21 The author examines the relationship between Marxism and critical realism. He problematises the suggestion that Marx implicitly utilised a critical realist theoretical framework. He does this by exploring three areas of inquiry: epistemology and ideology; the method of abstraction; causal powers and social form. By exploring these areas, the author demonstrates that critical realism in fact pursues a different theoretical project to that of Marxism. Moreover, by severing the link between theory and practice, critical realism commits fundamental theoretical problems and errors which it initially claimed to have surpassed. The author concludes by suggesting that these problems were inherent within the critical realist project from the outset. Marxism and Critical Realism: The Same,Similar,or Just Plain Different? 1 by John Michael Roberts 1. Introduction SINCE THEIR BIRTH in 1975 with the publication of Roy Bhaskar s A Realist Theory of Science, critical realists have sought to advance an emancipatory project for the social sciences. Attacking both empiricism and idealism, critical realism argues that through the abstraction of concepts from reality causal mechanisms and structures can be examined which, although seen as the outcome of human praxis, operate independently of human praxis. Thus individuals can critically understand the structures which constrain them. According to Andrew Collier therefore a theory is realist in a strong (critical) sense if it makes four claims about knowledge: 1) Objectivity, in the sense that something might be real without appearing at all. 2) Fallibility, in the sense that claims are always open to refutation by further evidence. 3) Transphenomenality, in the sense that there is always a need to go beyond appearances. 4) Counterphenomenality, in the sense that deep structures can contradict appearances (Collier 1994: 6 7).
22 Capital & Class #68 It also used to be suggested by critical realists that a great debt of inspiration was owed to the works of Marx. Marx similarly conceived science as a process of abstracting concepts in order to comprehend underlying structures (Issac 1990: 18). But such admiration is also tinged with enmity. For Marx is often seen to be a deficient realist because he advocates a form of historicism. At his worst, Marx s occasional adherence to a monistic hyper-naturalism champions a form of biological evolution for the social sciences (Manicas 1987: 116). Bhaskar has put the point more plainly. If Marxism is to make any progress as a research tradition and escape a form of historicism it has four stark choices: the neo-positivism of analytical Marxism, the neo-kantianism of Habermasian communicative action theory, the neo-nietzscheanism of post-marxism or dialectical critical realism (Bhaskar 1993: 352). In other words Marxism should seek out a realist meta-theory (Outhwaite 1990: 374) in order to explore the complex levels through which the abstract dynamics of capital are mediated (Marsden 1998: 318 19). Perhaps the disquiet today over Marx s realist status is to be expected. After all Marx was never consciously working as a critical realist. But if this is the case some important questions present themselves. To what extent could Marx ever be assimilated to the critical realist cause? How does Marx s realism differ from his materialism? Are they the same, similar or just plain different? These are certainly pertinent questions because many theorists still professing adherence to Marxism take the critical realist assault seriously and attempt to incorporate Marx to a critical realist theoretical/methodological framework. 2 My purpose in this paper is to problematise some of the assumptions underlying this incorporation. My argument proceeds as follows. In section 2 I examine some epistemological claims made by critical realists. It is my contention that their epistemological insights do not allow for the development of an adequate theory of knowledge for the simple reason that they do not nourish a critical theory of ideology in any objective sense. In section 3 I examine some methodological claims made by critical realists. It is my claim here that the critical realist method of abstraction seriously misinterprets the same methodological project of Marx and thereby undermines their assertion of uniting theory to practice. In section 4 I examine some ontological claims made by critical realists. It is my
Marxism and Critical Realism 23 contention here that they are not objective enough in their self-professed aim of discovering causal powers because they fail to discover at the same time the determinative force of those very same causal powers. I conclude by suggesting that critical realists today prioritise method over the theory-practice dialectic. Marxists therefore have a fifth, more obvious option than those heralded by Bhaskar: the Marxism of Marx, namely historical materialism. 2. Epistemology and Ideology In this section I wish to suggest that the critical realist theory of knowledge is inadequate. Its inadequacy stems from a debilitating subjectivism, itself predicated upon the separation of appearance from reality. The result of such a severance is a lack of any critical theoretical tools with which to understand how objective social structures can be transformed. This lack is the outcome of a weak theory of ideology-critique. I begin first by outlining a version of the critical realist theory of knowledge. A critical realist theory of knowledge Critical realists agree upon the basic premise that in order to gain adequate knowledge of the world there is a necessity to develop a naturalist theory of knowledge; the idea that there is an essential unity of method between the natural and social sciences (Bhaskar 1989: 67). But this statement is immediately qualified by Bhaskar. Just as similarities are evident between both natural and social scientific methods, so are differences. Bhaskar suggests three: i.) social structures, unlike natural structures, do not exist independently of the activities they govern; ii.) social structures, unlike natural structures, do not exist independently of the agents conceptions of what they are doing in their activity; iii.) social structures, unlike natural structures, may be only relatively enduring (so that the tendencies they ground may not be universal in the sense of space-time invariant) (Bhaskar ibid.: 79). According to Bhaskar, these principles demonstrate ontological, epistemological and relational limits to naturalism. But if this is the case, how do we gain social scientific knowledge? Bhaskar s answer lies with his transformational model of social activity (TMSA). This model revolves around the claim that society is both the condition and outcome of
