Further reading  —  riff-raff.se

Marxism, Metatheory and Critique

Richard Gunn

Post-Fordism and Social Form: A Marxist Debate on the Post-Fordist State Macmillan, London 1991, pp. 193–209

Quite often, major political differences can turn on apparently technical and esoteric conceptual points. Inasmuch as there is no such thing as ‘brute’ or sheerly immediate politics any more than there is a ‘brute’ fact, this should be unsurprising: how we see the world is bound up intrinsically with what we do. Nonetheless Marxists eager to address first-order social and political questions have been perhaps too ready to consign questions of concept-formation to a hinterland of ‘methodology’, whose problems are always on the agenda for tomorrow but which is never discussed today. Of course it was not always so: from Lukács’s 1923 declaration that Marxist ‘orthodoxy refers exclusively to method’ – a declaration which Lukács subsequent discussion proceeded to undermine1 - there developed a Marxist tradition according to which methodological concerns were prioritised over debates of a first-order or ‘empirical’ kind. A split arose between what Isaac Deutscher termed ‘political’ and ‘philosophical’ Marxism (Deutscher 1972). An exciting feature of debates in the 1960s and 1970s was that there emerged a promise that this split could be healed. But in the pragmatic Thatcherite 1980s Marxist theorising has too often taken a sheerly empiricist turn. Understandable political urgency has tended to marginalise conceptual questions, despite the circumstance that these questions are as political as may be. Thus it is more than welcome that Bob Jessop contends that ‘only by re-examining … methodological assumptions’ (Jessop 1988, p. 8) can a discussion of state theory proceed.

The ‘methodological assumptions’ upon which Jessop claims to rely are those of the ‘Critical Realism’ advertised by Roy Bhaskar in, most recently, two flagship-style articles summarising presentations to the UK Socialist Conference and published in the journal Interlink (Bhaskar et al. 1988, 1989; see Jessop 1988, pp. 8–14 and footnotes 15–18). Bhaskar offers a realism which is critical inasmuch as it challenges the fetishised claims of social appearances, and a criticism which is realist inasmuch as it turns upon identifying ‘enduring structures and generative mechanism’ which underlie and produce ‘observable phenomena and events’ (Bhaskar et al. 1988; the opening paragrapsh of the 1989 version reproduce the same views). The more complex statement of Critical Realism which Jessop invokes turns on a distinction between ‘real mechanisms’, ‘actual results’ and ‘empirical indicators’ (Jessop 1988, p. 9). For the purposes of the present discussion I set these complexities aside. Instead my topic will be the conceptual status of Critical Realism per se, this being a topic which raises quite directly the question of Marxist concept-formation and which therefore extends well beyond a discussion of Critical Realism’s specific claims.

Bhaskar opens his 1988 article by stating that ‘the left needs to take philosophy seriously’, which is certainly and uncontroversially the case. Throughout its two-thousand-years history philosophy, arrogating to itself the programme of a unique lust after wisdom, has reserved questions of concept-formation and concept-validation to its own discourse. The more modern the philosophy, the more hubristic has this arrogation become. From Descartes and Kant onwards, philosophy has devoted itself to evaluating the categories – the terms or truth-criteria – in and through which first-order discourse proceeds. Its quite reasonable entitlement for doing so is that vicious circularity is the outcome whenever first-order discourse, on its own behalf, lays claim to this role3: and so a second order discourse (viz. ‘philosophy’) is wanted to carry the epistemological can. This gives rise to both a narrowing of philosophical concerns3 and a problem. The problem is that the terms of a second-order discourse require, according to the same programme, to be validated by a third-order discourse … and so on, without hope of halt. In place of vicious circularity, infinite regress is called in to do the job. The twentieth-century philosopher who sees this consequence most clearly, and who most naively and engagingly accepts it, is Bertrand Russel (in his Introduction to Wittgenstein’s Tractatus and in his theory of logical types). But naive acknowledgements by no means amounts to resolution. And so ‘the left needs to take philosophy seriously’, brute facts and romanticist politics-of-immediacy being a nonsense. But – given the problems of infinite regress – it can only take philosophy seriously in the form of critique. It has to reappropriate philosophical questions (those of category-formation and category-validation) and avoiding the Scylla and Charybdis of vicious circularity and infinite regress, it has to do so by reopening the question of the relation between first-order theorising and theorising of a higher-order (or ‘metatheoretical’) sort. The difficulty with Bhaskar’s 1988 pronouncement is that he appears to construe the left’s need to take philosophy seriously to mean that the left needs a philosophy, namely, his own. The roots of Critical Realism lie in precisely the philosophy of science – the philosophy which claims to be able to demarcate natural-scientific truth-claims from others – which Russel’s Theory of Types inaugurated. In the old days, Critical Realism used to be termed ‘transcendental realism’ (Bhaskar 1975), thereby more directly indicating its Kantian source (which Russel and philosophy-of-science share). Quite apart from the question of the internal complexity of Critical Realism’s statement, there is a question about the conceptual – and political – status of Critical Realism itself.

