REPLY TO ELSTER ON “MARXISM, FUNCTIONALISM, AND GAME THEORY”
______________________________________________________________________
Jon Elster and I each worked sympathetically on Marxism for a long time, and each of us independently came to see that Marxism in its traditional form is associated with explanations of a special type, ones in which, to put it roughly, consequences are used to explain causes. In keeping with normal practice Elster calls such explanations functional explanations, and I shall follow suit here.1 He deplores the association between Marxism and functional explanation, because he thinks there is no scope for functional explanation in social science. It is, he believes, quite proper in biology, because unlike social phenomena, biological ones satisfy the presuppositions that justify its use. Elster therefore concludes that the Marxist theory of society and history should abandon functional explanation. He also thinks it should, instead, draw for its explanations on the resources of game theory.
I do not think that course is open to historical materialism. I believe that historical materialism’s central explanations are unrevisably functional in nature so that if functional explanation is unacceptable in social theory then historical materialism cannot be reformed and must be rejected. But I do not think functional explanation is unacceptable in social theory. My judgment that historical materialism is indissolubly wedded to functional explanation reflects my conception of the content of historical materialist theory. To display, then, the grounds of that judgment, I shall expound what I think historical materialism says. I shall provide a résumé of the theory that I attribute, on a textual basis, to Marx, and that I explicate and defend in my book Karl Marx’s Theory of History.2
In my book I say, and Marx says, that history is, fundamentally, the growth of human productive power, and that forms of society rise and fall according as they enable and promote, or prevent and discourage, that growth. The canonical text for this interpretation is the famous 1859 “Preface” to A Contribution to the Critique of Political Economy, some sentences of which we shall look at shortly. I argue (in section 3 of chapter 6) that the “Preface” makes explicit the standpoint on society and history to be found throughout Marx’s mature writings, on any reasonable view of the date at which he attained theoretical maturity. In attending to the “Preface,” we are not looking at just one text among many, but at that text which gives the clearest statement of the theory of historical materialism. The presentation of the theory in the “Preface” begins as follows:
In the social production of their life men enter into definite relations that are indispensable and independent of their will, relations of production which correspond to a definite stage of development of their material productive forces. The sum total of these relations constitutes the economic structure of society, the real basis, on which arises a legal and political superstructure.3 (emphasis added)
These sentences mention three ensembles, the productive forces, the relations of production, and the superstructure, among which certain explanatory connections are asserted. Here I say what I think the ensembles are, and then I describe the explanatory connections among them. (All of what follows is argued for in KMTH, but not all of the argument is given in what follows, which may therefore wrongly impress the reader as dogmatic.) The productive forces are those facilities and devices used in the process of production: means of production on the one hand, and labor power on the other. Means of production are physical productive resources; e.g., tools, machinery, raw materials, and premises. Labor power includes not only the strength of producers, but also their skills, and the technical knowledge (which they need not understand) they apply when laboring. Marx says, and I agree, that this subjective dimension of the productive forces is more important than the objective or means of production dimension; and within the more important dimension the part most capable of development is knowledge. In its higher stages, then, the development of the productive forces merges with the development of productively useful science.
Note that Marx takes for granted in the “Preface,” what elsewhere he asserts outright, that “there is a continual movement of growth in productive forces.”4 I argue (in section 6 of chapter 2 of KMTH) that the relevant standard for measuring that growth in power is how much (or, rather, how little) labor must be spent with given forces to produce what is required to satisfy the inescapable physical needs of the immediate producers.5 This criterion of social productivity is less equivocal than others that may come to mind, but the decisive reason for choosing it is not any such “operational” advantage, but its theoretical appropriateness: if kinds of economic structure correspond, as the theory says they do, to levels of productive power, then this way of measuring productive power makes the theory’s correspondence thesis more plausible.6 (I do not say that the only explanatory feature of productive power is how much there is of it: qualitative features of productive forces also help to explain the character of economic structures. My claim is that insofar as quantity of productive power is what matters, the key quantity is how much time it takes to reproduce the producers.)
