Zamtew (1988), XV (ii).ESSAY REVIEWA DIALOGUE WITH MARXISMANGELA CHEATER'S WORK deserves to be taken very seriously, both byAfricanists and those who have a special interest in Zimbabwe, and also by thosewho have a particular interest in the social sciences.Her output is prodigious. But it is not merely a question of quantity. Thiscorpus is informed by rigorous theoretical understanding and a familiarity with awide range of empirical data. Her own empirical work conveys the density andmulti-dimensional breadth of descriptive analysis. Furthermore it is characterizedby the degree of high moral seriousness with which she approaches the task ofanalysis and investigation. This moral seriousness is not reserved only for herattitude towards her own work and her profession. It spills over into her evidentconcern for her students and her desire to help them grasp the rigour of thediscipline of anthropology and its relevance to contemporary life. Furthermore, itmakes itself felt when she looks critically at questions of social policy and socialengineering. For it is one of her beliefs that social anthropology has somethingimportant to offer in this realm.Of the works under review, Social Anthropology: An Alternative Introductionstands head and shoulders above its older rivals in terms of contemporaryrelevance.1 This is achieved by a steadfast focus upon themes of intrinsic interestto the contemporary world Š those of work, production, and the relations andmodes of production Š in addition to the traditional focus on esoteric matters ofkinship, witchcraft and order in acephalous societies.2 A wide range of empiricaldata from over 100 different societies past and present in both the developed andthe developing world is introduced to illustrate and demonstrate varioustheoretical problems.For Cheater, the exotic past is of interest only in so far as it is relevant to ourunderstanding of the contemporary world. It is, therefore, relevant in two ways: totheorizing and to social policy. The relevance to theory emerges when, forexample, we try to model current knowledge of pre-class societies on Marxistlines Š a mistake, according to Cheater. The relevance to social policy emerges ina different way. All societies in Africa are currently undergoing rapid socialchange and experiencing more or less desperate attempts at social engineering bynational governments, aid organizations and international capital in order tomodify the worst aspects of that social change (or, sometimes, to benefit from theconfusion). But social policies which do not respect past modes of organization orhistorical and cultural beliefs which have survived into the twentieth century are,says Cheater, doomed to failure. Appropriateness and relevance consist, in part,of knowing and respecting people's customs, and to know people's customs onehas to examine them.1 A. P. Cheater, Social Anthropology: An Alternative Introduction (Gweru, Mambo Press,1986), vii, 288 pp, Z$5,80.2 See, for example, I. M. Lewis, Social A nthropology in Perspective (Harmondsworth, Penguin,1976); J. Beattie, Other Cultures (London, Routledge, 1965).181182 A DIALOGUE WITH MARXISMIn The Politics of Factory Organization and Idioms of Accumulation Cheaterpresents two full-length monographs based upon anthropological work con-ducted by herself in the last ten years.3 The former is a study of a weaving mill in asmall Zimbabwean town, the latter of small-scale commercial farmers in aPurchase Area. The Production and Marketing of Fresh Produce among Blacks inZimbabwe examines one aspect of the data derived from her research oncommercial farmers.4ORTHODOX MARXISM AND NEO-MARXISMThe organizational device I have chosen for examining her work entails focusingon the ways in which Cheater can be said to be writing within the Marxistparadigm. Now this is, I must confess, a difficult question. What we call theMarxist corpus is characterized by a widely varying range of styles with differentanalytical emphases, each often claiming to be the only legitimate and authenticheir to Marx's own work. Moreover, it has often been noted that the twentiethcentury writes within the parameters set by the nineteenth: Darwin, Marx andFreud having formulated the terms of the language which we now use for ourown scientific discourse. Perhaps Cheater's relationship with Marxist problematicis no stronger than this? But she conceives of her own stance as being positivelylocated within the perimeter of a neo-Marxism challenging 'orthodox' Marxism(SocialAnthropology, p. 16). She accepts, that is, the basic conceptual apparatusof Marxist thought as being at least heuristically useful for analytical purposes,just as other theoretical perspectives can be utilized in order to explain data (pp20-nNeo-Marxism can in general be characterized as a reformist response toperceived inadequacies within the 'orthodox' Marxist corpus. This reformismarises from a critique of Marxism as dogma, an openness to the possibility that atleast some major Marxist concepts are inadequate and a willingness to acceptnon-Marxist theoretical elements (to flesh out, for example, inadequacies oftraditional Marxism), together with a commitment to the ethical component ofMarxism currently conceived as a commitment to emancipation.For theoretical purists and 'conceptual hygienists' neo-Marxism is a puzzlingphenomenon. It is often hard to see what basis remains for the claiming that suchtheorists are Marxist at all, for fundamentalists cannot see how Marxist conceptscan be systematically challenged, shored up with concepts deriving from otherperspectives or even used within the framework of other theoretical perspectiveswithout this constituting an attack on the underlying paradigm,And there is a point to the notion of 'conceptual hygiene' within socialscience. Conceptual hygienists believe that ultimately social theories arecomposed (however loosely) of interconnected and interdependent propositions3 A. P. Cheater, The Politics of Factory Organization: A Case Study in IndependentZimbabwe (Gweru, Mambo Press, Zambcziana 18, 1986), xix, 156 pp. Z$ 13,50; A. P. Cheater,Idioms of Accumulation: Rural Development and Class Formation among Freeholders inZimbabwe (Gweru, Mambo Press, Zambeziana 16, 1984), xxii, 199 pp, ZS9.00.4 A. P. Cheater, The Production and Marketing of Fresh Produce among Blacks inZimbabwe (Salisbury, Univ. of Rhodesia, Supplement to Zambezia, 1979), v, 41 pp, Z$2,88.CAROLE PEARCE 183which stand or fall as a whole. They comprise, that is, a framework for a coherentworld view which makes possible the activity of explanation by providing a singlelens through which the social world can be viewed and mapped. Whileconceptual hygiene may be unappealingly fastidious the point which itemphasizes is, I think, a good one. For it is the lack of a central agreed theoreticalparadigm within the social sciences which is the single weakest factor in