24 Capital & Class #68 human praxis, while praxis is the (conscious) production and (unconscious) reproduction of society (Bhaskar ibid.: 92). By adopting the TMSA we can gain adequate knowledge of the motivated productions of society along with the unmotivated conditions necessary for these productions. In line with his limits to naturalism, however, Bhaskar strongly urges us to follow an epistemological relativism objects and structures can only be known under particular definitions (Bhaskar 1978: 249). This is an anti-foundationalist theory of knowledge insofar that being is only contingently related to knowledge (Pratt 1995: 65). Bhaskar consequently avoids the pitfalls of both extreme relativism and extreme foundationalism by insisting that there can only be relative degrees of truth or falsity, adequacy or inadequacy, better or worse knowledge because knowledge is a practical experience which presupposes an ontologically structured world. Being is only ever contingently related to knowledge and the latter alters in form as new structures of reality are discovered (Outhwaite 1987: 36 44). But many fellow critical realists are not convinced by Bhaskar s argument (e.g. Benton 1981a; Collier 1994; Keat and Urry 1982). Collier for instance argues that knowledge need not be characterised as an internal means of assessment through specific descriptions. All that epistemology demands is that a theoretical practice be an enquiry into reality, that is, aims to measure its propositions against reality (Collier 1979: 92). In this sense adequate knowledge of the world is arrived at through practice, through the testing of ideas against the real world. By invoking a correspondence theory of truth statements of the world correspond in some way or another to the real world a priori and relativist theories of knowledge are rejected in favour of a definition of truth. Correspond here is specially chosen to pick out the relation that holds when as it is said, so it is (Collier 1994: 240). But even though Collier is critical of Bhaskar s conception of epistemology he nevertheless sets out from the same epistemological assumptions, namely the need to situate knowledge within human praxis. The philosophy of praxis and a materialist theory of knowledge There can be no doubt that the philosophy of praxis maintains strong links within the Marxist tradition, the notebooks of
Marxism and Critical Realism 25 Gramsci being a clear example. Praxis theory seeks to bridge the gap between abstract and concrete by suggesting that through practical activity humans reproduce the very structures which in turn produce them. Reality can only be meaningful for human actors to the extent that they creatively come to know and understand it. Thus knowledge is generated through practical activity (Hoffman 1975: 16 19; see also Hoffman 1986). Yet although Marx undeniably does talk about the practical activity of humans does this imply also: i.) that Marx is an adherent to praxis theory?; ii) that the praxis theory pursued by critical realists is a development of Marx s ideas? 3 If Marx does speak incessantly about human activity, he does so in relation to a particular kind of human activity. Marx insists that human labour, in contradistinction to human praxis, is the fundamental category for the reproduction of social relations. Labour is a useful category because it can also incorporate relations of exploitation, surplus-extraction and a shared cognitive schema. Labour-relations denote a total social experience because they establish the main parameters of struggle (Rosenberg 1994: 51: cf. Clarke 1991). Let me explain further. At a meta-theoretical level we can say that the world upon which humans labour is not a passive object. As Engels says, (the world) does not move in the eternal oneness of a perpetually recurring circle, but goes through a real historical evolution (Engels, Anti-Dühring 1975: 36). Different layers of the world are a particular and relatively autonomous manifestation and development of a universal material substance. The world is in a state of motion. Consciousness exists as a qualitative manifestation of this motion and, as a form of matter, consciousness assists labour in thinking about the appropriation of this motion to meet human needs. Yet to think is to think about something. Thus although matter exists independently of consciousness, consciousness cannot exist independently of matter (Sayers 1983: 18). Thought itself is a particular, distinct manifestation of matter. So much so, in fact, that history itself is a real part of natural history of nature developing into man (Marx 1970: 143). Consciousness and human agency arise from natural-historical circumstances in the form of a being for self-creation (Marx and Engels 1968: 39 40; see also Hoffman 1986: 117; Woolfson 1982). Consciousness, as a material substance, is thereby a reflection
26 Capital & Class #68 reality. As Marx announces: For Hegel, the process of thinking is the creator of the real world With me the reverse is true: the ideal is nothing but the material world reflected in the mind of man and translated into forms of thought (Marx 1988: 102). A proviso must be inserted at this juncture. Materialist reflection theory does not entail an empiricist reflection theory. The latter argues that consciousness merely registers images via our senses in a mechanistic manner while the former argues that consciousness actively reconstructs the real, active properties of the object, as part of labouring activity, which it then reflects in consciousness. We do not merely reflect reality, we also actively create reality. Life gives rise to the brain. Nature is reflected in the human brain. By checking and applying the correctness of these reflections in his practice and technique, man arrives at objective truth (Lenin, Philosophical Notebooks 1972: 201; see also Lenin s Materialism and Empirio-Criticism 1964: 129). This is an important point, if for no other reason than the