1

MARX'S THEORY OF SOCIAL FORMS AND LAKATOS'S METHODOLOGY OF SCIENTIFIC RESEARCH PROGRAMS

Tony SmithDepartment of Philosophy

Iowa State University

A. INTRODUCTIONImre Lakatos's work has had an influence far beyond the

confines of the philosophy of science. As Deborah Redman writes,Lakatos is currently the most popular philosopher of science

among economists. According to Rosenberg, Milton Friedman'spositive methodology is being supplanted by Lakatos's methodology of scientific research programs (MSRP). At any rate, the Kuhnian wave of the seventies is being swallowed up by the Lakatosian program. (Redman 142)

There have been a number of attempts to comprehend mainstream (bourgeois) economics as a Lakatosian research program, or as a set of competing research programs. (Latsis, ed. passim; de Marchi and Blaug, eds.)i In contrast, the extent to which the Marxian study of capitalism can be interpreted from this perspective has hardly been explored.ii In the following sections some provisional steps in this direction will be taken.

I shall first introduce a reading of Marx's Capital that emphasizes the systematic dialectic of economic categories unifying that work. I shall then ask whether we can arrive at a

2

better understanding of Marx's systematic dialectic through considering how it contributes to a scientific research program in Lakatos's sense of the term.

Two other themes will arise along the way. Certain shortcomings in Lakatos's framework will be discussed. Also, a number of reasons for considering the Marxian research program superior to that of neoclassical economics will be mentioned. Itshould be stressed that the main topic of the present paper is the role of systematic dialectics in the Marxian research program. The other themes go far beyond the scope of single paper.

Lakatos's work can be understood as a response to the dead-end of naive falsificationism. According to this simplistic methodology, theories are to be tested by deducing predictions from them and then determining whether the events predicted occur. If the predicted events do take place at some later time,this does not verify the validity of the theory. That would be to commit the fallacy of affirming the consequent; there may be many other theories that also lead to the prediction in question.In contrast, if the predicted events do not take place, that is sufficient to falsify the proposed theory. Scientific progress on this view is a matter of abandoning theories that have been falsified, along with those that are in principle not falsifiableand those that involve only trivial predictions.

3

Naive falsificationism runs into insurmountable difficultiesat once. A prediction regarding concrete occurrences can only bederived from theoretical propositions if a set of conditions specifying the relevant concrete circumstances is added to the argument, along with a set of auxiliary hypotheses which must be taken as true for any test to be run. This means that disconfirmation of the prediction need not imply that the theory in question is mistaken. The problem could lie with the specifications of the background conditions or with the auxiliaryhypotheses. If one is willing to play with these specifications and hypotheses enough, it is possible to save any theory from disconfirmation, as the famous "Duhem-Quine Thesis" asserts.

For Lakatos, science is not simply about the formulation of theories, the derivation of predictions, and the confirmation/disconfirmation of those predictions. The fundamental unit of science is the research program, not the individual theory. (Lakatos 1970) A research program includes the following elements: a hard core, consisting of the basic postulates taken as inviolable by all those participating in the research program; a positive and a negative heuristics, which guide researchers towards the sorts of questions that ought to bepursued and the sorts of intellectual tools that ought to be used(and away from questions and tools that ought to be avoided)iii (Lakatos 1968 168-73); and, finally, a protective belt,

4

consisting of the actual theories, the auxiliary hypotheses, and the empirical conventions that make up the "body" of the researchprogram and that evolve over time.

From this perspective the notion of scientific progress is acomplex matter, involving a number of dimensions. A research program can be said to make theoretical progress if predictions of novel facts are generated within its protective belt. It can be said to make empirical progress if those predictions are then corroborated.iv (Lakatos 1970 118) Finally, there is heuristic progress, which occurs when changes in the protective belt are consistent with the hard core of the program, and not merely extraneous suppositions added ad hoc. (Lakatos 1978 33-4) All ofthese notions of progress are to be measured relative to competing research programs.

It is now time to consider Marx's Capital in the light of Lakatos's framework.B. MARX'S SYSTEMATIC DIALECTIC AND THE HARD CORE OF LAKATOS'S SCIENTIFIC RESEARCH PROGRAMS

1. A Lakatosian Reading of the Systematic Dialectic in 'Capital'

Marx's Capital is a complex work. A great variety of different sorts of projects are found within it. Proto-phenomenological descriptions of lived experience are found alongside comparative historical typologies; abstract growth

5

models are juxtaposed with detailed case studies of specific technologies. Nonetheless, the work as a whole does possess an overall architectonic. A systematic ordering of categories can be traced from the beginning of Volume I to the conclusion of Volume III. This ordering follows a dialectical logic.v For ourpurposes this simply means that it moves from simple and abstractcategories to those that are more complex and concrete. Or, in more Hegelian jargon, Marx proceeds from categories of simple unity to categories of difference, and then to categories of unity-in-difference.

It is not possible here to do more than to enumerate brieflythe main stages of this systematic ordering of economic categories; more will be said about certain of these stages below. All of the stages in Marx's systematic reconstruction of the capitalist mode of production are determinations of the valueform. Under this social form labor is undertaken privately, and proves to be socially necessary only subsequently, with sale of the produced commodity.vi The first determination of the value form is the commodity form, under which heading there is a systematic progression from the elementary form of commodity exchange, through the expanded form, to the general form. The next stage in the unfolding of the value form is the money form, in which the categories of money as a measure of value, as a means of circulation, and as an end of exchange are derived. The

6

third and last level of the value form is the capital form, whichis in turn broken down into three categorial levels. The first defines capital in production, which Marx then develops in terms of the notions labor power as a commodity, exploitation, and the production process proper. The next stage in the dialectical progression, capital in circulation, includes the concepts of simple reproduction, surplus, and expanded reproduction are derived. The final determination of the capital form is capital in distribution, which begins with the notion of distribution within industrial capital. The categories of cost prices and prices of production fall on this level. Next comes industrial capital and the consumer, where Marx discusses market prices. The systematic ordering concludes with the category distribution between industrial and non-industrial capital, where the social forms of rent, interest and merchant capital are defined.

From the standpoint of contemporary social theory, the systematic dialectic of economic categories is a strange sort of beast indeed. Most social theorists today, including many in theMarxian tradition, take dialectical logic to be a bizarre form ofhieroglyphics, a secret code that reveals nothing but a jumble ofmysticisms and half-baked science. This judgement, I believe, isunduly harsh. The procedures employed in systematic dialectical social theory are fully intelligible, and the specific details ofMarx's derivation of economic categories in Capital can be

7

defended for the most part. I have attempted to defend these claims in previous works. (Smith 1993a,b) Here I would like to pursue a different sort of issue: what role does the dialectical presentation of the basic social forms of capitalism play in the Marxian theory of capitalism as a whole? Lakatos's methodology of scientific research programs suggests one possible answer: Marx's dialectic of social forms provides the central "hard core"of the part of the Marxian research program devoted to the study of capitalism.

For Lakatos, the hard core of a research program consists ofbasic postulates that are taken for granted in work in the "protective belt" of the research program, and taken as relatively inviolable by those participating in the program. Thesystematic dialectic of social forms fulfills both criteria.