If indeed the left needs a philosophy (whether Critical Realist or otherwise) in order to address the issue of its concept-formation then, substituting infinite regress for vicious circularity, it professes itself willing to ascend a ladder of metatheories whose top rung rests nowhere but upon clouds. It is in the light of problems regarding this kind of idealism that, I suggest, Marx’s critique of philosophy in the 1840s advises us simply to ‘leave philosophy aside’ (Marx 1975, 5, p. xxx) and devote ourselves to the study of the empirical world. But such passages – veering back into vicious circularity from infinite regress – are atypical. The strength of Marx’s critique of philosophy is that it advertises a mode of theorising which, because it is at the same time both first-order and higher-order (both ‘theoretical’ and ‘metatheoretical’), escapes the dilemma of vicious circularity versus infinite regress.

In order to see how Marx achieves this, we have to turn to the thesis of the unity of theory and practice which his texts of the 1840s announce. Marx posits an internal relation as between theory and practice, and accepts (or rather stresses) that just such a relatedness obtains in the case of his own thought. In other words Marxist theorising is not just reflexive but practically reflexive (Gunn 1987a). Two definitions help to clear my argument’s way.

Theorising is reflexive when, and insofar as, it asks after validity of its own categories, concepts, truth-criteria and/or terms. Theorising is practically reflexive when it does this in the course of asking after its practical – its social or historical – situatedness. The phrase ‘in the course of’ reports the internal relatedness of theory to practice for which Marx contends. Were theory merely to ask after its conceptual validity and, separately, to reflect upon its social preconditions (as in bourgeois ‘sociology of knowledge’) then – as reported in the conjunction ‘and’ – it would regard theory and practice as linked merely in an external way. Practically reflexive theorising unites not just theory with practice but first-order theory with metatheory because (a) in reflecting upon society (its first-order object) it reflects upon itself, and because (b) in reflecting upon itself (upon its own categorial validity) it reflects upon the social practice which it challenges and within which it inheres. Its first-order object includes itself as a second-order object, and comprehension of its relatedness to social practice as a whole (see Lukács 1971 first essay; Horkheimer 1972, p. 229). Another way of stating this point is to say that practical situatedness impinges on theorising not just in a third-person – as in ‘sociology of knowledge’ – but in a first-person way. Yet another way of registering the argument is to say that practical reflexivity is by no means a ‘methodology’ discussion which might be confined to the introduction of a text which is informed by it; rather, it is the mode of practice of a practically reflexive text itself.

In sum Marxism, qua practically reflexive, is neither first-order nor sheerly higher-order because it is both together and in the same breath. The same body of theory (or movement of theorising) plays now a theoretical, and now a metatheoretical, role. Even the word ‘now’ in the preceding sentence is problematic, because the role played by any particular conceptual shift is a function only of the questions which, in relation to it, are posed. In other words instead of a dualism (‘theory’ contraposed to ‘metatheory’) we have a totalisation. Marx’s 1840s critique of philosophy is not a matter of turning away from the questions philosophy has reserved to itself towards empiricism, but of enunciating a mode of discourse which retains purchase on ‘philosophical’ issues (such as those of category-validation) while hurling the philosophical carapace of these same questions aside. It has good reason to do so since, as discussed above, nothing but stones and whirlpools – infinite ascent and circular regression – await the theoretical traveller who confronts either metatheory or first-order empiricism on their own, reciprocally distinguished, set of terms.

The argument just presented can be taken a stager further or, rather, it can be set into reverse gear. Sheer metatheory (philosophies and methodologies of science, for example) and sheer empiricism (positivism, sociology) are discourses which take in one another’s washing. Each relies upon the other to substantiate it: empiricism calls for the metatheoretically-prescribed demarcation criteria (for example, Popper 1972, p. 249) which might prise it away from mysticism, while philosophy of science is forced to humble itself by undertaking an ‘underlabourer’ role in relation to the theorising it claims to legislate for in order that it have something substantive to discuss. Russell, a philosopher by no means famous for his self-effacement, conceded this latter point when he remarked that any metatheory whose implications clashed with the results of physics could only be false.