fact that critical realists dismiss reflection theory exactly on the grounds that it is contaminated by empiricism. In the same breath they reject the monism associated with reflection theory because they see it as championing a sort of essentialism which seeks to reduce all cognitive relations to a single, ontologically, foundational cause (Bhaskar 1993: 217; Issac op cit.: 20; Lovering 1987: 290; 1990: 43; Sayer 1981: 14 15). Not only however is it clear that different forms of reflection theory exist, but it is equally clear that the easy dismissal of reflection theory invites critical realists to espouse a form of the structureagency couplet and the dilemmas which accompany it. It is difficult to perceive the TMSA in any other light. Just as structures constrain human actors so human actors produce consequences which go beyond their intentions. This is a symmetrical vision of society and the problems which Bhaskar discovered with similar theories, namely voluntarism and determinism, are embedded also within the TMSA (Bryant 1995: 84). I will expand upon the dualist problem here in later sections. For the moment I wish to stay with epistemology in order to see how the critical realist dualist understanding of this area translates into their ideology-critique. Let me first begin by saying something about the Marxist theory of ideology.
Marxism and Critical Realism 27 A tale of two ideology-critiques Marxism suggests that although a necessary relationship must exist between reality and how we think about reality, there also exists a degree of contingency with this relationship. Consciousness can always supply us with certain illusions about reality. Albeit, illusions still reflect truths about the objective world and thereby present us with illusionary forms of reality (Hoffman 1984: 101ff.; Sayers 1985: 134 41). Ideology is one manifestation of this illusion. Ideology, Marx consistently argued, relates to a limited social practice whereby the failure to solve contradictions in reality leads to an epistemological distortion concerning solutions about those contradictions. This distortion is not simply false but is itself a real manifestation of a contradictory essence. Essence and appearance are therefore both forms of the same reality. Correspondingly if ideology inverts the world in which we live, this is only because the social relations of which ideology plays a part are themselves already inverted. For example, the wage-form conceals the contradiction between the labourer s full day of work and the extraction of surplus-value. Ideology, by negating the inverted world of social relations in this manner, imbues appearances with an autonomous existence. Within specific class relations, therefore, ideology conceals objective contradictions which emanate from those relations. Under these conditions ideology can be said to benefit the interests of a ruling class. We arrive at a critical theory of ideology (Larrain 1979: 35 ff.). Critical realists, on the other hand, do not conceive reality and knowledge as unity-in-opposites. Separation between reality and appearance is accomplished because phenomenal forms are argued to be only realities of a causally secondary kind relative to real essences (Benton 1977: 178). This implies that appearances merely conceal reality, a reality which is only available to thought. Knowledge merely flows from the action of the subject, but is not acquired from nor reliant upon contact with the object through experience (Sayers 1985: 30). Hence critical realism s rejection of empiricism. By severing appearance from reality critical realists can only legitimately investigate ideology and, hence, knowledge, from a neutral standpoint. A neutral theory of knowledge confronts the world from a shaky epistemological platform because it has no objective reality with which to compare its
28 Capital & Class #68 own truth-claims from those made by competing truth-claims. Or rather, knowledge is not seen to be a distinct and unique form of an objective reality. Implicitly, therefore, the epistemological validity of competing truth-claims must be accepted and the chance to be critical diminishes considerably (Larrain 1994: 79 80). To give one illustration, Benton asserts that ideology can be divided into two facets: theoretical ideology and practical ideology. Specifically, (theoretical ideologies) may be thought of as articulations, in theoretical form, of the practical ideologies of the different social classes, the totality of these together in each case constituting the form of social consciousness peculiar to each class. Each theoretical ideology, then, is based on a corresponding practical ideology, embodied in the practices, experiences, struggles, etc. of the corresponding class and its individual members (Benton 1977: 179). What Benton seems to suggest is that each social class experiences, in its particular structural position, aspects of certain causal mechanisms, themselves situated within a mode of production, which then feed into the theoretical understanding which each social class gains of their social world. At first glance Benton s description sounds very Marxian. Yet on closer inspection it is decidedly theoreticist (Albury et al. 1981: 375 8). For ideology is situated within subjective experiences. Benton conflates the class struggle between workers and capitalists, a subjective struggle, with the objective contradiction between capital and wage-labour. It is the latter contradiction which Marx investigates in his epic, Capital, because this contradiction defines the social forms of human subjectivity by establishing the historical precondition for social existence within capitalism (Clarke 1994: 140). Benton examines the content but not the form of ideology. Ironically Benton, a naturalist, defends a position on ideology which many anti-naturalists would find no problem in supporting due to its subjectivist overtones. 4 More to the point, the sort of critique offered by Benton and critical realists generally resurrect some very awkward questions. How can normative (emancipatory), as opposed to descriptive, claims be supported? How do people develop discursive, reflexive and critical knowledge of underlying structures if a chasm is present between knowledge and the objective world? How can structures not only be reproduced but also transformed? (Baert 1996).