In Marx's words, the systematic dialectic of social forms traces "the intrinsic connection existing between economic categories or the obscure physiology, so to speak, of the bourgeois system." (Marx 1968, 165) This means that concrete Marxian research into capitalism, both theoretical and empirical,presupposes that the value form, the commodity form, the money form, and the capital form provide the over-arching context within which specific institutions and events are to be understood.vii The account of these forms provides the ultimate premisses for arguments regarding concrete institutions and

8

events in capitalism. In this sense, the systematic ordering of categories is a central part of the hard core of the Marxian research program.

Regarding the question of inviolability, it must be emphasized that Marx's dialectic of social forms is neither a revelation from God nor some sort of a priori transcendental deduction. It is instead "a reproduction of the concrete by way of thought." When Marx discussed the procedures he employed in constructing this dialectic, he took pains to stress that "the point of departure" was the "chaotic conception of the whole" latent in everyday experience in capitalism. (Marx 1973, 100-01) Systematic dialectics thus has an empirical basis, and is therefore open to revision. If, for instance, it could be shown that Marx's own account was distorted by extraneous influences,viii or that it was incoherent on its own termsix, or that it neglected matters that should have been incorporatedx, then revisions are in order.

This revisability of the hard core is compatible with Lakatos's understanding of a scientific research program. For Lakatos, the inviolability of the hard core is a relative rather than absolute matter. He insisted that the hard core hardens slowly, by trial and error. (Lakatos 1978a 48 n.4, 181). During this process of hardening, which in principle may last indefinitely, revisions in the hard core are possible.

9

It does not follow from this that propositions in the hard core have the same cognitive status as those in the protective belt of a scientific research program (SRP). As Blaug notes,Once we have accepted the notion that individual scientific

theories are necessarily linked together in a network of ideas called an SRP and that, being a program of promising work to be carried out in the future rather than a perfectlyfinished product, such an SRP must change and evolve in time, it follows that there is some core in the SRP that changes much less than the rest of it. (Blaug 1991 500)

The important point here is that this core is not subject to empirical testing in the same (relatively) direct fashion as the propositions making up the protective belt of the program are. Defenders of the program will give up the hard core only very grudgingly.

The categorial progression in Capital also claims this special cognitive status. In some cases, Marx's argument is thatcertain social forms are necessary conditions of the possibility of others. Money, for example, is a necessary precondition of capital. (Campbell, this volume) For a Marxist, this is an empirical claim, not a divine revelation or a transcendental deduction. Nonetheless, within the Marxian program it is difficult to conceive of any ordinary empirical test that could call this assertion into question. Another sort of claim made in

10

the systematic dialectic is that social agents operating within certain social forms necessarily tend to act in certain ways. Here too the force of the concept of necessity is so great that no ordinary empirical test calling the claim into question would be accepted by someone working within the research program. For example, if empirical evidence were presented purporting to show that the tension between money as a measure of value and money asa means of circulation had been abolished in capitalism, or that capitalism no longer involved the exploitation of wage labor, no Marxist would accept that evidence at face value. All would assume that the evidence in question was suspect in some fundamental way. This does not mean that one needs to have studied Capital from the standpoint of systematic dialectics prior to undertakingconcrete theoretical and empirical work in Marxism. It is certainly possible to contribute to a research program without ever explicitly reflecting on its hard core. Nor am I claiming that anyone who wishes to reflect on the hard core of the Marxianprogram must turn to systematic dialectics. Most theorizing requires no more than a brief list of the assumptions relevant tothe task at hand. However, if one wishes to consider the propositions making up the hard core assumptions of Marxian research regarding capitalism in a comprehensive fashion, systematic dialectics has no serious competitors.

11

2. The Hard Core and the Comparative Assessment of Scientific Research Programs

For Lakatos, the propositions making up the hard core have anecessary role to play in scientific research programs. They provide the ultimate assumptions without which concrete theoretical and empirical work cannot be undertaken. In Lakatos's framework, however, these propositions play next to no role in the comparative assessment of different research programs. One program is judged to be superior to another if it generates theories from which predictions of novel facts can be derived, and if those predictions are later corroborated. The theories and predictions must be consistent with the hard core assumptions of the program, but otherwise those assumptions play no role in assessing programs. I believe that this is mistaken. Ceteris paribus, one research program is superior to another if a) the hard core assumptions of one specify real mechanisms in the object realm under investigation, while those of the other donot, or b) the specifications of a set of real mechanisms in the hard core of one program are superior to those proposed in the competing program. I shall illustrate these points by means of acomparison between the hard cores of Marxism and neoclassical economics, considering in turn an instrumentalist and a non-instrumentalist reading of the latter.

a) The debate between instrumentalism and realism is one of

12

the central issues in the philosophy of science. If scientific progress is judged primarily by whether predictions of novel facts are corroborated, as Lakatos asserts, why should we care whether the ultimate assumptions underlying our theories are realistic or not? These assumptions are in effect instruments used to help derive predictions. The only relevant question would seem to be whether they effectively perform the task assigned to them. In this sense Lakatos's account of the role ofthe hard core in the methodology of scientific research programs appears fully compatible with an instrumentalist perspective. (Hands 1991 71)

Neoclassical economics is plagued by a deep schizophrenia onthe question of instrumentalism. (Reuten 1993b) As we shall see,there are those who take its postulates as an accurate representation of human nature. On the other hand, there are many others who willingly concede that its ultimate assumptions do not present anything like a realistic picture of economic life. Milton Friedman, for instance, finds a theory with completely unrealistic assumptions quite acceptable, as long as it generates corroborated predictions.

There are also those neoclassicists who cannot quite bring themselves to accept instrumentalism, even when the logic of their position is consistent with it. Blaug defends the Lakatosian position "that assumptions do not need to be tested

13

directly, although it might be useful if they could be, that in the final analysis only predictions matter, and that the validityof an economic theory is established when the predictions to which it gives rise are repeatedly corroborated by the evidence."(Blaug 1992 68; emphasis added) This clearly implies that any assumption, no matter how wildly implausible, should be accepted as valid if it underlies theories from which corroborated predictions are derived. As far as I can tell, this is quite inconsistent with the following assertion: "So long as the tests of the accuracy of predictions remain ambiguous - that is to say,forever - it will remain important also to test the descriptive accuracy of assumptions and to take the results of these tests seriously." (Blaug 1992 233) This implies that assumptions can have some explanatory force independent of their role in generating predictions. If this were the case, it would surely be relevant to comparative assessments of competing research programs. How this assertion coheres with Blaug's professed Lakatosianism is unclear.xi

When we turn to Marxism, little evidence of this sort of schizophrenia can be found. Those working within the Marxian economic program overwhelmingly accept a realist agenda.xii This extends to the propositions making up the hard core of the Marxian theory of capitalism, including the systematic ordering of social forms. This ordering has been defined as a progression

14

of categories, with each category delineating a social form. Each determination of a social form explicates real mechanisms operating in the capitalist mode of production. In brief, Marx'ssystematic dialectic purports to present the fundamental mechanisms at work in capitalism. It can be assessed in these terms apart from its role in providing background assumptions fortheories from which predictions of novel facts are derived.

