wide range of different activities he points to advances in the process of ' breaking up, dissolving and methodologically as well as critically reconceiving the unitary field ruled hitherto by Orientalism, historicism, and what could be called essentialist universalism '.
In this last phrase, Said thus links his critique of Orientalism to other critiques, such as those of racism or of patriarchy.
The more difficult question remains of what form this new kind of knowledge can take.
Here we return to the theoretical problem of how the other can be articulated as such.
How can we represent other cultures? asks Said, as Lvi-Strauss had done before him.
His own dismissal of deconstruction as a merely textual practice means that he is himself at a loss when faced with the complex conceptual dialectics of the same and the other.
As will be demonstrated with respect to Orientalism itself, Said can not get out of the Hegelian problematic that he articulates, and indeed tends himself to repeat the very processes that he criticizes.
His advocacy of an analytic pluralism in itself does not solve, or even address, the conceptual problems.
Nevertheless, Said's comments suggest the wider significance of his project.
The demise of an orthodox Marxism may have left theory with a sense that everything is now in flux, that the old verities have gone, but it has also involved the important realization, articulated so forcibly by writers such as Foucault or Said, of the deep articulation of knowledge with power.
The politics of poststructuralism forces the recognition that all knowledge may be variously contaminated, implicated in its very formal or ' objective, structures.
This means that in particular colonial discourse analysis is not merely a marginal adjunct to more mainstream studies, a specialized activity only for minorities or for historians of imperialism and colonialism, but itself forms the point of questioning of Western knowledge's categories and assumptions.
In the same way, Fanon suggests that at the political level the so-called ' Third World ' constitutes the disruptive term for the European political dialectic of capitalism and socialism.
Everyone feels the need nowadays to qualify the term ' Third World ', stating quite correctly that it should not be taken to imply a homogeneous entity.
The inadequateness of the term, however, insofar as it offers a univocal description of an extremely heterogeneous section of the world, also means that a suitable alternative general category can not by definition be produced.
In this situation, abject apologies in some respects remain complicit with the patronizing attitudes from which they attempt to disassociate themselves.
For the ' Third World ' was invented in the context of the 1955 Bandung Conference, on the model of the French Revolution's ' Third Estate ', and incorporating equally revolutionary ideals of providing a radical alternative to the hegemonic capitalist-socialist power blocks of the post-war period.
The Third World as a term needs to retrieve this lost positive sense  even if today the political order has changed so that to some extent the various forms of Islamic fundamentalism have taken over the role of providing a direct alternative to First and Second World ideologies.
' Third World ' will, therefore, be used in this book without (further) apology, or scare quotes, as a positive term of radical critique even if it also necessarily signals its negative sense of economic dependency and exploitation.
III THE PHILOSOPHICAL ALLERGY
Although Said rejects them, and Foucault characteristically does not mention them, the most effective ploys that have recently been played in this project of articulating another form of knowledge, of redefining the basis of knowledge as such, derive from a different although related body of work to that which Foucault describes  namely the phenomenological tradition of Heidegger, Levinas, and Derrida, which, seemingly like all twentieth-century European philosophy, also traces its apparent origins back to Husserl.
As we have seen, the fundamental problem concerns the way in which knowledge  and therefore theory, or history  is constituted through the comprehension and incorporation of the other.
This has led to a series of attempts to reinscribe a place for, and a relation with, the other as other, outside the sphere of mastery and therefore, logically speaking, both infinite and beyond the scope of knowledge.
Emmanuel Levinas, for example, whose career has been long enough to have introduced Husserl to Sartre in the thirties and to have been able to reply to Derrida in the seventies, proposes a rather different critique of such models of knowledge to those which we have encountered so far.
According to Levinas,
Western philosophy coincides with the disclosure of the other where the other, in manifesting itself as a being, loses its alterity.
From its infancy philosophy has been struck with a horror of the other that remains other  with an insurmountable allergy...
Hegel's philosophy represents the logical outcome of this underlying allergy of philosophy.
Levinas objects to the implicit violence in the process of knowledge which appropriates and sublates the essence of the other into itself.
But as we can see, he does not just blame Hegel here, for according to Levinas ontology itself is the problem.
Concerned to find a way to allow the other to remain as other, Levinas therefore rejects not only Hegel but Husserl, Heidegger and Sartre also, and abjures ontology altogether.
Because ontology involves an ethico-political violence towards the other, always to some degree seen as a threat, Levinas proposes ethics in its place, substituting a respect for the other for a grasping of it, and a theory of desire not as negation and assimilation but as infinite separation.
In Totality and Infinity (1961), a book self-consciously written under the shadow of two ' world ' wars in which Europe, at the limit of its attempt to devour the world, turned in on itself in two violent acts of self-consummation, Levinas questions the accepted relation between morality and politics.
It must always, he suggests, be possible to criticize politics from the point of view of the ethical.
As Althusser was keen to emphasize, according to Marx morality works simply as a form of ideological control, and Levinas concurs that ' everyone will readily agree that it is the highest importance to know whether we are not duped by morality ' (21).
But, he argues, the placing of politics  ' the art of foreseeing war and of winning it by every means'  before morality overlooks the extent to which war constitutes the philosophical concept of being itself.
For being is always defined as the appropriation of either difference into identity, or of identities into a greater order, be it absolute knowledge, History, or the state.
For its part, violence involves not just physical force, injuring or annihilating persons, but also
interrupting their continuity, making them play roles in which they no longer recognize themselves, making them betray not only commitments but their own substance...
Not only modern war but every war employs arms that turn against those who wield them.
It establishes an order from which no one can keep his distance; nothing henceforth is exterior.
(21)
War, then, is another form of the appropriation of the other, and underpins all ontological thinking with its violence.
Its corollary, or ' visage ', is the concept of totality, which, as Levinas observes, has dominated Western philosophy in its long history of desire for Unity and the One.
Through the totality, itself a kind of rational self writ large, the individual takes on meaning; the present is sacrificed to a future which will bring forth an ultimate, objective meaning when the totality of history is realized.
The objection therefore to totalization is not founded on any simple analogy with totalitarianism  though neither can this be excluded  but rather on the implicit violence of ontology itself, in which the same constitutes itself through a form of negativity in relation to the other, producing all knowledge by appropriating and sublating the other within itself.
As Levinas puts it, ' the idea of truth as a grasp on things must necessarily have a non-metaphorical sense somewhere '.
In Western philosophy, when knowledge or theory comprehends the other, then the alterity of the latter vanishes as it becomes part of the same.
This' ontological imperialism ', Levinas argues, goes back at least to Socrates but can be found as recently as Heidegger.
In all cases the other is neutralized as a means of encompassing it: ontology amounts to a philosophy of power, an egotism in which the relation with the other is accomplished through its assimilation into the self.
Its political implications are clear enough:
Heidegger, with the whole of Western history, takes the relation with the Other as enacted in the destiny of sedentary peoples, the possessors and builders of the earth.
Possession is preeminently the form in which the other becomes the same, by becoming mine.
(46)
Ontology, therefore, though outwardly directed, remains always centred in an incorporating self: ' this imperialism of the same ', Levinas suggests, ' is the whole essence of freedom ' (87).
For freedom is maintained by a self-possession which extends itself to anything that threatens its identity.
In this structure European philosophy reduplicates Western foreign policy, where democracy at home is maintained through colonial or neocolonial oppression abroad.
Levinas opposes freedom, based on self-interest, to justice, which respects the alterity of the other and can only be proposed through the asymmetry of dialogue.
This also implies an interrogation of the imperialism of theory itself.
For theory, as a form of knowledge and understanding of the spectator, is constitutively unable to let the other remain outside itself, outside its representation of the panorama which it surveys, in a state of singularity or separation.
This will also be true of any concept, because by definition the concept ' can not capture the absolutely-other '; and, to the extent that it must invoke a form of generality, of language itself.
Any conventional form of understanding must appropriate the other, in an act of violence and reduction.
This leads Levinas to denounce the inability of theory, in its drive to comprehension and representation, to do justice to any radical exteriority.
But how can we know and respect the other?
Is there a means of bridging the gap between knowledge and morality that avoids the problems of Kant's recourse to the aesthetic but also resists Lyotard's argument that the two are simply incommensurable?
How can Levinas' ethics work differently from ontology?
Against the egotism of the preoccupation of being with itself, he posits a relation of sociality, whereby the self instead of assimilating the other opens itself to it through a relation with it.
In the place of the correlation of knowledge with vision and light, the visual metaphor by which the adequation of the idea with the thing has been thought from Plato to Heidegger, Levinas proposes language, which in the form of speech enables a kind of invisible contact between subjects that leaves them both intact.
Language, however, should take the form of dialogue: whereas the universality of reason means that it must necessarily renounce all singularity, and whereas language's function in conceptualizing thought is to suppress the other and bring it within the aegis of the same, in dialogue language maintains the distance between the two; ' their commerce ', as Levinas puts it, ' is ethical '.
Dialogism allows for ' radical separation, the strangeness of the interlocutors, the revelation of the other to me ' (73).
The structure of dialogue, moreover, disallows the taking up of any position beyond the interlocutors from which they can be integrated into a larger totality.
The relation between them, therefore, is not oppositional, nor limitrophe, but one of alterity.
Dialogue, face to face conversation, maintains a non-symmetrical relation, a separation through speech.
In so doing it breaches any totality, including History:
To say that the other can remain absolutely other, that he enters only into the relationship of conversation, is to say that history itself, an identification of the same, can not claim to totalise the same and the other.
The absolutely other, whose alterity is overcome in the philosophy of immanence on the allegedly common plane of history, maintains his transcendence in the midst of history.
The same is essentially identification within the diverse, or history, or system.
It is not I who resist the system, as Kierkegaard thought; it is the other.
(40)
The thesis of the primacy of History, Levinas argues, forms part of the imperialism of the same.
For ' totalization is accomplished only in history, when the historiographer assimilates all particular existences and punctual moments into the time of universal history, whose chronological order, it is assumed, ' outlines the plot of being in itself, analogous to nature ' (55).
If History claims to incorporate the other within a larger impersonal spirit or idea, albeit the ruse of reason, Levinas contends that ' this alleged integration is cruelty and injustice, that is, [ it ] ignores the Other ' (52).
History is the realm of violence and war; it constitutes another form by which the other is appropriated into the same.
For the other to remain other it must not derive its meaning from History but must instead have a separate time which differs from historical time.
Whereas for Heidegger time and history are the horizon of Being, for Levinas' when man truly approaches the Other he is uprooted from history ' (52).
Time itself involves the ' relationship to unattainable alterity ', an absolute past.
It is in temporality, in anteriority, that we find an otherness beyond being.
Levinas calls the relation in which an infinite distance is maintained from the other ' metaphysics'.
Metaphysics, he writes, ' transcendence, the welcoming of the other by the same, of the Other by me, is concretely produced as the calling into question of the same by the other, that is, as the ethics that accomplishes the critical essence of knowledge ' (43).
Metaphysics therefore precedes ontology.
Though a troubling term, metaphysics for Levinas names a counter-tradition in philosophy in which the idea of infinity breaches all totality because ' it is a relationship with a surplus always exterior to the totality ' (22).
This surplus is the effect of the radical alterity of the other, whether as' face ' or as death, which prevents the totality from being constituted as such.
As might be expected, it is the possibility of this absolute otherness, and the ability to excise all violence in the relation with it, which Derrida questions in the first of his discussions of Levinas.
Whereas Levinas, like Habermas, posits an authentic language of expression which abhors the distortions of ' rhetoric ', Derrida argues that such alterity is constituted not through dialogue but rather through the operation of language itself: Levinas' transcendence-as-surplus is therefore redefined as a Derridean supplement.
This would mean that there can never be an authentic speech of the other as such, a position which certainly troubles Levinas' fundamental argument.
Despite their differences, Derrida's keen interest in Levinas, as Christopher Norris has argued in scrupulously non-Derridean terms, points to ' the ultimately ethical nature of his enterprise '.
The early essay on Levinas, dating from 1964, shows the extent to which Derrida has been implicated in such questions from the first  though, contra Norris, he has always shown that the conditions of ' writing ' that make ethics possible also makes them impossible.
Certain orientations of his work can be affiliated to Levinas' attempt to shift the relation to alterity from an appropriation by the same into its totality to a respect for the other's heterogeneity.
Derrida has even described the critique of logocentrism as' above all else the search for the ' other ' '.
This can be related to the concern in Derrida's work with the politics of feminism and other positions which contest institutional and political appropriation and exclusion.
In recent years Levinas has himself articulated more explicitly his account of the relation of the ethical to the political.
If there were just two, as in the face-to-face dialogue, then the ethical would preside as the injunction of responsibility for the other.
But as soon as there are three, he suggests, then the ethical moves into politics.
' We can never ', Levinas concedes to Derrida, ' completely escape from the language of ontology and politics'.
But this does not mean that the ethical has to renounce the moral order in the political world of the third person  of justice, of government, institutions, or the law.
The political can retain an ethical foundation: Levinas finds his example for this, unexpectedly, in Marx's famous comment on idealist philosophers  that the point is not to describe the world but to change it:
In Marx's critique we find an ethical conscience cutting through the ontological identification of truth with an ideal intelligibility and demanding that theory be converted into a concrete praxis of concern for the other.
From this perspective Levinas proposes the possibility that the much lamented ' subject ' be brought back not as the ontological subject which seeks to reduce everything to itself but as an ethical subject defined in relation to the other: ' Ethics redefines subjectivity as this heteronomous responsibility in contrast to autonomous freedom '.
We might compare this ethical relation to Cixous' remarks about the need to love the other or Kristeva's recent preoccupation with love which, from this perspective, hardly involves the sudden apostasy of which she has been accused, but rather as for Levinas consists of a way of formulating a ' responsibility for the Other, being-for-the-other '.
In each case these writers can be shown to be trying to place the other outside the sphere of mastery rather than in a relation of negation or of reduplication of the self.
Unlike a conventional ethics of altruism, such a relation remains one of alterity.
There will always be some return for the self in any gift  unless it can be articulated in an economy of ingratitude, a movement without return.
Unless, that is, philosophy can become dissemination: ' a work conceived radically is a movement of the same unto the other which never returns to the same '.
To the story of Ulysses, Levinas opposes that of Abraham who leaves his fatherland for ever, never to return.
This figure of the diaspora returns us to one of the most important aspects of Levinas' formulation of the relation of the ethical to the political, that is the connections which he makes between the structure of ontology and Eurocentrism, the latter ' disqualified ', as he puts it, ' by so many horrors'.
He connects the form of knowledge that is self-centred but directed outwards, philosophy as' egology ', quite explicitly with the appropriating narcissism of the West.
So in the past few hundred years Europe has been, as Gayatri Chakravorty Spivak has suggested, constituted and consolidated as' sovereign subject, indeed sovereign and subject '.
Just as the colonized has been constructed according to the terms of the colonizer's own self-image, as the ' self-consolidating other ', so Europe
consolidated itself as sovereign subject by defining its colonies as' Others', even as it constituted them, for purposes of administration and the expansion of markets, into programmed near-images of that very sovereign self.
It is this sovereign self of Europe which is today being deconstructed, showing the extent to which Europe's other has been a narcissistic self-image through which it has constituted itself while never allowing it to achieve a perfect fit.
This can be allied to Derrida's critique of ' a certain fundamental Europeanization of world culture '.
Derrida has sometimes been criticized for the generality of phrases such as' the history of the West ' or the claim that his work involves a critique of ' Western metaphysics'.
Although one of the earliest questions put to him in England concerned this category of ' the West ', in the subsequent fervour that accompanied the transformation of Derrida's work into the method of deconstruction, this problem tended to slip out of view.
In its largest and perhaps most significant perspective, deconstruction involves not just a critique of the grounds of knowledge in general, but specifically of the grounds of Occidental knowledge.
The equation of knowledge with ' what is called Western thought, the thought whose destiny is to extend its domains while the boundaries of the West are drawn back ' involves the very kind of assumption that Derrida is interrogating  and this is the reason for his constant emphasis on its being the knowledge of the West; in the same way Foucault also emphasizes that he is specifically discussing the ' Western episteme '.
The assertion that Derrida's work incurs a form of relativism is thus exactly to the point, though its implications are rather different from those generally assumed in such a complaint.
For we can say that deconstruction involves the decentralization and decolonization of European thought  insofar as it is' incapable of respecting the Being and meaning of the other ', and to the extent that its philosophical tradition makes' common cause with oppression and with the totalitarianism of the same '.
This has been the significance of Levinas' thought for Derrida.
As he puts it in Writing and Difference, at the very moment when the fundamental conceptual systems of Europe are in the process of taking over all of humanity, Levinas leads us instead to ' an inconceivable process of dismantling and dispossession '.
For Levinas' thought
seeks to liberate itself from the Greek domination of the Same and the One... as if from oppression itself  an oppression certainly comparable to none other in the world, an ontological or transcendental oppression, but also the origin or alibi of all oppression in the world.
This is the context in which to set Derrida's own intervention in Of Grammatology.
Everyone knows that that book is a critique of ' logocentrism '; what is less often recalled is that the terms of the critique with which it opens announce the design of focusing attention on logocentrism's ethnocentrism ' which, Derrida suggests, is' nothing but the most original and powerful ethnocentrism, in the process of imposing itself upon the world '.
It is this preoccupation which accounts for Derrida's choice, and forceful interrogation, of the privileged examples of Saussure  where he focuses on the ' profound ethnocentrism ' of his exclusion of writing  Rousseau and Lvi-Strauss.
In the case of the latter, Derrida's interest also focuses particularly on the way in which Lvi-Strauss produces his knowledge of a non-European civilization according to a doubled but non-contradictory logic which evades identity-thinking.
The well-known deconstruction carried out in ' Structure, Sign, and Play ' shows how the constitution of anthropological knowledge, though often paraded as scientific and objective, is nevertheless governed by a problematic of which it remains unaware: the philosophical category of the centre  which Derrida then proceeds to articulate with the problem of Eurocentrism.
The analysis of the dialectics of the centre and the margin can thus operate geographically as well as conceptually, articulating the power relationships between the metropolitan and the colonial cultures at their geographical peripheries.
This is not to suggest, however, that deconstruction in any sense brings another knowledge to bear: rather it involves a critique of Western knowledge that works by exploiting the ambivalent resources of Western writing, as if Marxism were to produce a critique of ideology without the advantage of its science (which, given the current ambiguous status of Marxist science is not a possibility to be dismissed lightly).
If one had to answer, therefore, the general question of what is deconstruction a deconstruction of, the answer would be, of the concept, the authority, and assumed primacy of, the category of ' the West '.
If deconstruction forms part of a more widespread attempt to decolonize the forms of European thought, from this perspective Derrida's work can be understood as characteristically postmodern.
postmodernism can best be defined as European culture's awareness that it is no longer the unquestioned and dominant centre of the world.
Significantly enough one of the very earliest uses of the term ' postmodern ', dating from the time of the Second World War, was that of Arnold Toynbee in his A Study of History.
He used it to describe the new age of Western history which, according to Toynbee, began in the 1870s with the simultaneous globalization of Western culture and the re-empowerment of non-Western states.
If this new period brought with it a phase of Spenglerian pessimism after the long years of Victorian optimism, Toynbee did not himself assume that the West was in decline as such, but rather that paradoxically the globalization of Western civilization was being accompanied by a self-consciousness of its own cultural relativization, a process to which Toynbee's own equally totalizing and relativizing history was designed to contribute.
Reviewing the genesis of his whole project, he recounts that his history was written
against a current Late Modern Western convention of identifying a parvenu and provincial Western Society's history with ' History ', writ large, sans phrase.
In the writer's view this convention was the preposterous off-spring of a distorting egocentric illusion to which the children of a Western Civilisation had succumbed like the children of all other known civilisations and known primitive societies.
Postmodernism, therefore, becomes a certain self-consciousness about a culture's own historical relativity  which begins to explain why, as its critics complain, it also involves the loss of the sense of an absoluteness of any Western account of History.
Today, if we pose the difficult question of the relation of poststructuralism to postmodernism, one distinction between them that might be drawn would be that whereas postmodernism seems to include the problematic of the place of Western culture in relation to non-Western cultures, poststructuralism as a category seems not to imply such a perspective.
This, however, is hardly the case, for it rather involves if anything a more active critique of the Eurocentric premises of Western knowledge.
The difference would be that it does not offer a critique by positioning itself outside ' the West ', but rather uses its own alterity and duplicity in order to effect its deconstruction.
In this context, we may note, attempts to account for poststructuralism in terms of the aftermath of the events of May 68 seem positively myopic, lacking the very historical perspective to which they lay claim.
Contrary, then, to some of its more overreaching definitions, postmodernism itself could be said to mark not just the cultural effects of a new stage of ' late ' capitalism, but the sense of the loss of European history and culture as History and Culture, the loss of their unquestioned place at the centre of the world.
We could say that if, according to Foucault, the centrality of ' Man ' dissolved at the end of the eighteenth century as the ' Classical Order ' gave way to ' History ', today at the end of the twentieth century, as' History ' gives way to the ' Postmodern ', we are witnessing the dissolution of ' the West '.
Marxism and the question of history
Marxist literary criticism has not produced a new theory in over twenty years.
Not since Macherey's A Theory of Literary Production of 1966 has there been any fundamental theoretical innovation.
This date and this gap are by no means fortuitous.
For much of this century Marxist literary criticism monopolized the realm of literary theory, for the simple reason that only Marxists consistently believed in its value and strategic necessity.