theirclaim to disciplinary seriousness. It is this lack which gives rise to eclecticism. Yeteclecticism remains unattractive, for ad hoc strategies are unlikely to lead to theconstruction of a unified body of theory with at least some of the characteristics oftheory in the natural sciences. This is particularly problematic when differenttheoretical perspectives constitute realms of discourse which differ not merely inscale and emphasis but also in kind at the level of formality, normativecommitment and methodological assumptions. Concepts borrowed from theo-retical frameworks which are essentially incompatible may lead to conceptualconfusion.Furthermore, there is no clear point at which such strategies can bedistinguished from the ordinary common-sense explanations of mundane actorsduring the course of their daily experiences. Ethnomethodologists point out thatsuch 'ethno-methods' of explication derive from explanatory 'recipes' which areculturally constructed and, therefore, part of the very social phenomenon that it isthe task of sociology and anthropology to explain. If social analysts cannot workwithin the clear parameters of unified theory then there is little to distinguish theirefforts from those of the 'naive agents' whose activities they describe, and thuslittle point in the social sciences at all.With this introduction it becomes clear that Cheater's adherence to an eclecticneo-Marxism is of interest for two reasons: firstly, in relation to mainstreamMarxism and its use in analysing social data and, secondly, in relation to socialtheorizing in general. And the point of the whole discussion is not the trivial one ofadjudicating Cheater's commitment to a particular theoretical perspective but oftesting the claim that social science can proceed as science under the programmewhich she is currently advocating.THE EMANCIPATORY AIM AND CRITICAL SOCIAL SCIENCEAccording to Habermas, there are three different types of interest which directand constitute the quest for knowledge. The empirical-positivistic sciences areinformed by an interest in control (in particular, the control of nature and of otherhuman beings). Hermeneutic disciplines are characterized by an interest inunderstanding, the paradigm case of which is intersubjective communication.Critical social science (a code-name for Marxist social science), however, ischaracterized by an interest in emancipation. This interest is expressed in thedesire to return humans to their authentic autonomous state, through self-reflection and self-understanding, by freeing us from those systematicallydistorted forms of action and understanding which keep us uncritical prisoners ofsocial reality. What this presupposes, therefore, is that, prior to critique, socialreality itself is ideologically experienced as an alien force which dominates andoppresses us.184 A DIALOGUE WITH MARXISM[The critical social sciences] are critical by constitution; it is this which distinguishesthem from the empirical-analytical sciences of the social order, as well as from thehistorical-hcrmeneutic sciences . . . The task of the critical social sciences is to discern,beneath the regularities observed by the empirical social sciences, those 'ideologicallyfrozen' relations of dependence which can be transformed only through critique. Thus thecritical approach is governed by the interest in emancipation, which Habermas also callsself-reflection?Critical social science must expose the ideological character of reality.6A necessary preliminary step towards this exposure of reality is, of course, anexamination of the instruments with which we approach such reality. Critiquerequires reflection on the very instruments of critique-methodological as well astheoretical. Critical theorists say that positivism, whether of the bourgeois orMarxist variety, has an interest in control and domination. It cannot, therefore, bean instrument of emancipation. The interest in emancipation and autonomy mustnecessarily work within a non-positivistic Marxist framework which engendersself-critique or self-reflection while at the same time allowing for a critique ofsocial reality.Habermas's theory of knowledge-constitutive interests has been challenged,7but is, nevertheless, useful for my present purposes, which are to situate the fourbooks under review. It is clear from her works that Cheater is motivated by aninterest in emancipation, an interest which, at the least, coincides with the generalframework of critical theory. This interest displays itself in its passionate respectfor the subjective rationality and autonomy of ordinary agents, the strategieswhich they take up in order to secure or defend their interests and the essentiallyreasonable nature of those strategies. Together with these beliefs goes acorresponding hatred of domination and injustice, whether the source of thatdomination is racist, colonialist or sexist, and whether the bureaucratic organs ofdomination are 'private' or 'public'. It is thus at the level of phenomenologicallyexpressed subjectivity and intersubjectivity that, for her, authentic agencyemerges. But it is also clear that her temper is empiricist, by which I mean,neutrally, to indicate the belief that knowledge of the social world is ultimatelydependent upon empirical observational statements. Her type of neo-Marxism,therefore, diverges both from the emancipatory interest of Western Marxism, inthat it is empiricist, and from orthodox Marxism in the ways described below.THE ADEQUACY OF MARXIST THEORYNow, to what extent and in what manner can one subject Marxist analytical toolsto critique and still be said to be operating within a Marxist framework? We makea start on these questions by considering Cheater's own brand of realism andempiricism.For Cheater, theory and description confront each other in mutual challenge.5 P. Ricoeur, Hermeneulics and the Human Sciences, ed. J. B. Thompson (CambridgeCambridge Univ. Press, 1981), 82.6 J. Habermas, Knowledge and Human Interests, trans. J. J. Shapiro (London Heinemann 2ndedn., 1978).7 R. Koat, The Politics of Social Theory (Oxford, Blackwell, 1981).CAROLE PEARCE 185Theories which cannot account for specific data must cede to other perspectiveswhich appear to have more explanatory power. Somewhat dismissive of Levi-Strauss (Social Anthropology, p. 20) she is, nevertheless, open-minded aboutmany different theoretical perspectives, an eclecticism which allows her to acceptthe importance of exploring microprocesses (Idioms, p. 176) transactionalism,social dynamics at face-to-face level and even to play with the idea ofethnomethodology and phenomenology (Social Anthropology, p. 25). Her mostimportant theoretical problem, however, concerns her dialogue with Marxism. Inevery case her choice of theory is governed by the pragmatic question: to whatextent does this theory help us to better understand the data? This suggests that shebelieves that, at least to some extent, data Š or description Š is 'innocent' ortheory-free. If it were not, there would be no possibility of the confrontationbetween theory and data ever taking place. As an empiricist, data is, for her, thebasis of theorizing and is therefore epistemologically prior to theory, whether it bethe data of subjective experience or that gleaned from her observations, reports,calculations and measurements, newspaper reports or archival research. Theory ismeasured to the extent to which it proves adequate to the data, in true empiricistand realist fashion. 