How is the phrase "real mechanism" to be understood? AndrewSayer writes, Within social structures there are particular 'positions'

associated with certain roles. It is particularly importantto distinguish the occupant of a position from the position itself. One of the most pervasive illusions of everyday thinking derives from the attribution of the properties of the position, be they good or bad, to the individual or institution occupying it. Whatever effects result, it is assumed that particular people must be responsible; there islittle appreciation that the structure of social relations, together with their associated resources, constraints or rules, may determine what happens, even though these structures only exist where people reproduce them. (Sayer 84-85)xiii

By "mechanisms" here I mean all the various ways structures of social relations may "determine what happens." Given a structure

15

of generalized commodity production, for example, the mechanism referred to as "the law of value" kicks in; socially average labor productivity will tend to govern exchange relations. Giventhe structure of the capital/wage labor relation, the exploitation mechanism kicks in; those who own and control capital will tend to appropriate a surplus value exceeding the value of the wages received by the work force. Given the structure of the relations among units of individual capital, themechanism of the redistribution of surplus value will kick in; units that successfully introduce process and product innovationswill win surplus profits. The derivation of a comprehensive ordering of the economic categories in Capital is simultaneously the derivation of the most basic real mechanisms in capitalism.

Marx employs abstractions throughout the progression of social forms in Capital. The discussion of the mechanisms underlying the labor process abstracts from the centrality of rawmaterial costs, fixed capital costs, and transportation costs in the process of accumulation. The discussion of the mechanisms underlying attempts to reduce circulation costs abstracts from the payment of interest and rent. The discussion of the mechanisms underlying the relations among many units of capital abstracts from the impact of the fiscal, monetary, and industrialpolicies of the state. The difference between the hard core of neoclassical economics (in the instrumentalist interpretation)

16

and the hard core of Marxian economics is not a matter of the onemaking use of abstractions and the other not. The difference lies instead with the sort of abstractions that are made. The abstractions of Marxism are enabling abstractions. On any given level of Capital the workings of certain real mechanisms are ignored in order to enable a consideration of the workings of another real mechanism on its own terms. This is quite differentfrom the introduction of abstractions whose lack of realism is deemed irrelevant.

The instrumentalist perspective does not allow a distinctionto be drawn between enabling abstractions, that is, abstractions that enable a comprehension of real mechanisms, and abstractions that distort the nature of the realm being investigated. For this reason it ought to be rejected. The hard core of the Marxian research program refers to real mechanisms in capitalism,while the hard core of neoclassical economics (in the instrumentalist reading) does not. This provides a reason for considering the former program superior, a reason that is not reducible to the corroboration of predictions of novel facts.

b) I now wish to compare briefly the hard cores of Marxism and neoclassical economics in the non-instrumentalist reading of the latter. Here too reasons can be given for the superiority ofthe Marxian research program that do not have to do with the predictive power of the theories derived from/consistent with the

17

hard core. They have to do instead with the different types of claims made in the two hard cores, and with the content of those claims.

For many neoclassical economists, the propositions of the hard core of their research program refer to a real realm of being, the economic realm. They define features of that realm that are taken to hold always and everywhere. Those making this sort of claim trace their heritage back to Adam Smith's assumption of an innate human drive to "truck and barter." (Brenner) A recent example is provided by Latsis. In his view the hard core of the neoclassical theory of the firm consists of four propositions: profit maximization, perfect knowledge, independence of decisions, and perfect markets. (Latsis 1972 209)The first and third of these assertions are usually taken to define a transhistorical disposition innate in human nature, a disposition of "rationality."xiv This disposition may be hamperedor even repressed in some institutional contexts; the fourth proposition implies that capitalist market societies provide the only adequate institutional context for the flourishing of this innate disposition. This too is a transhistorical claim.xv

How do things stand with Marxism? Certainly part of the hard core of Marxism includes transhistorical suppositions regarding human beings, as Geras has rightly insisted. But if werestrict our attention to Marx's dialectic of social forms in

18

Capital, there is no claim that the assertions formulated in the course of this dialectic hold always and everywhere. From the Marxian standpoint, the fact that the hard core assumptions of the Marxian research program are historically specific is very relevant to a comparative assessment with the neoclassical program.

Consider the following argument. In capitalism, as in all other modes of production, the continued reproduction of society depends upon co-ordinated labor. Yet in capitalism the value form reigns. Production is undertaken privately; producers discover whether they have contributed to social reproduction only later, with the subsequent sale of a commodity for money. In this manner the universal social power of labor bizarrely comes to appear in the form of commodities and the abstract symbols of money. When this money comes to function as capital, the bizarreness is taken a further step. The capital that purchases social labor in effect proclaims that this labor is merely part of itself ("variable capital"); it appears that capital alone is the power accomplishing the reproduction of society. And so commodities, money, and capital are fetishes in which social labor is manifested in inert and alien objects. Fetishism is a form of subjectivity necessarily generated by the objective material forms of capitalism.

The thesis that economic agents in capitalism are subject to

19

commodity fetishism, money fetishism, and capital fetishism can be taken as part of the Marxist response to the rationality assumption of neoclassical economics. But while the neoclassicalclaim (in the non-instrumentalist reading) is averred to be a transhistorical disposition rooted in human nature, the corresponding Marxist thesis is explicitly limited to a historically specific mode of production. In so far as the aim of the two research programs is to account for events, processes,and structures in capitalism, the historically specific nature ofthe claims made in the hard core of the Marxian program provides a reason to find this program prima facie superior to the neoclassical program. This reason is not reducible to the prediction and corroboration of novel facts.

Turning from the different types of claims made in the hard cores of the two programs to the content of those claims, here too we find much of relevance to a comparative assessment of competing programs, contra Lakatos. For the claims in question themselves have explanatory force.

For Lakatos, legitimate scientific explanations necessarily involve the prediction of novel facts. The propositions of the hard core may be necessary assumptions for such explanations. But since they themselves do not directly involve falsifiable predictions, they themselves do not directly involve explanatory claims. In the methodology of scientific research programs,

20

explanations in the proper sense of the term are found in the protective belt of a research program, not its hard core.

This view is consistent with the instrumentalist understanding of propositions in the hard core, an understanding I have already called into question. On the realist view of the hard core, it is mistaken to limit explanatory claims to the assertions in the protective belt from which predictions of novelfacts can be derived. The hard core propositions of Marxian social theory and non-instrumentalist neoclassical economics bothclaim to provide explanations of the fundamental mechanisms underlying economic activity under the capital form. If the claims made here by defenders of one program prove to be better warranted and more powerful in scope than those proposed by defenders of the other, then this is relevant to any comparative assessment. The implications of this point can be stated provocatively: even if neoclassical economics should turn out to have a better record at the prediction of novel facts than the Marxian program, the latter might still have greater explanatory force. This point is worth developing at some length.