2 But since the rise of structuralism in the sixties Marxist criticism has been more and more on the defensive.
Marxist humanists still often consider structuralism simply as an attack on Marxism; this theory would be more persuasive if so many structuralists had not also been Marxists.
By contrast, it is certainly the case that, with certain notable exceptions, there are few Marxist poststructuralists.
If by the late seventies the intellectual arguments of poststructuralism began to seem virtually unanswerable, resistance eventually crystallized around the question of history.
In its preoccupation with textuality, poststructuralism had apparently forgotten all about it.
The earliest such critique in the field of literary theory was that of Frank Lentricchia, who organized his entire account of modern criticism around the premise of a ' repeated and extremely subtle denial of history by a variety of contemporary theorists'.
Lentricchia's argument was quickly endorsed by Terry Eagleton who claimed first that poststructuralism represented a ' hedonist withdrawal from history ' (aestheticism) and, a year later, that it amounted to a more menacing holocaust-like ' liquidation of history '.
Moving beyond the confines of the literary, perry Anderson has similarly dismissed all poststructuralism on the grounds that it represents' the randomisation of history '.
But what, we might pause to ask, is this remark supposed to suggest?
Is Anderson using the word ' random ' in its mathematical sense, as one out of a series, or in its statistical sense in which each item has an individual chance?
It seems not.
More likely he is using it in the more everyday usage of ' not sent or guided in any special direction; having no definite aim or purpose ' (OED), which suggests that any such view of history must have no end, and therefore no teleology.
In other words, what we are really dealing with here is a defence of a belief in the rationality of the historical process.
Anderson's history must work according to a rational principle, the dialectic, and be moving towards an end which will reveal and enact its meaning.
Without such a purpose, it is assumed, history must be meaningless.
The terms of this argument repeat exactly those of the critical debate about univocal meaning, according to which the only alternative to the idea that history has a single meaning must be that it has none at all.
But of course no one has really been suggesting that history has no meaning, for the obvious reason that any interpretation of history as such must ipso facto assert meaning.
What is in dispute is whether history has a meaning as' History '.
One alternative would be that history may be made up of the multiple meanings of specific, particular histories  without their necessarily being in turn part of a larger meaning of an underlying Idea or force.
Anderson's accusation, which conflates the concept of the differential with the notion of the random, is therefore really directed against the possibility that different histories may have different meanings.
But why deny that history can have multiple meanings?
De facto, it already has  for even for a Marxist who believes in the possibility of ' science ' there exists the multiplicity of meanings of truth and of error.
We are therefore not so much talking about a single meaning as a true one versus all the others which are false.
In a similar way, even ' History ' as a metahistorical category achieves its single meaning by subsuming a range of ethico-political concepts, such as' progress', ' human freedom ', ' necessity ' and the like, which then form the basis of the regulation and authorization of historical interpretation.
This shifts the discussion away from any simple antithesis.
The question about history then becomes the more interesting one of the relation between different significations, and the ways in which such differences can, or can not, be articulated and unified under the same horizon of totalization to produce a single meaning.
Until the lonely hour arrives in which the philosophical proof of the truth of history is produced, then history will inevitably continue as a representation and interpretation of the past  rather than Marxist truth and the false or limited interpretation of all other historians.
As a form of understanding history will necessarily also be subject to the whole range of questions that surround interpretation, representation and narrative in any form.
This is the reason why in recent years theorists have turned their attention back to the question of the historicity of historical understanding, to its status as interpretation, representation or narrative, and, more radically, to the problem of temporality as such.
Lentricchia and Eagleton, by contrast, invoke history rather like ' the political ': an outside, a concrete, that somehow remains exterior to ' theory ', unaffected by it, capable of enclosing it and even swallowing it up  as if history were in a position to consume theory.
History here is self-evident and needs no elaboration.
We all know what it is (apparently).
But the problem with the idea that it is possible to dismiss structuralism and poststructuralism with the charge that they neglect history is that this argument itself neglects history.
For history, as Althusser noted, has always been a problematical concept.
Any examination of the history of ' history, will demonstrate that it has never had the immediate certainty that is implied in the all too frequent invocation of, concrete history '.
Far from being the concrete, it has always rather been the theoretical problem.
To acknowledge that amounts to something very different from simply excising history as such.
Nevertheless the suggestion that structuralism and poststructuralism have denied history is a persuasive one which now has wide currency.
Such an argument, in implying that the problem is simply a question of the lack of history or of its presence, as if history were some undifferentiated entity that could just be added or taken away, stepped into or got out of, skates over the fact that the real question has always focused on the much more difficult issue of what kind of history, and of what status can be accorded to historical thought.
The reproach that poststructuralism has neglected history really consists of the complaint that it has questioned History.
This becomes clearer if it is considered from the more general perspective of post-modernism which has been widely characterized as involving a return of history, albeit as a category of representation.
In fact it was rather modernism, as its name implies (from latin modo, just now, or hodie, today), that tried to awake from the nightmare of history, self-consciously setting itself against the past, and rejecting forms of historical understanding.
The argument against poststructuralism really just repeats Lukcs' reproach, set out in his 1957 essay ' The Ideology of Modernism ', that modernism involves a ' negation of history '.
If Lukcs' objections to modernism laid the basis for all contemporary objections to poststructuralism, his continuing influence can help us to understand why history in particular is privileged here.
II ' IN HISTORY '
For why, after all, ' history ' at all?
Why, from the point of view of Marxism proper, not the class-struggle, or economics, the state, or social relations?
The stress on history as paramount provides a straightforward indication as to where such arguments are coming from: not Marxism in general as a political practice, but the Hegelian Marxism of the philosophical tradition initiated by Lukcs' History and Class Consciousness (1923).
Criticizing the orthodox Marxism which regarded dialectics as an external law validated by natural science, Lukcs argued for the primacy of history over economics as the most significant element in the methodology of Marxism.
His stress on Marxism as a historical method that presupposed and required the idea of totality initiated a course that determined the history of Western Marxism to our own day.
That history, from one perspective, could be seen as a consistent struggle to retain Lukcs' legacy in which history, the dialectic and the totality are interdependent to the extent that each is essential to the operation of the other in the production of a Marxist science.
Poststructuralism, which in its own way also takes part in that history of Western Marxism, differs only insofar as it foregrounds the implications of the theoretical difficulties involved rather than repressing them in pursuit of the unrealized ideal.
For Lukcs' legacy could also be seen to have bequeathed a curse.
The insistence on history as a totality, necessary if historical materialism is to justify itself as true, left Marxism with a fragile category that from the very first was always on the point of breaking apart.
To characterize only recent French thought as' the logic of disintegration ', as Peter Dews has recently done, masks over the fact that such a logic is fundamental to Marxism itself, the unassimilable dark other to its' primacy of the category of totality '.
Yet if Lukcs laid this theoretical burden upon Western Marxism, he had, ironically, already written the narrative of its subsequent history in his earlier, pre-Marxist, Theory of the Novel (1920).
In that work it is possible to see clearly the relation between his insistence on totality and the Romantic aesthetic of totality as the inner necessity that moulds all works of art.
More significantly, totality is defined in terms of its absence for modern man who exists in a relation of alienation from the world.
The novel, according to Lukcs, consists of a striving for unity, a lost state of being represented for him as for many German Romantics by an idealized picture of classical Greece, punctuated by the continual intrusion of a heterogeneous discontinuity.
Lukcs argues that the novel can transcend this threat of dispersal through an assertion of a continuous temporality, but at the same time he formulates the process in the structure of a fall: ' Once this unity disintegrated, there could be no more spontaneous totality of being '.
This account, whose origins in German Romanticism are obviously comparable to Marx 's, points to the subsequent difficulty for Marxism itself: the lost ' spontaneous' totality of being can never be retrieved.
The objections to poststructuralism at one level therefore represent merely the latest version of the Romantic nostalgia for this unfallen totality of being.
The history of Western Marxism amounts to a history of attempts to provide a means for transcending the condition of alienation and achieving that lost totality, not through irony or a reified homogeneous temporality, as in Lukcs' novel, but, according to his formulation of Marxism, through history and class consciousness.
The problem, however, is that that reconstituted totality is never certain, its status always remains ambivalent.
The appearance of History and Class Consciousness at the height of the Modernist movement suggests the degree to which history was being totalized by Lukcs at the very moment when the process of detotalization had already begun  even in the aporetic possibilities of his own writings.
So in The Historical Novel (1937) Lukcs himself would reject as delusion what he calls' a specially strong temptation to try and produce an extensively complete totality ', advocating instead a dramatic concentration and intensification of, outwardly insignificant events'.
Their relation to the totality emerges not through the form of synecdoche  the typical detail which can then be generalized as metaphor  but should, according to Lukcs, be drawn out through the narrative which inscribes and extends a connection between such moments of empirical reality and the general laws of history as a totality.
Lukcs' historicism depends upon the resolution of this tension between the idea of history and the singular detail.
But the fundamental incommensurability of idea and event re-emerges in a precariousness in this narrativization of history where each disturbance in the writing punctuates it with the unassimilable, discontinuous and disjunctive temporality of the event.
Lukcs' early assault on Marxist economism, subsequently retracted under criticism from the Leninist orthodoxy of the Comintern, became particularly influential in the post-war period among Marxist intellectuals who sought to redefine a new Marxist humanism against the economism of Stalinism with which Lukcs' name had by that time itself become associated.
Most prominent among those was Sartre, who, in the spirit of Lukcs, declared that ' both sociology and economism must be dissolved in History '.
Today when the primacy of history above all else  the economic, even class conflict  is asserted within a Marxist discourse, together with an accompanying defence of humanism, it can usually be traced back to a Marxism of a Sartrean existentialist form.
It comes as no surprise, for example, to discover that the argument about poststructuralism neglecting history was initiated by those whose own intellectual formation can be identified with the existential Marxist humanism of the New Left of the 1950s and early 1960s.
But the case of Sartre poignantly demonstrates that the return-to-history argument can really only succeed through a form of historical amnesia which conveniently forgets that history was almost impossible to find.
Any historical investigation will show that history has always been a problematic concept for Marxism, not something that has the status of a ' concrete ' existence outside and beyond theory.
This could be demonstrated for Lukcs or indeed for Marx himself.
But in the context of the alleged neglect of history by poststructuralism, it will be more appropriate to confine our discussion to the Marxism of the post-war period.
Here too history has always been the problem not the solution  which is why both structuralism and poststructuralism can be positioned within the broad trajectory of a post-war Marxism that has taken the form of a sustained enquiry into concepts of history and even the very possibility of its conceptualization.
III THE CRISIS OF STALINISM
Sartre's existential Marxism, as Ronald Aronson has recently emphasized, was itself formulated in response to criticisms of the function of history in his work by the untimely post-war ' post-Marxist ', Maurice Merleau-ponty.
In Adventures of the Dialectic (1955), Merleau-Ponty accused the Sartre of The Communists and Peace (1952) of using an existentialist ontology to justify communism ' as a completely voluntary effort to go beyond, to destroy and to recreate history ':
Sartre founds communist action precisely by refusing any productivity to history and by making history, insofar as it is intelligible, the immediate result of our volitions.
As for the rest, it is an impenetrable opacity.
Through his' extreme subjectivism ', Merleau-ponty argued, Sartre avoided the traditional claims of Marxism to be the realization of history  and thus also the problem of the relation of Marxism to Stalinism.
With the growing recognition of the real nature of the Stalinist regime, the ideological divisions of pre-war Europe became too simplistic to sustain, and even the attentiste (wait-and-see) attitude taken after the war by many Marxists, including Merleau-Ponty himself in Humanism and Terror (1947), became no longer tenable.
Stalinism posed a crisis for Marxism from which it could, in a sense, be said never to have recovered, the conundrum being that if Marxism is true, as it claims that it is, how did the first Marxist state end up as Stalinist?
There are basically two positions that can be taken with respect to this irrevocable split between theory and practice: either Marxism has been shown to be not true, or the Marxism of Stalinist Russia was not proper Marxism.
The second alternative poses a serious problem, however, insofar as it leads to the further question, how could such Marxism not be true  in the sense of how could History, in the objective processes on which a scientific Marxism places so much faith, be undialectical enough to produce Stalinism from the October Revolution?
How could History have failed to bring about an end to history?
Merleau-Ponty's response was to argue that history itself had shown Marxist philosophy to have been flawed; such philosophy must therefore give up its claim to truth.
Ironically, then, the trajectory which today produces the Marxist argument that poststructuralism neglects history was itself initiated by the claim that Marxism itself had been invalidated by history.
In this context, the major challenge for Marxism became how to explain Marxism's ' detour ' from itself, a question on which Sartre's project was to founder in the second volume of the Critique of Dialectical Reason.
It was Stalin's unfortunate deviation which led to the whole series of attempts by Western European Marxists to return-to-Marx  the major examples in France being Sartre and Althusser, and in Germany the Frankfurt School.
Merleau-Ponty, on the other hand, arguing that only one Marxism, Stalinism, could be said to be the product of actual history, rejected any possibility of trying to define a new, more genuine Marxism as the return-to-Marxists tried to do.
According to Merleau-Ponty, Western academic Marxism, starting with Lukcs, amounted only to the production of ' ideas without historical equivalents' (204).
He therefore repudiated any attempt to begin the painful separation of an ' authentic ' from a ' false ', orthodox or official Marxism, choosing rather to stress the ambivalence of Marxism itself.
For Merleau-Ponty no sublation could resolve Marxism s own division between its theory and the history of its practice, itself, he argued, an acting out of its theoretical equivocation between history as a process of natural necessity and history as the product of human praxis.
These conflicts, he contended, although highlighted in the differences between the Western Marxism of Lukcs and the orthodoxy of Marxist-Leninism, could be traced back to the work of Marx himself.
Merleau-Ponty's argument thus meant acknowledging a certain equivocalness in Marxist claims to truth, which could no longer claim exemption from critical examination, and led inevitably, therefore, to an early form of what, even then, was characterized as' post-Marxism '.
Merleau-ponty, whose anticipations of later post-Marxists such as Foucault and Lyotard are readily apparent, therefore rejected not the dialectic as such, nor history, but the closed dialectic as an autonomous principle that was supposed to produce the grand narrative of History:
The illusion was only to precipitate into a historical fact  the proletariat's birth and growth  history's total meaning, to believe that history itself organized its own recovery, that the proletariat's power would be its own suppression, the negation of the negation.
(205)
In a provocative comparison of the failures of the French and Russian revolutions, he argued that the problem stemmed from the fact that no class, whether proletarian or bourgeoisie, can become the ruling class without taking upon itself something of the historical role of a ruling class  especially if at the same time it also considers that, history carries within itself its own cure':
To assume that the proletariat will be able to defend its dictatorship against entanglement is to assume in history itself a substantial and given principle which would drive ambiguity from it, sum it up, totalise it, and close it.
(221)
Instead, Merleau-Ponty proposed an open dialectic which would concede Marxism's equivocalness, and give up the claim to the dialectical logic of History as a process of objective truth.
Or to put it another way, he maintained that Marxism can only constitute its totality through its perpetual detours from itself.
Sartre's extravagances
Thus, the world and man reveal themselves by undertakings.
And all the undertakings we might speak of reduce themselves to a single one, that of making history.
Jean-Paul Sartre
Sartre took the opposite course to Merleau-Ponty and sought instead to define a new authentic Marxism.
His project in the Critique of Dialectical Reason (1960) amounted to a philosophical defence of Marxism and its dialectic through a reassertion of its claim to be the only valid interpretation of history, ' the untranscendable philosophy for our time '.
However, Sartre was no more prepared than Merleau-ponty to return to what he characterized as Marxist organicism, in which the laws of history unfold according to their own autonomous momentum.
Not wishing to propose History as a transcendent law outside the human, Sartre was careful to distinguish his claims from Engels' validation of the dialectic as the law of nature.
He sought instead to prove that history achieves its course and its meaning from the actions of men, even if men are also simultaneously the products of history.
Merleau-ponty had accused the existentialist Sartre of denying history and trying to hold everything together instead ' only by the hopeless heroism of the I '.
Sartre therefore tried to show that the two were not opposed: arguing that the dialectic is produced by human subjectivity rather than inscribed within history itself, at the same time he also asserted the truth of the Marxist account of there being one history with a single meaning  a history, that is, in which all differences return as the same.
Everything, therefore, hinged on Sartre's ability to substantiate the inter-dependence of these two claims.
' There is a crisis in Marxism ', Sartre announced at the beginning of the Critique; this crisis, not the first and not to be the last, Sartre characterized as the result of a paradox in which historical materialism had become ' at one and the same time, the only truth of History and a total indetermination of the Truth ' (I, 19).
Today we might say that it had been shown to be different from itself.
This was the aporia detected by Merleau-Ponty: but rather than question Marxism's truth, Sartre sought to remove its indetermination.
He argued that though historical materialism had satisfactorily explained the forms and conditions of human reality it had never established theoretically the validity of its own existence, never shown how it constituted not just the substance of reality but its logical form as well.
In order to prove the truth of History's prospective totalization into one meaning, therefore, Sartre's initial philosophical task was to ground his argument epistemologically by proving the legitimacy of the dialectic itself, thus demonstrating not only that history was dialectically intelligible but also why it necessarily should be so.
' The dialectic ', according to Sartre, ' is both a method and a movement in the object ' (I, 20): Marxism asserts simultaneously that both the process of knowledge and the structure of the real are dialectical, but it has never proved the former  basing its claim to truth instead on the ' dogmatic dialectic, of natural science.
Accordingly, much of Volume I of the Critique is taken up with the attempt to prove the dialectic a priori as the universal method and the law of anthropology, superseding that which Kant had provided for analytical reason.
Volume Two, Sartre announces, will then prove ' the Truth of History ' (I, 52):
it will attempt to establish that there is one human history, with one truth and one intelligibility  not by considering the material content of this history, but by demonstrating that a practical multiplicity, whatever it may be, must unceasingly totalise itself through interiorizing its multiplicity at all levels.
(I, 69)
The point, therefore, is to validate Marxism, to show that it is not simply a method of interpretation, nor even that it is the best method of interpretation that can most successfully account for the facts and the course of history, but to prove a priori that history works according to dialectical structures, and to demonstrate ' the moments of their inter-relations, the ever vaster and more complex movement which totalises them and, finally, the very direction of the totalization, that is to say, the ' meaning of History ' and its Truth, (I, 69).
Had he succeeded, Sartre would have established dialectical reason as successfully for the human sciences as Kant had established analytical reason for natural science.
But he was unable to do it.
The indetermination that he sought to banish increasingly returned to haunt him; he could never demonstrate how the unceasing totalization of multiplicity would ever reach its promised moment of finality.
Sartre's claim for the continued validity of Marxism as a method of understanding necessarily meant that he had to respond to the problem of Stalinism.
His was the first attempt to explain Stalinism as an aberration through the forms of Marxist analysis itself.
Rather than reject Marxism on the grounds that it had been disproved by history, as Merleau-ponty had done, he sought to account for Stalinism through a dialectical analysis of the specific history of the Soviet Union since the Revolution: theory and practice, he argued, had become separated with the result that the former had become ' sclerosed ' while the latter had become ' blind ' and ' unprincipled ' (I, 50).
According to Sartre, if the dialectic had become blocked, an understanding of its detour could nevertheless only be achieved through the use of a dialectical logic.
He therefore sought to prove both that the structures of history were necessarily dialectical and that the course of actual history could be shown to be so.
But in attempting to discover why the Russian Revolution followed the path it did, he found himself arriving at the conclusion that, far from being a ' false ' deviation, in the circumstances of history Stalinism had been necessary.
Having set out to demonstrate how, through its understanding of the structure of history, Marxism was in a position to forecast the future, he came to the conclusion that there was no guarantee that history could promise anything better that what had just passed.
II THE FIRST CRITIQUE
Unity had been easy during the Resistance, because relationships were almost always man-to-man.
Over against the German army or the Vichy government, where social generality ruled, as it does in all machines of state, the Resistance offered the rare phenomenon of historical action which remained personal.
The psychological and moral elements of political action were almost the only ones to appear here, which is why intellectuals least inclined to politics were to be seen in the Resistance.
The Resistance was a unique experience for them, and they wanted to preserve its experience in the new French politics because this experience broke away from the famous dilemma of being and doing, which confronts all intellectuals in the face of action....
It is only too obvious that this balance between action and personal life was intimately bound up with the conditions of clandestine action and could not survive it.
And in this sense it must be said that the Resistance experience, by making us believe that politics is a relationship between man and man or between consciousnesses, fostered our illusions of 1939 and masked the truth of the incredible power of history which the Occupation taught us in another connection.
If by 1945 Merleau-ponty had here already indicated the historical basis for what was to become Sartre's project in the Critique, the theoretical problem that followed was how to unite orthodox Marxist concepts of economic determinism and historical necessity with the individualism of Sartre's earlier work, that is the existentialist notion of the authentic self and the possibility of choosing one's own freedom: how to relocate human practice within historical determinism, to reconcile the individual with the social, the idea of agency with that of necessity, or freedom with history?
For Marxist or Hegelian thought, such an age-old contradiction ought to be resolvable through the operations of dialectical logic.
Citing Marx's famous remark in The Eighteenth Brumaire, ' Men make their own History... but under circumstances... given and transmitted from the past ', Sartre suggested that this implied ' the permanent and dialectical unity of freedom and necessity, (I, 35).
Accordingly, he attempted to shift Marxism away from orthodox theories of an absolute determinism towards the primacy of a concept of ' History ' which, while still a totality as it had been for Lukcs, a process with a determinate meaning and end, could also include a concept of human agency and thus articulate the individual with the social, freedom with determinism.