Marxism and Critical Realism 29 3. The Method of Abstraction Another substantial area in which critical realists initially claimed to follow Marx is with the method of abstraction. Indeed the declaration that critical realists pursue social scientific knowledge is usually grounded within this method. 5 Nowadays abstraction can be referred to without any discussion of Marx s use of it. 6 Of course there is nothing inherently wrong about this. But it does indicate the extent to which critical realists have shifted their concerns from the attempt to link theoretical and practical issues to the separation of each from one another. Thus in this section I will suggest that critical realists utilise a method of abstraction which is not grounded within the totality of social relations. As a result the critical realist method of abstraction, far from solving methodological problems, actually reasserts them albeit in a new form. Rational abstraction as Marx s abstraction? Clearly critical realists go beyond abstracting from mere surface appearances. They are against bad abstractions and chaotic conceptions. A bad abstraction is one which is based upon a non-necessary relationship between certain objects (Sayer 1981: 9). Instead of identifying the internal powers associated with an object, a bad abstraction investigates contingent and external relations among different objects. An example would be Max Weber s connection between the Protestant ethic and the development of capitalism. I will term this abstraction a chaotic abstraction, or abstraction (1). A rational or good abstraction, by way of contrast, isolates the necessary and internal properties of an object, namely its generative or causal powers. Once identified the diverse but contingently combined determination of those properties can be examined at a more concrete level. This move is particularly important because only then will we be able to establish the activation of the causal mechanism in question. An example might be the internal relationship between landlord and tenant, a relationship which assumes many guises in different contexts. In this way a precise definition of the object can be arrived at so that when a move is made back to the concrete one can gain a more accurate understanding of the object s interaction with a diverse range of elements. The finished product is the movement: concrete abstract, abstract concrete (Sayer
30 Capital & Class #68 1994: 87; see also Benton 1977: 165 9; Dickens 1996: 52 70; Keat and Urry op cit.: 111 14; D. Sayer op cit.). I will term this abstraction a rational abstraction, or abstraction (2). Critical realists claim to derive abstraction (2) from Marx s recommendations on method presented in the Grundrisse (1973: 100 109) and subsequently reproduced in A Contribution to a Critique of Political Economy (1972: 205 14). Marx also stresses the isolation, in thought, of an object from its concrete surroundings. But critical realists are prone to leave Marx s method of abstraction just at this point. Yet a difficulty presents itself if this is the case. In the Grundrisse and the Contribution Marx wishes to address the problem of how we come to have a theory. Marx was concerned primarily here with the methodological problem about how we take simple definitions and build them up over time so that we gain a clear sense of concrete reality. He makes no epistemological claim as to whether or not these definitions correspond to reality. He obviously couldn t do so because his remarks are shorn of any specificity. Therefore Marx is not interested in his methodological description given in the Grundrisse and the Contribution in trying to ascertain whether or not concrete reality arrived at through abstract definitions produces correct knowledge about the capitalist world (Ruben 1977: 150 62). However critical realists, unlike Marx, do tend to use abstraction (2), an ahistorical abstraction, to investigate specific social relations, such as capitalism, and then claim that adequate knowledge is produced. Marx s descriptive point is turned into a proscriptive agenda. The systematic nature of Marx s abstraction So how does Marx proceed? If external method must be avoided it would seem to be the case that method must inhere within, emerge from and reflect upon specific social relations and the social forms which those social relations produce. Its theorisation of its object, of its presence within its object and the validity of its categories (as categories appropriate to the theorisation of precisely that object) are not three separate conceptual moves but a single totalisation (Gunn 1989: 97 101). By reflecting the dynamic nature of the world, immanent analysis endeavours to reveal the practical nature of theory such that the potential for change is disclosed in the actual analysis (Postone 1996: 89). Let me term this abstraction as a systematic abstraction, or abstraction (3). Capital provides us with an example.