For the non-instrumentalist neoclassical economist, the propositions in the hard core of the neoclassical research program define a mechanism in which the maximizing behavior of atomized rational individuals underlies economic processes. Economic agents are atomized in the sense that the independence

21

of decision making is assumed. They are rational in that economic agents are assumed to chose "in accordance with a preference ordering that is complete and transitive, subject to perfect and costlessly acquired information." (Blaug 1992 229) And they are maximizers in the sense that the pressures of the market competition lead them to optimize in the face of their constraints.

The Marxian dialectic of social forms proposes a quite different explanation of the most basic mechanisms at work in capitalism. This can be seen from an examination of the circuit of capital, a device Marx employed to summarize his initial discussion of the commodity, money, and capital forms:

CapitalM - C - P - C' - M'

Capital first takes the form of money (M) to be invested. It is then invested in the purchase of two sorts of commodities (C), means of production and labor power. Labor power is set to work in a production process (P), resulting in the output of goods andservices in the commodity form (C'). These commodities are then sold in the market. If they can be sold for a money price exceeding the initial money invested (M'), capital may be accumulated and the circuit can begin again. Reference to this circuit explains why seeing the maximizing activity of atomized rational individuals as the basic mechanism at work in capitalism

22

is mistaken.Regarding atomism, it is true that economic decision making

is "atomized" under the capital form. Nonetheless, the identity of economic agents in the circuit of capital is defined in terms that imply internal relations with other agents, as opposed to the external relations of atomism. The categories "buyer of labor power" and "seller of labor power," for instance, cannot bedefined apart from each other.xvi This is not merely a logical point resting on semantical features of our language; the point rests on the real material relations involved. (Ollman 1976)

The atomistic categorization of economic agents is also suspect in that the circuit of capital provides a material basis for various forms of collective identity. All those in the position of being buyers of labor power objectively share certaininterests as a result of this; all those in the position of having to sell their labor power objectively share interests as well. In this sense class is necessarily constitutive of economic agents' identity. Any attempt to consider individuals without taking this into account is necessarily inadequate.xvii

Finally, individuals in capitalism always already find themselves situated in a context defined by the commodity form, the money form, and the capital form. This undercuts the possibility of isolated decision making. The dialectic of socialforms explains why and how the subjective psychology of

23

individual economic agents (reflected in their "preference sets")is determined within a specific class context, and so cannot claim ontological independence. (Arthur) The immediate preference of the owners and controllers of capital is to accumulate capital; that of the owners of labor power is to obtain access to the means of subsistence. The holders of capital tend to develop a preference to reduce the time it takes for their capital to transverse the circuit of capital; the preference of wage laborers is generally to resist speed-ups in the labor process. These and other preferences are not best seenas private decisions made in isolation from other social agents. They are rooted in the social forms defining capitalism.

Regarding rationality, I have already argued that the Marxist dialectic of social forms purports to explain why the actions of even the most rational economic agents are simultaneously permeated by irrationality. Commodity fetishism, money fetishism, and capital fetishism inevitably distort the reasoning process of even the most "rational" economic agent in this system. Further, nothing in the logic of the capital circuit implies that the preference orderings of economic agents need be complete and transitive. (The more we reflect on the psychic manipulations of advertising in capitalism, the less plausible this assumption becomes.)

Perhaps the most obvious way the assumptions of the hard

24

core of neoclassical economics distort the real mechanisms at work in capitalism has to do with the supposition that the agentsin the capital circuit possess perfect information. Each step inthe circuit is characterized by radical uncertainty, due to the strategic concealment of information. (Laibman 238) Also, there are no guarantees that means of production purchased will providean adequate return on investment before technical change has subjected them to "moral obsolescence." Neither are there guarantees that the labor power hired will meet or exceed the social average of labor productivity, that the commodities produced will find buyers, or that money will retain its value over time.xviii If economic agents lack perfect information, if they must make decisions in situations of radical uncertainty, then there is no way for them ever to know in advance what the correct maximizing behavior might be, even if we counterfactuallysuppose both the capacity and the inclination to solve foot-long Lagrangian equations before acting.xix As a result, economic agents in capitalism are often not content to operate "under given constraints." Analysis of the capital form reveals that constraints need not be accepted as given by economic agents. Rather than accepting technology as exogenously given, research and development is incorporated within the capital circuit, thereby changing the constraints under which capital operates. Rather than accepting the constraint of the given wage rate, for

25

instance, labor-saving innovations can be sought. Nor are spatial constraints passively accepted; capital flight to low-wage areas can occur. Likewise, wage laborers need not accept the constraint of the given balance of class forces. They can attempt to organize in order to shift that balance in their favor. (Storper and Walker) There is thus an extensive range of phenomena that can be accounted for within the Marxian hard core,phenomena that fall outside the scope of the hard core of neoclassical economics.

It is possible to derive predictions of novel facts from thehard core assumption that the basic mechanism at work in capitalism is the maximization of given preference sets by atomized rational individuals. It is also possible to derive predictions of novel facts from the hard core of the Marxian program for studying capitalism, that is, from the Marxian dialectic of social forms. Lakatos is not wrong when he notes how the hard core of a research program contributes to scientificexplanations in this manner. But he is, I believe, mistaken whenhe limits the explanatory role of the hard core to this. The hard core of neoclassical economics and the systematic ordering of social forms in Capital offer competing explanations of the basic mechanisms at work in capitalism. The relative cogency of these competing explanations, and the relative scope of the phenomena explained, can in principle be weighed apart from the

26

relative success of the two programs at predicting novel facts. A convincing case can be made that the neoclassical program restson an inadequate grasp of the social forms that make up the capitalist mode of production, social forms that generate the basic mechanisms at work in this system. It is possible to make this case without introducing predictions of any sort.

It is now time to summarize this over-long section. In Capital Marx presented a systematic reconstruction of the social forms that make up the capitalist mode of production. What is the point of this project? A working hypothesis was formulated based on Lakatos's methodology of scientific research programs: the dialectic of social forms can be seen as an attempt to articulate the hard core of a research program in Lakatos's senseof the term. This hypothesis is helpful to a certain extent. The ordering of social forms in Capital does provide a basic vocabulary and framework that orients Marxian studies of capitalism. Also, the assertions making up this systematic progression are not subjected to direct empirical testing in the same manner as other propositions formulated within the Marxian program. Approaching Capital with Lakatos's methodology in mind allows us to appreciate Marx's accomplishment in a new way. Has anyone working in any research program in any discipline ever gone to the lengths Marx did to make the hard core assumptions ofhis program explicit?xx

27

There are, however, two aspects of Marx's approach that do not fit Lakatos's framework so easily. First, there seems to be a disposition towards instrumentalism latent in Lakatos's position, while Marx's systematic ordering of social forms claimsto reconstruct in thought the inner nature of the capitalist modeof production. Second, consideration of the hard core plays nextto no role in the comparative assessment of competing research programs for Lakatos. This too can be called into question. Thehard core of the neoclassical program (on the non-instrumentalistinterpretation) consists of statements taken to hold always and everywhere, while the hard core of Marxism is by and large historically specific. I have argued that this consideration counts as a major advantage of the Marxian program. And then there is the question of explanation. For Lakatos, the propositions of a hard core are not in themselves relevant to a comparison of competing research programs. The hard core provides background conditions for theories from which predictions of novel facts are derived, and only the success of these predictions is ultimately relevant to comparative assessments. But the Marxian dialectic of social forms and the hard core of the neoclassical program (on the non-instrumentalistreading) both purport to explain the fundamental mechanisms at work in capitalism. If one of these accounts is more cogent thanthe other, or if one can explain a greater scope of phenomena in

28

capitalism, this is surely relevant to any comparative assessmentof the competing programs.