Today, Sartre's voluntarism is to some extent returning to favour as the result of a desire to retrieve the categories of agency and the subject, which goes together with the wish to get out of the apparently totalizing systems of Adorno, Althusser or Foucault.
Sartre's stress on the role of the subject also finds approval because many of those no longer prepared to argue for a general theory of history as the progress of a single narrative of class-struggle, have begun to argue instead for a return of its correlative, the subject, almost as if it was the next best thing in the absence of history itself.
But it still leaves us with the crucial problem that Sartre had to solve, namely how to link human consciousness with the processes of history so that the former can be said to be the agent of the latter.
How could this take place?
As Sartre had put it in What is Literature?:
' Never has homo faber better understood that he has made history and never has he felt so powerless before history '.
How can man make history if at the same time it is history which makes him?
Sartre considered that orthodox dialectical materialism takes the easy way out by merely eliminating the first in favour of the second, making man a passive product entirely determined by economic circumstances.
Unwilling to follow Merleau-Ponty by dropping the second in favour of the first, Sartre argued that it was possible for man to be both at once through the movement of praxis, that is intentional actions which produce material effects.
Its corollary, whereby praxis becomes determining history, he called the ' practico-inert ', by which he meant the material circumstances that have themselves been created by previous praxes and which form the conditions for new praxis.
The individual is thus both totalizer and totalized, deftly uniting freedom with necessity.
Though this neatly solves the problem of how man can make history while at the same time history makes him, it does not answer the larger question of how a multiplicity of the products of individual acts, ' totalizations', can themselves be totalized into the overall totalization required by the logic of dialectical rationality  rather than being the arbitrary, blind and self-cancelling movements of, say, Hardy's immanent will.
Sartre's fundamental thesis, that ' History continually effects totalisations of totalisations' (I, 15), does not in itself answer what Ronald Aronson has rightly characterized as the question for Sartre: ' how do separate, antagonistic actions yield a history; how do individual totalizations lead to Totalization (and also progress, the direction of history, its truth and its meaning)? '
For the whole basis of Sartre's argument is that the dialectic of history is not a metaphysical law, ' some powerful unitary force revealing itself behind History like the will of God ', but the continuously produced effect of individual conflicts; each action is in its turn subsumed as a part of the whole in an ever broader, developing totalization (I, 37).
What Sartre needs to demonstrate, therefore, is that if the law of the dialectic works from the individual level, overall it produces nothing less than the intelligibility or the meaning of History as such:
The dialectic is the law of totalization which creates several collectivities, several societies, and one history  realities, that is, which impose themselves on individuals; but at the same time it must be woven out of millions of individual actions.
We must show that it is possible for it to be both a resultant... and a totalizing force... how it can continually bring about the unity of dispersive profusion and integration.
(I, 36)
But what produces the overall direction of History?
How does it ' continually bring about the unity of dispersive profusion and integration '?
Sartre does not attempt to answer this question, shifting rather to the notion of a lack of self-consciousness.
The dialectic is the law that remains hidden.
At the moment, ' history is made without being known (l'histoire se fait sans se connaitre) '  history constitutes, we might say today, a political unconscious.
' Marxism ', on the other hand, Sartre claims, ' is History itself becoming conscious of itself ' (I, 40): as for Lukcs, it is by becoming conscious of itself as the subject of history that the working class will understand history's meaning  and so recognize itself as the meaning of history.
While he suggests in apocalyptic tones in The Problem of Method that this process of self-consciousness is at last beginning to take place, and that civil, foreign and colonial wars are becoming apparent as different forms of a single class struggle, Sartre also admits that the divorce between theory and praxis which ensued under Stalinism has generally prevented any clear self-consciousness among the masses.
The ' detour, of Stalinism is thus formalized as Marxism's own descent into its unconscious, resulting in a dream-work of heterogeneous histories that have eluded systematization and subsumption into the single meaning of Marxism's own reading of history.
It is therefore, Sartre argues, our historical task to make it known, promoting not just the historical process as such, but also the general recognition whereby the plurality of the meanings of individual histories can be seen to combine to make one history, with one meaning  the ' Truth of humanity ' (I, 822).
Sartre's account thus sets up the articulation of history, univocal meaning, and totality as the indissoluble set of elements required for the validation of Marxism, necessary in order to save it from its detour from itself.
The problem, however, remains how such universals are produced from the multiplicity of initiating individual praxes.
If History is a history of conflict, how could it be both one and internally diversified without the inner moving principle of the dialectic?
How can History be a unity if it is also conflictual, if each action is aimed at destroying the other and results in a double negation in which the original aims of each action have been destroyed by the other?
If each action negates the aim of the other, where is the ' unity ' totalized in the conflict?
How does History constantly totalize itself?
(I, 817  18) Sartre arrives at what he calls' the real problem of History ', that is how there can be totalization without a totalizer, only at the very end of Volume I. It is not until the next volume, however, that he intends to show how individual actions, separate multiplicities, make up ' one human history, with one truth and one intelligibility ' (I, 69).
Throughout, Sartre has nevertheless taken it ' for granted that such a totalisation is constantly developing both as History and as historical Truth ' (I, 822).
His History, therefore, is always in process: but its teleology of a final totalization always has to be assumed.
He thus asserts the Truth of History while constantly projecting forwards and deferring its proof.
But when the proof comes it also turns out to rest on the assumption that it is already true.
III THE SECOND CRITIQUE:
' We are trying to establish the Truth of History'
Sartre produces a persuasive formulation of the relation of the individual to the determining historical circumstance through his concepts of praxis and the practico-inert; it is also obvious to him how the larger question of History ought to be solved.
He asserts it often enough: ' History continually effects totalisations of totalisations, (I, 15).
But in Volume I of the Critique, as we have seen, he continually defers the demonstration of his proof.
For the ' dogmatic dialectic ', as he describes dialectical materialism, the whole question naturally poses no problem, for each person or group simply constitutes a partial moment of an already operative movement of totalization that produces them and then goes beyond them.
But Sartre's dispensing with all historicist schemas creates the problem of how, between two autonomous and contradictory totalizations, there could be ' one dialectical intelligibility of the ongoing process' (II, 13).
He attempts to solve this through matching his account of contradiction with a new concept derived from the conjunction of work and labour: ' anti-labour ' (anti-travail), a category of negation.
The products of a conflict, as well as its residues, may seem incomprehensible insofar as they differ from the original intentions of any of the combatants.
Nevertheless they constitute the basis and conditions for further actions and history.
Dialectical reason encounters such products as undecidables:
aporias  because they seem to be at once the results of a communal enterprise while at the same time bearing witness to the fact that this enterprise never existed except as the inhuman reverse side of two opposed actions in which each aims to destroy the other.
In the dialectical perspective, we encounter these objects as productions which are human and provided with a future... thus they seem by themselves to be totalizations in process.
(II, 20)
This means that first, they lead to unforeseeable results, and second, that the internal structure of social objects contains' the double negation of themselves and of each constituent part by the other ' (II, 22).
There is, therefore, before any understanding of historic agents and movements, a certain aporia in all social ensembles: from afar they may appear whole, but close to, they can be seen as riddled with holes.
This increasingly comes to resemble nothing so much as Sartre's own account of History.
In the context of the negativity of this formulation, Sartre is obliged to pose once again the problem of whether there can be a totalization, without any independent totalizer or totalizing force such as the dialectic, or whether the structure of negation that he has described means that, as for Adorno or Bakhtin, history does not develop positively but, rather, negatively, and is therefore instead detotalizing:
Marxism is rigorously true if History is totalization; it is not so any more if human history decomposes into a plurality of particular histories, or if, in any case, within the relationship of immanence that characterizes struggle, the negation of each adversary by the other is on principle detotalizing.
Certainly, we have neither the design nor the actual possibility of showing here the full truth of dialectical materialism...
Our goal is solely to establish whether, in a practical ensemble torn apart by antagonism (whether there are multiple conflicts or whether they are reduced to one) the breaks themselves are totalizing and carried along by the totalizing movement of the ensemble.
(II, 25)
At this point Sartre confronts the possibility that the whole project of the Critique, stated so confidently on its very first page  ' it must be proved that a negation of a negation can be an affirmation ' (I, 15)  may break down.
perhaps the Truth of History can not be proved; its direction can not be discerned.
Sartre's reaction to the threat that the negation of a negation may produce a detotalizing effect is to introduce a new category, a unifying force of a ' singularization ' which incarnates the universal: ' If totalization really is an ongoing process', he writes, ' then it operates everywhere.
That means both that there is a dialectical meaning of the practical ensemble... and that each singular event totalizes in itself the practical ensemble in the infinite richness of its singularity ' (II, 26).
Unlike Lukcs' insignificant event from which the universal is precariously drawn out through the narrative, Sartre's singularity works synecdochally in a conventional antinomy with the universal, the relation between the two structured according to the familiar nineteenth-century model of organic growth or process in which each singular event makes up the whole while, as he puts it, ' the whole is entirely present in the part as its present meaning and as its destiny '.
Sartre's ' singular universal ', therefore, begs the question, for it is predicated on the assumption that, if there is not a totality as such, then there already is an overall totalization: ' If totalization really is an ongoing process... '.
That developing process is itself totalized through Sartre's assertion that each totalization is the totalization of all struggles, as it must be if the singular is to incarnate the universal.
The concept of the singular universal thus facilitates a circularity in the argument whereby Sartre can avoid the question which he began by posing.
So the singular universal presumes the totalization he can't (yet) prove.
This strategy marks a structure of repetition in Sartre's text: each time he poses the question of how there can be totalization of History without a totalizer, he retreats to a more limited example whose unity is already evident, but which in the end only brings him back to the original question again.
In this case Sartre demonstrates how the singular universal works, by appealing to the concrete example of a boxing match  a random exemplification that just happens to reproduce the single adversaries of the master/slave dialectic.
Its totalization takes place through its incarnation, as part of the totality of boxing, in the overall framework within which each individual fight occurs.
Each boxing match, Sartre claims, must be both a unique event and also in some sense the incarnation of all boxing, whose rules and conventions it follows, and whose past and future history it sets itself against.
But it is also apparent that that totality is not completely known, nor is its future shape even presumed.
By totalization, therefore, Sartre does not here mean anything like a predetermined end or final closure of a totality but rather a process of mediation among the parts, where each is determined by the other.
No one can predict the future of boxing, either at the level of particular victories or defeats, in effect a structure of repetition, nor at the more general level of possible modifications or developments of the sport.
Yet each individual bout articulates itself within the framework of the history of boxing, known and unknown.
This allows Sartre to include chance and contingency in his scheme.
Since it can not be known as a concept that will realize itself in the future, Sartre argues instead that the totality only produces itself in the moment: ' The incarnation as such is at once unrealizable except as totalization of everything and irreducible to a pure abstract unity of that which it totalizes' (II, 58).
The universal, the totality, can only be known through the singular.
But its future direction remains unknown and indeterminate.
At the same time Sartre makes a larger, though less persuasive claim that the boxing-match also works synecdochally and re-exteriorizes a more fundamental violence, namely the interiorized condition of scarcity that he considers to be the basis of conflict in general.
The boxing match ' is the public incarnation of all conflict ' (II, 32), totalizing in its own struggle the whole of ' contemporary irreducibilities and fissures' (II, 26):
thus one can and must say... that each fight is the singularisation of all the circumstances of the social whole in movement and that by this singularisation, it incarnates the enveloping totalization which the historical process is.
(II, 58)
But then he adds, ' I have said and I repeat that we have not yet proved that this enveloping totalization exists'.
So, although it is possible to conceive of any event as an incarnation of the totality, insofar as it must itself make up a part of that totality in its determination, unlike the case of the boxing match, where we can define the overall entity ' boxing ', it still remains unproven that an overall entity, ' History ', can be said to exist at all.
For all his use of the model of the boxing match, Sartre has really got no further with the fundamental question of how there can be totalization without a totalizer.
It is simple enough for him to show that a boxing match takes place within a general totality of ' boxing ', which is itself only ever present in any individual incarnation, for as a game each match is both an individual bout and something conducted according to general rules and a specific social tradition.
But what about those conflicts that do not take place within such a constituted social system, such as conflictual bourgeois societies which can not be said to be unified, except, as Sartre suggests dismissively, by appeal to a lost paradise before the class struggle?
Such societies make the conflicts of history irreducible to Sartre's logic unless, like the boxing match, they can be shown to take part in a larger totalizable category.
However, when he poses the question of whether his method could work as well for the class struggle as for the boxing match, Sartre admits that he finds it impossible to answer and returns instead to the much easier case of sub-groups.
Once again he backtracks and assumes a larger unity in which conflict takes place.
The major question thus always remains unanswered in the Critique: every time that Sartre announces that he is about to proceed with the fundamental problem of how History can be a totalization without a totalizer, he turns back to a previous, more easily intelligible stage on the way.
His difficulty is accompanied by a no doubt symptomatic increasing distrust of universals so that, in championing specificities against them, he seems to give up the attempt to validate the universals  History as Totalization  that originally formed the object of his project.
At this point a further contradiction in Sartre's whole enterprise begins to open up: for someone so deeply distrustful of universals it seems curious that he has involved himself so emphatically with the notions of totality, History, and the dialectic.
As Sartre insists more and more upon the virtues of specificity (II, 200  05), such is his distaste for Marxist or any other universalizing categories that he refuses even to countenance them, attempting to replace concepts with specificities, universals with singulars.
His scepticism goes to the extent of even denying the existence of ' society ' or ' the nation ' as such (II, 24, 61)  the very totalities which his own general argument for totalization requires.
Here certain similarities with later ' post-Marxist ' theorists such as Foucault or Lyotard, begin to become apparent.
But if Sartre anticipates such later thinkers we should not assume too quickly that they have simply taken his insights further.
For Sartre's singular universal could never be expected to solve the formal theoretical problem of how individual existential existence can be related to History insofar as it simply renames, in one oxymoronic category, the original antithetical terms.
Without an adequate theory of their articulation, they simply begin to separate again.
IV THE AMBIVALENCE OF HISTORY
The ' singular universal ', with its organicist and essentialist overtones whereby the part incarnates the whole, is thus the product of Sartre's hesitation between the singular, which remains privileged as the existential basis of history, and the universal which as a Marxist he feels is required for its intelligibility and validation.
The effect of Sartre's oscillation between these two poles is that his argument is increasingly drawn towards the very positions that ostensibly he wishes to refute.
So, for example, he compulsively returns to the idea that history might consist of several totalizations rather than one:
Might not History, at the level of the grand ensembles, be an ambiguous interpenetration of unity and plurality, of dialectic and anti-dialectic, of sense and nonsense?
Might there not be, according to the circumstances and the particular ensemble, several totalizations without any links between them except coexistence or no matter what other exterior link?
(II, 131)
Up to this point, Sartre's notion of conflict has operated in direct antithesis to that of Bakhtin: whereas for Sartre, struggles are totalizing, always creating a larger, meaningful and developing whole out of multiplicities, for Bakhtin they are detotalizing, dissolving ' previous totalisations'.
But Sartre's own text here develops a dialogism in the tension between these two possibilities which becomes an increasingly dominant characteristic of his regressive-progressive method and ends up by detotalizing the very totalization which he sought to prove.
Here we find the theoretical corollary of what Althusser was to characterize as the tragic ' double thesis' of Sartre's ' historicist humanism ', and which produced, in Michael Sprinker's words, a ' ceaseless rebounding between the poles of revolutionary optimism and historical pessimism '.
Sartre himself theorizes this growing ambivalence as the difference between his own method and that of orthodox Marxism, elaborating the distinction between concept and incarnation as two possible paths of a dialectical understanding of the same social reality.
The concept, as employed by orthodox Marxism, goes from the singular to the universal and therefore, Sartre claims, detotalizes in a movement of ' decompressive expansion ', whereas incarnation involves' a way of totalizing compression which, on the contrary, seizes the centripetal movement of all the significations drawn in and condensed in the event or in the object ' (II, 59).
But it is rather Sartre's own text that seems to be caught between these two dialectical possibilities of expansion and compression in a double logic.
Thus the very concept of totalization, distinguished from totality, must always be refused its prospective closure, for if ' History continually effects totalisation of totalisations' (I, 15) it must necessarily also mean that by definition it can never be absolutely totalized.
As he puts it in The Problem of Method: ' For us the reality of the collective object rests on recurrence.
It demonstrates that the totalization is never achieved and that the totality exists at best only in the form of a detotalized totality. '
How then can history totalize totalization if totalization is never accomplished?
This is the point at which the dialectic, as a unity of method and movement, of subject and object, knower and known, requires the writing subject who must effectively hold them together.
Thus the critical investigation
takes place inside the totalisation and can be neither a contemplative recognition of the totalising movement, nor a particular, autonomous totalisation of the known totalisation.
Rather it is a real moment of the developing totalisation...
(I, 48)
This means that the process of totalization must be kept moving by the critical investigation itself on which it comes to depend but which by the same token it can never subsume.
Every time that Sartre asserts the enveloping movement of the historical process, while adding emphatically that he has yet not proved that such a totalization exists, he must always simultaneously introduce a counterstructure of repetition, so that his argument seems to fluctuate, like the groups that he describes, ' in a state of perpetual detotalisation ' (I, 579).
This is why although he dares to pose his overwhelming question, he can never be in a position to answer it with more than an assertion and the promise of more totalizations.
The indetermination he sought to excise returns to govern him.
The Critiques can do nothing but proliferate extravagantly into the writing of a perpetual process of deferral.
If the singular universal increasingly punctuates the forward movement of Sartre's text instead of providing it with a dialectical meaning and direction, the problem that it was invoked to solve meanwhile takes its own aberrant course.
Denying that conflict is an a priori structure actualized in historical struggle, we have seen that Sartre does explain, through the category of the ' practico-inert ' (determining material conditions which have themselves been created by previous praxis), how individuals or classes in conflict produce a historical movement to which they are then subject.
Actions, in turn, create the conditions for their own contradiction when the successful fulfilment of praxis is prevented by the action of another praxis whose aims conflict with it.
The fact that conflict produces anti-labour, or a contingent situation that was not the original aim of either conflictual group, must mean that in this schema History has a negative unity, notwithstanding the troubled formulation of the singular-universal.
If Sartre denies all possibility both of an underlying historical structure and of a larger unity, a ' hyper-organism ' as he puts it, within which conflict takes place, then history as he conceptualizes it here has no specific or necessary direction.
More seriously, it suggests that history will always deviate from any intended route and take an unforeseen one instead.
If political struggle takes place for specific purposes and anticipated results, here Sartre seems to condemn it to an unending series of detours that will never arrive at their destination.
Insofar as' History ' names the horizon of totalization of those ethico-political meanings that point in the direction of social change, it here not only loses its single meaning, but threatens to lose even the bases on which its meaning is constituted.
Sartre, therefore, can not after all be said to substitute a simple voluntarism for an underlying historical structure.
In the second Critique Sartre seems remarkably equivocal with regard to his central question of how struggles which have no controlling totalizer or underlying structure of totalization can be intelligible.
He begins by allowing the possibility of schism, such as that which took place in the Roman Empire (II, 84), which implies that humans can choose not to allow totalization.
He then admits that the result of any conflict, though it may be intelligible, may not necessarily imply any progress.
Progress or improvement is no more likely than decline: he now repeatedly disowns any necessary historical logic of progress; its incarnations will always, strictly speaking, be accidental ones in relation to the objectives which were at the origin of praxis:
[ History ] is not rigorous because it always proceeds by faults and corrections, because it is not in any way a universal schema but a unique adventure that unfolds on the basis of prehistoric circumstances which constitute in themselves, and in relation to all the objectives and all the practices, a heavy and badly understood legacy of fundamental deviations.
(II, 238)
History therefore does have a structure of sorts  a legacy of continual aberration.
By abandoning any logic of progress for the negative synthesis of anti-labour which produces deviation, Sartre is, on the other hand, able to account for the major detour that provides the context for his whole debate with Merleau-Ponty, and indeed for all post-war Marxism: the spectre of Stalinism.
The problem, however, is that though Sartre produces a provocative analysis of why Stalin was successful, this imperceptibly slides into an account of why Stalin had to do what he did, to the extent that Sartre appears to justify his actions according to the particular necessities of his historical circumstances.
Thus Sartre argues that it was necessary to choose between the break-up of the Revolution and its deviation.
Stalinism, in fact, accords all too perfectly with the perverse structures of Sartre's own theoretical argument.
It also serves, appropriately enough, as a way of closing down the aporias which have opened up in Sartre's text.
Sartre takes the slogan ' socialism in one country ' as an example of the unintended but necessary product of the anti-labour of the Stalin-Trotsky conflict.
He translates the ' socialism in one country ' of the one versus the ' internationalism ' of the other into a conflict between his own, now reified and separated, terms of concrete incarnation and abstract universalism:, the revolutionary incarnation chose the singular against the universal and the national against the international' (II, 223).
Stalin therefore becomes the authentic Marxist, able to deal with specific historical circumstances, as against Trotsky who is regarded as having been hopelessly caught up with the a priori universalism of an abstract Marxism.
According to Sartre, the Revolution's ' incarnation directly contradicts its universalization, (II, 116).
If he therefore demonstrates why socialism in one country was historically necessary, he ends up by apparently justifying Stalinism: in showing how anti-labour produces deviation, he seems to endorse its course while rejecting any overall schema which can provide the basis of a claim that it will be ultimately corrected.
The analysis stops at the point where even Stalin's anti-semitism begins to appear necessary (II, 281).