'The very layout of this text reflects my concern to"domesticate" Marxist perspectives' (Social Anthropology, p. v). By this shemeans to do two things: firstly, by working from within Marxist theory, to bringthese perspectives 'home', to make them familiar tools for analysis and thus helpus understand our own situation with greater clarity. The second aim is clearlysubversive. By 'domesticating' Marxist perspectives we tame them and put themto use. To apply critical instruments is at the same time to test their usefulness.Often, according to Cheater, we find these instruments inadequate for the criticaltask for which they were created.Cheater's critique of the instruments of analysis is not reserved for Marxismalone, nor does it emerge from a functionalist standpoint defending itself againstMarxism. On the contrary, she points out the essential similarities betweenMarxist and functionalist methods of analysis: 'far from being mutuallyantagonistic, then, Marxist and structural-functionalist approaches may beentirely compatible and even complementary, for functionalism especially isholistic' (Social Anthropology, p. 15). She notes, quite rightly, that thefunctionalist perspective is neither necessarily conservative (p. 11) (though it maybe inadequate) nor incapable of dealing with social change. Theory is the meansby which we achieve an understanding of the empirical world but, for Cheater,the success of theory depends upon the extent to which it can capture a pre-existing reality. What we must here note, however, is that functionalism, aspositivistic social method, is compatible with Marxism only when the latter, too,is positivistic.In Social Anthropology she criticizes Marx's 'evolutionism' (pp. 12,48,201).She demonstrates that the theory of pre-class societies is over-simplified in anumber of ways. 'The demise of orthodox Marxism is related mainly to its limitedability to explain societal data in non-western systems' (p. 17). Firstly, in pre-classsocieties, the 'base' in the relations of production is often not 'economic':economic activities are organized on the 'ideological' basis of kinship andmarriage (p. 13). Similarly, the 'ideological superstructure' often provides theconditions necessary for the economic base: i.e. the relationship of dependencebetween base and superstructure should, she suggests, be inverted (p. 214).186 A DIALOGUE WITH MARXISMSecondly, Engels's (hopeful) notion that in pre-class societies relations betweensexes in marriage (p. 144) and other social relations (p. 140) are characterized byequality cannot be sustained by ethnographic description of such societies.Thirdly, the characterization of pre-class societies as being based on 'communalownership* is too crude, for it masks the fact that land in such societies was often'owned' by individuals or families, if we accept that rights of usufruct arecontained in the notion of ownership (pp. 48-9). Lastly, the notion that the statenecessarily represents the interests of the ruling class has been undermined bystudies of societies which have a 'state' but not a formal 'ruling class' (p. 174).Turning her attention to applications of Marxism in contemporary classsocieties, Cheater notes that there are major problems in using Marxist conceptsin the whole sphere of worker consciousness, class consciousness and falseconsciousness (pp. 95-7,241,259). She notes that the analysis of religion as anideological manifestation of man's alienation from himself fails to explain how, insome circumstances, religion is not necessarily quietist in form but may becomethe vehicle of political protest (p. 240). She points out that not all forms ofinequality derive from social classes created in the relations of production, sincethe major form of inequality Š that of gender Š is not based on class (p. 98). Shesuggests (p. 169) that emergent classes in Africa may be more accuratelyconceptualized as Weberian 'status groups' rather than Marxist 'classes'.Furthermore, in the political sphere, she notes (p. 95) that structural relationshipsbetween independent governments, labour and capital, are more complex than isusually admitted in Marxist theory.It is clear that the very skill of its structure and immediacy of its appeal andrelevance is dependent on the way in which this text is organized around theMarxist concepts of modes and relations of production and its focus on Marxisttheoretical problems. Reading through the text one finds in microcosm anintroduction to almost every significant Marxist theoretical problem and, ofcourse, to a lot more besides. Thus, for example, on the concept of ideology sheoutlines the Marxist theoretical problem of whether ideologies belong, conceptu-ally, to the economic base or the ideological superstructure. She examines theHindu caste system which provides, as gender does, an 'ideological basis forallocating roles in production and society' (p. 218). After discussing thelegitimation and generative mechanisms of ideologies she turns to the relationshipbetween ideologies and social structure among the Kalahari of the Niger delta andcomments:the belief system as a whole went beyond simple refraction to conceptualise limitedalternatives to the existing social order. These alternatives were strictly controlled. Theydid not seek to to replace so much as to maintain this unitary system, by providingsafety-valves to release dangerous pressure on it (p. 227).There is, of course, more to religious and cosmological beliefs than this. Suchbeliefs attempt not only to explain the misfortunes and irregularities of social lifeby answering the essentially causal question 'Why?', they also (according tofunctionalist theory) represent ethno-theories about the cosmos. New religionsare formed when traditional cosmologies can no longer relate to a changed socialorder. An extended discussion of the Cargo Cults follows, bringing us back tothe question of why 'ideological mystification so often chooses a religiousCAROLE PEARCE 187idiom' (p. 234). This question highlights one of the difficulties of functionalistand Marxist analysis. For, if religious beliefs are merely epiphenomenalreflections of other elements of the base (or superstructure) how can we accountfor their content as specifically religious experiences and, furthermore, how canwe account for the force of the belief? To pursue this line