In Lakatos's framework there are two other parts of a research program besides the hard core: the heuristics of the program and its protective belt. Can a consideration of these themes add anything further to our understanding of the contribution made by the dialectic of social forms to the Marxianresearch program?C. THE THEORY OF SOCIAL FORMS AND THE HEURISTICS OF THE MARXIAN RESEARCH PROGRAM

Marx's account of social forms can be used to formulate a sets of precepts making up both a positive and negative heuristicfor concrete theoretical and empirical research. This includes both the questions to be asked and the tools to be employed in concrete research.

Once again, it is important to be clear about the claim being made. I am not asserting that one must have previously studied Marx's systematic ordering of categories in order to grasp the heuristics of the Marxian research program. Nor am I claiming that this heuristics can only be comprehended in terms of this ordering. The point is simply that a number of central features of the positive and negative heuristics of Marxian research can be derived from the systematic dialectic found in Capital. This counts as another significant contribution to the

29

Marxian program.Concrete theoretical and empirical research begins with some

basic question or set of questions. What determines whether a specific question fits within the Marxian research program? It is possible to go through Marx's systematic reconstruction of capitalism step-by-step, and derive a set of general questions for concrete research from the analysis of the various social forms. Part of the positive heuristic of Marxism is that the questions asked in concrete research address the issues raised inthe analysis of social forms.

It is obviously not possible to provide a comprehensive listof all the questions for concrete research that can be derived from Marx's systematic ordering of social forms. A quick survey of the main categories in the three volumes of Capital leads to the following representative sample. In a given historical period or geographical region, what new areas of social life are being commodified? What is the relationship between social laborand money? What are the various functions of money that come into play in the given context, and how are they in tension with each other? What is the role of paper money? What impact does agiven change in the labor process have on the rate of exploitation? What is the balance of class forces at the point of production? What sectors are experiencing the most growth in rates of accumulation? How are organizations restructuring in

30

response to accumulation? Which stages in the circuit of capitalprovide the biggest barriers to increases in the rate of circulation? To what extent are the rhythms of the different sectors of the economy in sync? How are process and product innovations creating surplus profits for certain firms/sectors incertain regions? What are the trends in the rate of profit, and what are the key factors determining those trends? What are the causes of economic crises? How do prices respond to shifts in technology and in patterns of consumer demand? Where do the interests of industrial firms, merchant capital, financial capital, and landlords coincide? Where do they diverge? How do the workings of the commodity form, the money form, and the capital form manifest themselves in the everyday lives of social agents? Where are the points in the social order where strugglesare most likely to break out? What are the likely results of these struggles?

As incomplete as this list is, it does suggest how the Marxian ordering of social forms generates a set of questions that define an on-going research program. Of course it is possible to pose a number of these questions within competing programs. Nonetheless, three features make this set of questionsunique, taken as a whole. The first and most significant is the centrality of class issues, ranging from direct class conflict atthe point of production to the manner in which changes in

31

technology, investment patterns, corporate structure, and so on, affect the balance of class forces.xxi The second has to do with time. The above questions are formulated in a way that implies that capitalism has a history, and that this history matters. Capitalism is based on ceaseless experimentation, a ceaseless process of overcoming certain barriers to accumulation only to come across new ones, whose shape depends on the paths chosen in the past. (Storper and Walker) Third, these general questions are relevant no matter what specific shape capital may take. They hold whether we are investigating separate firms operating in purely competitive markets, or networks of firms cooperating together in a keiretsu. In these and other possible cases the same fundamental class antagonisms are found, the same tensions between the various functions of money arise, the same stages in the circuit of capital must be transversed, and so on.

This takes us to the negative heuristic of Marxism, the sorts of questions that should be avoided in order to remain consistent with the hard core of the research program. Any set of questions regarding the socio-economic realm that obscure the class dynamic are to be avoided. Any set of questions that leadsresearchers to approach capitalism as if it were not developing over time is to be rejected. And any set of questions that does not allow for capitalism to take on a variety of shapes are suspect as well. Once again, the contrast with the neoclassical

32

research program is sharp. Its basic questions concern the decisions of isolated individuals, ignore path dependencies, and are particularly unsuited for an understanding of the so-called Japanese model of capitalism (Gerlach). They are thus to be avoided, given the negative heuristic following from Marx's dialectic of social forms.

Besides generating a set of questions, the heuristics of a research program also define the sorts of intellectual tools thatare compatible with the program, as well as the sorts of tools that are to be avoided.xxii It is not possible here to do more than mention briefly a few ways in which the systematic ordering of social forms contributes to this dimension of the Marxian program.

* The necessary connection between the social forms constituting the capitalist mode of production and the various types of fetishism rules out uncritical use of methodological tools that simply reflect surface appearances. The production function calculus, which assumes that "capital" and "labor" are two independent "factors of production," is an example of a tool of analysis that is to be avoided for this reason.

* The hard core claim that capitalism necessarily generates various types of fetishism also implies that the methodological tools of ideology critique ought to be employed. Systematic dialectical method is itself an example of ideology critique.

33

The systematic relation between the categories "exploitation" and"market price," for instance, provides a basis for a critique of the ideology of consumer sovereignty. Other methods provide tools for ideology critique as well, and are equally within the positive heuristic of the Marxian program.xxiii

* A hard core whose propositions emphasize the historical nature of social forms rules out methodological tools that do notallow for the possibility of historical change. This is why so much of sociobiology is incompatible with the Marxist research program.

* A hard core emphasizing the possibility of historical change through the social agency of working men and women is incompatible with an uncritical use of survey questionnaires thathamper this agency. Consider a survey questionnaire asking respondents whether they would support reducing the tax on capital in order to spur investment and create jobs. This question presupposes capital fetishism, in that capital is treated as a thing, not a social relation. The question is formulated to obscure the real mechanisms at work in investment. It implies that capital investment naturally flows to where it isneeded, unless the "unnatural" interference of state taxation distorts the flow. The question does not challenge the pre-understandings of the addressees, or treat them as subjects capable of changing their views on the capital relation. Within

34

the Marxian program methodological instruments should be instruments of social change as well; survey questionnaires that reinforce dominant world views do not meet this criterion.xxiv

* The dynamism of capitalism as analyzed in Marx's progression of social forms is more compatible with certain formsof mathematical analysis than others. The mathematical techniques of static equilibrium theory may be helpful in certaincarefully defined investigations. In general, however, there is a great danger that use of this methodological tool will lead to a distortion of an essential dimension of social life. Simultaneous equation systems cannot represent the passage of time, especially the temporal dynamic set off by technological change.xxv In contrast, difference equations can depict significant aspects of the trajectory of an economy over time.xxvi

For example, it is possible to distinguish how prices at the end of a period of production are distinct from those at its beginning, due to the way prices respond to technological changes. (Carchedi) The endogeneity of technical change in capitalism is a fundamental claim of Marx's dialectic of social forms, and so a positive heuristic encouraging the use of mathematical tools incorporating technical change can be directlyderived from that dialectic.