We turn to relations of production. They are relations of economic power, of the economic power7 people enjoy or lack over labor power and means of production. In a capitalist society relations of production include the economic power capitalists have over means of production, the limited but substantial economic power workers (unlike slaves) have over their own labor power, and the lack of economic power workers have over means of production. The sum total of production relations in a given society is said to constitute the economic structure of that society, which is also called—in relation to the superstructure—the basis, or base, or foundation. The economic structure or base therefore consists of relations of production only: it does not include the productive forces. The “Preface” describes the superstructure as legal and political. So it at any rate includes the legal and state institutions of society. It is customary to locate other institutions within it too, and it is controversial what its correct demarcation is: my own view is that there are strong textual and systematic reasons for supposing that the superstructure is a lot smaller than many commentators think it is.8 It is certainly false that every noneconomic social phenomenon is superstructural: artistic creation, for example, is demonstrably not, as such, superstructural, for Marx. In these remarks I shall discuss the legal order only, which is uncontroversially a part of the superstructure.
So much for the identity of the three ensembles mentioned in the “Preface.” Now relations of production are said to correspond to the level of development of the productive forces, and in turn to be a foundation on which a superstructure rises. I think these are ways of saying that the level of development of the productive forces explains the nature of the production relations, and that they in turn explain the character of the superstructure copresent with them. But what kind of explanation is ventured here? I argue that in each case what we have is a species of functional explanation.
What sort of explanation is that? It is, very roughly, an explanation in which an event, or whatever else, if there is anything else that can have an effect, is explained in terms of its effect. But now let us be less rough. Suppose we have a cause, e, and its effect, f. Then the form of the explanation is not: e occurred because f occurred—that would make functional explanation the mirror image of ordinary causal explanation, and then functional explanation would have the fatal defect that it represented a later occurrence as explaining an earlier one. Nor should we say that the form of the explanation is “e occurred because it caused f.” Similar constraints on explanation and time order rule that candidate out: by the time e has caused f, e has occurred, so the fact that it caused f could not explain its occurrence. The only remaining candidate, which I therefore elect, is: e occurred because it would cause f, or, less tersely but more properly, e occurred because the situation was such that an event of type E would cause an event of type F.9 So in my view a functional explanation is an explanation in which a dispositional fact explains the occurrence of the event-type mentioned in the antecedent of the hypothetical specifying the disposition. I called the laws justifying functional explanations consequence laws. They are of roughly this form:(E > F) > E (a more precise specification of their form is given in section 4 of chapter 9 of KMTH). If this account of what functional explanations are is correct, then the main explanatory theses of historical materialism are functional explanations. For superstructures hold foundations together, and production relations control the development of productive forces: these are undeniable facts, of which Marx was aware. Yet he asserts that the character of the superstructure is explained by the nature of the base, and that the base is explained by the nature of the productive forces. If the intended explanations are functional ones, we have consistency between the effect of A on B and the explanation of A by B, and I do not know any other way of rendering historical materialism consistent.
I now expound in greater detail one of the two functional explanatory theses, that which concerns base and superstructure. The base, it will be recalled, is the sum total of production relations, these being relations of economic power over labor power and means of production. The capitalist’s control of means of production is an illustration. And the superstructure, we saw, has more than one part; exactly what its parts are is somewhat uncertain, but certainly one bona fide part of it is the legal system, which will occupy us here. In a capitalist society capitalists have effective power over means of production. What confers that power on a given capitalist, say an owner of a factory? On what can he rely if others attempt to take control of the factory away from him? An important part of the answer is this: he can rely on the law of the land, which is enforced by the might of the state. It is his legal right that causes him to have his economic power. What he is effectively able to do depends on what he is legally entitled to do. And this is in general true in law-abiding society with respect to all economic powers and all economic agents. We can therefore say: in law-abiding society people have the economic powers they do because they have the legal rights they do.