of argument is, of course,to threaten the whole fabric of the base/superstructure distinction together withthe epistemological priority of the economic sphere, as Cheater points out. For, ifreligion is to be seen merely as false consciousness:We need to ask ourselves whether the concept of false consciousness is useful... Thisis a major problem area for social anthropology, for most of our theoretical approachesassume that, below the surface of individual conciousness and motivation, lies a deeperanalytical reality. We should take care not to assume that such reality renders our surfaceconsciousness false by definition (p. 259).In The Politics of Factory Organization two of the three objectives of herstudy unfold from within Marxist theory. These are: to study the 'contradictions'emerging from the relationships between labour, capital and the state incontemporary Zimbabwe; and secondly, to examine the data for the evidence ofthe growth of'class consciousness' among the urban proletariat of Zimbabwe at aparticular period of time and within the context of micro-political conflict: 'Thisissue Š indeed, all of the material presented here Š should, therefore, be seenagainst the broader background of crises of capitalism and Zimbabwe's attemptedtransition to socialism' (p. 142).It is to be noted that in all of these works Cheater pointedly refuses to use the'normal anthropological preferences for the "ethnographic present" '(p. xiv) andmakes it clear that her findings do not extend in time and space beyond the periodand context of her actual research. In this factory, then, as was the case throughoutZimbabwe after Independence, a Workers' Committee designed on theGramscian model was established by the state with the intention of replacing'non-political' trade unions. The new Workers' Committee, however, found itselfin competition and conflict with a number of other organizations designed toprotect labour. This, in addition to institutionalized practices of management(including racial prejudice, organizational inefficiency and financial incompe-tence) in the context of the changing economic circumstances which Cheatercharacterizes as crises of capitalism and the phenomenon of dependency,provided the context for the conflict between labour and management.Power struggles played out within the work situation are of particular interestto Marxists, for it is in this context of conflict that the conditions for the breedingof class consciousness and class action are said to exist. If, therefore, classicalMarxist theory is correct, conflict between managers (or owners of capital) andworkers ought to generate class consciousness among the workers, manifested, atleast, by a solidarity amongst workers.Political conflict within the workplace can, however, also be seen inWeberian terms as a conflict, not over the means of production, but overconsumption and status. From this perspective, conflict is not a prerequisite (anecessary and sufficient condition) for the breeding of class consciousness, but ischaracterized by individualistic strategies in jockeying for improvements toconditions of work, pay, hours, housing, and so on.188 A DIALOGUE WITH MARXISMShe shows that the conflict was resolved with the agreement of the workforceand the government agencies involved, by laying off 80 men out of a workforce of540 (p. 135), rather than by a show of solidarity amongst labour. The irony of thisis that the state and its agencies, despite their overt commitment to the protectionof labour, ultimately supported both management and the status quo againstlabour, thereby contributing to the latter's fragmentation.What does this have to say about class consciousness? Cheater suggests twothings. Firstly, this study provides no evidence for the development of classconsciousness at this time of crisis, contrary to findings elsewhere (andparticularly in Zambia) (p. 144). Does this mean that crises of capitalism do not(necessarily) lead to class consciousness? Cheater is non-committal on this score,pointing out that these textile workers were reacting to, rather than initiating, thecrisis. Her own pragmatic explanation is that workers robbed of initiative (p. 145)may be more likely to resort to individualistic strategies to protect themselvesrather than present a unified face to management.Secondly, it is clear that there was an internal conflict between socialism as ameans of development and socialism as the representative of labour. This conflictexpresses itself in practical terms as follows. Under socialism the state has a dutyto protect labour against exploitation by capital. Yet under 'African socialism',too, the state is economically dependent upon the success of capital, for wages canbe paid Š in the private and the public sector Š only if profit is generated, thusgenerating the basis for taxation. In a conflict, therefore, between capital andlabour the state may decide not to bite the hand that feeds it, and to support theinterests of capital rather than labour. Emasculation of an independent tradeunion which, though 'non-political', is not an agent of the state, can be conceivedof as a pre-condition for the emasculation of the labour movement itself.8 Thissuggests two readings, one from within Marxism, one from without. Either theState deludes itself by believing that it can promote socialism, for according toMarxist theory, universal emancipation is predicated upon the destruction of theState. Or Marxist theory is wrong, for conflict between labour and managementdoes not give rise to class solidarity.In her Idioms of Accumulation Cheater comments:In a very real sense, economic development is the vehicle of class formation. Equality inpoverty is not really equality at all. As wealth is generated through production, so classdifferentials emerge, in part through patterns of appropriation of the product itself (p. ] 35)We cannot understand this process of class formation at the national level, withoutknowing in detail how it operates in local societies, and how these two levels interlock Myaim here is to show how economic, social and political processes in Msengezi not onlyshape this local society but are equally part of shaping the collectivities and theirinterrelationships in Zimbabwe (p. xxi).In this work the paradox of Cheater's attitude becomes clearer. Workingwithin familiar Marxist terms (relations and modes of production) and, as in allher works, providing depth to these concepts with others borrowed from main-8 See South African Labour Bulletin (1987). XII. vi vii.CAROLE PEARCE 189stream social anthropology (kinship networks, family relationships in the divisionof labour, ethnic groupings) she draws in dense and rigorous detail a lucid pictureof the two main ways in which these small-scale capitalist farmers have managedto become successful accumulators. Cheater uses the metaphor 'idioms' ofaccumulation to indicate that one of