All this said, two crucial theses are in order: (a) whenever theory and metatheory are prised apart then only on the one side positivism and, on the other, infinite ascent through metatheories (‘idealism’ in on sense of the term) can obtain. And (b) unless practical reflexivity is the condition of theorising the course of thought reported in my first thesis is inescapable. To be sure this ‘unless’ awaits rigorous proof. But notice its conditions: to avoid it we have to be able to unite first- and third-person discourse and theoretical and metatheoretical discourse minus practice reflexivity. Not a single utterance in the two-thousand-years-old tradition of Western philosophy suggests that this can be done. Practical reflexivity therefore appears as the catharsis of theoretical truth and of practical emancipation (involving demystification) alike.

Perhaps my argument may strike a reader as a top-heavy condemnation of what Bhaskar and Jessop have to say. Perhaps either of them distinguishes theory from metatheory in so rigorous a fashion as to qualify as a ‘philosopher’ in my above-criticised sense. Possibly practical reflexivity lurks somewhere within their works. In fact the opposite is the case; in demonstration of which I shall discuss a passage from each in turn.

Firstly, Bhaskar et al. (1988) reports that ‘social theory and social reality are causally interdependent’ in the sense that ‘social theory is practically conditioned by, and potentially has consequences in, society’. This sounds like practical reflexivity until we notice the causalism by means of which each of these two citations establishes the connections upon which they rely. What is missing is the in-the-course-of condition which, in my definition, allowed the thesis of an internal relation as between theory and practice to stand. And it was through this internal relatedness that first-order theory and metatheory were enabled to form a unity. Causal relations can only be external relations. Bhaskar’s hostage to causalist fortune severs first-person from third-person reflexivity and supplies exactly the conceptual gap within which sociologies of knowledge thrive. In other words a unity of theory and metatheory is far from being Bhaskar’s profferment. What obtains even in such encouraging formulations as the one just quoted is faithfulness to his philosophy-of-science (his Russellian and ultimately Kantian) roots.

Secondly, Jessop (1988, footnote 18) turns instead to the notion of a ‘double hermeneutic’ as developed in Anthony Gidden’s work. ‘A hermeneutic framework about a reality,’ says Jessop, is ‘already pre-interpreted by its participants’ (loc. cit.). Here too something like practical reflexivity appears to be signalled: interpretations arise from reality and inform reality (or practice) in their turn. However, if causalism undermines practical reflexivity so too – from as it were the opposite direction – does a sheerly hermeneutical approach. From a third-person standpoint (that of causalism) we are translated over into a first-person standpoint (that of meaning). Practical reflexivity for its part requires a totalisation of these standpoints in such a way that, reflecting on the presence of our theory in our practice and of ourselves as subjects in society (our object), we can see ourselves as socially conditioned (‘causalism’) and as socially conditioning (‘hermeneutics’) in, once again, the same breath. To draw two breaths here instead of one is to sever sociological from categorical self-reflection; thereby, it is to sever first-order from higher-order theorising and theory from practice itself.4 The ‘dialectic between structure and strategy’ announced by Jessop (1988a) signifies his endorsement of just such a dualistic approach. Jessop, like Bhaskar, writes ‘and’ where practically reflexivity writes ‘in-the-course-of’. External relations supplant internal relations, as is usual when sociology (or indeed any other first-order discourse) calls for metatheoretical – methodological or philosophical – support.

Thus I conclude that it is fair to tax Bhaskar and Jessop with missing the point of practical reflexivity and so reproducing, in different fashions, a theory/metatheory split. And a further point in this connection is to be noted: such a split has consequences not merely for the conceptual status of, but within, the first-order dimension of theorising itself.