That seems to refute the doctrine of base and superstructure, because here superstructural conditions—what legal rights people have—determine basic ones—what their economic powers are. But although it seems to refute the doctrine of base and superstructure, it cannot be denied. And it would not only seem to refute it, but actually would refute it, were it not possible, and therefore mandatory (for historical materialists), to present the doctrine of base and superstructure as an instance of functional explanation. For we can add, to the undeniable truth emphasized above, the thesis that the given capitalist enjoys the stated right because it belongs to a structure of rights, a structure that obtains because it sustains an analogous structure of economic power. The content of the legal system is explained by its function, which is to help sustain an economy of a particular kind. People do usually get their powers from their rights, but in a manner that is not only allowed but demanded by the way historical materialism explains superstructural rights by reference to basic powers. Hence the effect of the law of property on the economy is not, as is often supposed, an embarrassment to historical materialism. It is something that historical materialism is committed to emphasizing, because of the particular way it explains law in terms of economic conditions. Legal structures rise and fall according as they sustain or frustrate forms of economy that, I now add, are favored by the productive forces. The addition implies an explanation why whatever economic structure obtains at a given time does obtain at that time. Once more the explanation is a functional one: the prevailing production relations prevail because they are relations that advance the development of the productive forces. The existing level of productive power determines what relations of production would raise its level, and relations of that type consequently obtain. In other words: if production relations of type R obtain at time t, then that is because R-type relations are suitable to the development of the forces at t, given the level of their development at t.10
Now to say that A explains B is not necessarily to indicate how A explains B. The child who knows that the match burst into flame because it was struck may not know how the latter event explains the former (because he is ignorant of the relationship between friction and heat, the contribution of oxygen to combustion, and so on).11 In this sense of “how,” we can ask: how does the fact that the economic structure promotes the development of the productive forces (or that the superstructure protects the base) explain the character of the economic structure (or the superstructure)? Consider an analogy: to say, correctly, that the species giraffe developed a long neck because of the utility of that feature in relation to the diet of giraffes (acacia tree leaves) is not to say how the utility of that feature accounted for its emergence or persistence. To that question Lamarck gave an unacceptable answer and Darwin an excellent one. To the corresponding questions within historical materialism no one has given excellent answers. I make some unexcellent attempts in chapter 10 of my book. This seems to me an important area of future research for proponents of historical materialism, because the functional construal of the doctrine cannot be avoided.
Let me now summarize my argument for the thesis that the chief explanatory claims of historical materialism are functional in form. Historical materialism’s central claims are that:
(1) The level of development of the productive forces in a society explains the nature of its economic structure, and
(2) its economic structure explains the nature of its superstructure.
I take (1) and (2) to be functional explanations, because I cannot otherwise reconcile them with two further Marxian theses, namely that
(3) the economic structure of a society promotes the development of its productive forces, and
(4) the superstructure of a society stabilizes its economic structure.
(3) and (4) entail that the economic structure is functional for the development of the productive forces, and that the superstructure is functional for the stability of the economic structure. These claims do not by themselves entail that economic structures and superstructures are explained by the stated functions: A may be functional for B even when it is false that A exists, or has the character it does, because its existence or character is functional for B. But (3) and (4), in conjunction with (1) and (2), do force us to treat historical materialist explanation as functional. No other treatment preserves consistency between the explanatory primacy of the productive forces over the economic structure and the massive control of the latter over the former, or between the explanatory primacy of the economic structure over the superstructure and the latter’s regulation of the former. I did not come to associate historical materialism with functional explanation because I thought functional explanation a good thing and I therefore wanted Marxism to have it. I began with a commitment to Marxism, and my attachment to functional explanation arose out of a conceptual analysis of historical materialism. I do not see how historical materialism can avoid it, for better or for worse. Contrast Jon Elster’s attitude to Marxism and game theory. He wants Marxism to liaise with game theory because he admires game theory and thinks Marxism can gain much from the match. He wants to put Marxism and game theory together. I would not say that I want to put together Marxism and functional explanation, because I think functional explanation is inherent in Marxism.
At the beginning of his article Elster complains that Marxist social analysis has been contaminated by the principles of functionalist sociology. I am sure that claim is both historically and conceptually incorrect. Marxists do not indulge in functional explanation because they are influenced by the bad bourgeois science of functionalist sociology, and it is not open to them to use the better bourgeois science of game theory instead. They indulge in functional explanation because they are committed to historical materialism. Because functional explanation cannot be removed from the center of historical materialism, game theory cannot be installed there in its stead. But it might be thought that game theory could also figure at the center of historical materialism, not as a replacement but as an addition. Yet that, too, I argue, is false. Game theory may be, as Elster says, “tailor-made for Marxist analysis,”12 but it is irrelevant to historical materialism’s central theses, which are propositions (1) and (2). Its relevance, as I now explain, is to theses immediately peripheral to (1) and (2).