Marxism and Critical Realism 31 In Capital Marx claims that a commodity possesses both abstract and concrete labour. In other words labour possesses a dual function: it produces a useful product for others and it provides a means to procure a good. Goods are obtained through the exchange of commodities via the medium of money. In order for a good to carry exchange-value the individual labour embodied in the good must be abstracted from its concrete confines and subject to the socially necessary labour-time taken to produce the commodity in question. Concrete labour must be transformed into abstract, homogeneous social labour, namely value (Marx 1988: 125 ff.). Thus the labour represented in the commodity is not individual labour but social labour. And social labour is mediated through the objective form of money. A double-sequence is achieved. Exchange-value first serves as the appearance of the quantitative relationship between commodities. Marx then identifies socially necessary labour-time and abstract labour. From this base Marx returns to exchange-value, but does so with a new vantage point. For exchange-value can now be identified in its qualitative, objective form as the money-form (Banaji 1979: 31 2). Under capitalism labour serves as the main socially mediating activity which produces a set of structures that in turn dominate labour. This is the value theory of labour (Elson 1979). But how can Marx be certain that the commodity is the correct starting point at which to abstract? Marx starts the opening chapter of his epic by noting: The wealth of societies in which the capitalist mode of production appears as an immense collection of commodities ; the individual commodity appears as its elementary form (Marx 1988: 125). Why does Marx start with a form of appearance if his task was to comprehend an underlying reality existing beneath distorted appearances? The reason is simply because the most basic form of appearance in capitalist society lies with the exchange of commodities. Correspondingly the most basic concept to be developed by bourgeois ideologues is that of the commodity. The commodity, then, is the economic cell-form of bourgeois society. Or rather, the commodity represents the most abstract form of the double character of labour in capitalist society. Needless to say, by starting with the commodity Marx has as yet said little about the dynamics and reproduction of capital. For instance he has not at this stage of his analysis
32 Capital & Class #68 mentioned the extraction of surplus-value. However the commodity is the ideal point from which to begin a critique of capital because it presupposes that the society in which we are interested is one dominated by an immense collection of commodities. Marx thus grounds his analysis at the earliest opportunity within bourgeois social relations and simultaneously fixes a point of entry from which to understand capital and the commodification of labour power (Mattick Jr. 1997). Even so, Marx does not wish to suggest that a complete understanding of the commodity can be acquired at this level. After all the commodity still assumes a form of appearance so that its determination still resides in a relatively underdeveloped stage. Three important observations are worth stressing at this point in the argument. First, abstraction (3) is reliant upon a distinction between the method of presentation and mode of inquiry. Of course the method of presentation must differ in form from that of inquiry. The latter has to appropriate the material in detail, to analyse its different forms of development and to track down their inner connection. Only after this work has been done can the real movement be appropriately presented. If this is done successfully, if the life of the subject-matter is now reflected back in the ideas, then it may appear as if we have before us an a priori construction (Marx Capital I: 102). Why is the distinction between the two so important? The reason is related to the essence-appearance divide. If we wish to explain the developmental tendencies of capitalism we must be careful not to conflate those tendencies with actual historical development, such as the emergence of capitalism in England or in France. To do so would merely collapse the divide. Instead we must demonstrate the inner connection between forms of existence, forms peculiar to capitalism, and unravel those tendencies in actual history. Categories derived are validated by their ability to explain and express the inner movement between social forms. As each category is unfolded it is reflexively checked. Social relations are the context of the argument itself. In this respect Marx advances retroactively, by the ability to explain the developmental tendencies of capitalism (Postone op cit.: 141).
Marxism and Critical Realism 33 Second, the commodity presupposes advanced capitalism and advanced capitalism presupposes its historical precondition, namely the separation of the labourer from the means of production. And since we now understand the social relations which determine and fuse together the entire circuit of capital we alter our understanding accordingly of the commodity, including its determining properties. Abstraction and determination are therefore not the same thing (Clarke 1994: 137). Third, by beginning from a form of appearance from the commodity Marx begins from the point of political economy from bourgeois forms of thought and automatically critiques those forms of thought at the same time by situating them within the historical limits of capitalist social relations. The commodity is therefore not the start of the analysis but its result. From retroduction to retroaction, or, the necessity to start correctly The foregoing discussion suggests that the question of where to begin the process of abstraction is absolutely fundamental for method. How do critical realists address this problem? In one respect their starting point is arbitrary. It all depends upon the object of investigation. No presupposition about the object is made. All that is presupposed