In regard to Jessop something of this has already been indicated: hermeneutics distinguishes itself from causalism, and methodology from sociology, in the same movement as structure and strategy become first-order truth-claims controlling terms. In Bhaskar’s work it appears that the theory/metatheory severance is so severe that there is no relation at all between Critical Realism as a metatheory and the first-order political prescriptions Critical Theorists wish to raise. Bhaskar et al. (1989) opens with a summary of the philosophical claims set forth in Bhaskar et al. (1988) and then offers a list of political desiderata which he and his associates would like to see the left endorse: the connection between the first and the more lengthy second part of this article remains thoroughly unclear. Arbitrariness within first-order theory seems to be the primary consequence of Critical Realism’s theory/metatheory split. On closer inspection, however, an interrelation between the levels of Bhaskar’s discussion can be discerned at least in regard to the form of the first-order political theory his 1989 article presents. What this first-order theory amounts to is a defence of certain ‘models’ of democratic socialism held to be praiseworthy. Model-building is of course a central activity of social theory, from neo-classical economics through Weberian sociology, which understands itself as independent of philosophical dictats-from-above. Models, unlike philosophies, can be built and torn down in and through first-order theorising itself; they (or something like them) become needful under the sign of a theory/metatheory severance inasmuch as guidance has to come from somewhere in the matter of which categories fist-order theory is to employ. In other words, a totalisation as between theory and metatheory does obtain in Bhaskar; but it is totalisation which obtains only in the mode of being denied. It is atotalisation which goes forward behind the backs of the Critical Realists who are unconsciously its bearers. The theory/metatheory separation has consequences for, and within, both metatheory and theory: it constitutes the former as ‘philosophy’ or ‘methodology’ and the latter as a positivism which uncritically derives the models which guide it from the way of the existing social world. The model which it makes most sense to employ is the model in closes conformity to this world’s ways.

This may seem an unfair charge since, after all, a remodelling of existing society might amount to confronting society with its failure to live up to the rationale which animates it: it might amount to immanent critique. To such an objection I have two replies. The first is that immanent critique has to challenge not just the content of social relations but their form, and, theory being seen as a moment within social practice, not just the content of first-order truth-claims but their form as well. That is to say, it cannot merely be a matter of the empirical or ethical accuracy of models: and in the end it is only some such notion of accuracy upon which model-building relies. To be sure Weber says that his models (or ideal-types) are not to be understood normatively; and to be sure Kant says that the regulative ideas (in effect the building-blocks of models) which control his ethics are not to be construed in an empirical way. In each of these cases, the qualifying clause is intrinsic to the coherence of the model-building programme. But in the case of immanent critique it would have to be an ethical-empirical unity which enabled it to go forward. In other words the ‘models’ supporting an immanent critique would have to be inaccurate, empirical accuracy being infected with ethical desiderata and ethical accuracy being infected with empirical concerns. Precisely the is/ought distinction respected by Weber and Kant alike would have to be overcome. I suggest that the best way to resolve these conundrums is to say that immanent critique has nothing at all to do with model-building. And, on the basis of this suggestion, my charge of positivism against Bhaskar stands.

The second of my replies is ad hominem, but to the same effect. In fact Bhaskar understands the relation of metatheory (Critical Realism) to first-order theory in a way which precludes the notion of immanent critique. Bhaskar et al. (1989) announce ‘a distinction between (a) emancipatory action or practice and (b) emancipated action or practice, which are often unwittingly confused… [This distinction] is the distinction between socialism as a movement (or programme) and socialism as a society (or way of life)’.

I leave aside, here, the question of the egregious and ultimately bureaucratic elision of ‘programme’ with ‘movement’. Instead I should like to focus on the distinction between emancipatory action and its goal. And on this score I should like to confess myself as one who is confused, unwittingly or not, by the circumstance that Marx reported communism not to be a future ideal but as the real or actual movement of the working class. ‘Communism is neither the teleology of the capitalist system nor its catastrophe. It is … a concept that we can only formulate within the form of transition’ (Negri 1984, p. 165). Once emancipatory action is separated from its goal, the notion of self-emancipation which Marx celebrated throughout his life’s work is undermined. The matter is all the more strange inasmuch as Hilary Wainwright, one of the signatories to Bhaskar et al. (1989), was a co-author of a work entitled Beyond the Fragments in which the notion of ‘prefigurative’ political organisation – or in other words political organisation which quite precisely refused to separate the goal of action from action itself – was deployed against Leninist vanguardists, male intellectuals and other rightly-or-wrongly identified riff-raff. Prefiguration (both theoretical and practical) is in the event necessary to critique if critique is to be immanent, inasmuch as such critique has to be able to discover a basis within the present on which an emancipatory future might be built. Bhaskar’s ‘(a)’-and-‘(b)’ distinction severs action from the future toward which it aims and which (see writers as different as Hegel, Marx, Heidegger, Kojave, Bloch and Sartre) self-determines and self-defines it. His distinction alienates action from itself and places the question of the rightness or wrongness of such action (I understand these terms both epistemologically and normatively, and in the face of positivism’s is/ought distinction I would stress these latter forms’ unity) in the hands of metatheorists who, qua metatheorists, perforce construe philosophy and methodology as a practice-independent ‘realm of their own’ (Marx 1975, 5, p. 447). Stated the other way round: the alienation of action – the severance of itself from its goal, and the undermining of its self-determination – is a first-order consequence of the theory/metatheory split, whose roots lie in the theory/practice split (the subversion of practical reflexivity) itself. Marx reports the same thing when, in the terminological formulae of The German Ideology, he traces the alienation of theory from practice and the consequent constitutions of philosophy on the one hand and positivism on the other to the division between mental and manual labour. For Marx immanent critique turns upon practical reflexivity, since otherwise neither the rootedness of present theory in past practice nor the internal (‘prefigurative’) relation between present practice and a theory of the future – this ‘neither … nor’ comprising once again a single totalisation – can be understood. Thus, once again, I conclude that my charges against Bhaskar et al. (in this case to the effect that model-building and immanent critique count as entirely different conceptual procedures) are indeed fair.