Elster makes deft use of game theory in a discussion of the dialectics of class struggle that I greatly admire. And it is not surprising that game theory illuminates class behavior. But Marxism is fundamentally concerned not with behavior, but with the forces and relations constraining and directing it. When we turn from the immediacy of class conflict to its long-term outcome, game theory provides no assistance, because that outcome, for historical materialism, is governed by a dialectic of forces and relations of production that is background to class behavior, and not explicable in terms of it. Game theory helps to explain the vicissitudes of the struggle, and the strategies pursued in it, but it cannot give a Marxist answer to the question why class wars (as opposed to battles) are settled one way rather than another. The Marxist answer is that the class that rules through a period, or emerges triumphant from epochal conflict, does so because it is the class best suited, most able and disposed, to preside over the development of the productive forces at the given time.13 That answer may be untenable, but I cannot envisage a game-theoretical alternative to it that would qualify as historical materialist.
Elster says that “game theory is invaluable to any analysis of the historical process that centers on exploitation, struggle, alliances, and revolution.”14 But for Marxian analysis those phenomena are not primary but, as it were, immediately secondary, on the periphery of the center: they are, in the words of the 1859 “Preface,” the “forms in which men become conscious of the conflict [between forces and relations of production] and fight it out.”15 To put the point differently, we may say that the items on Elster’s list are the actions at the center of the historical process, but for Marxism there are also items more basic than actions at its center.16 By “revolution” Elster must mean the political phenomenon of transfer of state power, as opposed to the transformation of economic structure political revolution initiates or reflects. Many facts about political revolutions are accessible to game theoretical explanation, but not the world-historical facts that there was a bourgeois revolution and that there will be a proletarian one. Elster urges that game theory bears on strategic questions of great importance to Marxists. I accept that contention, which is amply supported by the excellent illustrations in his article. When faced with a strategic problem, such as how to transform society, we need strategic, not functionalist, thinking. But when Marx called on the workers to revolutionize society, he was not asking them to bring about what would explain their doing so: the exhaustion of the progressive capacity of the capitalist order, and the availability of enough productive power to install a socialist one.
The concepts exercised in the previous sentence take us away from game theory to the fundamental context of historical materialism, that of forces and relations of production. There exists a splendid unpublished essay by Jon Elster entitled “Forces and Relations of Production.” The essay makes no use of game theory. That is striking confirmation of my view that it is irrelevant to the foundational claims of Marxism: it shows that Elster himself agrees, in practice, with that view. Having constructed a rigorous theory of contradiction between forces and relations of production, Elster says that “the great weakness of the theory is that it is very difficult to link it to action.” Now despite my insistence on the centrality in historical materialism of things that are not actions, I do appreciate that actions are prominent proximate causes of social effects. If links with action cannot be forged, if the question how the functional explanations of historical materialism explain cannot even in principle be answered, then that would have lethal significance for historical materialism. And this brings me to Elster’s critique of functional explanation.
I remarked earlier that even when A is functional for B, A’s existence or character need not be explained by that fact. Thus to confer credibility on the claim that B functionally explains A, one must supply evidence in excess of that needed to show that A is functional for B. Elster and I disagree about what sort of further evidence is necessary. He demands that the claim that B functionally explains A be supported by a plausible story that reveals how B functionally explains A. I think that is sufficient, but not necessary. For I think one can support the claim that B functionally explains A even when one cannot suggest what the mechanism is, if instead one can point to an appropriately varied range of instances in which, whenever A would be functional for B, A appears.17 This is an application to functional explanatory claims of a general truth about explanatory claims. There are always two ways of backing them up. Suppose, for example, that Elster and I notice a dead body in the library of the country house the morning after the dinner party, and that we hypothesize that its owner died because of something he ate the night before. Further research can take either of two forms. We might open him up to see whether there are any poisons in him, which would be analogous to what Elster thinks we must do to back up functional explanations, or we might find out what he ate, what other guests ate, and which other guests took ill or died, and that would be analogous to the way I say we can proceed with functional explanations. In my procedure we look for appropriately consonant and discrepant parallel instances. In Elster’s we rely on preexisting knowledge about parallel instances at a more basic causal level and we look for a mechanism in the given case that is consonant with that knowledge.