the features of the language of self-perceptionis that it provides us with ways of conceptualizing strategies open to us. For thefarmers of Msengezi, two alternative strategies were, theoretically, open: the'traditional' idiom of accumulation, via the mechanism of polygyny, and the'modern' idiom, modelled upon the White capitalist farming peer group withwhom the Msengezi farmers interacted. In Marxist analysis the mode ofproduction (and therefore of accumulation) is intimately interdependent with therelations of production. Successful accumulators could choose to accumulatethrough polygynous marriage, using wives and children as a source of cheaplabour (where labour power is the source of value and therefore of profit). Or theycould choose the (more difficult, but more prestigious) method of hiring labour,often using non-agricultural sources of income for original capitalization of thefarm and relying upon 'modern' methods of planting and fertilizing to accumulatehigh yields. Cheater shows that the most successful farmers tended to use the'traditional' rather than the 'modern' idiom, but points out that the success of thisstrategy depended to a large extent on the farmer having previously dislocatedhimself from those aspects of 'tradition' which would require him to redistributeany surplus. Accumulation depends in the first instance, among both types offarmers, on the strategies seized to shrug off 'extended family' obligations.As a result of this research Cheater claims that there is a great deal more classdifferentiation within Black rural areas than Marxists would have us believe. Theclass structure of a country like Zimbabwe cannot legitimately be said to besimply superimposed on the racial structure. Nor can it be said that rural farmerscould or would present a united alliance with the proletariat. Purchase Areafarmers explicitly see their class interests as capitalist in nature.From these considerations emerges the following conclusion: 'Farmers in thePurchase Lands may be Black, and they may be undercapitalized, but in myopinion they are part of the capitalist class in Zimbabwe. From the material that Ihave examined, I do not see how any other conclusion may be drawn' (p. 176).And, as part of the capitalist class they contribute to the generation of wealth aswell as to the process of class differentiation.In The Production and Marketing of Fresh Produce among Blacks inZimbabwe, Cheater challenges Davies's Marxist analysis of the informal sector9from the point of view of the data which she collected during her research onMsengezi. The main points at issue are whether the 'informal sector' is a productof 'formal sector' capitalist practices, if the two can indeed be analyticallyseparated, and what policies a socialist government should adopt towards'integrating' the former into a socialist economy. Cheater says that the two cannot9 R. Davies, The Informal Sector: A Solution to Unemployment? (Gwelo, Mambo Press;London, Catholic Institute for International Relations. From Rhodesia to Zimbabwe 5, 1978).190 A DIALOGUE WITH MARXISMbe analytically separated, since the same individuals practise both kinds ofeconomic activity, that the 'informal' is not a consequence of the 'formal' and,lastly, that no government, socialist or otherwise, could incorporate (i.e. control)the informal sector. 'My own view is that we need a new kind of analysis, one thatcan handle the analytical question of how indigenous . . . economic activityarticulates with industrial production more satisfactorily than simply relegating itto the non-capitalist periphery . . .' (p. 30).Core Marxist ConceptsTo consider which (if any) of Marx's concepts are 'core' and which are not wouldbe fruitless, for this matter is quite possibly one that can be determined only byfiat, according to one's intellectual tastes. Some points of relevance can, however,be raised. One is that many Marxists have had difficulties with the conceptscriticized by Cheater.It has been denied, for instance, that the base/superstructural problem (atleast, when conceived as economic determinism) and the idea of historicaldeterminism are part of Marx's thought, let alone central to them. It has also beendenied that Marx's 'materialism' involves the same dichotomy (but this timeinverted) as Hegel's dichotomy between materialism and idealism: it is thedichotomizing tendency itself which is the source of the error and this error hasbeen imported by his epigones into Marx's own works.In addition, it has been claimed that Marx, in emphasizing the importance ofconcrete historical investigation, would not have pre-empted investigation bypostulating a priori what kinds of structures and relationships one would find ineach particular society and epoch. As far as the problem of consciousness isconcerned, Lukacs Šand many other Marxists writing in Europe Š find greatand ultimately insoluble theoretical difficulties with the whole problem ofconsciousness, subjectivity and objectivity in Marxism.10 Cheater is, therefore, ingood (Marxist) company by being satisfied with relatively few Marxist concepts.One of these is that complex of propositions called 'the labour theory of value',which she utilizes in Idioms of Accumulation. This cluster of concepts, too, hasbeen heavily criticized both within and outside the Marxist corpus. She does notfocus (critically or uncritically) upon what many take to be a core concept of theMarxist critique of social reality: the concept of alienation. Nor does she directlyallude to her interest in emancipation. Yet she consistently adheres, throughoutthese works, to the notion of production, underpinned by the idea of relations ofproduction as analytically central to the understanding of social reality.It is this adherence to a minimalist Marxist conceptual apparatus, togetherwith her critical acceptance of the Marxist theoretical problem, which entitlesCheater to describe herself as a neo-Marxist. But we must agree that this is aneo-Marxism so attenuated as to be almost invisible. It also renders problematicher empiricism and pragmatism, as I shall now show.10 J. Elster, Making Sense of Marx (Cambridge, Cambridge Univ. Press, 1985); A Giddens,Capitalism and Modern Social Theory (Cambridge. Cambridge Univ. Press. 1971), esp. 22 ff;L. Kolakowski, Main Currents of Marxism (London, Oxford Univ. Press. 