In the preceding paragraphs I have offered a discussion of the consequences of the theory/metatheory separation (stemming from a theory/practice separation) within first-order theorising in a more-or-less unsystematic way. However, it is easy enough to draw the threads together and to show that one central difficulty underlies all that has been said. Once theory is separated from metatheory, in such a way as to permit the latter to dictate terms to the former – despite all the modesty of philosophical and methodological ‘under-labouring’, and despite first-order theory’s attempt to claw concept-formation back into its own domain through the idea of ‘models’ – universals are privileged over particulars in the constitution of first-order thought. Categories and terms are necessarily universals, in the sense that they amount to lenses through which a whole series of reciprocally distinct particular objects are to be seen. All thought proceeds in terms of universals so that a thought wishing to focus on particulars has to be a thought willing to encompass paradox, and (see Adorno 1973) to think against itself. On this score the basis for complaint against metatheory as a discrete discipline is not that it insists on the cognitive and normative-political role of universals – it could not do otherwise without succumbing to brute facticity and romanticism – but that it prioritises universals above particulars in such a way as to deny the latter a cetegorial and political voice. All that metatheory can prescribe to first-order theory is the set of spectacles (the universals or categories) which such a theorisation should wear, it its truth-criteria are to be defensible. Metatheory as separate from first-order theory constitutes the latter as a domain in which particularity has no rights. And yet it is particulars which first-order theory experientially encounters. I can stumble over an object in the dark, without my spectacles and half-asleep and before I can find the light-switch: that is, I can encounter it as a particular before (in terms of universal categories) I know what it is. Unsurprisingly, the history of philosophy records a debate between the cognitive roles of particularity and universality respectively. The universalism of Plato’s Theory of Forms declares against the particularism of the Sophists in the ancient world, Plato (in his Parmenides) averring for reasons best known to himself that there are no universal Forms of such undignified materials as ‘hair’ and ‘dirt’.5 In the early-modern period, the same battle was renewed in the form of a struggle between deduction (starting from universals: Descartes) and induction (starting from particulars: Bacon), deduction winning – as is consistent with the theory/metatheory severance upon which modern philosophy turns – the theoretical day. (Just this victory is celebrated in Karl Popper’s notion of refutability.) In our own times, philosophically marginalised figures like Nietzsche, Bataille and indeed Adorno line up on the side of particularity as a reference-point against philosophers of science, from Russell onwards, comfortable in their purely metatheoretical role.

One point to note here is that a metatheory prescribing terms to first-order theory separates universals (which have priority) from particulars (which count as secondary) within first-order theory itself. The lens and its object, not to mention the optician and his/her client, are after all two different things. Hence the need for ‘models’ or ideal-types or regulative ideas – Kant in his Critique of Judgement was one of the first to see the difficulty clearly – to mediate between categories and the objects which categories claim to know. Returning closer to home, the invocation of ‘intermediate’ categories in Jessop (see ch. 4) belongs within the same problematic. The programme of mediating between universal categories (for example, ‘capitalism’) and more particular categories (for example, modes of regulation, Fordism and post-Fordism and so on) emerges in the light of the argument just presented as a function of theory’s dissociation from practice by way o the theory/metatheory split.

Perhaps Jessop might reply to this that the problem of integrating general points with more specific ones inheres in the activity of theorising itself. Discussion of capital-in-general is a necessary rather than a sufficient condition for understanding the quite particular conditions of our 1980s capitalist world. Were he to say this his point about specificity should be well taken, but the terms in which his point is stated – a genus/species distinction as between capital-in-general and the modes in which it is articulated – would remain open to challenge in Marxist terms.