* Another point regarding the methodological limits of mathematics has to do with the issue of realism. The dialectic

35

of social forms uncovers mechanisms that have a material existence in the social world. The use of mathematical techniques is to be avoided whenever those techniques obscure theoperation of real mechanisms in capitalism.

An example of the problems that can arise if this last methodological injunction is ignored is given by Fred Moseley. Define "P" to represent total profits; "S," total surplus value; and "U," the total expenses of unproductive labor. From the mathematical point of view the equation

P = S - U is equivalent to the equation

S = P + U.However from the perspective of the material forces at work in capitalism, the two could not be more different. There are real mechanisms at work that lead the amount of profit to equal the total amount of surplus-value minus the costs of unproductive labor, as expressed in the first equation. There are no mechanisms that operate along the lines of the second equation; with the level of profits held constant, an increase in unproductive labor does not result in a greater amount of surplus-value. (Moseley 1994 87) Mathematical techniques, no matter how convenient their use might be to the social theorist, are never a substitute for the qualitative comprehension of real mechanisms.xxvii

36

D. MARX'S THEORY OF SOCIAL FORMS AND THE "PROTECTIVE BELT" OF THEMARXIAN RESEARCH PROGRAM

Lakatos's Popperian heritage comes most into play in his discussion of the protective belt of a research program. The theories and empirical assertions found in the protective belt are measured by three standards: the degree to which predictions of novel facts can be derived from them, the extent to which these predictions have been corroborated, and the consistency of the theories and empirical assertions with the hard core of the program. This insistence on the centrality of prediction comes from Popper.

Marxism more than holds its own against competing research programs when measured by the standard of the corroborated prediction of novel facts within the protective belt of the program. Marx's prediction of recurrent crises in capitalism, ofthe persistence of conflict regarding the length of the working day and the intensity of labor, and of an ever-increasing concentration and centralization of capital, to mention just a few examples, has been corroborated again and again.xxviii Nevertheless, here too the prediction and corroboration of novel facts ought not to be the sole criteria for evaluating whether a research program is progressive or degenerate (given the consistency of the protective belt with the hard core of the program). A consideration of the relationship between Marx's

37

systematic dialectic of categories and concrete work in the protective belt of Marxian theory can illustrate this point (which can be made from many other approaches as well).

Each category in Capital defines a determination of a socialform, and each determination specifies a real mechanism at work in the capitalist mode of production. Activity within the protective belt of the Marxian research program attempts to specify how these basic mechanisms operate in specific concrete contexts. However important the role of predictions may be here,that role is a restricted one for three reasons.

First, the mechanisms do not operate in closed systems whereconditions remain identical and effects are uniform. And so any attempt to investigate how these mechanisms operate concretely must be formulated in terms of tendencies. In any given case thenon-occurrence of a certain event does not show that there is nota tendency for an event of that sort to occur; it does not even show that there is not a necessary tendency in that direction. In any given case, for example, it is always possible for a substantial money reserve to be invested in non-profit organizations; this does not change the fact that once the circuit of capital has been established it is necessarily the case that there will be a structural tendency for money to be invested in order to accumulate capital. And so the corroboration or non-corroboration of a prediction regarding a

38

given concrete event is not an especially interesting matter fromthis perspective. "Falsifications of predictions of contingencies in open systems whose initial state is incompletelyknown need not be treated as theoretically significant." (Sayer 190; see also 111-114)

A second point emerges from consideration of the architectonic of Capital. Marx's systematic dialectic is organized linearly, progressing step-by-step from simple and abstract social forms to more complex and concrete forms. Each subsequent stage in the dialectical progression "sublates" those that have gone before, that is, it incorporates earlier ones while going beyond them in some fashion. This might be taken to imply that each time the social theorist reaches a new categoriallevel he or she should go back to the beginning, and show how allof the mechanisms considered earlier are now modified as a resultof the addition of the new determination. This, however, would be an impossible task. Marx never even came close to completing the straightforward linear ordering of the social forms making upcapitalism. If he had gone on to attempt to show how the variousforms and the mechanisms they generate are modified when they operate simultaneously, he would never have finished more than a small portion of Volume I of Capital. There is simply too much complexity here, generating too high a degree of indeterminacy.

This has direct implications for concrete research in the

39

protective belt of a research program. In the concrete realm a whole range of mechanisms operate simultaneously. Whatever difficulties arise in the course of formulating specific predictions regarding the workings of a given mechanism are dwarfed by those that arise when a variety of mechanisms operate simultaneously. In the face of a much higher degree of indeterminacy, the corroboration or lack thereof of specific predictions has even less significance.

Third, in many cases the social forms of capitalism generatetendencies that in specific concrete contexts may come into tension with each other, leading to a higher level of indeterminacy yet again. The capital form necessarily involves atendency for the owners and controllers of capital to seek to increase the rate of exploitation; everything else being equal, this raises the rate of profit. In Marx's view, the capital formalso generates a tendency for the organic composition of capital to rise; everything else being equal, this lowers the rate of profit. There is a tendency built-into the capital form to further capital accumulation through lowering wage costs by deskilling; there is also a tendency built-into the capital form to further capital accumulation through productivity increases demanding higher levels of skill. There is a tendency in the capital form for technical innovation to be spurred by the fear of the "moral obsolescence" of fixed capital; there is a tendency

40

in the capital form for previous investment in fixed capital to retard technical innovation. In these and similar cases the tendency that dominates in a given concrete context will depend on a variety of contingent historical factors.xxix And so the corroboration/non-corroboration of predictions regarding any given concrete context will have a contingent quality as well.xxx

From the standpoint of a Lakatosian such as Blaug, these three points simply reveal the immunizing strategies that undermine Marxism's claim to be taken seriously as a scientific research program. (Blaug 1992 60-61) The criticism could not possibly be more misguided. Methodology must adjust to the nature of the objects being studied. If capitalism is an open system rather than a closed system, if a variety of distinct mechanisms operate simultaneously in it, and if it is beset by conflicting tendencies, all of this must be acknowledged in any adequate methodology. Issues of social ontology cannot be removed by methodological fiat. The social ontology of capitalism implies that the corroboration/non-corroboration of predictions will have only restricted significance in any plausible research program, even when we restrict our attention to what Lakatos termed the protective belt of programs. CONCLUSION

I hope to have established that the systematic dialectic of social forms in Capital plays a central role in the Marxian

41

research program. It makes explicit much of the hard core of theMarxian study of capitalism; it generates a set of general questions to orient concrete theoretical and empirical inquiries;and it provides guidance regarding which sorts of intellectual techniques are generally suited to the study of events, processes, and structures in capitalism.