Sartre never broaches the question of the other crimes committed by Stalin (which is just as well in view of the way the argument is going) nor does he address the obvious problem that his own identification of Stalin as the practical Marxist of specificities against Trotsky's abstract universalism runs directly counter to the fact that Stalinism produced the most clear-cut example of a Marxism which internalized an abstract schema within itself.
Instead, at this point Sartre at last abandons all hope of proving History as a totalization without a totalizer.
He moves into a long attempt, which takes up the rest of the book, to revise the earlier ontology of Being and Nothingness into a new ontology of action and even of History, as if, after all, he is investigating the prospect of accrediting the latter with ontological status  a possibility which has always haunted his text in its insistent negation.
This analysis, though lengthy, remains incomplete.
In studying the deviation, Sartre himself deviated from his original project.
But this was not just the effect of Stalin per se, despite his perverse attraction as a sort of singular-universal who did in effect unite ' man ' with ' history ': it was also the result of the theoretical collapse of the singular-universal as a totalizing concept that could save history from the aberrant consequences of anti-labour.
Sartre's philosophical grounding of ' History ', therefore, foundered in the second volume when the logic of history inexorably brought him, not to totalization without a totalizer, but to the very reverse: the figure of Stalin and the conclusion that Stalinism had been indispensable for the development of socialism in the Soviet Union.
Although the Critique had been intended to rescue Marxism from the sclerosis of Stalinism, Sartre found that his theory of history, far from explaining what had gone wrong when the most radical political theory turned out in practice to be one of the most oppressive, had rather shown why it had necessarily happened that way.
Volume II was never completed, and only published, unfinished, in 1985.
Thus Sartre's attempt to prove the truth of Marxism and of History at a philosophical level was abandoned.
The basic theoretical problem had been to show how two or more autonomous and contradictory totalizations make up one dialectical intelligibility: to do this he needed to totalize the classes in struggle, and to discover the synthetic unity of a conflictual society.
The articulation of individual praxis with History stood or fell with the concept of history as totalization without a totalizer.
The stakes had been painfully clear to Sartre himself: ' Marxism is rigorously true if History is totalization; it is not so any more if human history decomposes into a plurality of particular histories' (II, 25).
The problem was that the premises of his own theory required the answer to this question to remain always in abeyance, while his text enacts rather than resolves the equivocality of the choice which it sets up.
If Sartre began by attempting to unite man with history, in the notes at the very end of the manuscript he poses once more the fundamental ontological question, and immediately follows it with a decidedly unequivocal answer:
IS HISTORY ESSENTIAL TO MAN?
No.
(II, 454)
Sartre's argument for History as totalization, then, was already caught up in interminable difficulties by the time he was drafting Volume II of the Critique in 1958.
No sooner had he published Volume I in 1960 than the whole status that he claimed for History, for man, and for their articulation, came under attack.
V HISTORY AND ETHNOCENTRICITY
The case of Sartre demonstrates why the assertion of ' history ' against structuralism and poststructuralism must always remain problematical.
In this context it is ironic that it was a structuralist critique of Sartre's claims for history that, in historical terms, brought about the abrupt intellectual demise of existential Marxism itself.
Lvi-Strauss' famous objections to Sartre, which appeared in the last chapter of The Savage Mind (1962), are sometimes represented as if they were merely a structuralist attack on Marxism.
But their immediate occasion was as a response to the use that Sartre himself had made of Lvi-Strauss' The Elementary Structures of kinship (1949) in the first Critique.
Sartre had stated that
If there is to be any such thing as the Truth of History (rather than several truths, even if they are organized into a system), our investigation must show that the kind of dialectical intelligibility which we have described... applies to the process of human history as a whole
(I, 64).
For Sartre ' human history ' was identified with the history of the West, and it was for this reason that Lvi-Strauss contested Sartre's claim to have established the human foundation of ' a structural, historical anthropology ' for Marxism.
In retrospect, it was highly significant that resistance to Sartre's ' History ' began by drawing attention to its ethnocentrism and Eurocentrism.
Lvi-Strauss' critique of Sartre effectively takes the form of a deconstructive analysis: he begins by arguing that the Critique works by a double movement  just as, only four years later, Derrida was to suggest that the anthropologist's own work functions in exactly the same way.
Lvi-Strauss focuses on the ambivalence which we have already seen to be such a distinctive feature of the Critique, pointing in particular to Sartre's vacillation between two concepts of dialectical reason.
In the first he regards it as antithetical to analytical reason, as truth to error, while in the second he sees it as complementary to it.
If this is the case, then Lvi-Strauss argues that Sartre disqualifies his own Critique, which establishes the truth of dialectical reason partly through the exercise of analytical reason, as any critique, which separates subject from object, knower from known, inevitably must.
But then ' it is difficult to see how analytical reason could be applied to dialectical reason and claim to establish it, if the two are defined by mutually exclusive characteristics'.
Moreover, if the two eventually arrive at the same truth anyway then there seems to be little need for dialectical reason at all.
' Sartre's endeavour seems contradictory in the one case and superfluous in the other ' (246).
While stressing that they both take their point of departure from Marx, Lvi-Strauss suggests that for him Marxism implies that
the opposition between the two sorts of reason is relative, not absolute.
It corresponds to a tension within human thought which may persist indefinitely de facto, but which has no basis de jure.
(246)
The two forms of reason do not exist independently of each other, as different forms of reality, but exist in relation to each other in an economy comparable to Derrida's differential ' stricture '.
Unknown to Lvi-Strauss, Sartre himself had already proposed that dialectical understanding itself works along two paths, one of ' expansion ' and one of ' compression ', which suggests that having separated dialectical from analytical reason, he had then to reintroduce the latter under a new guise.
The argument about the relation of the different kinds of reason to each other will doubtless (and necessarily) continue without resolution; what is important in the present context is that, in rearticulating the possibility that there is no absolute difference between the two, Lvi-Strauss puts into question Sartre's definitions of man, human history, and finally his concept of history in toto  all of which are predicated on the dialectic.
If conflict is the motive force of history, and history a series of progressive dialectical totalizations, then Sartre claims that what he calls' backward societies' have existed in a state of equilibrium and are therefore without history.
He writes:
There is no logical (dialectical) absurdity in the idea of a country with no History, where human groups would vegetate and never break out of a cycle of repetition, producing their lives with primitive techniques and instruments and knowing absolutely nothing of one another
(I, 126).
In the background here is Hegel.
According to Sartre, history is only born from a sudden imbalance of scarcity which disrupts all levels of society and initiates conflict and therefore progression.
Many have objected that this analysis is hardly Marxist  insofar as Marxism takes the form of an analysis of the misappropriation of surplus value.
But Lvi-Strauss rather objects to the ethnocentrism of Sartre's argument, for it means that ' man ', the constituents of whom should have been the result and product of Sartre's anthropology, is defined in advance in terms of the dialectic as historical humanity, with history effectively restricted to societies of the West.
The rest are excluded.
It is exactly this assumption, however, that shows the extent to which Western society is indistinguishable from all other cultures, for each, according to Lvi-Strauss, has always assumed that it represents the full meaning and significance of human society:
a good deal of egocentricity and naivety is necessary to believe that man has taken refuge in a single one of the historical or geographical modes of his existence, when the truth about man resides in the system of their differences and common properties.
(249)
Sartre's ethnocentricity derives from the whole project of his existentialism and his phenomenological definition of man in terms of the experiencing self defined against an other.
Lvi-Strauss suggests that as a science anthropology should be attempting not to provide a definition of man as he is known experientially in our own society  an unscientific subjectivism  but should rather begin by ' dissolving ' him.
Here we encounter the first attempt to undo the category of ' man ', an enterprise which has caused more distress than most in the recent history of the social sciences.
Lvi-Strauss' point here, however, is simply the objection that Sartre defines' man ' in advance, predetermined by the particular experience of what it is to be a man in twentieth-century post-war French society.
The assumption that one's own experience, one's own gender, one's own society, constitutes the sole reality is in fact the very antithesis of the whole concept of any anthropology.
Sartre, as Simon Clarke argues, ' takes the conscious rationalization of his own culture for the ultimate meaning of humanity '.
But how can Sartre claim to found a general anthropology when he defines it solely in terms of his own society?
As soon as Lvi-Strauss shows that the experience on which Sartre bases his philosophy is not a universal one, then the general inferences for all humanity that he draws from it can no longer be justified.
Sartre's existential consciousness can not be dehistoricized into a general foundation for a concept of History.
Having questioned the basis of Sartre's definition of man and the human according to the categories of civilization and primitivism, Lvi-Strauss focuses on this relation to history.
It can not, he argues, be tied to the notion of ' man ' as if they were images of each other:
We need only recognize that history is a method with no distinct object corresponding to it to reject the equivalence between the notion of history and the notion of humanity which some have tried to foist on us with the unavowed aim of making historicity the last refuge of a transcendental humanism: as if men could regain the illusion of liberty on the plane of the ' we ' merely by giving up the ' I's that are too obviously wanting in consistency.
(262)
In other words, Sartre's attempt to combine Marxism with existentialist subjectivity can not resolve its difficulties through a shift from economism to history, for history implicitly continues to fulfil the same function anyway.
Here Lvi-Strauss argues that behind the original question from which Sartre began  how can man make history if history makes him?  lurks another: if it is man who makes history, how does' History ' gain its exorbitant status as the desired, unachievable object of Sartre's text?
His emphasis on the primacy of history means that he forms' an almost mystical conception of it '.
Sartre seems to have remembered only half of Marx's and Freud's combined lesson.
They have taught us that man has meaning only on the condition that he view himself as meaningful.
So far I agree with Sartre.
But it must be added that this meaning is never the right one: superstructures are faulty acts which have ' made it ' socially.
Hence it is vain to go to historical consciousness for the truest meaning.
(253)
Historical consciousness is necessarily ideological, even if dialectical, and can never in itself provide the one true meaning of history.
Despite Sartre's reiteration that it is man who makes it, history increasingly assumes its own ontological status in the Critiques.
Above all, Lvi-Strauss questions its status as totalization.
The idea of history as the totalization of totalizations does not work.
How total, he asks in a critique of totalization which Derrida would characterize as his first ' classical ' formulation, can the totalization be?
Can it include every historical fact?
History names the process of constituting historical facts, and, particularly, of their selection.
A history that included everything would amount to chaos:
In so far as history aspires to meaning, it is doomed to select regions, periods, groups of men and individuals in these groups and to make them stand out, as discontinuous figures, against a continuity barely good enough to be used as a backdrop.
A truly total history would cancel itself out  its product would be nought.
(257)
Meaning works through a form of metonymy, distinguishing between elements in terms of significance and insignificance, and that is why it must always be partial in relation to any text it interprets.
(We might compare this to the way in which, in Sartre's later text, the synecdoche of the singular-universal insistently slides into the singularity of the event).
Totalization in the first Critique, Lvi-Strauss suggests, can only create its meaning by selection through such metonymic devices of exclusion, that is, by founding itself on an  ethnocentric  absence.
Historical consciousness dehistoricizes by shifting diachrony into a single synchronic totality:
And so we end up in the paradox of a system which invokes the criterion of historical consciousness as a means for distinguishing the ' primitive ' from the, civilized' but  contrary to its claim  is itself ahistorical.
It offers not a concrete image of history but an abstract schema of men making history of such a kind that it can manifest itself in the trend of their lives as a synchronic totality.
Its position in relation to history is therefore the same as that of primitives to the eternal past: in Sartre's system, history plays exactly the part of a myth.
(254)
History, far from constituting a privileged form of (historical) knowledge, is simply the myth of modern man, and merely amounts to a method of analysis.
For all its stress on history and totalization as process, Sartre's argument is structured through the creation of a form of synchronicity.
Lvi-Strauss demonstrates this through his well-known discussion of the function of the category of time.
Like all models of knowledge, he suggests, history requires a code to analyse its object, and for most historians this code consists in chronology.
The use of chronology in historical writing, or in literary history, gives the illusion that the whole operates by a uniform, continuous progression, a linear series in which each event takes its place.
History is thus a process of a continuous unfolding.
But in fact, Lvi-Strauss argues, the process model of history is an illusion, for even dates do not work like that.
Dates only work by being a member of a class.
In itself, a date tells us nothing: it only takes on meaning when it is part of a series.
However that class does not necessarily correspond to other sets of dates, periods, millennia, ages, etc.
History is a discontinuous set composed of domains of history, each of which is defined by a characteristic frequency and by a differential coding of before and after.
It is no more possible to pass between the dates which compose the different domains than it is to do so between natural and irrational numbers.
Or more precisely: the dates appropriate to each class are irrational to all those of other classes.
(259  60)
So-called historical continuity is therefore often fraudulently constructed out of discontinuous sets which each have different temporalities.
History can neither be total, nor a simple series of facts, nor a continuity.
The fact that it must always involve codification and therefore also interpretation means that ' historical knowledge has no claim to be opposed to other forms of knowledge as a supremely privileged one ' (263).
Historicity is a mode of knowledge for some societies but can not ipso facto claim to be the fundamental basis of knowledge for all of them  a point that was to be re-emphasized by Foucault in The Order of Things.
It is thus far from the case that the search for intelligibility comes to an end with history as such.
Lvi-Strauss therefore makes three major criticisms of Sartre's concept of history.
In the first place, he contests its equation with any anthropological definition of ' man '; in the second, he argues that Sartre's description of history as making up one ' History ' with one meaning is only achieved through the exclusion of all other histories with other meanings: the totalization can only totalize if everything which remains other to it is excluded.
Sartre's definition is therefore tautological: it is one history because it is (only) one history.
Specifically, Sartre creates a single history by excluding all histories except that of the West; his history as totalization can therefore only work through a determined ethnocentricity.
In the third place, even in its own terms, the concept of history as a continuous development, a progressive totalization of totalizations, is dependent on a notion of chronology which assumes a synchronic homogeneous notion of time.
But even historical chronology works through discontinuous sets, the elements in each class being defined only differentially in relation to the others in its class.
This final criticism amounts to the challenge that Sartre's history can only ever be theorized as totalization insofar as it has been conceptualized as a synchronic form.
While the criticism that Lvi-Strauss, structuralism emphasizes the synchronic at the expense of the diachronic has assumed the status of a critical truism, this in fact repeats the substance of his critique of Sartre, namely that the latter attempted to transform history into a space of synchronicity.
Indeed totality in both Critiques presupposes a homogeneous temporality: the singular universal could only ever work by presupposing what Althusser would call an ' essential section '.
Lvi-Strauss himself, like Althusser and Foucault, was preoccupied with the conceptualization of forms of heterogeneous temporality that consistently elude and trouble all theorization of history as a homogeneous diachrony.
Lvi-Strauss' attack, together with Sartre's own theoretical difficulties which have already been charted, was effective enough to be quite devastating to the project of the Critique  which was never completed.
Sartre took four years to respond, and when the reply eventually came even his most sympathetic admirers agreed that it did not succeed in answering the main criticisms.
Lvi-Strauss' objections to Sartre's theory of history on the grounds of its ethnocentrism was certainly to the point, demonstrating that one lesson of Sartre's Herculean attempt to make history truth, and to give it one meaning, was the relation of such history to Western cultural imperialism.
At the same time we should also recall that, as Levinas puts it ' the best thing about philosophy is that it fails'.
To the extent that Sartre's ontology does not succeed in totalizing history, it remains open to its irreducible otherness.
perhaps this is one reason why, despite the Eurocentrism of his intended historical schema, Sartre himself could certainly not be accused of ethnocentrism in his politics; already the first Critique contains substantial analyses of the political and psychological structures of colonialism and racism, even if they have to be, as always, ' dissolved in History ' (I, 716).
But this still allows Sartre to argue, for example, for the legitimacy of the use of violence by colonized peoples against their oppressors, thus anticipating Fanon's dictum that colonization was achieved by violence and must therefore be overcome with it.
Aronson sees Sartre's turn to the cause of resisting colonial oppression as an act of ' self-flagellation ' and a product of politico-theoretical despair, but this passes over the historical and political context in which Sartre, and Les Temps modernes, played a leading role in the opposition to the Algerian war.
In retrospect these concerns now seem a much bolder move, constituting an important political lead which was to be continued in the resistance to the war in South-East Asia in the sixties and early seventies.
But Sartre's courageous intervention against French and other colonialisms could not have a corresponding theoretical impact so long as he retained his historicist Marxist framework.
For his unitary theory of history has the effect of disallowing radical attempts at rewriting or retrieving other histories excluded by the West.
Although it was taken by many as an attack on history as such, it was the critique of Marxist historicism initiated by Althusser that enabled new political possibilities in this direction.
The scientific critique of historicism
Before Sartre had even replied to Levi-Strauss criticisms, Althusser had published for Marx (1965), a work which, together with the later volume Reading Capital (1968), far from defending Sartre's Marxism from the structuralist challenge, completed the move against it, and offered a new interpretation of Marxism from which humanist existentialism and Hegelianism had been resolutely purged, with Sartre's voluntarism replaced by a more mediated form of economism.
It is easy to represent Althusser's intervention purely in terms of structuralism, but to suggest that this was its only significant intellectual context is misleading.
If his work needs to be considered in the context of the contemporary politics of the Communist Party, as several commentators have stressed, it also requires reference to work done in the history of the sciences, particularly that of Gaston Bachelard in the history of physics and chemistry, his pupil Georges Canguilhem in the life sciences, and Jean Cavaills in mathematics.
All four, together with Althusser's pupil, Michel Foucault, worked within an epistemological tradition which was critical of the positivism which, up to that time, had dominated the history of the sciences.
To represent Althusser's work purely as a ' structural Marxism ' therefore passes over the fact that there were distinct intellectual traditions in France whose difference was particularly pronounced in their respective philosophies of history.
Much of what has been considered to be poststructuralism's wild disregard for history can be accounted for by the fact that it was operating within this  largely unknown outside France  anti-empiricist and anti-positivist tradition.
In 1978 Foucault suggested that post-war French philosophy divides according to a line ' which separates a philosophy of experience, meaning and the subject from a philosophy of savoir, rationality and the concept ': in other words, Sartre and Merleau-Ponty, against Cavaills, Bachelard and Canguilhem, or the return-to-Hegel of Kojve in the thirties, versus the return-to-Rant initiated in the philosophy of the sciences by Lon Brunschvicg in the late nineteenth century.
Long before Sartre developed his own form of Marxist-Hegelian history, Bachelard had been working in a critical relation to Hegelian historicism.
What was new in Althusser was that for the first time this epistemological tradition was developed for a Marxism.
Against Sartre's claim to establish Marxism's truth philosophically, Althusser reasoned that if Marxism is a science, then the history of Marxism ought to conform to the kind of history that had been developed for the sciences.
The question then simply became whether it did or not.
Or so it seemed.
The difficulties in which Althusser subsequently became enmeshed were the result of his ignoring Canguilhem's warning that although the history of science takes science for its object, it is not itself a science, and therefore can not claim to be value-free (or, in Marxist terms, non-ideological).
As Gregory Elliott has recently emphasized, although Althusser always presented himself as the figure of the rigours of orthodoxy against the eclecticism of the existentialists, in his own work he was just as catholic, allying Marxism with non-Marxist philosophy, even if it was a history of science to which, he claimed, ' French philosophy owes its renaissance in the last thirty years'.
II BACHELARD AND THE HISTORY OF SCIENCE
Against the Hegelian synthesis of all kinds of history within the same developmental schema, Bachelard argued that the history of science can not be assimilated into the progressive evolutionary form commonly ascribed to other kinds of human history, and nor can it be mapped on a one-to-one basis against the history of its age.
In certain pure sciences, mathematics for instance, although discoveries may enable changes and developments of a material kind, their occurrence can not ipso facto be explained by being related to the allegedly determining political and economic history of their immediate era.
As Levi-Strauss was to argue, different histories have different temporalities: the time scales of the sciences do not work at the same pace as other forms of history: they have their own dynamic, their own rhythm, their own times, sometimes fast, sometimes slow, that do not operate by the ordinary round of the year; Bachelard was fond of pointing out that from a scientific point of view the ten years from 1920 to 1930 were as long an era as the previous five hundred.
This unevenness of development means that there can be no general history of science as such: it is uneven but it can not be combined.
Many sciences moreover share the characteristic that a major discovery means that all other models and theories are simply out of date and have to be discarded.
From the point of view of the present, the past has to be excised.
Contemporary science is able to designate itself, through its revolutionary discoveries, as a liquidation of a past.
Here discoveries are exhibited which send back all recent history to the level of a prehistory.
Here we do indeed find an example of a ' liquidation ' of history which, it will be recalled, is exactly the accusation that Terry Eagleton makes against poststructuralism; but the actual example in Bachelard demonstrates how much more complex the issue turns out to be.
Science's negative relation to the past emerges as one of its most distinctive and significant aspects, one which marks a complete break with the cumulative structures of the arts and social sciences.
It means, in particular, that the temporality of science can not be accommodated to the rhythms of traditional historiography, which has not, however, prevented positivistic historians of science from writing its history solely in terms of precursors and anachronistic anticipations of modern ideas in early thinkers, as if science unrolled smoothly and inevitably from year to year.
The problem with this approach is that it overestimates the extent of narrative continuity in the history of science which, according to Bachelard's examples, works rather by sudden disruptions, discontinuities, and entire reorganizations of its principles.
Science is forever remaking its own history.
As early as 1934 Bachelard had argued that the revolutionary changes in physics, such as relativity theory and microphysics, meant that science itself was currently defined by its reaction against the past, and had become a ' philosophy of the non,  non-Cartesian, non-Euclidean, non-Newtonian, and non-Baconian.
These major transformations can not be mapped onto the model of a continuous history, for its stress on putative anticipations fails to account for the way in which the whole form of knowledge can be transformed and a new understanding created.