is a transcendental claim that the world is structured in a certain way. Once a phenomenon is detected which requires us to identify and explain the mechanism responsible for its existence, so critical realists argue, then it is necessary to build a model of the mechanism via the cognitive materials of knowledge about the phenomenon already gained. Information is collected about the mechanism which, if it was to exist, would account for the phenomenon in question (Bhaskar 1989: 19 20; Collier 1994: 22, 161, 163, 166; Sayer 1994: 107, 158 9, 207). A threephase scheme emerges: science identifies a phenomenon (or a range of phenomena), constructs explanations for it and empirically tests its explanations, leading to the identification of the generative mechanism at work, which then becomes the phenomenon to be explained; and so on (Bhaskar op cit.: 20). Correspondingly the intransitive realm (the real entities and structures of the natural world) can only be explored through the transitive realm (models and concepts of the natural world) (Bhaskar 1978: 21 4; Bhaskar 1989: 15 21: see also Keat and Urry op cit.: 62). However in normal conditions closed, experimental systems do not exist. 7 Indeed the social sciences,
34 Capital & Class #68 whose object of investigation revolves around unpredictable human behaviour, do not have the luxury of experimental closed systems. Mechanisms and causal powers cannot survive in a vacuum but only within open systems. Critical realists term this procedure retroduction. Some critical realists claim that Marx employs the method of retroduction. 8 Once we stress that the main problem with retroduction lies in its failure to posit a necessary cut off point in retroducing generative mechanisms, namely that pertaining to specific, historical social relations, then we have no choice but to dispute the critical realist claim here. More specifically we can argue that retroaction holds three advantages over retroduction. First it is entirely legitimate to retroduce another generative mechanism from the previous generative mechanism and so on. Infinite regress is thereby encouraged and we go round in circles trying to figure out the determining ideology (Baert op cit.; Gunn op cit.). This is not the case with retroaction. Second, the retroduction of structures do not reflect the movement and development of a complex whole, despite Bhaskar s claim to the contrary (Bhaskar 1989: 20). Retroduction instead prefers to draw upon models, metaphors and/or analogies of mechanisms which are familiar (on metaphor and realism see Lewis 1996). To retroduce is not to retroduce from reality. This is because the causal power discovered by abstraction (2) is not in any sense related to its objective precondition. In effect abstraction (2) conflates the mode of presentation with the object of inquiry by examining the determining properties of a causal power from the level at which it is abstracted. Retroduction is therefore an analytical move but not a synthetic one. Retroduction isolates the causal properties of an object from its contingent surroundings. But retroduction does not take the next step of drawing out all possible determinations of the causal power linked to its socially mediated development. It cannot do so because there is no sense of universality in the immediate phenomenon identified. So when the model of the causal power in question has been constructed critical realists pass directly to the next level of retroduction. In other words there is a sense of isolation between concepts. Critical realists develop concepts which do not inhere within each other. And so, for example, we do not have any ground to say that the
Marxism and Critical Realism 35 commodity inheres within capital. As a result the relationship between concepts is non-contradictory and ahistorical. The tendencies of a causal mechanism become self-ascribed and self-referential and are seen to evolve in a linear manner (Arthur 1997; Banaji op cit.: 39 40; cf. Harris 1979: 352). Third, and somewhat ironically, the insistence that an unobservable mechanism can only be identified through the construction of a hypothetical model reinstates the empiricist notion of sensory experience. According to critical realists, constructing the model in question is accomplished by analysing an observable object immediately presented to the senses. To the extent that this technique works in the natural sciences is not the direct concern of this paper. What can be said, however, is that it is just not the case that objects and relationships in the social world are unobservable for sensory experience. Rather, objects and relationships reside in distinct ideological forms of existence which demand that we reach an understanding of the specific (communicative) social relations which connect these forms to one another (Stockman 1983: 207). In order to see how these different methodological starting points impact upon an investigation of capitalism I turn to ontological issues. 4. Causal Powers and Social Forms So far I have reproached critical realists on both epistemological and methodological grounds. I now wish inquire into their ontological claims. I will argue that although critical realists outline a theory of causality which has superficial similarity to that utilised by Marx, they in fact pursue a different ontological project altogether. Indeed, I shall demonstrate that critical realists confuse two issues determination and causality and that such a confusion results in an argument for a neofunctionalist and neo-atomist theory of social structures and causal powers. I will demonstrate this by firstly outlining the theory of causality advanced by critical realism. Then I will show how such a theory has been transferred into a characterisation of a mode of production. I follow this through to see how critical realists understand the causal power of capitalism. I conclude the section by suggesting that problems implicit in the critical realist theory of causality imply that the main tenets of the critical realist argument collapse.