Two other themes were also discussed along the way. Both were treated in a preliminary manner, and point to the need for further study. The first concerned Lakatos's proclivity to assess competing research programs simply in terms of their relative success at predicting novel facts. I have argued that this criterion is too restricted, and that other factors are relevant as well. Far more significant than prediction is retroduction, defined by Sayer as "(the) mode of inference in which events are explained by postulating (and identifying) mechanisms which are capable of producing them." (Sayer 97) Perhaps equally important is the question of scope. In principle, a research program that accounts for more phenomena within the object realm in question than its competitors could bejudged superior, even if those competitors had a somewhat better record at the prediction of novel facts. In the present work no attempt has been made to fix the precise weight that should be given to prediction, retroduction, and scope in the assessment ofcompeting research programs.

42

The second topic for further research that has been raised is the comparative assessment of neoclassical economics and the Marxian program for the study of capitalism. I have suggested that the Marxian program is superior on all three relevant grounds, the prediction of novel facts, the account of explanatory mechanisms employed in retroductions, and the scope of the phenomena explained. To establish these claims in detail,however, would go far beyond the scope of a single paper.

43

NOTES

44

BIBLIOGRAPHYArthur, Chris (1993) [ref. to last year's paper]Blaug, Mark (1991) "Afterword," in De Marchi and Blaug, eds.

---. (1992) The Methodology of Economics: Or How Economists Explain. Second Edition. New York: Cambridge

University Press.Brenner, Robert (1977) "The Origins of Capitalist Development: A

Critique of Neo-Smithian Marxism." New Left Review, No. 144.Buck-Morss, Susan (1991) The Dialectics of Seeing: Walter Benjamin and the Arcades Project. Cambridge, MA: MIT Press.Campbell, Martha (199?) [ref to second paper]Carcedi, Guglielmo (1993) "Marx's Logic of Inquiry and Price

Formation." In Moseley.de Marchi, N., and M. Blaug, eds. (1991) Appraising Economic Theories. Aldershot, Hants: Edward Elgar.Friedman, Milton (1953) Essays in Positive Economics. Chicago:

University of Chicago Press.Geras, Norman (1982) Marx and Human Nature: Refutation of a Legend. London: New Left Books. Gerlach, Michael (1992) Alliance Capitalism: The Social Organization of Japanese Business. Berkeley: The University

of California Press.Glass, J.C., and W. Johnson (1989) Economics: Progression,

45

Stagnation, or Degeneration? Ames: Iowa State University Press.

Hands, D. Wade (1991) "The Problem of Excess Content: Economics,Novelty and a Long Popperian Tale." In de Marchi and Blaug,eds.---. (1993) "Popper and Lakatos in Economic

Methodology." In Maki, et. al.Laibman, David (1993) Value, Technical Change, and Crisis.

Armonk, N.Y.: M.E. Sharpe.Lakatos, Imre (1968) "Criticism and the Methodology of Scientific

Research Programs." Proceedings of the Aristotelian Society Vol. 69 (149-86)---. (1970) "Falsification and the Methodology of Scientific Research Programmes." In Lakatos and Musgrave.---. (1978a) Philosophical Papers, Vol. 1. Cambridge:Cambridge University Press.---. (1978b) Philosophical Papers, Vol. 2. Cambridge: Cambridge University Press.

Lakatos, Imre, and A. Musgrave, eds. (1970) Criticism and the Growth of Knowledge. New York: Cambridge University Press.Latsis, Spiro (1972) "Situational Determinism in Economics." British Journal of the Philosophy of Science Vol. 23 (207-

45).---, ed. (1976) Method and Appraisal in Economics. New York:

46

Cambridge University Press.Leijonhufvud, Axel (1976) "Schools, 'Revolutions', and Research

Programes In Economic Theory." In Latsis.Machlup, F. (1978) Methodology of Economics and Other Social Sciences. New York: Academic Press.Maki, Uskali, B. Gustafsson, and C. Knudsen, eds. (1993) Rationality, Institutions & "Economic Methodology." New

York: Routledge Press.Marx, Karl (1968) Theories of Surplus Value, Volume 2 Moscow:

Progress Publishers.---. (1973) Grundrisse. New York: Vintage.---. (1978) Capital. Three volumes. Moscow: ProgressPublishers.

Marx, Karl, and F. Engels (1983) Letters on 'Capital'. London: New Park.

Moseley, Fred, ed.(1992) New Directions in Marxian Theory: Form and Logic in Capital. New York: Humanities Press.

---. (1994) "Unproductive Labor and the Rate of Profit: A Reply." Science and Society Vol. 58, No. 1.

Murray, Patrick (1994)Ollman, Bertell (1976) Alienation. Cambridge: Cambridge

University Press.---. (1993) Dialectical Methodology

Pratt, V. (1980) The Philosophy of Social Science. London:

47

Tavistock.Redman, Deborah (1991) Economics and the Philosophy of Science.

New York: Oxford University Press.Reuten, Geert (1993a) "The Difficult Labor of a Theory of Social

Value: Metaphors and Systematic Dialectics at the Beginning of Marx's Capital." In Moseley---. (1993b) [ref. to last year's paper]

Reuten, Geert, and M. Williams (1989) Value-Form and the State. London: Routledge.

Sayer, Andrew (1984) Method in Social Science: A Realist Approach. London: Hutchinson.Smith, Tony (1990) The Logic of Marx's Capital. Albany: State

University of New York Press.---. (1993 a) Dialectical Social Theory and Its CriticsAlbany: State University of New York Press.---. (1993 b) "Marx's Capital and Hegelian Dialectical Logic." In Moseley, ed.

Storper, Michael, and R. Walker (1989) The Capitalist Imperative: Territory, Technology, and Industrial Growth. New York:

Blackwell.Vanberg, Viktor (1993) "Rational Choice, Rule-Following and

Institutions: An Evolutionary Perspective." In Maki, et. al.Vogel, Lise (1983) Marxism and the Oppression of Women: Toward a Unitary Theory New Brunswick, N.J.: Rutgers University

48

Press.Walker, Dick (1989) "In Defense of Realism and Dialectical

Materialism." Berkeley Journal of Sociology. Vol. 34.

i. This, of course, is not to imply that all mainstream economists have accepted Lakatos. (See Hands 1993; de Marchi and Blaug)

ii. To my knowledge the only extensive discussion of Marxian theory from a Lakatosian perspective is found in Glass and Johnson. Their account does not emphasize the main theme of the present essay, Marx'sordering of social forms.

iii. The heuristics of a program could be considered as part of its hard core. I have found it more convenient to consider them separately.

iv. The term "protective belt" emphasizes that the theories and empirical assumptions making up this belt are subjected to the empirical testing, thereby "protecting" the hard core of the researchprogram from such testing.

v. Marx himself acknowledged the importance of Hegel's Logic for an understanding of Capital: "I leafed through Hegel's Logic again and found much to assist me in the method of analysis. (Marx and Engels 1983, 50) I have attempted to substantiate this reading of Capital in detail in Smith 1990. See also Reuten and Williams.

vi. This is the so-called "qualitative" dimension of the labor theory of value.