Bachelard's work on the formation of scientific disciplines led him to argue that the proper form of historical analysis should focus not upon an empirical history but upon the cognitive or epistemological status of concepts that distinguish a new science from an old one.
Those concepts have not evolved from the old ones, for it is precisely their radical difference that constitutes the new science, the new ' positivity ', produced by what he termed an ' epistemological rupture '.
Bachelard himself preferred to give examples of such ruptures rather than theorize how they took place.
Althusserians subsequently placed much emphasis upon the ' epistemological break ', and were in turn castigated for being unable to explain how it occurred.
But this neglects the force of Althusser's emphasis on Marxism as itself a theoretical practice with its own history of epistemological self-correction, a possibility derived from the work of the mathematician Jean Cavaills, who stressed the degree to which the history of mathematics, particularly set theory, could be accounted for by the dialectical development of the concept.
Much of the emphasis on set theory in Lacan and others, as well as the similarity of certain of their ideas to those of G?del, whose work Cavaills utilizes, can be attributed to his influence: indeed it would be possible to argue that the whole emphasis in post-war French thinkers on a non-contradictory heterogeneity in which incompatible or incommensurable elements are juxtaposed against or as part of each other is derived as much from set theory as from Freud.
Cavaills developed these ideas into a theory of science as such, which, he argued, changed not through empirical discovery but through the theoretical reworking of its own concepts in the ' pure ' sciences.
This view, whereby science progresses through the dialectic of its concepts rather than by testing its hypotheses against ' experience ', was developed significantly by both Bachelard and Althusser.
If it enabled the latter the crucial theoretical move of being able to reject the classical empiricist conception of knowledge, it was also to put him in the position of even castigating as' historicist ' any attempts to account for theoretical discourse in terms of its historical conditions of production  perhaps one of the major ways in which he differed from Canguilhem and Foucault.
Bachelard and Cavaills agree that the distinguishing characteristic of modern science is the degree to which it has become separated from common-sense knowledge so that consciousness and its concepts are now opposed:
The break between ordinary and scientific knowledge seems to us so clear that these two types of knowledge could not have the same philosophy.
Empiricism is the philosophy which corresponds to ordinary knowledge.
There empiricism finds its origin, its evidence, its development.
By contrast, scientific knowledge is bound up with rationalism and, whether one wishes it or not, rationalism is allied to science, and demands scientific goals.
The use of scientific instruments in particular means that scientific perception is constantly at odds with the experience of everyday perception.
Because scientificity is achieved through a break with common-sense forms of thinking, termed ' epistemological obstacles', Bachelard, like Lvi-Strauss and Althusser after him, argues that any philosophy such as existentialism that is founded on the basis of the truth of the experience of the knowing subject is bound to involve illusory or ideological forms of thought.
For this reason, Bachelard refers ironically to Sartre's phenomenology as a belated form of alchemy.
But if the new science is produced through a rupture with the ' errors' of the old, the tenacious hold of common-sense forms of thought means that any given text may simultaneously embody aspects of the old and new ways of thinking, theoretical and ideological frameworks that Bachelard, and Althusser after him, term ' problematics':
Instead of the parade of universal doubt, scientific research requires the establishment of a problematic.
It takes its real departure from a problem, even if it is badly set up.
The scientific I is then a programme of experiments, whereas the non-scientific I is already a constituted problematic.
Bachelard tries to think through the problem of how epistemological obstacles operated both before and after scientificity.
In particular, he suggests that the educational system has a marked effect on the production and reproduction of scientific knowledge, and criticizes it for the ahistorical way in which it teaches scientific problems, theories, experiments and proofs.
While pedagogy continues the myth of an elementary or easy science, language itself can also produce difficulties.
The nomenclature of science does not refer to definitive concepts:
It is ceaselessly adjusted, completed, varied.
The language of science is in a state of permanent semantic revolution.
Obstacles such as these lead Bachelard to formulate a theory of ' material psychoanalysis' which offers a ' psychoanalysis of objective knowledge ' to account for and think through the problem of epistemological obstacles.
It is also supposed to have a therapeutic effect of a ' brutal, surgical ' separation of unconscious and rational convictions in order ' to cure us of our images or at least to limit their power '.
This separation between scientific and common-sense knowledge in turn produces a significant effect on Bachelard's thinking about history.
For, as Lecourt describes it, he attempts
to elaborate a system of concepts which will make it possible to think the intrication of two histories: that of the scientific and the non-scientific in the practice of the scientists.
Hence this project culminates in Le rationalisme appliqu with the project of an epistemological history which is presented as a dual history; a ' ratified history ' (or history of the scientific in scientific practice) and a ' lapsed history ' (or history of the interventions of the non-scientific in scientific practice).
This means that science can have two histories, which constantly intertwine but never resolve, with one evaluated as positive, the other as negative and therefore silently suppressed even though it may remain determining.
Instead of upholding one at the expense of the other, Bachelard offers the possibility of a deconstructive history which would reinscribe that which had been excluded; this could also enable a differential history of science and ideology, accounting for the perpetuation of ideology after the production of science.
To do so, Bachelard, as Althusser was to do later, invokes the use of psychoanalysis for the study of history and ideology.
The notion of the epistemological break, while offering a theory of scientificity, and even of ideology, implied a very different view of history from that of Sartre.
In some sense, it went to the other extreme, for from a totalizing history it projected a form of history in which there was no attempt to link different histories at all.
Bachelard assumed the necessary division of the scientific from the non-scientific, even if de facto he was continually encountering their imbrication and finding himself in the position of trying to keep them apart.
Lecourt has pointed to the absence of ' a concept that would enable him to think together several histories with different statuses; in short, the concept of a differential history '.
Such a general concept was to be provided by Althusser's theory of relative autonomy within a structure in dominance.
III ALTHUSSER AND THE SCIENCE OF HISTORY
Althusser's rereading of Marx can thus be as usefully considered in the context of theories of the history of science as of structuralism; a certain conflation between the two has been possible because both were opposed to humanistic and phenomenological theories of knowledge in general and to historicism in particular.
Within the general framework of his attack on the humanistic Hegelian tradition of Western Marxism, Althusser's specific objection to Sartre's attempt to mediate Marxism with existential subjectivity was that such a move went against the crucial discoveries which had founded Marxism in the first place; in an extension of Lvi-Strauss' argument, he maintained that the notion of ' man ' that Sartre used was derived from a particular ideological definition of the human subject which represses Marx's insight that the human subject is not the centre of history, together with Freud's that the subject is not centred in consciousness.
For Althusser both history and the subject are equally decentred: his attack on Sartre's claims for a unitary history (however complex its textual elaboration may have proved to be) was therefore accompanied by a critique of the notion of the unitary human subject that constitutes it.
One way of characterizing Althusser's intervention would be Martin Jay's observation that he effectively destroyed the Lukcsian notion of totality.
Another would be to say he attempted to theorize the very position that Sartre had fought to get out of.
If Sartre's argument depended on a logic of history as totalization but broke down when he could not combine the praxis of the individual with the general logic of ' totalization without a totalizer ' except through the proliferation of his own writing, Althusser, by contrast, exploited the possibility of history as a ' process without a subject ', a history characterized by radical breaks and discontinuities, distinct from each other and not totalizing.
It was this aspect of Althusserian theory, perhaps more than any other, that led many orthodox Marxists to consider that Althusser was not really a Marxist at all.
The accusation that poststructuralism neglects history undoubtedly harks back above all to the work of Althusser who, more than anyone else, appears to have attempted to eliminate history.
Notwithstanding their philosophical and political differences, there are nevertheless certain similarities between Althusser's and Sartre's projects when viewed in relation to orthodox Marxism.
Althusser shared Sartre's opposition to Stalinism's emphasis on economism and technical determinism, dissociating himself from it not through an assertion of individual agency but through a reformulation of the Marxist thesis of determination by economic relations  redefined as a causal rather than a historical relation.
At the same time Althusser argued that Sartre had not isolated the central problems of orthodox Marxist theory, and as a result continued to work with some of its more questionable preconceptions.
In particular, in spite of his attempt to avoid positing history as an a priori transcendent law, in the published first volume of the Critique he had still utilized an organicist teleological model of history which assumes that the end is already implicit in the beginning, and that history rolls forward to a determined end.
Both Stalinism and Sartreanism assume that history is an emancipatory process of self-realization, even if the forces of production in the one are replaced by the praxis of the self-conscious human subject in the other.
Althusser termed such a view ' historicism ': an abstract philosophical scheme that imposes an overall process of transformation upon historical events.
At first glance, it might seem as if Althusser himself gave up history and tried science instead.
Like Sartre, he sought to constitute Marxism as a form of truth, but attempted to prove its truth not through the dialectic of history but rather as a science, authenticating Marx's ' immense theoretical revolution ' epistemologically through a demonstration of its scientificity.
In order to claim a scientific status for Marxism as knowledge rather than ideology, or non-knowledge, Althusser invoked Bachelard's historical epistemology which allowed him to posit the idea of a radical discontinuity between the two, with Marxist science separated from earlier forms of non-knowledge in Marx's texts by an ' epistemological break '.
This necessarily meant that Althusser endorsed Bachelard's arguments about empiricism and rejected the concept of history as a system of progression or evolution.
However Althusser found it difficult to maintain a Marxist theory of history while avoiding its customary Hegelian form.
His basic argument, that Marx proposed a new conception of knowledge defined against Hegelianism, implied an accompanying revision of the Hegelian concept of history, which, as we have seen in the cases of Lukcs and Sartre, had hitherto provided the dominant model of Marxist historicism.
Althusser's theory of history has been more widely attacked and denigrated than any other aspect of his work, largely because he dared to argue that, far from providing the unassailable foundation of Marxism, history was a problematic concept even in Marx's own texts:
this apparently full word is in fact theoretically an empty word, in the immediacy of its obviousness  or rather, it is the ideology-fulfilment which surfaces in this lapse of rigour.
Anyone who reads Capital without posing the critical question of its object sees no malice in this word that ' speaks' to him: he happily continues the discourse whose first word this word may be, the ideological discourse of history, and then the historicist discourse.
As we have seen and as we understand, the theoretical and practical consequences are not so innocent.
(143)
Althusser argues that Marx's intervention did not merely amount to the historicization of the formal categories of the classical economists.
If this was the case all he would have done would have been to Hegelianize Ricardo.
Marxists often imply that Marxism simply involves the introduction of a historical framework  ' always historicize! '  but this assumption can only be made, according to Althusser, because of ' the confusion that surrounds the concept of history ':
In reality, it is to introduce as a solution a concept which itself poses a theoretical problem, for as it is adopted and understood it is an uncriticized concept, a concept which, like all ' obvious' concepts, threatens to have for theoretical content no more than the function that the existing or dominant ideology defines for it.
It is to introduce as a theoretical solution a concept whose status has not been examined, and which, far from being a solution, is in reality a theoretical problem.
(93)
Historicism presumes that a concept of history can be borrowed for Marxism from Hegel or from the practice of empiricist historians without difficulty and without asking how such a concept is specific to Marxism: for Althusser the crucial question is to ask ' what must be the content of the concept of history imposed by Marx's theoretical problematic? '
That problematic can not be identical to Hegel's because both Hegel and Marx define historical time in terms of the social totality: insofar as their definitions of the social totality can be shown to differ, so also will their concepts of historical time.
For Flegel ' historical time is merely the reflection in the continuity of time of the internal essence of the historical totality incarnating a moment of the development of the concept ' (93).
Historical time is the existence of the essence of the social totality and will therefore indicate its structure.
Althusser isolates two key characteristics of Hegelian historical time, ' its homogeneous continuity and its contemporaneity '.
The first is well known: here time is the ' continuum in which the dialectical continuity of the process of the development of the Idea is manifest ' (94).
In this schema the science of history consists of the problem of the division of this continuum into the periods that constitute successive dialectical totalities.
The second characteristic, the contemporaneity of historical time, that is, the category of the historical present, is more complex but an essential element in the whole Hegelian model of history, indeed is its condition of possibility.
If historical time is the existence of the social totality then the relation between the two must be one of immediacy, allowing what Althusser calls an ' essential section ', that is' a break in the present such that all the elements of the whole revealed by this section are in an immediate relationship with one another, a relationship that immediately expresses their internal essence '.
This section is possible precisely because the unity of the whole is an expressive totality, that is' a totality all of whose parts are so many ' total parts' each expressing the others, and each expressing the social totality that contains them, because each in itself contains in the immediate form of its expression the essence of the totality itself ' (94).
Hegel's conception of historical time, then, reflects his conception of the intrinsic unity between all parts of the social totality, each a part of the whole and the whole present in each part, so that history too partakes of a self-reflective immediacy which paradoxically makes it ahistorical.
Perhaps surprisingly, given his alleged ' structuralism ', Althusser argues that the structuralist distinction between synchrony and diachrony rests upon this Hegelian version of historical time which is both continuous and contemporaneous with itself.
Against this Althusser maintains that time, even chronological time, is not just an empirical entity, but a concept; that though history may be articulated in general with chronological time, each history has its own temporality, which can only be found by establishing the conceptual nexus of the history in question.
In order to determine what an event is, we must know the concept of the history in which the event is to occur.
For example, in order to understand the history of physics, we have to know the concept or problematic of physics in order to establish what an event in physics consists of.
Of course a chronological narrative can be constructed, but the history of physics in fact has its own temporality in which the first event after Aristotle was Newton, the second was Einstein, the third black holes, etc.
Now this specific temporality will have no homogeneous relation to, say, the history of literature.
However, it can be articulated with other histories, indeed is articulated, according to the overall but decentred totality of the particular mode of production.
Sometimes Althusser seems to imply that different histories may range through different modes of production, at other times it appears that they are specific to each, an effect of the overdetermination of the social formation.
But in either case, unlike the Hegelian essential section, where each event can be shown to be in an essential articulation with the whole in a continuous and homogeneous spatio-temporality, a cross-section at any particular moment will show a heterogeneous array of presences and absences.
Althusser elaborates his thesis that Marxism is not a historicism at some length, presenting a critique of the historicist and humanist traditions which he takes back from Sartre to the beginning of the century, even to the Russian Revolution itself, and in which he also includes the ' absolute historicism ' of Gramsci and the Frankfurt School.
Above all, he criticizes the ways in which Hegel reduces the diverse historical totality of a society to a single internal principle, so that history occurs only through the principle of contradiction in the dialectic.
For Althusser, the apparently simple contradiction is always overdetermined: the inversion of Hegel's single principle to a dialectic generating successive modes of production amounts to economism.
This objection holds equally for Sartre's Hegelian historicism, the effect of which is to reduce the multiplicity of different practices to a single practice, ' real ' history ' (136).
Sartre's historicism shares the common tendency of all historicist interpretations of Marxism which transform ' the Marxist totality into a variant of the Hegelian totality ': they all share the structure of the contemporaneity of a temporal presence and continuity which allows the possibility of an essential section.
paradoxically therefore, Althusser argues, humanist Marxism shares the same basic theoretical principles with the orthodox economist Marxism of the Second International which politically it was its aim to oppose: for whether passive or active, fatalist or voluntarist, both reduce Marxist analysis to a single theoretical problematic.
All that has happened is that the relations of production have been turned into historicized human relations.
IV THE LONELY HOUR
Althusser's procedure has been to show that, within a notion of history that seemed as if it could be invoked on its own as self-evident, there rests an entire presupposition about the conception of the social whole that is not derived from Marxist theory.
The Marxist concept of historical time must instead, he argues, be thought through ' on the basis of the Marxist conception of the social totality ' (97).
That Marxist conception is, of course, the Althusserian one: in Althusser's reading of Marx the unity of the whole is precisely not that of the Hegelian  and Sartrean  expressive totality; rather it is constituted through overdetermination, through:
a certain type of complexity, the unity of a structured whole containing what can be called levels or instances which are distinct and ' relatively autonomous', and coexist within this complex structural unity, articulated with one another according to specific determinations, fixed in the last instance by the level or instance of the economy.
(97)
The Hegelian model of the coexistence of presence which allows the possibility of the ' essential section ' is incompatible with this description, but Althusser continues to assume nevertheless that there is a totality and that it has' the structure of an organic hierarchised whole ' (98).
This structure, made up of a succession of different levels or instances, is dominated by one form of production which forces the unity of any conjuncture, the non-economic structures determined ' in the last instance ' by the economic (99).
This last description is often misunderstood: the point is that the economic is never a simple causal function that operates alone:
the economic dialectic is never active in the pure state; in History, these instances, the superstructures, etc.  are never seen to step respectfully aside when their work is done or, when the Time comes, as his pure phenomena, to scatter before His Majesty the Economy as he strides along the royal road of the Dialectic.
From the first moment to the last, the lonely hour of the ' last instance ' never comes.
The significance of the allusion to Freud in this famous passage is to suggest that to conceive of the economic as operating in isolation is as illusory as to imagine that the ego can operate without the unconscious: they are both the reciprocal products of the other.
The ' last instance ', the economic, never operates in isolation separate from all the other instances of the social totality: it is the lonely hour that never comes.
If one were to pursue the analogy with Freud rigorously, Althusser is even suggesting here that the primacy of the economic (ego) is a delusion, and that the superstructure (unconscious) is the more fundamental determining force, or at the very least that they are equally overdetermined.
Whereas Sartre's totality was never totalized because it was always still in process and could never be closed, Althusser's totality is never totalizable because it is decentred and displaced in time.
The different structured levels do not coexist in a temporal present which coincides with ' the presence of the essence with its phenomena '.
This means that the model of continuous and homogeneous time which Lvi-Strauss had also argued against can not here be regarded as the time of history.
This has nothing to do with turning diachrony into synchrony; the point is rather that
it is no longer possible to think the process of the development of the different levels of the whole in the same historical time.
Each of these different ' levels' does not have the same type of historical existence.
On the contrary, we have to assign to each level a peculiar time, relatively autonomous and hence relatively independent, even in its dependence, of the ' times' of the other levels...
Each of these peculiar histories is punctuated with peculiar rhythms and can only be known on condition that we have defined the concept of the specificity of its historical temporality and its punctuations (continuous development, revolutions, breaks, etc.).
(99C100)
These histories, their temporalities defined according to the specific concepts of particular domains, are not, however, independent of the whole: they are dependent on it, but in a structure derived from the ' differential relations between the different levels within the whole... the mode and degree of independence of each time and history is therefore necessarily determined by the mode and degree of dependence of each level within the set of articulations of the whole ' (100).
So if there are different histories, they must nevertheless be related to those other histories from which they differ but with which they, articulate in a structure of relative effectivity.
Such histories are constituted according to a Saussurian differential relation.
Althusser therefore criticizes the Annales historians for merely arguing that periodizations differ for different times, and that each time has its own rhythms.
This is not enough, for it is also necessary to ' think these differences in rhythm and punctuation in their foundation, in their type of articulation, displacement and torsion which harmonizes these different times with one another '  though it must be added that this begs the question of how such harmonization is achieved.
For such times are not even necessarily the obvious ones, ' the visible sequences of events recorded by the chronicler ', they may be invisible, ' a complex ' intersection ' of... different times, rhythms, turnovers, etc. ', only visible when their particular concepts are constructed and produced ' out of the differential nature and differential articulation of their objects in the structure of the whole ' (101C03).
As an example of what he means, Althusser refers to Foucault 's, remarkable studies', Madness and Civilization and The Birth of the Clinic, two instances in which the historian has had to construct the concept of their history.
This is antipodal to the empirically visible history in which the time of all histories is the simple time of continuity and in which the ' content ' is the vacuity of events that occur in it which one later tries to determine with dividing procedures in order to ' periodise ' that continuity.
Instead of these categories, continuity and discontinuity, which summarize the banal mystery of all history, we are dealing with infinitely more complex categories specific to each type of history, categories in which new logics come into play.
(103)
This means that if one tries to take an ' essential section ' there is no essence revealed which is the present of each level; indeed the break valid for one history would not necessarily correspond to that valid for any other which will live in a different time and in a different rhythm.
The present of one level is, so to speak, the absence of another, and this co-existence of a ' presence ' and absences is simply the effect of the structure of the whole in its articulated decentricity.
(104)
Any determinate mode of production, therefore, will evidence such a form of historical existence in its social formation: dislocated, uneven, absent and present.
There is, Althusser stresses, no ' single ideological base time, to which all these different temporalities can be related, no ordinary ' single continuous reference time ' which they can be seen to dislocate (105).
This differential account of history requires us to rethink a whole series of common notions such as unevenness of development, of survivals, backwardness, even, Althusser claims, the contemporary economic practice of under-development'  notions that provide the very basis of Western ethnocentrism.
It also means that there can be no history in general, only specific structures of historicity.
Nor is history an evolving totality; each mode of production is made up of differentiated histories.
These differentiated histories form a specific historical totality, for each history operates within the general totality of the mode of production.
The economic therefore determines each history or level to the extent that its history is structured differentially against the totality which is defined as a specific mode of production.
If for Hegel historical time is the reflection in time of the essence of the historical totality, for Althusser it is a function of the structure of the totality arising from a particular mode of production.
No totality has a necessary transcendence embodied within it; as for the later Sartre, the course of historical change is open and will work only through the overdetermination of particular historical conjunctures.
Althusser suggests that although Marxist history is defined as a theory of the modes of production, Marx did not give us any theory of how the transition was effected from one mode of production to another, nor of how each mode of production was constituted.
Much of the effort of Althusserian Marxism was taken up with trying to produce such a theory, and it was the failure to produce it that perhaps was the main reason for its subsequent collapse.