36 Capital & Class #68 A critical realist theory of causality A good way to approach the critical realist theory of causality is to reconsider empiricism. Empiricists assert that causality exists between two events when a regular, constant succession can be observed between them. Critical realists are sceptical of this causal theory on two grounds. Firstly, empiricists reduce a statement s empirical content to the actual or hypothetical regularity between independent objects. This is an atomistic conception of causality. Secondly, if causality can only be arrived at through observation then causal production as an independent phenomenon must be an illusion, an illusion which realists wish to reject (Harré and Madden 1975). 9 According to Bhaskar the empiricist error can be readily highlighted once we divide the world into three constituent domains: the empirical, the actual, and the real. While the empirical domain refers to the experience of events, the actual domain refers to the events themselves and these can be observed through experimental activity. The third domain is the most important and interesting of the three. The real domain refers to the intrinsic powers of objects which exist irrespective of whether they generate events (Bhaskar 1978: 56). As such objects can be said to comprise structures which cause powers to emerge. And in talking about the structure generating some power, you are also enquiring after a mechanism generating an event (Collier op cit.: 43). What critical realists defend here is a moderate essentialism, the claim that some things possess an evolving essence which, in turn, will enable us to isolate the causal properties of the object in question and relate those properties to external, contingent and non-essentialist objects of investigation. Causality alludes to the structure of an object and what it can do and only derivatively what it will do in any particular situation (Sayer 1997: 462 3; 1994: 104). Water for example is not only composed of H2O. Water is composed of a complex number of levels hydrogen and oxygen atoms, protons, electrons, neutrons and so on which interact in a variety of ways (Manicas 1987: 255). Consequently objects possess differentiated powers and tendencies depending upon the contingent conditions within which they are operating (Sayer op cit.: 107 110). The causal power of a mode of production Many critical realists claim that Marx works implicitly with a realist theory of causality. 10 Keat and Urry, for example,
Marxism and Critical Realism 37 conceive a mode of production as comprising a number of complex relationships which function in a particular way because they exist within the total structure of which they play a part. In any mode of production the single, causally dominant structure is tied to the relations of production. This dominant causal mechanism unifies each element of the mode of production within an overall structure. (T)he causes and effects of social phenomena are aspects, or applications, of the structural relationships between the elements (Keat and Urry, op cit.: 97). For capitalism, the defining structure is that between capital and labour. This relationship is functionally necessary for the capitalist mode of production (CMP) because it establishes the essential prerequisite for that mode as well as establishing the main contradiction. One such prerequisite is the establishment of value. Value, on Keat and Urry s estimation, is an underlying central structural mechanism of capitalism which requires elucidation if we wish to understand the surface phenomena of prices, profit, rent and interest (Keat and Urry op cit.: 104 6). But can Marx be assimilated to the realist theory of causality so easily? If we follow the methodological advice given by historical materialism then the critical realist theory of causality is found wanting. In Capital vol. III Marx argues: the specific economic form, in which unpaid surplus-labour is pumped out of direct producers, determines the relationship of rulers to ruled, as it grows out of production itself and, in turn, reacts upon it as a determining element. Upon this, however, is founded the entire formation of the economic community which grows out of production relations themselves, thereby simultaneously its specific political form (Marx 1966: 791). Marx locates the movement of history as one in which distinctive modes of surplus-extraction can be readily identified between different classes. In turn this implies that the means of producing a surplus exist as a contradictory unity with the relations of production. In order to ensure the continual reproduction of surplus-labour, the class character of the economic form must stamp its dominance upon all of society so that human activity is mediated by distinctive forms of exploitative existence, such as the state.
38 Capital & Class #68 Undoubtedly the critical realist exploration of Marx s conception of a mode of production covers similar terrain. However critical realism encourages a neo-functionalist reading of this term. It conceives the driving contradiction of each mode of production as being defined primarily between the forces of production and the relations of production one is functionally dependent upon the other and vice versa. Such a position does not, however, understand the relationship as a contradictory unity. Instead the relationship is posited as a systematic disequilibrium and class struggle is posed as the functional resolution (Stockman 1983: 218). The problem is rendered more precise once we begin to explore the specific form within which each causal power inheres. The value of determination and social form Social forms of existence under capitalism reflect the extraction of surplus-value because they comprise an underlying material content related to capital accumulation. However social forms also refract these social relations because forms exist as a qualitative and ideologically distinct moment of them. Therefore each social form simultaneously constitutes an object of the ideological environment with autonomous value and character (Medvedev 1978). This complex intrinsic-extrinsic dialectic is not functionalist because form does not arise as a necessity of class relations. Form is placed at a lower level of abstraction than that of the class character of the capitalist mode of production because we explore how form has been historically structured by those class relations (Clarke 1983; Holloway and Picciotto 1978; Jessop 1982; Pashukanis 1989). Depending upon the object of investigation, form and content can lead to different questions being asked. Emphasis upon form leads to questions about the kind of social relation being invoked whilst emphasis upon content leads to questions about what the social relation is about (Larrain 1983: 158). Certainly, as Tony Smith suggests, causal powers and generative mechanisms are useful concepts to help us understand the qualitative nature of each social form. For example Smith insists that generative mechanisms can aid us in identifying tendencies operating in different social forms which, in different concrete circumstances, might come into conflict with one another. Likewise a mechanism can refer to the various ways structures of social relations may determine