For it is not just a matter of universals tending to overshoot particulars, in the sense of different flowers being able to grow in any field universal categories might circumscribe. Nor (pace Adorno) is it just a matter of the uprooting of hedgerows undermining the survival-basis of flora and fauna although, to be sure (Adorno 1973, p. 362), the ‘philosopheme of pure identity’ connotes ‘death’. Rather the question turns on the circumstance that universals can be particulars in their turn. More strongly stated: in social life, all universals are particulars (or potentially so). In the self-understanding of social life, therefore, a genus/species distinction can have no place. Nor can a theory/metatheory distinction, since if universals are particulars then there can be no question of prescribing to social theory the universal categories in terms of which particulars are to be seen. Nor, finally, can there be a theory/practice distinction – in the sense of an external separation between the two – since the very abstractions and generalities and universals in which theorising goes forward are ones which have a vivid social (a practical) mode of life.

It remains to clarity the claims which my preceding paragraph has raised. Jessop styles his conception of an interaction between ‘structure’ and ‘strategy’ dialectical, but an entirely different conception of ‘dialectics’ will be presented below. Pace Jessop, it will turn not just on the notion of an interrelation (in the end a causalist one?) but of mediation, understanding this latter term in the sense of mode of existence: two terms reciprocally mediate one another when each is the mode (or form) in which the other subsists. Once again internal rather than external relatedness: universality and particularity subsist in and through one another in the dialectic of social life. To be sure they can subsist through one another contradictorily, that is, in the mode of being (each of them) denied, and this is what Adorno’s epigram concerning the ‘philosopheme of pure identity’ reports. Universality existing as particularity can countermand itself, and vice versa. For the present we can set these complexities aside, however, and confine ourselves to enunciating two theses: (a) an understanding of theory and metatheory as forming two discrete realms of conceptualisation – for example, philosophy and sociology – cuts the ground from under the idea of universals existing in and through practice while conversely (b) only practical reflexive theory can admit the possibility of universals existing in this way.

Practical reflexivity can admit this possibility inasmuch as its understanding of its own categories in the course of reflecting on its practical situatedness implies a similar understanding (an understanding in what I have called the ‘first person’) of the object, that is, society, upon which it reflects. In the course of theorising theories, or ‘ideologies’, it theorises practice as well. In other words it problematises not just theory and practice but the relation in which theory and practice stand. It unites its own subjectivity with social objectivity in the same movement as it construes ideology as a material existence while equally construing practice as shot through with potentially mystificatory ideas. Conversely, a sufficient condition of blinding oneself to the possibility of universals existing in and through practice (however contradictorily) is to remove metatheory from theory: once this is effected universals can only be prescribed to the understanding of particulars or, to state the same point the other way round, they can only be understood as abstracted from rather than as inhering in the particulars whose significance they claim to report. A comprehension of universals as particulars is precluded. Furthermore, if indeed as contended earlier a separation of theory from metatheory and practical reflexivity are mutually exclusive, it follows that my theses (a) and (b) report from two different viewpoints – those of theory and practice – one and the same thing.

All this, it seems to me, was sharply perceived and registered by Marx. His 1857 Introduction to the Grundrisse, distinguishing as it does between labour as production-in-general, that is, as an activity upon which all social formations depend, and labour as an abstraction which qua abstraction obtains only in a society of generalized commodity-production (Marx 1973, Introduction; Gunn 1989, pp. 105–108) announces nothing else. Abstract labour (a universal) is practically real, and indeed murderous, as anyone whose life depends on the sale of their labour-power quite directly knows. But it is practically real only within a particular social formation (a capitalist one) and only as a particular condition of such a social formation (inasmuch as for example unsold and unpaid domestic labour is no less a condition of the reproduction of capitalist relations of production from day to day). Marx’s critique of political economy turns on the refusal to elide abstract labour as production-in-general with abstract labour which, as a socially and practically existing category, appears in and through generalised commodity-production alone. To construe labour as a generally applicable category for all societies is to abstract from particular social conditions; to construe labour as a peculiarly capitalist abstraction, both in theory and in practice, is to approach it as an abstraction in and of the specific social world of the ‘doubly free’ labour whose historical preconditions the final section of Capital Volume I reports. To confuse these two senses of the category ‘labour’, and these two meanings of abstraction, is to render oneself complicit in the eternalisation of commodity production against which Marx, from the beginning to the end of his life’s work, so tirelessly declares.