Althusser's influence declined in France after 1968, partly as a result of the role played by the PCF, of which Althusser was a member, in the events of May 1968, and partly as an effect of a number of critiques of his work, some of them by Althusser himself.
In Britain by contrast Althusser's greatest impact occurred in the decade that followed: from the late sixties his work constituted something like a hegemonic ' theory in dominance '.
After 1978, however, the influence of Althusser declined rapidly after a series of books by two Marxist sociologists, Barry Hindess and Paul Hirst, who developed and extended the implications of those critiques already made in France.
Their work made it impossible henceforth to invoke the work of Althusser in Britain without reference to the problems which they had articulated.
With Hindess and Hirst criticism of Althusser's theory of history shifted to the problematic of the relation of the concept (increasingly assimilated to representation) to the real, thus reducing it to a question of epistemology.
Where, though, does this leave the concept of history in Althusser?
Does it have its own specificity aside from the problems of the subject, representation, narrative and interpretation, with which many have now shown that it is necessarily involved?
Does this mean, as Hirst suggests, that the only concept of history must in fact still be the Hegelian one?:
Teleology and spirituality are essential mechanisms of all philosophies of history.
And, save for mindless antiquarianism or utter scepticism, there can be no history without a philosophy of history.
It is in the philosophy of history that the past becomes a possible and rational object of knowledge.
It is through the conception of historical time as a continuum that the past becomes a coherent object.
For Hirst this means that there can be no Marxist ' science of history ' that can be opposed to the essentialism and teleology of a philosophy of history.
Althusser's mistake, according to Hirst, was to attempt to construct another philosophy of history; but any theory of history, he contends, can not do without teleology, spirituality (the realization of the Idea), and the continuum.
On the other hand, we would add that the lesson of Hegel 's, Marx 's, and Sartre's attempt also suggests that history can not be coherently essentialist and teleological either.
Hirst's straightforward characterization of Althusser's work as' a failure ' does not acknowledge the constant tension in the historical project itself.
If Sartre's endeavour to ground the Marxist science of history suggests that the Hegelian totality and continuum can only work by a continual labour of excluding the partial and discontinuous, Althusser's effort to constitute a differentiated history shows that you can not do it without a teleology.
What both demonstrate is that any history as such has to think both, simultaneously.
As long as history is assumed to operate according to the protocols of a conventional logic, where a contradiction simply means you can not think it or do it as Hirst supposes, then it remains at the impasse he describes.
Althusser's significant contribution was to problematize the concept of history by addressing its presuppositions about temporality  an area which Hindess and Hirst altogether neglect.
In order to make their claim that his decentred totality is still expressive and therefore essentialist, they have to ignore the arguments about temporality in the critique of the Hegelian essential section as' the co-existence of presence ', and thus fail to do justice to the way in which Althusser constructs, as Foucault puts it, ' a counter-memory  a transformation of history into a totally different form of time '.
If the Althusserian mode of production is made up of differential times and histories, ' a complex ' intersection ' of the different times, rhythms, turnovers, etc. ', then each element can not express the whole because the whole is only accessible as a concept, which is precisely not expressed at all.
For the concept, ' like every concept, is never immediately ' given ', never legible in visible reality: like every concept this concept must be produced, constructed '  by the analyst.
This formulation enabled Althusser to theorize a decentred totality which allowed the possibility of differences without reducing each instance to the operation of an essence or a single principle, such as the dialectic.
If his notion of the mode of production as such a totality could not be sustained in a differential relation to other modes of production (notions of residual and emergent forms notwithstanding), Althusser nevertheless offered a particularly interesting theorization of the problems involved in the concept of the historical, articulating the paradoxical conditions of any theorization of history.
If he showed it to be an impossible concept, he continued to demonstrate that it remains a necessary one, and acknowledged that we must learn to live with that impossibility.
How does this take place?
It is at this point that the subject of history re-enters; history may be a process without one, but the subject is nevertheless inscribed within history.
Once again, as for Sartre, the subject operates at the pivot of the paradox.
For the subject does not understand history according to its scientific formulation, but undergoes the process of inter-pellation at the level of ideology, and thus experiences it through the formulas of historicism.
History, which can now no longer be considered a concept as such, is therefore made up of the incommensurable relation between these two disjunctive set-ups.
It therefore operates for Althusser both at the level of science and of ideology, not in terms of truth to falsity, but as an irresolvable dialectic between the differential relations of the mode of production and the historicism of the ideological notion of history.
Hence his argument that ' the knowledge of history is no more historical than the knowledge of sugar is sweet '.
What then is the relation of science to ideology, of Althusserian history to its ideological historicist formulation?
The gap between them, according to Althusser, is mediated by art.
This statement is often regarded as a curious relic of the values of bourgeois culture.
But it is also possible to see it as an attempt to formulate the way in which the sliding incompatibility of the two can only be perceived through an ' internal distantiation ' in which the problem of that ' relation ' is enacted by its relation, in the sense of the telling of a story  which is how we get history.
If history is a process, Althusser remarks, ' there is no such thing as a process except in relations [ sous des rapports ] '.
History is a matter of relations, and thus of writing reports.
Althusser thus suggests that history can only be thought through as a permanent contradiction: it is a totality, but that totality is a decentred structure in dominance in which each history's history is defined not through its identity with, or difference from, a general history but by being differentiated from every other history, on which it is necessarily also therefore dependent, in a kind of negative totalization.
Within the realm of ideology, on the other hand, history is experienced by subjects as a purposive continuum, with themselves as its subject.
Another way of putting this would be to say that Althusser demonstrated that according to the protocols of conventional logic, history is impossible.
If you try to think of it as a closed totality you get into the problems of historicism; if you try to think of it as entirely differentiated, then it becomes meaningless since there is no necessary connection, positive or negative, to anything else, nor would any one history produce any effect on another.
Any differential theory of identity must think both totality and difference simultaneously.
It is this structure which both Sartre's ' singular universal, and Althusser's relative autonomy within a structure in dominance attempt to formulate, and in both cases the in-between of such simultaneity emerges as the process of writing itself.
V ' IF THERE IS HISTORY ': HISTORY, HERMENEUTICS AND HISTORICITY
With Sartre and Althusser we encounter the two poles of post-war French return-to-Marxism.
Although both could be said to be reacting against the ossification of Stalinism, the subjectivist and ultra-objectivist paths that they followed were antithetical.
This schematic characterization, however, does not do justice either to the manner in which they attempted to keep both poles in play at once, nor to the way in which they came to concede the impossibility of the theoretical projects which they undertook.
To dismiss them as failures on this account, however, would be merely reductive.
Their real force can be discerned from the extent to which the problematics they set up have continued to exercise subsequent writers.
Since Sartre and Althusser no one has attempted a new theorization of a Marxist history.
We are not, therefore, now looking at another paradigm, another system.
Rather than attempting to repeat and surpass the exalted projects of Sartre or Althusser in a new guise, later writers such as Foucault or Derrida learnt a lesson from history, and stopped to ask why such comprehensive theories, like so many of those which preceded them, could not hold together.
For the grand narrative of History was always too big for its boots.
Instead, therefore, of excluding history as has so often been claimed, such thinkers begin, in a rather more conventional way, from the anti-historicist perspectives of Althusser, even, arguably, of Sartre.
The generic label ' poststructuralist ' is here useful merely as a shorthand to designate those contemporary writers who share not a hostility to history as such but a distrust of simple historicisms.
Lyotard, for example, is best known for his scepticism towards historicist universal narratives, advocating instead the possibility of a multiplicity of heterogeneous, conflicting and incommensurable histories.
Foucault works from exactly the same tradition in the philosophy and history of science as Althusser: like him he utilizes Bachelard's concept of differentiated histories, which still stands as the major alternative, epistemic or otherwise, to historicism.
If such a theorization of history, as diacritical and singular, articulated according to breaks and ruptures, was by no means new even with Althusser, it has been continued as a self-conscious derivation from him by Derrida.
In Positions Derrida comments:
Althusser's entire, and necessary, critique of the ' Hegelian ' concept of history and of the notion of an expressive totality, etc., aims at showing that there is not one single history, a general history, but rather histories different in their type, rhythm, mode of inscription  intervallic, differentiated histories.
I have always subscribed to this.
Contrary to the claims of certain American commentators, Derrida's uncharacteristically emphatic endorsement here of Althusser's project with regard to the possibility of ' intervallic, differentiated ' histories, suggests the problems involved in invoking Althusser as an ' answer ' to Derrida; both are equally distrustful of any form of Hegelian historicism.
This does not, however, mean that they then take up identical positions in relation to the problem of history.
Derrida has not been concerned to formulate a new philosophy of history; nor has he attempted to specify new methodologies in the manner of Foucault.
Describing himself as' very wary of the concept of history ', Derrida has rather attempted to shift the problem away from the conceptual analysis of history as an ' Idea ' which, perhaps more than anything else, has proved to be Hegel's most enduring legacy, towards an analysis of the interstices in the implications of the general system in which it operates.
In the first instance, therefore, he focuses not on history as such but on the related problem of hermeneutics and historical understanding: history here becomes a problem of meaning and interpretation  ' the age already in the past is in fact constituted in every respect as a text '.
This also involves an inquiry into the ways in which the critique of the sign affects historical representation, and the dependence of history on the genetic metaphor and on narrative.
Important though these issues are, Derrida's major contribution has been his insistence that history is a metaphysical concept according to which the meaning of history always amounts to the history of meaning.
As he himself puts it:
From the first texts I published, I have attempted to systematize a deconstructive critique precisely against the authority of meaning, as the transcendental signified or as telos, in other words, history determined in the last analysis as the history of meaning, history in its logocentric, metaphysical, idealist... representation.
The Derridean critique of logocentrism necessarily includes the concept of history insofar as it depends on notions of presence and meaning determined as truth.
For all its frequent invocation as the ' concrete ', history must by definition entail a problematic represencing of an absence; Derrida therefore argues that, even in its' materialist ' conceptualization, it can not avoid a certain metaphysics.
As early as the 1966 essay ' Structure, Sign, and Play ', for example, he maintains that his analysis of the paradoxical metaphysics of ' the centre ' is equally applicable to the historical notions of ' origin' and ' end '.
History consists of
a concept which has always been in complicity with a teleological and eschatological metaphysics, in other words, paradoxically, in complicity with that philosophy of presence to which it was believed history could be opposed.
The thematic of historicity... has always been required by the determination of Being as presence...
History has always been conceived as the movement of a resumption of history, as a detour between two presences.
Derrida's deconstruction of the notion of presence by the logics of the ' always already ' and ' originary repetition ' inevitably conflict with a history constructed in terms of a teleological movement from an origin, which can always be reawakened, towards the self-realization of an idea ' whose end may always be anticipated in the form of presence '.
The meaning of such a history will always be determined through appeal to a transcendental signified, whereas a deconstructive analysis will demonstrate the simultaneous effects of the lack of such an authority.
More radically still, Derrida works at the limits of any possible philosophy of history, arguing that it is not just that the problems of hermeneutics, specifically of interpretation and language, affect historical understanding, but that what in a broad sense he calls writing, or diffrance, determines history.
At the very opening of Of Grammatology, for example, he sets out the thesis that writing constitutes the condition of emergence for all forms of historicity as such:
Historicity itself is tied to the possibility of writing...
Before being the object of a history  of a historical science  writing opens the field of history  of historical becoming.
' Language as the origin of history ': on what basis does Derrida make the startling and radical claim that writing as transcendence constitutes the condition of all historicity?
In order to try to elucidate this controversial argument, we can return to the well-known remark in the essay ' Diffrance ':
if the word ' history ' did not carry with it the theme of a final repression of difference, we could say that differences alone could be ' historical, through and through and from the start.
Here we revert to the dialectic of the same and the other: as Derrida puts it, ' that the same... is never the identical, means first that Being is history '.
It is only through difference, by which the same becomes other and produces a tissue of differences, that history could ever take place: for if full presence were possible, then there would be no difference, and therefore no time, space  or history.
Diffrance means precisely that you can never get out of  and therefore have no need to get back to  history.
It also means that if difference in its sense of non-identity sets up the possibility of history, then difference in its sense of delay means also that it can never be finally concluded, for such deferral will always inhibit closure.
It is in this sense that Derrida argues that Husserl's Origin of Geometry sets up ' the possibility of history as the possibility of language ' whereby ' difference would be transcendental ': writing, in the general significance which Derrida gives it of a differential marking, must be the condition of any historicity.
' History as diffrance then means that history will itself always be subject to the operations of diffrance, and that diffrance names the form of its historicity.
The same conditions hold for totalization.
Derrida argues that though history is given the form of a totality by Hegel, his Aufhebung shows that in order to achieve that totality it must constantly transcend itself in a movement of excess.
History
is the history of the departures from totality, history as the very movement of transcendence, of the excess over the totality without which no totality would appear as such.
History is not the totality transcended by eschatology, metaphysics, or speech.
It is transcendence itself.
History, in fact, works by exactly the same structure of supplementarity as Derrida charts in Of Grammatology.
Here Rousseau himself demonstrates in his oscillating interpretations a history that can not be linear, nor operate according to a single temporality, disturbing instead, the time of the line or the line of time ':
We may perceive here the strange workings of the historical process according to Rousseau.
It never varies: beginning with an origin or a centre that divides itself and leaves itself, an historical circle is described, which is degenerative in direction but progressive and compensatory in effect.
On the circumference of that circle are new origins for new circles that accelerate the degeneration by annulling the compensatory effects of the preceding circle, and thereby also making its truth and beneficence appear.
Such a history can not be represented by the movement of a linear progression of unfolding time.
Its' transcendence ' marks the way in which it can similarly never be limited to a finite totality, nor, conversely, to an infinity:.
it exceeds itself.
This process of supplementation, this' overabundance of the signifier ' that always goes beyond itself, is the result of a lack, or absence at the centre or origin, which must always be supplemented.
Thus the argument that ' history ' is what ' poststructuralism ' lacks, itself repeats totalization's own structure of supplementarity according to which history functions both as an excess and a lack in the origin.
Insofar as it sets up such a process of necessary and constant supplementation, we could say that the impossibility of totalization produces a writing-effect whose process of perpetual deferral unremittingly provokes more writing.
It is thus no longer a question of being able to produce a new concept of history, which, as Derrida puts it, ' is difficult, if not impossible, to lift from its teleological or eschatological horizon.
History can not be done away with any more than metaphysics: but its conditions of impossibility are also necessarily its conditions of possibility.
This means, as Rodolphe Gasch observes, that ' the mimicry of totality and of the pretension to systematicity is an inseparable element of deconstruction, one of the very conditions of finding its foothold within the logic being deconstructed '.
Derrida himself, therefore, does not in any sense abjure history (or totality) but rather attempts to reinscribe it by writing histories that set up supplementary figures whose logic simultaneously invokes and works against historical totalities.
From this perspective, even Sartre's Critique looks rather different from the way it appeared in the early sixties when read in the context of Lukcs' History and Class Consciousness.
Today it seems to evince a growing recognition that totalization can not be achieved without a movement involving the transcendence of itself.
In Sartre's terms, a totalization needs a totalizer.
It must always involve an excess beyond the totality without which the totality could never be totalized, which must mean that it can never in fact be closed.
Lyotard points to exactly the same structure in Marx, who could likewise never complete Capital.
This endlessness of perpetual deferral, he argues, is later formalized in ' a tragic political party... the negative dialectic of the Aufkl?rung; it is the Frankfurt School, demythologized, Lutheran, nihilistic Marxism '.
But instead of the perpetual threat of totalization, a spectre which the Frankfurt School anticipates endlessly, Derrida suggests rather that the problem in any structure is rather how it achieves closure.
Even if this appears to have occurred, a deconstructive account will show how such a text had to dissimulate in order to cover over its own openings, or, to put it the other way round, it will show how history must always be organized by an attempted occlusion of its own conditions of historicity.
This means that it is not necessary to reject totalization as such  because such a rejection assumes its very possibility, whereas all attempts at totalization such as Sartre's demonstrate rather its impossibility.
Peter Dews has recently claimed that ' post-structuralism can be understood as the point at which the ' logic of disintegration ' penetrates into the thought which attempts to comprehend it, resulting in a dispersal into a plurality of inconsistent logics'.
So poststructuralism is itself the ' shattered mirror of the logic of disintegration.
But poststructuralism precisely does not try to comprehend disintegration, for grasping heterogeneity together is exactly what all totalizing theories have unsuccessfully attempted to do.
Having elaborated the paradoxical conditions of historicity, of any history or totalization, Derrida himself has been particularly concerned to analyse those such as Husserl, Heidegger or Levinas, who have been involved in investigations of time and temporality.
Although it would be possible to pursue the question of history in terms of such analyses of the forms of historicity, such an enquiry would take us on a very different path from that prompted by our original question, namely if poststructuralism can apparently be faulted by reference to a history which it neglects, where in Marxism can this history be found?
Since Sartre and Althusser there have been a number of possibilities: for some, the absolute historicism of the Frankfurt School has become increasingly attractive, although in its current manifestation in the work of Habermas we might say that history has been eclipsed far more effectively than by any comparable French philosopher.
If Althusser gave us history without a subject, Habermas gives us subjects without history.
Otherwise there have been two possibilities which effectively continue the lines of descent from Hegelian historicism and the history of science.
The most notable representative of the latter has been Michel Foucault, who has remorselessly continued the critique of totalizing forms of history and the disavowal of a general philosophy of history in favour of strategic ' genealogical ' analyses.
The alternative to this has amounted to a reaffirmation of historicism almost as if nothing had happened.
This has been the course of Perry Anderson, who after his Althusserian moment has returned to espouse the virtues of empirical historical studies, under the general aegis of a historicism anchored in the recasting of orthodox historical materialism by G.A.
Cohen.
More influential in the realm of literary and cultural theory has been another overt defence of historicism  that of Fredric Jameson.
It is to Foucault and to Jameson that we now turn.
Foucault's Phantasms
It is in the work of Michel Foucault that we find the most unrelenting offensive against historicist theories of history.
But Foucault at least can hardly be accused of neglecting history as such.
The demise of Althusserianism has meant that the extent to which subsequent writers, such as Foucault or even Derrida, continued to work within the problematic that his work had established, has tended to be overlooked.
It was Althusser who, after Sartre, problematized the very concept of history and laid the basis for much subsequent theoretical investigation.
In the post-Althusserian context of today it is nevertheless somewhat startling to find the Althusser of Reading Capital citing his debt to Foucault (along with Bachelard, Cavaills, and Canguilhem) as one of our masters in reading learned works'. perhaps even more unexpected, in the light of the fact that a popular British Marxist position on Althusser is that he simply turned history into theory, is the choice of Foucault's Madness and Civilization (' that great work') and The Birth of the Clinic as examples of the kind of history, focused on the necessity of the production of a concept, that he was advocating.
In certain respects Foucault always remained close to the general positions from which Althusser worked, particularly in relation to the influence of Bachelard and his scepticism towards progressivist and homogeneous histories.
All were concerned to establish the possibility of discontinuity in a history, as Althusser described it, no longer ' steeped in the ideology of the philosophy of the Enlightenment, i.e. in a teleological and therefore idealist rationalism '.
But whereas Althusser followed Bachelard's and Cavaills' preoccupation with the formal systems of the pure sciences and set theory, Foucault rather pursued Canguilhem's focus on the medical and biological sciences, which emphasized their cultural frame and the historical and institutional (both discursive and non-discursive) conditions of their emergence which, according to Canguilhem, ' chronicle-history ' neglected.
From Bachelard and Canguilhem, Foucault derived his emphasis on the history of science as an epistemological study of the formation of concepts, a process of separation and discrimination which defines both the object and the problem that is being made intelligible.
For Canguilhem the history of a concept will have its own specific temporality, demonstrating less a process of epistemological self-correction as in the pure sciences than the persistence of the problem within all the contradictory solutions and ideological values that have been given to it and which make up its history.
This struggle is the inevitable result of disciplines which have no claim to the validation procedures of the pure sciences and involve non-discursive as well as discursive practices and forms of knowledge.
Here the history of such sciences does not consist in the gradual unfolding and emergence of scientific truths, but rather of a history of ' veridical discourses'.
Truth, like historicity, is derived from particular discursive practices; it operates internally as a form of regulation, as well as being the historical product of the battle between different discursive regimes.
Foucault's contribution was to adapt this type of analysis from the life-sciences to the human and social sciences.
Typically, his history takes the form of establishing a concept of something that chronological history would assume had no history, for example, madness, or sexuality.
In relation to the almost antithetical Marxist positions of Sartre and Althusser, Foucault does not, however, simply follow the latter rather than the former: he articulates through a historical perspective the problems that their work encountered and attempts to produce a new method of historical enquiry  though not a general theory of history as such  that is both theoretically coherent and politically effective with respect to the particular problems under examination.
That History and Marxism as such are casualties of this process is one obvious reason why attacks upon Foucault have been particularly virulent.
Foucault objects to historicism and Western humanism to the extent that they assume a continuous development, progress, and global totalization.
To this he adds his own more individual list of suspect conceptual categories: the subject, class, ideology, repression, the science/non-science distinction, as well as any general theory of society, causality, or of history itself.
It is interesting that despite this some writers are still prepared nevertheless to claim Foucault as a Marxist  perhaps less an indication of his Marxism than of his discursive power and the lack of alternatives within Marxism today.
Foucault emphasizes that his work does not lay claim to universal or general categories, nor is it even homogeneous, a presupposition that, as he has shown, has less to do with the work as such than the critical construction of its' author '.