I can illustrate what is at stake by reference to the case of Lamarck and Darwin. Darwin showed how functional facts about the equipment of organisms contribute to explaining why they have it: the answer lies in the mechanism of chance variation and natural selection. Now I claim, and Elster denies, that, before Darwin thereby advanced the science of natural history, the belief that the useful characters of organisms are there because they are useful was already justified, by the sheer volume of evidence of adaptation. The belief was certainly widely held, by people who had no idea how to elaborate it and by others, such as Lamarck, who had what proved to be an unsatisfactory idea of how to elaborate it. And I contend, and Elster denies, that it was a justified belief. This debate is pursued elsewhere, and I shall not take it further here.18
Now because I concede that Marxists have not yet produced good elaborations of their functional explanatory theses, I concede that historical materialism is at best in a position like that occupied by natural history before Darwin transformed the subject. But I am not convinced that it has got even that far. For whereas Elster and I disagree strongly about what would confirm functional explanations, we disagree less about whether Marxists have actually produced well-confirmed functional explanations. The essays in Marxist functional explanation which he discusses are sadly representative, and I have no desire to defend them against his criticisms. Here we can make common cause. Many Marxist exercises in functional explanation fail to satisfy even the preliminary requirement of showing that A is functional for B (whether or not it is also explained by its function(s)).19 Take, for example, the claim that the contemporary capitalist state functions to protect and sustain the capitalist system. Legislation and policy in the direct interest of the capitalist class can reasonably be regarded as confirming it. But what about putative counterexamples, such as social welfare provision and legal immunities enjoyed by trade unions? These too might be functional for capitalism in an indirect way, but that is something which needs to be argued with care, not just asserted. But those who propound the general claim about the state rarely trouble to say what sort of evidence would falsify or weaken it, and therefore every action of the state is treated as confirmatory, because there is always some way, legitimate or spurious, in which the action can be made to look functional. Methodological indiscipline is then compounded when, having established to his own satisfaction that state policy is functional, the theorist treats it, without further argument, as also functionally explained. He proceeds from “A is functional for B” to “B functionally explains A” without experiencing any need to justify the step, if, indeed, he notices that he has taken a step from one position to a distinct and stronger one.20
Most Marxists are methodologically unself-conscious. If they were more sophisticated, they might provide a better defense of the functional explanations they offer. And then, again, they might not. I do not know how to be confident about this, one way or the other. But I maintain my insistence, first, that historical materialism cannot shed its commitment to functional explanation, and, second, that there is nothing inherently suspect in it. Elster’s philosophical criticisms of historical materialist functional explanation still strike me as without force, by contrast with his polemic against particular essays in functional explanation. Our philosophical disagreement is pursued in Political Studies and Inquiry. In fn. 8 of his present contribution Elster offers two new objections to my own theory of functional explanation, both of which are misguided. His first objection is that even when it is true that whenever A would have favorable consequences for B, A appears, A might not be explained by its possession of such consequences, because a third factor, C, might both cause A to have favorable consequences for B, and cause A to appear, without causing the latter as a result of causing the former. That is so, but it is not an objection to my theory.21 The form of an ordinary causal law is: whenever A occurs, B occurs. Once again, this might be caused by a third factor, C, so related to A and B A does not qualify as causing B. But there are tests which, when appropriate results are forthcoming, render the hypothesis that there exists such a C implausible, and suitably analogous tests may be conducted in the case of consequence laws.22 Elster’s second fresh objection rests on the premise that I do not mention time in my characterization of consequence laws. It is true that I do not mention particular amounts of time when describing the form of such laws in general terms, just as one does not when one describes the form of ordinary causal laws as “whenever A occurs, B occurs.” But causal laws are not therefore “vacuously confirmable,” because particular causal laws include appropriate temporal specifications. All that need be said in general terms about consequence laws and time will be found on pp. 260–61 of KMTH.