Abstraction from is compatible with all manner of genus/species, universal/particular, metatheory/theory and practice/theory distinctions and dislocations; it entitles all manner of models and intermediate categories, and in fact requires them so as to prevent the universality of theorising and the particularity of theorisation’s object from floating reciprocally free. Abstractions in and of (otherwise known as ‘determinate’ or ‘substantive’ abstraction: see Gunn 1987b and Bonefeld 1987) is the preserve of practically reflexive theorising alone. Indeed practically reflexive theorising has to be able to construe its abstractions as determinate ones since, otherwise, it would require a metatheory standing outwith first-order theory as the court of appeal before which its ‘abstractions from’ were supposed to receive justification and make sense. Abstraction-from amounts to an induction which all philosophers of science from Plato (‘Don’t tell me which actions are virtuous; tell me what virtue is’) to Popper (Don’t announce to me a logic of scientific discovery, but describe to me the methodology in terms of which one is able to refute a point’) have agreed requires philosophical or methodological back-up of some sort. Abstraction-in-and-of confounds, no less than does Marx’s refusal to separate emancipatory and emancipated action, the theory/metatheory distinction upon which such a back-up turns. And in fact it is the category oof action, or practice, which is all-important. Once its moments, for example its project and its goal, are held apart from one another practical reflexivity’s totalisation implodes. Categorial appeal to an alienated practice reproduces the alienation of practice rather than the (possible) practice which, to use an ancient Marxist expression, might stand alienation on its head. If the theory/metatheory split has its roots in practice, it has practical consequences as well. And these consequences are of a murderous kind.

Bhaskar and Jessop agree. Labour as a contradictory relation between use-value and value-production – between its particularity and its universality, each existing in a capitalist world through the other but only in what I have called the ‘mode of being denied’ – can never be made sense of on a Bhaskar/Jessop account, since once universality is construed as genus and particularity as something which lurks beneath the notion of species everything becomes properly ordered: contradiction is marginalised as incoherence, instead of being construed as the movement of the universal/particular relation itself.

Between them, in other words, Bhaskar and Jessop by allowing philosophy to complement sociology subtract class struggle from the categories of Marx’s mature work (see Cleaver 1979; Negri 1984). One way of seeing Marx’s intellectual development – Althusser’s way – is to counterpose his youthful humanist zeal against the scientific historical materialism which he announced in his first exile and as he approached middle age. A quite different way – that which the argument of the present article suggests – is to regard his early critique of philosophy as being pursued through his later studies into a critique not just of political economy but of sociologism in all its guises: that is to say, in the name of practical reflexivity, he hunts down the consequences of a theory/metatheory separation to its (and his) end. First of all (the anti-philosophical polemics of the 1840s) comes his condemnation of metatheory as a ‘realm of its own’. Subsequently, in positivist England and as announced in the 1857 Introduction, comes his critique of first-order theory (positivism) which blushes whenever metatheoretical claims are invoked. Hence a deepening of an original insight rather than an abandonment of it. What confuses this record is the ‘historical materialism’ – the general theory of historical and social change – enunciated in The German Ideology of 1845–46 and summoned to memory in the form of a ‘guiding thread’ in the famous or notorious Preface of 1859. Historical materialism can sound all-too-like the kind of sheerly first-order theory of society which Jessop (with complexities) wishes to apply and which Bhaskar (with no less a sequence of complexities but in a metatheoretical register) offers philosophically to underpin. Perhaps we should get used to seeing historical materialism as a temporary aberration on Marx’s part. Nothing in his critiques of philosophy or political economy turns upon it, and indeed the notion of historical materialism as a general theory of society which awaits only its historically rich application reproduces the genus/species distinction against which the notion of determinate abstraction (the notion of the presence of the universal within the particular and vice versa) contends. Like any other sociology, historical materialist sociology is placed by Marx out of conceptual bounds.

This said, it would be wrong to conclude the present article on a textual note. I set out from the premise that apparently technical difficulties within concepts can carry potent political points. Marxists are used to seeing concepts pertaining to Taylorism (in the production process) and to market rationality as fatal to millions. My proposition is to the effect that concepts deriving from a theory/metatheory severance are fateful, and fatal, no less. If not the theory/metatheory split itself then, at least, the theory/practice split from which it arises can kill. Self-emancipation – initially a particular point within my article but now, construing my own points as practically reflexive, quite properly becoming a universal and conditioning one – counts as torpedoed at the moment when such a split occurs. In effect this is the débâcle of what used to be termed Western Marxism. The unity of theory and practice enunciated by Lukács in 1923 has fallen asunder; antiquaries with names like Bhaskar and Jessop peruse the ruins; and who it is that will reap the whirlwind remains to be seen.


Adorno, T.W. (1973), Negative-Dialectics (London: Routledge & Kegan Paul).

Althusser, L. (1969), For Marx (London: Allen Lane).