Nevertheless we can say at least that his preoccupation with history has been consistent throughout: the articulation of repressed history in Madness and Civilization (1961), the historicity of history in The Order of Things (1966), the epistemic mutation of history and the theoretical difficulties of historiography in The Archaeology of knowledge (1969), and the attempt to write a different kind of history, ' genealogy ', that demonstrates the emergence of new forms of power in Discipline and Punish (1975) and The History of Sexuality (1976C84)
Like many of those discussed in this book, Foucault endorses the ethico-political project of establishing forms of knowledge that do not simply turn the other into the same: as he put it in 1968, he wishes to find another politics than that which ' since the beginning of the nineteenth century, stubbornly persists in seeing in the immense domain of practice only the epiphany of a triumphant reason, or in deciphering in it only the historico-transcendental destination of the West '.
This focus on the link between the structures of knowledge and of power is increasingly related to an accompanying analysis of the discursive and technological mechanisms of repression and domination.
These ethico-political concerns formed the basis of his early works on madness: as he announced in his first book, Mental Illness and Psychology (1954):
One day an attempt must be made to study madness as an overall structure  madness freed and disalienated, restored in some sense to its original language.
Madness and Civilization was such a study, but the assumption that it was possible simply to lift the repression and to let otherness speak for itself was questioned by Derrida in his well-known critique, ' Cogito and the History of Madness'.
It is in this context that we can see the importance of the so-called Derrida-Foucault debate which is often misrepresented, not least by Foucault himself, as a confrontation between ' textuality ' and ' history '.
But Foucault's own subsequent work shows that it could not really be a question of choice on these terms, for the simple reason that, as he himself is at pains to point out in The Order of Things, history is itself a discursive practice: while the latter can not be simply equated with the textual, it can not be crudely opposed to it either.
The dispute between Derrida and Foucault was less a question of text versus history than an argument about history itself.
Derrida focuses on Foucault's claim that in Madness and Civilization he is writing a history of the Other.
His critique centres on two related problems: in the first place if, as Foucault argues, the expulsion of madness by reason constitutes the possibility of history as such, so that this gesture of exclusion produces the fundamental structure of historicity, then the ' classical ' moment of this proscription that he describes must be an example rather than an originary moment.
In the second place, if reason is defined by its elimination of madness, and if history is a rational concept, the question then follows, how can you write a history of madness?
Derrida comments: ' It is the meaning of ' ' history ' or archia that should have been questioned first, perhaps'.
Implicitly he is criticizing Foucault's understanding of the relation of the same to the other which posits madness as outside the sphere of reason.
To regard the latter as systematically oppressing madness and forcing it to the margins  to literature, or the hospital  is simply to repeat, in reverse, the very structure that is being criticized.
Derrida contends that if madness is constituted as madness, as other, by reason, then this means that reason is itself defined through it and therefore already contains and depends upon it.
Madness can not be considered to exist outside the historical conditions of its production  and could never therefore be ' restored... to its original language ' as Foucault hopes.
The problem for Foucault is that this argument involves more than just madness as such, for it really amounts to a questioning of the very possibility of critique.
In his early work we can see that Foucault's position involves a remarkable development of Althusser's hints that art can function as a privileged category that provides an ' internal distance ' from ideology by relating histories, writing reports.
Something of the same special place of ' retreat ' for art, especially literature and painting, is to be found in Foucault.
David Carroll has recently emphasized the significance of, a transgressive aesthetics or poetics of self-reflexivity' in his work.
Such self-reflexivity, far from involving a turning inwards as is so often generally supposed, comprises rather an edge or void from which a critical perspective can be opened up: a ' thinking of the outside '.
So madness, or certain radical forms of writing, such as that of Sade, Bataille, Blanchot or Roussel, by transgressing the limits of order and turning back to reflect upon it, enable a space from which a critique can be made.
This self-reflexivity operates at the limit of reason or history, eluding even the structure of the epistemes.
In this schema, Foucault is therefore proposing a fundamental, enabling separation between writing and history.
But after Derrida's essay, Foucault abandoned his claims that certain forms of literature could effect such a critical, reflective detachment.
This led to two problems: is there then any space from which a critique can be established? and how does Foucault reformulate the relation of writing to history?
In his subsequent work, the connection between the two would remain profoundly equivocal.
The seriousness of Derrida's intervention can be discerned from Foucault's subsequent repudiation of the central thesis of Madness and Civilization and the change of direction that his work took thereafter.
This is apparent from his rethinking of the ' age-old distinction between the Same and the Other '.
While he retained his criticisms of rationality, Foucault substituted the idea of an otherness at work within reason for that of a repressed alterity existing outside or beyond it.
If madness or the other is always inside, then this means that it is always already a part of reason or the same; but it will also be exactly the element that reason is unable to comprehend, and will therefore work disruptively.
This reformulation shows why it was necessary for Foucault to free himself from what he later termed ' the repressive hypothesis' of liberation.
It also provides the context in which to consider both the claim that for the later Foucault knowledge is absolutely determined, leaving him in the impossible situation of requiring something outside this for any prospect of critique, as well as the question of exactly how power and resistance are interdependent and to what extent they are separable.
Any answer to these problems must take Foucault's own resituating of the dialectic of the same and the other into account.
Even so, it is not at all obvious how he relocates writing in relation to the history to which it had until then been opposed.
If it had previously operated critically outside history's limits, offering the only available possibility of critical distance, how might writing work transgressively within it?
II DIFFERENTIATED HISTORIES
In The Order of Things, published three years after Derrida's critique, Foucault re-examines the links between Enlightenment rationality and history and poses the questions about the latter which Derrida had suggested that Foucault's work invites.
In the Preface he resituated his earlier formulations as follows:
The history of madness would be the history of the Other  of that which, for a given culture, is at once interior and foreign, therefore to be excluded (so as to exorcise the interior danger) but by being shut away (in order to reduce its otherness); whereas the history of the order imposed on things would be the history of the Same  of that which, for a given culture, is both dispersed and related, therefore to be distinguished by kinds and to be collected together into identities.
(xxiv)
Foucault thus modifies his argument that reason simply excluded madness and initiated the possibility of history by suggesting instead that the histories of the Other and of the Same are necessarily implicated within each other.
But at the same time he refuses Derrida's equation of historicity with difference as such, instead reformulating his former thesis so that now history itself takes part in the epistemic shifts that he traces.
Initially Foucault argues that the stasis of what he calls the ' Classical Order ' gave way to ' History '  which took over both as the form of knowledge and as the fundamental mode of being for empirical phenomena.
By the end of The Order of Things, however, he revises this somewhat conventional thesis to suggest that what was involved was not so much a move from a static to a historical view of things as the break-up of a common, unified historical time-scheme in which every phenomenon had had its place in the same space and chronology.
In place of a unity of time came the notion of discrete temporalities, with a recognition of a historicity proper to each discipline or area of knowledge.
Instead of a great historical narrative common to ill, everything now had its own chronology and its own history.
The effect of this was to dehistoricize man himself.
Whereas formerly he had been the subject of history, taking pride of place in God's historical scheme from creation onwards, now ' the human being no longer has any history: or rather, since he speaks, works, and lives, he finds himself interwoven in his own being with histories that are neither subordinate to him nor homogeneous with him ' (368  9).
Man was no longer the measure of all things.
Two possibilities were developed to deal with man's new homelessness in the world, his exposure, as Foucault puts it, to the finitude of the event, which attempted to draw all these heterogeneous histories together again: either men sought for some fundamental law through which a new general history could be constituted, or historicity as such was defined in terms of ' man '  whether as progress, economic laws, or cultural totalities.
The creation of man as centre was effected by defining him against other, now marginalized groups, such as women, the mad, or, we would add, the allegedly sub-human ' native '.
This move formed the basis for the human sciences which, through their organization around the figure of man, once more brought about a unity to History.
We shall examine the relation of this new humanism to the history of Western colonialism in a later chapter: it is not a question that Foucault himself elaborates in the course of what is claimed to be an, ethnology of Western culture'.
Rather he points to the theoretical paradox involved, namely that the human sciences' very emphasis on historicity as a mode of being was equally applicable to themselves as forms of knowledge, and inevitably destroyed any attempt to formulate universal laws comparable to those of the natural sciences.
Here we can see a rather different perspective on the human alienation which Lukcs describes; in Foucault's account, Marxism's attempt to reproduce the totality in itself falls within the terms of his argument by which totalizing theories of history are based on an anxiety derived from the break-up of its unity.
Marxism's claim to the status of a science proper thus functions as a device to avoid the historicity through which it accounts for other phenomena.
The Order of Things by contrast points to the emergence of historicity as a mode of understanding, and argues that ' History ', both as a form of knowledge and as the primary state of being of empirical phenomena  broadly speaking, the assumption of all Marxisms  is itself a historical phenomenon.
This means that history can not provide an unquestionable ground for knowledge, and that historicity can not claim an a priori privilege as the fundamental mode of being either.
As the two ' counter-sciences' ethnology and psychoanalysis have suggested, history is simply one possible discursive form of understanding  even if its problematic of temporality spills over into many others.
In making this argument, however, Foucault is not concerned to dispense with history  rather to make history itself an object of historical investigation and to question its presuppositions.
In particular, he challenges Hegel's equation in the Phenomenology of the evolution of history with the developing consciousness of man himself.
Although this has been particularly influential for many forms of Western Marxism, it also represents a deviation from Marx's original claim that history is the effect of material conditions rather than of human consciousness.
In this way Foucault could be said to be returning to Marx in removing the subject from the centre of history, were it not for the fact that he dispenses with the consolations of Marx's historicism also.
In this situation he asks, how do we come to terms with the event, with continuities and discontinuities, in short with history as difference and not just the story of sameness?
Foucault adopts a strategy, obviously indebted to Bachelard, designed to restore the otherness that History by definition must disallow: he produces an account of epistemic shifts, with prior epistemes presented as altogether estranged from the present.
In order to come to terms with the past, the initial gesture must be to confront its strangeness, rather than to seek for similarities and continuities so that it can be equated with the present and thus, in effect, dehistoricized.
Following Althusser's suggestion that a historical problematic might be altogether invisible even to experiencing subjects, in The Order of Things Foucault analyses what he provocatively calls' the historical a priori ' according to which the knowledges of grammar, natural history and wealth, and their epistemic replacements, philology, biology, and political economy, were structured.
But as with Althusser's modes of production, so Foucault's attempt to elaborate a ' positive unconscious' of knowledge was criticized by Sartre, and subsequently by many others, for being unable to give any account of change and for implying a total discontinuity between periods.
The objection that Foucault neglects history because he does not attempt to give reasons why the epistemic shifts he describes occurred is perhaps inevitable but also begs the question: for conventional historiography has in general done nothing but account for such shifts  which has meant that it has consistently failed to recognize alterity and incommensurability in its insistent search for continuities with the past.
Without history as a form of mediation the differences return as the problem, and it was this which, according to Foucault, The Order of Things was really trying to address.
Here we might recall Derrida's criticism of Levinas's history ' as a blinding to the other, and as the laborious procession of the same '.
In the face of a history which obscures such discontinuities, the first stage for Foucault, therefore, is to defamiliarize it by reconstituting it without the mythology of a continuous History which has turned difference into identity.
It is often assumed that Foucault is simply the philosopher of discontinuity, merely substituting it where previously there had been continuity; but the discontinuous is emphasized only because so much stress is normally placed on the continuous.
It is not just a question of exchanging one for the other.
This has a methodological implication as well: all too often it is assumed that Foucault's stress on disruption can be taken as equivalent to randomness, as we have seen with Perry Anderson.
However, it is more helpful to consider it in the context of Canguilhem, who emphasized that the life sciences, like the natural sciences, require their own specific mode of history; they show that historical method itself must be heterogeneous, in the sense that there is no single method applicable to the whole range of different histories.
Even the history of discontinuities is itself impermanent and discontinuous.
If certain forms of history stress continuity, try to account for change, the history of science by contrast must be disjointed, for it is made up of a series of corrections in which the errors of the past have to be simply discarded.
Foucault, as we have seen, is also criticized on the grounds that he can not give a cause for the shifts he describes, but this criticism itself begs the question insofar as it assumes a certain kind of history, which itself presupposes that there was a cause in the sense of a single uniform causality, rather than a disconnectedness in the scientific mode.
In the context of conventional historiography, Foucault argues that the point is to analyse the different kinds of transformation, the complex ' play of dependencies', links and redistributions, rather than to provide yet another account of change, succession and its causes.
The Order of Things, although in many ways Foucault's most influential work, remains, however, an oddity in certain respects: first, that in arguing for an a priori common to a (limited) number of knowledges, Foucault at times seems to be advocating a structural key between different levels within the episteme, thus restoring the form of the essential section so criticized by Althusser.
Foucault has even been accused of returning, in this work, to the concept of a totality in the episteme; it has certainly been somewhat hastily assumed that the latter can be appropriated more or less as a new way of describing a historical ' period '.
This fails to recognize, however, the extent to which it articulates only the structure of certain specific forms of knowledge rather than some single overarching principle.
The constant emphasis on its being the Western episteme suggests immediate problems for any assumption that it constitutes a totality.
The episteme rather delineates what Foucault calls a ' cluster of transformations'; these, he suggests, enable the substitution of ' differentiated analyses for the themes of a totalising history:
They allow us to describe, as the episteme of a period, not the sum of its knowledge, nor the general style of its research, but the deviation, distances, the oppositions, the differences, the relations of its multiple scientific discourses: the epistemic is not a sort of grand underlying theory, it is a space of dispersion, it is an open field of relationships and no doubt indefinitely specifiable.
They allow us furthermore to describe not the great history which would carry along all the sciences in a single trajectory, but the types of history  that is to say, of retentivity and transformation  which characterize different discourses... the episteme is not a slice of history common to all the sciences: it is a simultaneous play of specific remanences.
finally they allow us to situate the different thresholds in their respective place: for nothing proves in advance... that their chronology is the same for all types of discourse...
The episteme is not a general stage of reason, it is a complex relationship of successive displacements.
This is very different from the integral paradigms of Thomas Kuhn to which Foucault's epistemes are sometimes compared.
But if The Order of Things is thus concerned to analyse ' ensembles of discourses' that do not form a totality, it does not focus on the ways in which such forms of knowledges relate to the institutions in which and through which they are produced.
In The Archaeology of knowledge Foucault corrects this by addressing the difficulties of making knowledges intelligible as a part of their institutional, social and political practices.
III ARCHAEOLOGY
Foucault's distrust of conventional forms of history, as we have seen, is a consistent thread in his project, as is his insistence that his historical method is limited to addressing specific problems  often comparable to those posed by the social sciences  to historical documents and practices in order to make them intelligible.
Only once, however, does he devote a whole book to the problem of historical methodology.
In The Archaeology of knowledge he develops his suggestion in The Order of Things that an epistemological mutation is taking place today with regard to the very concept and methodology of history.
The book is both an account of and an intervention in that process, veering between the descriptive and the prescriptive.
It begins by delineating the epistemological shift: the old order could be termed ' History ', a continuous and chronological historiography, including Hegelianism and related forms of Marxism, with its philosophies of history, its assumptions of a rational, progressive and teleological historical development, its desire to discover a meaning in history, its questioning of the relativity of historical knowledge, and its use of categories such as tradition, the history of ideas, the oeuvre, the author and the book.
Against this Foucault contrasts the work of the Annales school which analyses long continuities in certain social forms:
The old questions of the traditional analysis (What link should be made between disparate events?
Flow can a causal succession be established between them?
What continuity or overall significance do they possess?
Is it possible to define a totality, or must one be content with reconstituting connections?) are now being replaced by questions of another type: which strata should be isolated from others?
What types of series should be established?
What criteria of periodisation should be adopted for each of them?
What system of relations (hierarchy, dominance, stratification, univocal determination, circular causality) may be established between them?
What series of series may be established?
And in what large-scale chronological table may distinct series of events be determined?
(3  4)
Foucault notes that, at the same time as the Annales school and others were constructing a history according to the long dure, in the history of science, philosophy, and literature, attention was turning in exactly the opposite direction, that is away from vast unities towards phenomena of rupture, discontinuity, displacement and transformation, towards different temporalities as well as architectonic unities.
Here he instances in particular the work of Bachelard, Canguilhem, Serres, Guroult, and Althusser.
He argues that, given the proliferation of both discontinuities and long periods, the problem is now to constitute the series, its elements, its limits, and its relation to other series.
In other words, rather than set these antithetical approaches against each other as one might have expected, Foucault suggests that they are part of the same mutation.
Foucault characterizes this epistemic shift in terms of the difference between what he calls' total ' and ' general ' history, or, History and archaeology.
He explains this as follows.
Total, or, elsewhere, global, history assumes a spatio-temporal continuity between all phenomena, and a certain homogeneity between them insofar as they all express the same form of historicity  Althusser's essential section  whereas in general history the problem is precisely to determine the relation between different series: whereas a total history draws everything together according to a single principle, a general history analyses the space of dispersion and heterogeneous temporalities.
Total history seeks to reconstitute the overall form of a society according to some fundamental principle, law, or form, be it metaphysical or material, while general history despite its name is by no means concerned to produce a general theory of history, nor even a cohesive or comprehensive view, but rather to conduct a historical investigation according to particular problems, opening up a field ' in which one could describe the singularity of practices, the play of their relations'.
Foucault specifies the difference between them by contrasting their relation to the document: the historians of total history have been engaged above all in the interpretation of documents, attempting to reconstitute the past, to give it an inner meaning (always available only to the historian), to recover a voice and allow it to speak.
The historians of a general history, on the other hand, reject interpretation as such in favour of making the evidence of documents intelligible by posing questions to them, constituting through such questions what counts as the series of historical events, its elements, its limits, and its relation to other series, or what rules operated for particular discursive practices.
This stress on intelligibility is a useful corrective to the widespread current assumption  for which Foucault himself is also partly responsible  that history is just another form of interpretation.
Unexpectedly, perhaps, Foucault traces the beginnings of this epistemic mutation in the thinking of history to Marx.
This implies a challenge that contemporary Marxism has yet to address.
As we have seen, for some Foucault can apparently be dismissed with ease as merely the philosopher of discontinuity, a description which is hardly adequate; for others, criticism takes the form that he simply relativizes history, but this is really no better, for history is itself a mode of demonstrating the relativity, temporariness, and temporality of phenomena.
Unless it is possible to prove the identification of History with Truth, as Sartre tried to do, then the claim that a single univocal History is not relative only works by removing all other elements in the set to which it could be compared, thus making a set of one.
Effectively, however, all this does is to turn all history into a single event  which, paradoxically, dehistoricizes it.
What has not been adequately considered is Foucault's characterization of an epistemological shift within the theorization of history itself.
He suggests not only that it begins with Marx but that he himself forms part of this mutation, which means that, according to his argument, in order to dismiss him you would have to dismiss Marx also.
Foucault thus does not merely set up an alternative history, but contends that that alternative is part of a displacement that is in the process of replacing the history that preceded it.
It is therefore necessary to address not only his theoretical and methodological but also his historical arguments  something that his critics have singularly failed to do.
The Archaeology of Knowledge thus analyses the shift in historiography of which Foucault himself is the most powerful contemporary representative.
As epistemic breaks go, however, it seems to be a slow one: Foucault attributes its hesitancy to a fundamental reluctance to think difference rather than the reassuring form of the identical: it is, he comments, ' as if we were afraid to conceive of the Other in the time of our own thought '.
Here we encounter the continuing dominance of the philosophy of the subject, which, according to Foucault, was specifically introduced in order to provide a ' shelter for the sovereignty of consciousness' against the intrusion of heterogeneity.
Continuous history and the subject are thus dependent on each other.
Instead of centring his analyses on the knowledge derived from the experience of the subject, Foucault investigates the conditions of emergence of the subject as the basis of knowledge; he argues that at the same time as it was widely proposed as the one saving good of human civilization it also facilitated a more sinister operation.
Just as History involved the legitimation as knowledge of certain forms of political power, so the production of the subject by the human sciences as an object of knowledge also enabled a new form of political control: ' The individual is not a pre-given entity which is seized on by the exercise of power.
The individual, with his identity and characteristics, is the product of a relation of power exercised over bodies'.
In Discipline and Punish Foucault demonstrates how the individual is constituted through specific technologies of power; in the History of Sexuality, he shows the extent to which the human sciences, including medicine and the practice of psychoanalysis, facilitated the extension of control beyond the limits of the body in the construction of the inner realms of subjectivity, consciousness and experience.
Foucault argues that a whole series of movements since the nineteenth century, including various anthropologizing Marxisms, have developed complicitly with this so as to preserve the sovereignty of the subject against Marx's and others' decentrings: positivism, the Hegelian Marxism of Lukcs, the Marxist humanism of Sartre, as well as various theories of cultural totalities such as that of the Frankfurt School.
Dispensing with the subject necessarily also means the end of the use of the category of ideology; in this context Foucault's later work counters Althusser's influential essay on ideological state apparatuses.
Many of those now hostile to Althusser continue to endorse this essay, which offers a theory of institutional power dependent on the categories of the subject and ideology.
Foucault is critical of such a theory not just because it is based on a science/non-science distinction which for him is simply the product of a particular discursive formation which claims access to the real, rather than involving any epistemological questions of truth or objectivity, but also because it produces the notion of ideology as a secondary mediation (as in Althusser's interpellation) in an inside/outside structure between the determinants of power and the individual subject.
For Foucault, the tendency of theories of ideology to entrammel themselves in the categories of psychoanalysis, even with the eternal in Althusser's case, means that they themselves begin to utilize the very procedures of individuation that they ought to have been analysing.
The questions of the subject and of ideology raise as their corollary the problem of what position of enunciation the historian can claim in relation to his or her own work.