vii. A quick note on the ambiguity of the term "concrete" is called for. In one sense, the ordering of socio-economic categories in Capital moves from the abstract to the concrete. The categories of Volume III are much closer to the surface level of appearances than those of Volume I. Nonetheless, a concrete social form (defined by a

concrete category) still remains a form of social life, that is, a structure within which concrete events unfold, in a different sense of the term "concrete." There is an unbridgeable gulf between the logic of capital - including the most "concrete" stages of that logic- and the concrete processes of capitalism.

viii. See Reuten 1993a on the pernicious impact of the residues of Ricardian value theory in Capital.

ix. See Arthur's proposed changes to Marx's ordering of economic categories, made from the standpoint of Marx's own principles. (Arthur 1993)

x. See the work of materialist feminists, one implication of which isthat room must be found for gender-specific categories in any systematic reconstruction of the capitalist mode of production. (Vogel)

xi. Blaug gives his own reasons for rejecting instrumentalism and its "poor man's version," descriptivism, in the following passage: "[Their] weakness is that of all black-box theorizing that makes predictions without being able to explain why the predictions work: the moment the predictions fail, the theory has to be discarded in toto because it lacks an underlying structure of assumptions, an explanans that can be adjusted and improved to make better predictions in the future." (Blaug 99) But in principle instrumentalists/descriptivists can provide elaborate stories of whatis "inside the black box." What distinguishes their view is the claim that there is no way to assess these stories apart from their role in generating corroborated predictions. And this is precisely

what Blaug (and Lakatos) elsewhere assert.

xii. This does not imply that Marxists need reject instrumentalist laws in contexts where calculating devices are all that is called for. (Sayer 116-17) Nor does it imply that Marxist theorists never utter statements inconsistent with the underlying commitment to realism implicit in Marxian theory. (Some examples are discussed in Walker.)

xiii. The concept of mechanism defined in this passage is closely related to the notion of formal causality explored by Patrick Murray elsewhere in this volume.

xiv. "The classical rational choice paradigm seems to owe its appeal [to the fact that] it provides a unified theory which applies to all human behavior, independent of the particularities of time and place." (Vanberg 183)

xv. The assumption of perfect knowledge is a special case, to which I shall return below.

xvi. Of course, two people holding these position may be externally (contingently) related in many other dimensions. (Sayer 84)

xvii. This is not to deny that other factors (such as gender and race)are also constitutive of individual identity.

xviii. Blaug writes, "No doubt, the postulate that economic agents act rationally in their own interests with perfect knowledge and correct expectations only makes sense when we are in equilibrium, while complicating the story of how we get there from a situation of

disequilibrium: in equilibrium, market price carry all the knowledge we need but out of equilibrium they systematically mislead us." (Blaug 159) Just one question: in the entire history of capitalism when have "we" ever been in this equilibrium?

xix. Much of the new institutionalism redefines the notion of "rational" to refer to the methodical following of institutionalized routines. These routines are strategic responses to situations whereaccumulating all of the relevant information is either impossible in principle or excessively costly to obtain. In other words, the rational individuals of neo-institutionalism are "satisficers" ratherthan maximizers. As far as it goes, this is compatible with the Marxian emphasis on the dynamism of capitalism, although the continued emphasis on individualism brings the neo-institutionalist perspective closer to the neoclassical position. See Maki, et. al., Part IV.

xx. Only the beginning sections of Max Weber's Economy and Society come to mind as a comparable achievement, although Weber's presentation of the fundamental categories of his research program lacks the breadth, depth, and logical rigor found in Capital.

xxi. I do not mean to suggest that non-class issues such as race or gender fall outside the scope of Marxian research. But the class-specific manner in which racial and gender oppression occurs must always be made explicit.

xxii. This dimension of a research program was of great importance to Lakatos. In one place, at least, he even suggests that the introduction of new problem-solving techniques is the most important

factor in determining whether a research program is progressing, more important than even the prediction of novel facts: "progress is not even so much in the actual novel predictions . . . but in the mathematical and physical novelty of problem-solving techniques which later lead to, and form a part of, a progressive research program." (Lakatos 1978b 101) It is not clear to me how this statement is consistent with the stress on novel predictions elsewhere in his writings.

xxiii. Walter Benjamin's use of montage is relevant here. This is a method of juxtaposition that aims to show how the promise of fulfillment, and the hell of the endless return of the ever-the- same, co-exist in commodity society. Susan Buck-Morss terms this method "the dialectics of seeing."

xxiv. None of this should imply that survey questionnaires are ruled out in the Marxist research program tout court. But certain types ofquestionnaires most certainly are.

xxv. "Neo-Walrasian closed system models have so far been inadequate -or, at best, grotesquely cumbersome - vehicles for representing the role of ignorance and the passage of time in human affairs . . . Bothproblems are, it would appear, rooted in the hard-core heuristic routine of modelling the behavior of each individual agent so as to portray his every action as part of a comprehensively planned 'optimal' time-path." (Leijonhufvud 107)

xxvi. Mathematical catastrophe theory has great promise in this context as well. It seems especially suited for representing phenomenon such as a sudden stampede into liquidity. (Laibman 207)

(Is catastrophe theory mathematics finally catching up to the dialectic principle that "quantity becomes quality"?)

xxvii. "(T)he logical or mathematical manipulation of a hypothetical closed system represented by symbols is a poor guide to causal structure, for the rules governing these kinds of manipulation need not correspond to the laws governing the possible ways-of-acting of real objects; models may be run backwards, effects can be used to calculate ('determine') causes and hence an ability to calculate and 'predict' may rest upon mis-specifications of even the basic asymmetries of causal dependence." (Sayer 118-19)

xxviii. There has been a quite extensive debate on what exactly the term "novel facts" means. The conclusion of this debate appears to be thatthe term can be unpacked in a variety of ways, no one of which is reducible to the others. (Hands 1991b 96 ff.) Towards the end of hislife Lakatos himself stressed that predicting what is already known, but merely "fortuitous" within competing theories, also counts as a prediction of a novel fact. (Lakatos 1978a 179, 183, 184-5, 186) While economic crises, capital/wage labor conflict, and the concentration and centralization of capital, are phenomena that have been known for quite some time, their persistence is certainly "fortuitous" from the standpoint of neoclassical economics. The fact that Marxian economic theory successfully predicted their persistence is strong evidence for the progressiveness of the Marxian program, as measured by Lakatos's standard.

xxix. In these sorts of cases I believe it is arbitrary to grant one member of the pair priority by terming it "the tendency," reducing

the other to the secondary status of a mere "counter-tendency." If both are equally immanent in the capital form, the language of "conflicting tendencies" is less misleading. The term "real contradictions" could even be used, if one did not mind sending the lovers of formal logic into a state of apoplexy.

xxx. Other matters outside the scope of this paper complicate matters even further. Gender, race, aesthetic experiences, ethical values, and religious commitments, are just some of the features that would have to be introduced to get a full categorization of economic agents. The more dimensions are introduced to the categorization of individual economic agents, the more indeterminate their behavior in any given specific situation.