3 vols., 1978). esp. Vol. II.CAROLE PEARCE 191THEORY, DESCRIPTION AND THE STATUS OFTHE EMPIRICALLY OBSERVABLEMarxist realist materialism transcends what is scornfully called 'simple-mindedbehaviourism' in that it makes allowances for the existence of at least three typesof entities which are not observable apart from this method. These are ideas,customs, and beliefs which are not necessarily reducible to physical things,although they are the material products of man's labour and material forces in thesocial world; the entirely theoretical constructs of false consciousness andideology; and structures and forces.Enormous problems now arise for the Marxist undertaking empiricalresearch. For the Marxist materialist must, in addition to conducting empiricalinquiries, investigate the gap between appearance and reality. The problem ishow to study empirically and respect theoretically concrete social reality while atthe same time making the distinction between appearance and reality andbetween true and false consciousness which Marxist theory itself demands as acrucial methodological distinction.The ordinary empiricist, of course, has no difficulty with this. All substantialknowledge, for the empiricist, derives from experience. Advances in scientificknowledge must, therefore, be predicated upon observation, and statementswhich cannot in principle be related to observational statements are simplyunintelligible.Marxists, but not only Marxists, would be unhappy with this as a conclusivestatement of method for it leaves no room either for theorizing about a realitybehind ideologically distorted appearance or for analytically grounding theconcepts of ideology and false consciousness and nor, therefore, for a moretough-minded theory of emancipation than that available to the non-Marxisthumanist. The appearance/reality distinction entails that emancipation isnecessary precisely because authentic autonomy exists only in potentiality andnot in actuality. Human actions can be considered rational and autonomous onlywhen the potentiality coincides with the conditions necessary for its expression.Mere empiricism cannot handle a potentiality within which reality is to belocated, especially since theory locates this reality as quite external to our actualexperience.If this is so then it is very difficult to undertake empirical sociological andanthropological work as a Marxist, for one has to obey at the same time twocontradictory methodological precepts. The first is that the material world marksthe limits of reality. The second is that the material world is composed ofappearance and is thus itself a mask of reality. Interesting consequences followthis conflict. Firstly, our ability to accept as true people's subjective understandingof themselves or of the world is undermined. Secondly, in proportion asappearance is severed from reality, so diminishes the possibility of any non-trivialdiscoveries being made in the social sciences. If there is no way in which anempirical observation can falsify any single Marxist proposition, when theempirical is conceived as the world of appearances, there is equally no way inwhich empirical observations can lead to new discoveries. If theory is free tochoose its own terms of reference then Marxist observations of the world are192 A DIALOGUE WITH MARXISMincorrigible to the extent that they describe a world already pre-given by Marxisttheory.The Empirical Status of ConsciousnessCheater discusses this fundamental problem in 'Contradictions in modelling"consciousness": Zimbabwean proletarians in the making?'11 In this importantarticle she points out that the notion that (subjective) consciousness is essentiallyflawed can be taken to much more radical lengths than Marxists are prepared togo while still retaining faith in the essentially rational and autonomous nature ofconcrete actions and beliefs. For we do not have to stop, with the Marxists, at thenotion that only certain kinds of consciousness (non-revolutionary or non-classamong the working class and all 'petty bourgeois' consciousness) is flawed. Wecan deny altogether that there is one reality and thus that there is only one(privileged) way of characterizing it.By denying that there is a reality (usually conceived of as sets of structuralrelations or functional systemic needs) which lies beneath and generatesappearances we gain a number of advantages and at least one loss. The firstadvantage is that to deny the ontological basis not only of the Marxist paradigmbut of all 'grand theory' as well is to deny the epistemic as well as ontologicalprimacy of the base, however characterized. The second is to admit that subjectiveexperience may be partial and correspondingly flawed, but it is not for that reasoneither irrational or the epiphenomenal product of underlying generativemechanisms or forces. This leads to a further possibility, which Cheater embraces,which is that it is people who generate ideologies through interaction, rationalprocesses and beliefs and that ideology itself does not merely rationally reflect butalso actually produces that reality. The humanist advantage of this move is that itprotects our intuitive beliefs in the rationality of everyday agents by locating thegenerative source of ideology in human agency rather than in disembodiedstructural relations. In addition, of course, it legitimates the primacy of empiricaldescription and observation.But, most importantly, it makes possible the notion that reality is essentiallyrelative to some particular point of view and therefore essentially contestable.And this is an objective reality which it is the task of social science to describe andexplain. Not only may the views of ordinary agents, whether politicians orpeasants, be conceived as relative to their particular position, but also may theviews of the researcher/observer/theorist herself.Cheater returns to her argument in The Politics of Factory Organization toillustrate the problem. She points out that workers' strategies were informed bybeliefs about the nature of the social reality, beliefs which were entirely rationaland which were generated by the workers' perceptions of themselves and theirantagonists. Their understanding of their relationship with one another,management, labour representatives and with the state were complex, cuttingacross official versions of the state's commitment to socialist objectives; but thestate itself presented, created and generated contradictory views of what that11 A. P. Cheater, 'Contradictions in modelling "Consciousness": Zimbabwean proletarians in themaking?', Journal of Southern African Studies (1987-8), XIV, 291-303.CAROLE PEARCE 193commitment entailed and demanded of the workers (p. 298-9). Such situationalsocial construction of reality (and, of course, of consciousness) has consequenceson the possibilities and nature of political control (p. 299) and also of legalcontrol.The theoretical problem is not so much that the class model is crude, but thatit could be made responsive after refinement. Cheater's critique suggests that themodel is congenitally inadequate for it cannot account, by virtue of its flawednotion of subjective consciousness, for the situational way in which interests areperceived, created and developed by rational agents. This is to say that theMarxist model of consciousness largely misses the mark by postulating anunderlying generative reality instead of locating reality as a construct of people'sconsciousness.Cheater then proceeds