Althusser, L. (1976), Essays in Self-Criticism (London: New Left Books).

Althusser, L. and E. Balibar (1971), Reading Capital (London: New Left Books).

Bhaskar, R. (1975), A Realist Theory of Science (Leeds: Books of Leeds).

Bhaskar, R. et al. (1988), ‘Philosophical Underlabouring’, Interlink 8.

Bhaskar, R. et al. (1989), ‘Philosophy: More Underlabouring’, Interlink 13.

Bonefeld, W. (1987), ‘Marxism and the Concept of Mediation’, Common Sense 2.

Cleaver, H. (1979), Reading ‘Capital’ Politically (Brighton: Harvester Press).

Deutscher, I. (1972), Marxism in Our Time (London: Jonathan Cape).

Giddens, A. (1981), A Contemporary Critique of Historical Materialism (London: Macmillan).

Gunn, R. (1987a), ‘Practical Reflexivity in Marx’, Common Sense 1.

Gunn, R. (1987b), ‘Marxism and Mediation’, Common Sense 2.

Gunn, R. (1987a), ‘Marxism and Philosophy’, Capital & Class 37.

Horkheimer, M. (1972), Critical Theory: Selected Essays (New York: Seabury Press).

Jessop, B. (1988), ‘Regulation Theory in Retrospect and Prospect’, International Conference on Regulation Theory, Barcelona, June 1988, subsequently published in ‘Printed-Serie’ der Zif-Forschungsgruppe ‘Staatsaufgaben’ 1, University of Bielefeld, West Germany, Zentrum für interdisziplinäre Forschung, Bielefeld.

Jessop, B. (1988b), see Chapter 4 in this volume: ‘Regulation Theory, Post-Fordism and the State’.

Lukács, G. (1971), History and Class Consciousness (London: Merlin Press).

Marx, K. (1973), Grundrisse (Harmondsworth: Penguin).

Marx, K. (1974), Collected Works (London: Lawrence & Wishart).

Negri, T. (1984), Marx beyond Marx (Massachusetts: Bergin & Garvey).

Popper, K. (1972), Conjunctures and Refutations (London: Routledge & Kegan Paul).


1. Lukács 1971, p. 1. Lukács goes on to argue that dialectical method is available only from the class standpoint of the proletariat, and that indeed an external relation between method (or form) and content is a bourgeois deficiency in Kant’s thought.

2. See Althusser 1971, p. 59. Althusser announces that theoretical practice (science) is its own truth-criterion, but then adds that this thesis applied only to sciences which are ‘truly constituted and developed’ for their part. Let anyone make sense of this who can. I quote Althusser because, although I do not discuss structuralism explicitly in the present paper, so much of Jessop relies (via Poulantzas) on quite-rightly-unfashionable Althusserian thought. Notice, in the context of my later argument, the shiftiness as to the status of philosphy as between Althusser (1969) and Althusser (1976). It is by no means clear to me that comtemporary Marxism has sorted out the difficulties which Althusser’s intellectual demise left unresolved.

3. Philosophy used to reserve to itself questions not just of concept-formation (or ‘metatheory’) but of cosmology. These latter were both first-order and higher-order, and could be both because truth was seen as pre-inscribed in the very order of things. The ratchet-effect of enlightenment has (despite Green romanticism) once-and-for-all undermined any such programme, but – see below – it is far from clear that either sociology or philosophy-of-science (or both in combination) can perform the task which cosmological philosophy, as in Plato’s employment of mimesis, undertook. Our reckoning has to be with the movement of contradiction rather than with any static and harmonious whole.

4. See Giddens 1981, pp. 53–5. Gidden’s view is that se are in some respects conditioned by our circumstances and in other respects free. But unless I am in one and the same respect self-determining the notion of my freedom (my self-determination) falls. A bit of causalism and a bit of ‘human action’ undermines, through its eclecticism, the latter. Two breaths always prevent the necessary single breath when, from alienation towards freedom, we are required to totalise our resources and take just one leap.

5. Marx was less prurient. In a letter to Engels of the early 1850s he looks forward to having put paid to ‘all this economic shit’. My suspicion is that Marx’s shit (Scheiße) reports – even perhaps by way of classical allusion – Plato’s dirt. In other words Marx was prepared to dwell, for decades, within the dirt – the particularism as opposed to universalism – which Plato categorically and categorially renounced. Which ‘hair’ Plato most mistrusted remains an open question, but already in this martinal encounter we can see the battle between universalism and particularism breaking out.

Transcribed in July 2005, by PHFK.
Further reading  —  riff-raff.se