Marxism and Critical Realism 39 what happens (Smith 1997: 181). Even though Smith is not a critical realist per se., he nevertheless comes dangerously close in committing a common critical realist mistake, namely the conflation of causality and determination. We will now see that while causation is certainly a particular kind of determinative relation, it is not the only such determinative relation nor the most important (Ruben 1990: 231). The idea that an object of investigation possesses particular causal powers which exist in contingent and open conditions of existence says very little about the delimited and structured form of the power in question. It tells us little in the sense that the causal power has yet to be situated within the totality from which it is abstracted. For example the TMSA stipulates that, within the social world, the causal power of social structures are dependent upon unpredictable, interpersonal human behaviour. Causal powers, on this understanding, are relational. But this is vague and question begging. What does it mean to say that causal powers are relational? How are they grounded? As Suchting suggests, if causal powers are conceived as being inherent within each person, to the extent that structures only exist as the outcome of human behaviour, then we adopt an individualist and voluntarist position, contrary to the theoretical underpinnings of the TMSA (Suchting 1992: 28) Similarly the idea that value is a causal power, the critical realist assertion, in fact imbues value with a metaphysical status as something which is real before the development of advanced capitalism. As a result it is not at all clear how value can be said to determine anything, the reason being that as capital assumes ever more complex forms the determinative capacity of value recedes into the background. In other words critical realists tend to conceive value as existing outside of the totality of capitalist social relations (Arthur 1997). Plainly therefore to concentrate upon the causal power of value alone will not be sufficient for a full understanding of the determinate capacity of value. We require a supplement to the theory of causal powers (Suchting 1992: 28). Such a supplement can be discovered once an additional question is posed: what historical conditions does value presuppose? This is not just a petty point but of immense importance. If we start our analysis with the production of value we take production as being independent of exchange and circulation. We thereby cease to view production, exchange and circulation as moments in the totality of objective class
40 Capital & Class #68 relations. Critical realists, particularly those writing today, tend to downplay the unifying force of these moments, namely the historic separation of the labourer from the means of production. This historic separation is the very foundation, or the very essence, of our social relations. Or rather this foundation lies not outside the circuit of capital, it suffuses the circuit as a whole (Clarke 1994: 140). The reason for its importance is simple. Before technological innovation can be achieved capital must ensure that a supply of labourers exist who are free to sell their labour-power and free from ownership of the means of production. Securing both the socialisation of productive forces and the private appropriation of a surplus, this double-form of freedom also generates the main contradiction of capitalism because it has the power to actualise potential contradictions, such as the contradiction between use-value and exchange-value (Larrain 1983: 150-157). Obviously a critical realist theoretical framework can also highlight the importance of the capital-labour relationship. However the neo-functionalism pursued by critical realism implies that this crucial relationship cannot be placed within its proper social (class) relation. Keat and Urry, for example, point out that the capital-labour relationship is the most significant relationship in the CMP. Yet they stress that this relationship is essentially one based upon exchange. Labour and capital, the two central elements within the CMP, are defined functionally in terms of the dependence of labour on capital. The relationship between labour and capital is structured exploitatively through the capitalist appropriation of surplus-value created by labour. The nature of wage-labour is determined by this mode of appropriation (Keat and Urry op cit.: 107). If we take Keat and Urry s definition at face value, then class relations cannot be discerned. For, as Clarke suggests, within circulation capitalists and workers enter as individuals engaged in a free and equal exchange of commodities. No objective class interests are present. If we turn our attention to production, the problem still remains. Here we see only individual relationships between capitalists and workers. Certainly workers can act collectively to assert their claims over the product
Marxism and Critical Realism 41 produced. But workers can also act in unison with the capitalist if they feel threatened by competing capitals. Therefore we can only conclude that class interests are not defined individually, contra Keat and Urry (Clarke 1994: 139). The ramifications for the critical realist project are disconcerting to say the least. In order to see why I must return momentarily to the commodity. Essence as appearance and the return of atomism In section 3 we saw that use-value cannot serve as the main causal power of capitalism because it is ahistorical. Throughout history people have produced useful objects in order to survive. So if we wish to establish the specific properties of capital we have to look to exchange-value. Starting with the commodity, therefore, reveals to us the initial determinations of the commodity and hence the initial determinations of capital. They provide, what Banaji terms, the dialectical-logical basis for the derivation of capital (Banaji op cit.: 31), or what was previously called retroaction. However, and this is where Marxism differs significantly to the critical realist theory of causality, at each stage of the dialectical-logical process we return to exchangevalue, to a form of appearance. But this time the appearance represents a new development in the stage of capital. For example Marx moves from use-value and exchange-value to value and money. As such our initial starting point use-value and exchange-value is now discerned as being mediated rather than immediate (ibid.: 30). As section 2 demonstrated, the return to appearances is crucial if we are to avoid dualism in the hope of obtaining a critical grasp of reality. If the movement and transformation of an essence is to be explored by starting with and returning to its appearance then viewing the world as comprising three domains the empirical, the actual, the real collapses. It collapses, on a retroactive understanding, because the three domains are implicitly united through a fourth domain the historical. Or rather, the three domains appear as qualitative manifestations of the movement of the fourth domain. Three further consequences follow. First each generative mechanism is viewed externally by critical realists, severed from its internal and open nature to capital. Subsequently critical realism presents us with a new form of objectivism or atomism to the extent that the notion of causality invoked is unable to be internally linked with a social and ideological environment. And by prising apart structures