truly to set the cat among the pigeons by suggestingthat observer consciousness is similarly situational and thus similarly partial,although by the same token similarly rational (p. 303). It is at this point that thetheoretical gains of relativism reveal their price, and that price is nothing less thanthe justification and validation of social science itself. For one does not have to bea positivist to realize that science must make some kind of claim to objective truthin order to count as science at all. Perspectivalism and relativism may seem toprovide a solution, not the least because contemporary philosophers of sciencelike Feyerabend and Kuhn embrace it,12 and because the practice of physics, atleast, justifies both the view that nature can only be understood relative to one'sposition and the view that the observer is essentially implicated in the nature andconclusion of the observation.But the intellectual problem vis-d-vis Marxism remains. For we cannotconsistently on the one hand say that all points of view are relative and thusprobably equally true (or false) and on the other hand ride roughshod overMarxist work. Nor can we criticize Marxism for idealism while at the same timemaintain an essentially idealistic and even solipsistic theory of knowledge. Thecorollary of this statement is equally true: no Marxist (or any other theorist) canaccept the reality appearance distinction without also being forced to accept theessential relativism, and thus partiality, of his own view (not the least when it is aquestion of consiousness, true or false).13The proposition that all reality is socially constructed either undermines itselfor is pointless. For if we want to use that proposition in order to demonstrate thefalsity of a set of views Š for example, that common sense and subjective needsare of no theoretical importance Š we cannot do so, since the truth of thatproposition is itself implicated in a relativist and thus essentially partial orsubjective point of view and carries no conviction. The proposition thusacknowledges its own ideological character. If, however, we merely want toexpress the view that all knowledge is relative, then we deny the possibility ofscientific knowledge and render our own attempt at sociological or anthropo-logical work useless.This stance does not, therefore, undermine only social science's claim to': P. K I eyerabend. Philosophical Papers I Cambridge, Cambridge Univ. Press. 2 vols., 19X1 );T. Kuhn, The Structure of Scientific Revolutions (Chicago, Univ. of Chicago Press, 2nd edn., 1970).13 See Kolakovvski. Main Currents <ij'Marxism, lor versions of "revisionarv" Marxism.194 A DIALOGUE WITH MARXISMscientificity and also its justification for involvement in matters of social policy: itfundamentally undermines the possibility of social science as critique by failing toprovide any kind of solid theoretical ground from which critique can proceed.Habermas's intuition that the interest which constitutes our knowledge-seekingactivities is relative to our purposes is significant here. For Habermas does notdoubt that there are, underlying the relativity of our scientific interests andactivities, truths about capitalist society Š truths upon which agreement ispossible and which we need to know in order to emancipate ourselves from'ideologically frozen relations'.The Empirical Status of Marxist RealityIn my view, the fundamental problem of Marxist theory is how to characterize thestatus of the empirical and to ground the 'reality' underlying appearances,independently of the theory which posits this reality. If we cannot ground itindependently, then such a reality is theoretically vacuous since we possess nocriteria which help us to know, for example, at what points reality coincides withappearance, and at which other points it does not. It is clear that Cheater herselfgrapples with this problem throughout her work. It is equally clear that herintellectual preferences lead her to grant priority to empirical data. Yet she cannotuse empirical observations to disprove or challenge Marxist propositions, for noamount of empirical evidence can prove Marxism wrong.This is true for two reasons. Firstly, Marxism itself forbids it by characterizingcrucial aspects of the empirical world, viz. consciousness and belief, as appearance(in so far as they are false or ideological or, in Gramsci's works, the result ofhegemonic control). Secondly and in general, it is widely held by philosophers ofscience that theory is necessarily underdetermined by data.14 What this means isthat no single proposition can be proved true or false with reference to empiricalevidence. This is partly because no proposition, not even the most apparentlystraightforward description, is 'theory-free', just as no perception is untheoretical.Secondly, because no observational statement exhausts the phenomena, forphenomena can always be characterized in a different way by an observationalstatement coming from a different theoretical perspective. Lastly, because the unitto be tested by the evidence is not, and cannot be, one statement but must be 'thewhole of science'. This is, if true, nothing but bad news for empiricists, but it isalso, in my view, bad news for theory, which then becomes necessarily idealistand, therefore, immune to the lessons of experience and observation.THEORY AND PRACTICEI have several times pointed out in this essay Cheater's interests in the sphere ofsocial policy. In Idioms of Accumulation, she challenges policy which ignores thecomplexity of class formation in the countryside, by stressing that the ruralcapitalist class should not be absorbed into the peasantry, administratively, forwealth generation ultimately depends upon the growth of class. In The Politics ofFactory Organization, she condemns state intervention and the destruction of the14 J. Dancy,/f« Introduction to Contemporary Epislemology (Oxford. Black well. 1985), chs. 6and 7.CAROLE PEARCE 195trade union on the grounds that such steps weaken labour and its bargainingpower. In The Production and Marketing of Fresh Produce among Blacks inZimbabwe, she objects to the notion that the informal sector could, or should, beincorporated in a socialist Zimbabwe. Her introductory text is full of livelycriticisms of socialist or quasi-socialist policy, starting with the critique of landreform undertaken in various African countries, particularly in Tanzania.I have also noted that she believes that one of the tasks of anthropology is tohelp guide policy-makers. This can be done by helping to provide them with theknowledge of how the societies in which we live actually work, how rights andduties are organized and distributed, and what the ecological and politicalconsequences are of attempting to alter this existing mode of organization.Her approach demands, however, that we grant priority to descriptiveanalysis over Marxist theory. Reality, however