Foucault is particularly critical of the appropriating structure of totalization, Marxist or otherwise, insofar as it implies the superiority of the theorist who produces the totalization of knowledge; in the same way, he distrusts the use of history as an encompassing framework because it works as a power structure that enables the expropriation and control of the past according to the perspective and truth of the present.
It elides the fact that the historian will always also be historically located and therefore can not be in a position to produce a final totalization, a dialectical situation anticipated by Sartre.
Foucault argues that just as there can be no general theory of history, but only particular answers to particular questions which make individual practices intelligible, so the intellectual can best hope to be specific rather than universal (universal in the sense of proposing transcendent values, systems, totalities, narratives or teleologies).
This does not mean that the intellectual therefore nihilistically celebrates dispersion, fragmentation or relativity: rather she or he is the person who, facing such dispersion but without conceding to the nostalgic desire for totalization, poses the questions and constitutes the series and continuities for analysis  and thus for transformation  while attempting to respect its heterogeneity.
This suggests why, contrary to the way in which certain commentators tend to represent them, it is clear that for Foucault, as for Lyotard or Derrida, total fragmentation would be as counter-productive  and as impossible  as total synthesis.
As Gayatri Chakravorty Spivak puts it: ' Theoretical descriptions can not produce universals.
They can only ever produce provisional generalizations, even as the theorist realizes the crucial importance of their persistent production '.
Such arguments will have no pretensions, by definition, to knowledge-claims that affect to rise above the historical conditions in which they are made.
IV HISTORY AND THE EVENT
Despite his programmatic emphasis in the Archaeology on the discursive formation as a means of making intelligible those knowledges that are formulated through their institutional components, Foucault turned away from this kind of historical enquiry because it was too ' clean, conceptually aseptic '  in other words, too apolitical.
Foucault therefore abandons analysis of epistemes for the more Nietzschean ' genealogy ', which allows him to articulate conflict in terms of differentiated histories with their own conceptual specificities and their own times, while retaining the possibility of the formulation of aims and intentions:
I would call genealogy... a form of history which can account for the constitution of knowledges, discourses, domains of objects etc., without having to make reference to a subject which is either transcendental in relation to the field of events or runs in its empty sameness throughout the course of history.
Genealogy develops the possibility broached in the Archaeology that in a general history different significances can be accorded to events, depending on ' their correlation with other previous or simultaneous events, discursive or not '.
Here it is the problem the historian poses that determines what constitutes an event and what status it has.
Foucault's genealogy means that by asking a question, posing a problem, you set up a generality against which you constitute events and arrange them in a series.
The construction of that generality does not pretend to be the only possible one  the same event could operate in all sorts of different ways in different series, temporalities, which would mean that, strictly speaking, it was no longer the same event, for it would have been dispersed in their different rarefactions.
This is not supposed to suggest that events can not be said to occur straightforwardly in the real but rather that when set up in any series, narrative, or history they are constructed as such events retrospectively by the historian.
If all history attempts to conceptualize the event, to wean it from its finitude, then Foucault like many of his contemporaries is concerned to respect its singularity.
In the essay ' Theatum philosophicum ' (1970), written shortly after the Archaeology, Foucault attempts to avoid the snares of the problem of the relation of the event to the totality, or the particular to the general, that beset both Sartre and Althusser, by arguing that the event as event is only constituted through its repetition in thought as a ' phantasm ': ' it makes the event indefinite so that it repeats itself as a singular universal '.
This argument, which Foucault derives from Deleuze, although at the same time he tellingly invokes Sartre's theoretical formulation designed to solve exactly the same problem, provides a way of avoiding the incommensurability of the relation of the event to the concept by allowing ' the disjunctive affirmation of both '  thus solving the problem that the concept, as a part of the language of generality, will inevitably travesty the event's singularity:
Logique du sens causes us to reflect on matters that philosophy has neglected for many centuries: the event (assimilated in a concept, from which we vainly attempted to extract it in the form of a fact, verifying a proposition, of actual experience, a modality of the subject, of concreteness, the empirical content of history); and the phantasm (reduced in the name of reality and situated at the extremity, the pathological pole, of a normative sequence: perception-image-memory-illusion).
After all, what most urgently needs thought in this century, if not the event and the phantasm?
Well, it might be hard to see it as that important.
Foucault's gesture, however, marks the radical scope of such an anti-essentialist project which exploits Plato's admission that there can not only be a good copy but also a bad one (phantasma)  characteristic of the secondary representations of the poets.
Deleuze reverses Plato by validating this simulacrum of the good copy on the grounds that, precisely because it is a bad copy, it breaks down all adequation between copy and model, appearance and essence, event and Idea.
As plato realized, it is dangerous because its effect is to dethrone the Idea from its position of Truth.
Moreover, because the bad copy by definition can not claim to be copying anything but itself, it creates its' original ' retroactively, so that the copy precedes the original in a ghostly originary repetition.
Descombes describes a comparable paradoxical structure in his account of ' originary delay ': a first event can not be the first event if it is the only event; it can not be said to be a first until it is followed by a second, which then retrospectively constitutes it as the first  which means that its firstness hovers over it as its meaning without being identifiable with it as such.
Thus, the phantasm, rather than constituting the event, hovers over its surface like a cloud, as an effect of meaning not identifiable with anything in the event as such.
Deleuze's best-known example of how this works is the battle: ' Where ' ', he asks, ' is the battle?
What we call a battle consists of a vast, heterogeneous array of individual actions in the field  fighting, firing, charging, wounding  none of which constitutes' the battle ' as such.
' The battle ' hovers over the individual actions like in incorporeal cloud, distinct from them, but at the same time making up a surface of their meaning-effect, a simulacrum that brings the event into being at the moment when language and event coincide.
Foucault recognizes in Deleuze's account of events as singularities, points or intensities on a surface ready to be actualized in any particular form or meaning, the potential for pushing further his own notion of history as a genealogical series.
Such genealogy works by repeating the (non) event, as an event, in thought  in a structure comparable to Freud's ' deferred action, or ' retroactivity '.
No more than the latter does it seek to lay claim to ' the real, or Truth as such.
Foucault inflects this model by focusing on the possibility of constructing the series so as to repeat the disruption and discontinuity of the (non-original) event.
In freud, the point is similarly not just the question  on which most attention gets focused  of whether the event ' really ' happened (a good copy) or was subsequently fantasized by the experiencing subject (a bad copy), but rather that it is repeated as a disruptive event that fissures ordinary forms of psychic continuity and therefore gains analytic attention in the present.
The same structure can be utilized by the historian so that the writing of history can itself become a disruptive event and consequently a form of political intervention.
V FOUCAULT'S PHANTASMS
If Foucault proposes the philosophy of the phantasm as a way for the historian to produce the meaning-effect of the event while still doing justice to its singularity, it also offers a way of thinking through some of the paradoxes that we have encountered in the problematic conceptualization of history.
From Sartre to Foucault history has repeatedly emerged as a contradictory concept, both totalizing and detotalizing, essentialist and non-essentialist.
Such contradictions can be productive: the attempt to reject historicism absolutely results either in an utter particularism or in a surreptitious return of historicism in a different form.
Only an understanding that recognizes that an irresolvable tension works within the historical schema itself will be in a position to make its contradictory claims productive.
This possibility is outlined by Foucault himself as early as The Order of Things.
He comments:
The more History attempts to transcend its own rootedness in historicity, and the greater the efforts it makes to attain, beyond the historical relativity of its origin and its choices, the sphere of universality, the more clearly it bears the marks of its historical birth, and the more evidently there appears through it the history of which it is itself a part... inversely, the more it accepts its relativity, and the more deeply it sinks into the movement it shares with what it is recounting, then the more it tends to the slenderness of the narrative, and all the positive content it obtained for itself through the human sciences is dissipated.
(371)
History becomes the impossibility between this Scylla and Charybdis  in Lyotard's terms, it contains within its own project an incommensurable difference.
It can only be described in terms of its organization according to an economy of logical tensions or strictures, of demands and constraints.
It is, as Derrida comments, a question of showing:
that history is impossible, meaningless, in the finite totality, and that it is impossible, meaningless, in the positive and actual infinity: that history keeps to the difference between totality and infinity.
Thus both historicism or entirely differentiated histories are in themselves impossibilities: history will always involve a form of historicism, but a historicism that can not be sustained.
It is thus a contradictory (quasi) concept  a phantasm  in which neither the elements of totalization nor difference can be definitively achieved or dispatched.
This means that history can be theorized not so much as a contradictory process but as a concept that must enact its own contradiction with itself: ' this difference is what is called History '.
In this context, we may recall that Althusser, rather than simply criticizing the notion of history as a totality as Foucault often tended to do, argued for the rearticulation of different histories within a decentred totality, on the assumption that history can not do without one.
Was it simply incoherence on his part when he suggested that his' science of history ', even though it allows for differentiated histories, still demands to be considered within a general concept of history?
We have seen the ways in which Sartre's Critique shows how totalization can not work without a movement of self-transcendence, a repeated interpolation of an excess beyond the totality which paradoxically then means that the totality can no longer be a totality.
Although Sartre's inability to effect self-totalization is often presented as a failure, the movement of history that he describes is increasingly enacted through his own writing.
Such a shift is not immediately discernible in Foucault.
His Deleuzian notion of the phantasmatic event constituted a brief attempt to reformulate the relation of the particular to the general as a problem of history.
But he made little effort to develop this outside his own definitions of the genealogical method, while his shift into the problematics of power seemed to lead him into a labyrinth from which it was virtually impossible to extract himself.
The philosophy of the phantasm may, however, help us to do justice to the event of Foucault himself.
One of the oddities of Foucault's work is that it seems riven by an internal tension  for example, as peter Dews notes, while on the one hand Foucault lays claims to a form of objectivity in his archaeology, and eschews interpretation in favour of ' intelligibility ', on the other hand throughout his life he was also prone to endorse a Nietzschean insistence on the interminability of interpretation.
For Dews this equivocation over the epistemological status of his own discourse is a sign of an uncontrollable oscillation: ' the shifting perspectives of Foucault's work ', he writes, ' do powerfully illuminate, but at the same time fall victim to, the contradictory processes which they address'.
But such a self-undermining of the epistemological status of his own work is also characteristic of Lacan, and of Freud  indeed, Dews argues that this constitutes Freud's greatest strength.
We might therefore wonder whether its appearance in Foucault, far from being the result of theoretical ineptitude, does not involve simulacra, or ghostly bad copies, similarly designed to undermine the claims of theoretical mastery, and to produce in his texts surface-effects of the kind of heterogeneity we might expect from someone who had contested the unifying function of ' the author '.
With respect to history, the vacillations of Foucault's writing enact the impossibility of its simultaneous finitude and infinitude, the irresolvable conflict between history as meaning and history as difference, between history as a teleology and eschatology and history as the event, as finitude and mortality.
Here we once more encounter the recognition that at a conceptual level the idea of history can not be taken further: rather it can only be addressed through a tension in the writing itself.
The early opposition between writing and history which enabled a critique from an outer limit has thus been replaced by a dissension within Foucault's own discursive practice.
Instead of locating the other elsewhere in the transgressive writing of literature or madness, Foucault himself becomes plato's banished poet, and enacts in his language a supplementary simulacrum of the delirium of history.
As Maurice Blanchot has put it: ' And were not his own principles more complex than his official discourse, with its striking formulations, led one to think? '
VI HISTORY AS POWER
In later work Foucault veers away from his most radical philosophical insights towards an analysis of discursive and extra-discursive systems of domination and exploitation, with increasing focus on the apparently more political question of power.
But perhaps this shift is not so distinct: for the interaction of power and resistance, which mimes the mutual contamination and transmutation of Freud's death-drive and pleasure principle, also operates is a simulacrum of the ungraspable, vacillating manoeuvres of Foucault's formulations of history.
Whereas Foucault was inclined to remark that in retrospect he considered that all his work had been about power, it seems almost equally possible that his analyses of power constitute a continuing meditation on the phantasm.
Although in no sense a general theory, Foucault's notion of power reformulates the problem of agency and determinism which had beset Sartre and many others by focusing on the possibility of making intelligible the strategies and techniques of local operations of power without relying on the dialectic of ideology and the consciousness of subjects, or on their corollary, the assumption that power operates globally and homogeneously.
The absence of the category of consciousness inevitably plays down the role of individual subjects and thus of individual agency and resistance as a result of specific acts of will.
If this causes difficulties for some, Foucault's scepticism with regard to the tendency to inflate the effect of individual agency can only be compared to the position of many Marxisms in which resistance and revolution are hardly the privilege of the individual as such, but rather of collective class action.
Those who forget the virtues of solidarity in order to protest against the downgrading of individual agency might recall that it has been intellectuals who have been most prone to inflate the significance of individuals  particularly intellectuals  to the same degree that their theories propose universal categories and claim universal effects.
Moreover the exclusive focus on ' resistance ' as a privileged political category is itself open to question.
As with his genealogical history, Foucault's meditations on power are not themselves without problems, but reactions to them can also be too hastily dismissive, often because it is assumed that they are proposed as a general theory.
For example, after the History of Sexuality much has been made of Foucault's analysis of power as a form of totalizing paranoia; but the analyses in that book, of, for example, the shift from the Catholic confession box to the confessional psychoanalytic couch, are both culturally and historically specific, and Foucault's remarks about them need to be put in the same perspective.
After all, if ' the system ' really was bent upon the form of totalizing control that according to Foucault psychoanalysis, for example, enables, it is worth recalling that psychoanalysis has never been adopted by the state as such and that its activities remain confined to a few very limited districts in a handful of prosperous cities round the world.
Similarly, those who claim that Foucault removes the possibility of resistance as such miss the point: all that he downgrades is a theory of resistance centred on the individual subject as sovereign agent.
He argues that to focus analysis in this way neglects the disciplinary forms and technologies through which power operates.
The claim for a straightforward oppositional kind of resistance also assumes that subjects can resist from a position outside the operations of power, according to the dominant inside/outside model of conventional politics.
This is the context in which to place Foucault's own recognition after Derrida's critique that he could no longer postulate madness or the other as outside, after which he maintained that the other is also always inside; he formulates the structures of power in exactly the same way, so that the forces of domination and resistance are caught up, sometimes indistinguishably, within each other.
Acts of resistance may of course continue to be initiated through individual acts of will, but as for Sartre there is no guarantee that they will produce intended effects.
An awareness of this structure does not mean that strategic intervention is either useless or impossible; but it does mean that analysis of how resistance actually operates, in what conditions it succeeds or fails, needs an altogether more complex model.
Foucault's account of power is thus difficult to the degree to which he argues that the exercise and resistance of power work in a disruptive rather than a dialectical relation to each other, suggesting that ' points of resistance are present everywhere in the power network '.
Where there is power, there is resistance: contrary to what is often assumed, it is the absence of resistance which is impossible.
power is a two-way process.
Just as the exercise of power is heterogeneous, so is resistance; Foucault's point is simply that ' there is no single locus of great Refusal, no soul of revolt, source of all rebellions, or pure law of the revolutionary '.
Resistance does not operate outside power, nor is it necessarily produced oppositionally: it is imbricated within it, the irregular term that consistently disturbs it, rebounds upon it, and which on occasions can be manipulated so as to rupture it altogether:
Just as the network of power relations ends by forming a dense web that passes through apparatuses and institutions, without being exactly localized in them, so too the swarm of points of resistance traverses social stratifications and individual unities.
And it is doubtless the strategic codification of these points of resistance that makes a revolution possible, somewhat similar to the way in which the state relies on the institutional integration of power relationships.
It is in this way that Foucault can return to the possibility of doing historical work that has political force through his notion of genealogy, which means, as he puts it, ' that I begin my analysis from a question posed in the present '.
The question enables the tactical use of historical knowledge in contemporary political situations which necessitate the posing of the question with which genealogy begins.
Such politics stress the local or the specific without assuming that they constitute the starting point for a global hegemony into which they will be subsumed.
Foucault does not aim to produce ' a ' politics any more than ' a ' history.
It is this factor which, perhaps above all, has enabled the critical use of Foucault's analyses of power in demarcated areas of analysis.
Ironically, Foucault's most problematic ' theory ' has generated the most successful and probing historical work among his followers.
VII CODA: THE NEW HISTORICISM
The relation of Foucault's politics to history leads to the question of the tactical use of Foucault himself in current forms of criticism; in addition to his extensive use by historians and sociologists, he is also associated with those movements that have become known as' new historicism ' and ' cultural materialism '.
The former is identified closely with Foucault, while the latter owes its allegiance to Raymond Williams, and really only amounts to a way of describing British ex-Marxists.
New historicism is most concerned with the late Foucault, in particular with representations and mechanisms of power, the means through which cultural artefacts can be shown to be not passive reflectors of the history of their time but active producers of it.
This charting of the circulating relations between aesthetic and other forms of production works best in those historical periods, such as the Renaissance, where there was no modern concept of Literature, thus allowing literary texts to be mapped against the political and other discourses of which they formed a part.
New historicism emphasizes the way in which certain rituals and practices, for example of kingship, are equally a part of the processes and representation of power  whether at a beheading, a masque or in a poem.
At the same time, it can also bring out the complex ways in which such forms of power also produce their own forms of resistance; as critics like Stephen Greenblatt demonstrate, these are not separable processes but are simultaneous effects of power.
The illustration of such a double logic working at a textual level often comes to seem remarkably close to a deconstructive analysis: in the same way as Derrida or de Man could be said to be deconstructing received readings that have institutional purchase, so the new historicists shift our understanding of institutionalized historical accounts.
However, the very historical focus poses the problem of agency and containment in a more immediate way that demonstrates some of the dilemmas that follow from Foucault's genealogical history.
The real difficulty, as Greenblatt shows, is what status can be accorded the category of subversion or resistance: what is the historical status of the ' subversive, elements of a text?
If the process of the construction of knowledge can be shown to work against itself, this can operate because the ideas of the dominant order are not in fact threatened by alternatives which, with hindsight, may appear radical to us now.
As the cultural materialists concede, subversive thoughts are not subversive until they become a practice.
Can they be shown to have had historical effects?
From the point of view of a radical politics, the onus on the critic, therefore, is to show the ways in which such subversions can be shown to have produced specific instances of change.
Correspondingly, as Greenblatt argues, Renaissance ideas which might be subversive today are ignored in favour of ' radical ' ideas which seem to echo our own.
Both groups follow Foucault to the extent that they neither propose, nor utilize, a general theory of history as such; but unlike Foucault they simply tend to shelve the whole problem so as to avoid its theoretical difficulties.
So, despite the Lukcsian resonance of its name, new historicism abandons both the notion of history as a readable background to literary texts, and the Marxist dependency on reflection, in its effort to recontextualize literary texts in a more immediate way with other forms of social production.
For their part, the British cultural materialists quickly adopted a name which tactfully removes the suggestion of Marxism as such.
In the context of the present discussion of history, moreover, it is highly significant that traditional ' historical materialism ' has been jettisoned in favour of the anthropological ' cultural materialism '.
The cultural materialists abandon the traditional Marxist use of history as a ground for truth in favour of bringing history into the present day in order to intervene in their own institutional and academic political context.
In this case it becomes less a matter of addressing contemporary political problems than of resisting institutional power, namely those critical readings which have recently claimed cultural hegemony, and of providing instead alternative accounts which insist on heterogeneity and resistance in historical texts.
An identification is then implicitly or explicitly made with parallel forms of political struggle in our own day.
In other words, where yesterday's historian looked for the history of an oppressed working class, today's historian looks for marginalized groups, and those who have transgressed social norms.
Whether they actually were subversive becomes irrelevant to the extent that they can now be retrieved to offer a potential that has a contemporary, that is twentieth-century, political relevance.
To this extent we could say that the cultural materialists re-assert a form of reflection theory, where history has become a mirror in which contemporary political priorities have been substituted for the former certain ground of Marxist analysis.
The cultural materialists are inclined to separate the self-contradictory differences isolated by the new historicists into a more conventional political paradigm of opposing classes, of hegemony and subversion.
Similarly, though Foucault is often invoked, there is a marked tendency to continue to utilize theoretical categories such as ideology, consciousness and the subject.
Here it is salutary to recall Foucault's scepticism with regard to the ' perilous ease ' with which politics quickly assumes positions that provide intellectual guarantees rather than specific analyses of particular relations or transformations; he reacts in the same way to political analogies, and correspondences, or to hasty links with current political practices.
To the last, Foucault himself remained suspicious of any ' progressive politics, if that meant that it continued to be linked to a (hidden) meaning, origin, or the subject.
But if the cultural materialists, unlike the more fastidious new historicists, cheerfully ignore the theoretical consequences of Foucault's work for many Marxist concepts, in other ways they are closer to him insofar as they make clear in their work their own political priorities and commitments.
Nevertheless it is worth recalling that Foucault never starts at the political, but rather begins with a contemporary problem and then addresses questions to politics about it.
He argues that intellectual work need not always be measured against immediate political ends; rather the question to ask is what project is it undertaking, what problem is it analysing?
It should always be possible for the intellectual to justify and to show the grounds on which any particular activity has been constituted.
At the same time, it is too easy to condemn certain forms of work for being apolitical simply because they do not correspond to a certain paradigm of immediate political effectivity.
Political interventions can also work according to different time scales.
But if the cultural materialists tend to place a politics rather than a problem as the starting point of their enquiry, they do emphasize the deployment of specialized knowledges in the service of the popular political struggles of today.
This suggests a closeness to Foucault's notion of genealogy that the more strictly academic new historicists, whose own politics remain more carefully hidden, ignore.