characterized, has, for her, anirreducible obdurate character ('Contradictions in modelling "consciousness" ',p. 303) and is thus the necessary basis and sine qua non for all theorizing. Herconcern with obdurate reality is fundamental to her whole method andtheoretical stance. Throughout her works she suggests that if we take the troubleto document the social world in scrupulous, clear-sighted and dispassionatedetail, we are less likely to make ignorant blunders about its character and thetypes of policy we create for it. Basically, Cheater is saying that arm-chairMarxism is no substitute for trying to find out what people actually do, think andbelieve. And in this I agree that she is correct, although it is hard to find a goodphilosophical basis for defending this view, for it is difficult to see how we canseparate 'reality' from our description of it (which is not quite the same as sayingthat reality is, therefore, necessarily relative).The second conflict between her approach and those of Marxist policy-makers emerges from the positivism which often goes hand-in-hand withempiricism in differentiating between questions of value and questions of fact.The task of social theory, according to this view, is exhausted by the extent towhich it can answer the question, 'What is there?' The answer to this questioncannot lead us to determine what ought to be, much less how we can bring thelatter about. There are traces of this positivism to be found in Cheater's work,particularly when she draws distinctions between the data itself and theinterpretation of data. Questions of value, in this view, are radically external tosocial theory.Yet Marxism insists that social analysis, at the same time as it is analysis, isalso critique which has direct normative implications and which must shape ouractions.15 It may be true that social analysis necessarily involves critique and thushas normative implications, but I am not convinced that the connections betweencritique and remedial action can be clearly drawn. The intuition of positivism, ifnot its justification for this intuition, is, I believe, correct. In critical respects, thehiatus between theory and action cannot be bridged, either because the empiricalworld is too rich or because it is too indeterminate to be captured by any onetheory, however powerful. To fail to see this is to allow ourselves to be blinded by15 But see also M. Weber, 'Science as a vocation', in H. H. Gerth and C. W. Mills (eds.). FromMax Weber (London, Routledge, 1948). 129-56.A DIALOGUE WITH MARXISMgical beliefs, to be deceived by those theories which we hope willit is dangerous to deceive ourselves wilfully.of Cheater's thesis that it is inappropriate and perhaps even%ory to use socialism as a vehicle both for generating social wealth and attime for promoting equality (The Politics of Factory Organization,Jthis is a compelling critique because it derives its strength from theotion that the mode of production is analytically central. A State whichupon the capitalist mode of production for its own economicot overthrow either modes or relations of production: socialistj g, therefore, cannot be a function of the State in these circumstances. So& ^ socialist state overthrow the capitalist mode of production and turn tov %ive production? In Idioms of Accumulation, Cheater sees this as a; "^fcntradiction: an equality in poverty can scarcely be a rational goal for ag. JSfcolicy, and class formation is necessary to accumulation. If this double-ts* tique of Marxism as a vehicle for social transformation holds good, then.*^ pivots of the appeal of Marxism falls away, for it is central to that appeal- ^tism can provide a way out for the Third World from its dependence onr^sive capitalist and imperialist West.* "Interest in social control and the interest in emancipation run dangerously^»irxism, as a social theory which is both holistic and normative, can becither goal, but it is more likely to be used for social control, whileTiding as emancipation. Marxists have, therefore, to be particularly-KD ensure that the one goal is not transmuted into the other. It is the virtuewori( mat she helps us maintain that vigilance.CONCLUSION1 have used as a heuristic organizational device the question ofSi commitment to a neo-Marxist perspective. What has emerged duringrSe of this discussion is a commitment which is itself a sophisticated^ device for organizing anthropological material (as in Social A nthro-w which is used to browbeat Marxists less knowledgeable about^* phic data than she herself is. as well as to confront socialist policy-"Vvith its own inner contradictions. Hence Cheater's wry confession, which* ill a number of different levels, to her own 'heresy' (ibid., p. vii). But there*"* Ig elements in her work w hich lead her further and further away from theparadigm, at least in its most doctrinaire form. Cheater's intellectual"*c:es, presented in 'Contradictions in modelling "consciousness" ', areg luiding her towards the small-scale interactionist perspective which castsci oubt on the ability of grand theory to provide explanations at any serioussophistication, vhile respecting the autonomy of living rational agents- complexity of empirical data. Interaciionsim or, more properly,- i cjnalism is, of course, compatible with the Marxist problematic and so, asIn own above, is the radically relativistic stance which she embraces in her'tide.I * is also clear that this relativistic stance brings its own problems which toduplicate rather than solve the problems which Marxism poses for_ Thus what also emerges from Cheater's work, at a level of generalCAROLE PEARCE 197seriousness, are the following questions. How much more mileage can be wrungout of the Marxist paradigm, especially if social theorists wish to make themselvesavailable to fresh insights about the social world? Is the foray into relativism anecessary move and does it solve the problem which Marxism presents? If not,how are social scientists to resolve their general anxiety over the inadequacies ofsocial theory and the relationship between theory and empirical description? Myown view, still in embryonic form, is that social science must emerge from arelativistic and idealistic stance which robs it of all claims to truth and, therefore,of the possibility of argument, discussion and comparison. How this is to be doneis the main problem. But social scientists should take courage from the fact thatrelativism, as a philosophical position, is only one of several attempts tocharacterize the conceptual nature of our enterprises, not necessarily a truedescription of them. Cheater's works highlight all these problems. If, in the end,she places priority on empirical description, we can only be grateful for the vigourand quality of the new ethnographic material about our much under-describedsociety which is represented in this collection.University of Zimbabwe CAROLE PEARCE