I now take up two issues in the part of Elster’s article in which he successfully conjoins Marxism and game theory. In a highly original account of the ideology and practice of social democratic capitalism, Elster sets the stage by describing the dissolution of the marginalist illusion, and the action unfolds along lines scripted by Zeuthen and Nash on the one hand and Lancaster on the other. I have two criticisms of this treatment. The first is that Elster misidentifies the illusion that survives after the marginalist one has been dissolved. He calls it “the presentist illusion,”23 and attributes it to “diachronic alienation.”24 Workers are alienated “from their own history, i.e., from past generations of workers who produced the means of production currently used,” and they overcome that alienation “by taking possession of their history.”25 Elster would agree that unrevolutionary workers believe that the capitalist is entitled to a return because he is the morally legitimate owner of the means of production. He thinks the presentist illusion explains why they think the capitalist’s ownership is legitimate. But in what does the illusion consist? In a false belief that the means of production were not produced by workers in the past? But workers know better than that. They know, if they reflect on the matter, that means of production were produced by earlier workers, but just as they believe that their own employer is entitled to a return, so, in parallel, they think the employer of earlier workers was; whence, in particular, employers of workers producing means of production came to possess them legitimately and passed them on, directly or indirectly, through market exchange and gift (especially inheritance), to the employers of today. If there exists any kind of presentist illusion, why should workers not project it backwards when they think about their predecessors?
My second criticism of the game theoretical part of Elster’s article concerns his remarks on the locus of exploitation. He writes
that the exploitation of the working class … does not consist only in the capitalists’ appropriation of surplus-value but also in the workers’ exclusion from decisive investment choices that shape the future.26
Much the same sentence occurs in an earlier version of Elster’s article, except that the word “mainly” occurs where the word “only” appears in this final version. This reply was originally composed in response to that earlier version. Having read my response, Elster changed “mainly” to “only,” thereby partly spoiling some criticisms I had made of the original version. I shall nevertheless enter the following paragraph of criticism of his original formulation (the one with “mainly”) here, not only out of vanity but also because it still applies, if with reduced force, against his revised formulation, and most importantly because I think it is useful to try to identify rather precisely what exploitation consists in.
I do not doubt that workers are excluded from investment decisions, but I deny that they are thereby exploited. If someone robs me of the power to control my own life, he does not ipso facto use me unfairly to his own advantage, which is what, very roughly, exploitation is. Authoritarian parents do not, by virtue of being authoritarian, qualify as exploiters of their children, and authoritarian parenthood is a good analogue to the relationship Elster highlights here, which is one of subordination, not exploitation. That subordination is, moreover, a consequence of exploitation in the traditional sense, which is therefore not displaced by (what is anyway wrongly considered) a further form of exploitation. It is because capitalists appropriate surplus value that they are able to decide what to do with it, to consume and invest in whatever proportions they choose. And the exploitation of the worker lies in the appropriation, not in the subsequent disposal over what has been appropriated. Part of what moved Elster to make his (original) statement was the fact, which he emphasizes elsewhere, that only a small proportion of total social product remains for capitalist consumption after workers’ income and capitalist investment have absorbed their shares.27 But because there are relatively few capitalists, that small proportion enables them to enjoy a life of comfort and freedom inaccessible to workers. The difference in per capita personal income remains massive, and it matters a great deal to the self-perception and sense of dignity of working people. Working-class existence, even in America, is full of strain unknown to wealthy people. Elster’s (original) formulation overlooks that sheer difference in standard of living between the classes remains a major part of the injustice of capitalism.
My present view about the matters in contention between Elster and myself is as follows: (1) Functional explanation lies at the heart of historical materialism. (2) Game theory therefore cannot replace functional explanation within Marxist social analysis. (3) Nor is there a place for game theory at the heart of historical materialism, alongside functional explanation. (4) But game theory is very helpful in relation to claims near, but not quite at, historical materialism’s heart. (5) There is no methodological error in historical materialism’s functional explanatory theses. (6) But Marxists have not done much to establish that they are true. If Marxian functional explanation remains as wanting in practice (as opposed to high theory) as it has been, the foundational claims of historical materialism might need to be severely modified. Positions of great traditional authority might have to be abandoned. One of Elster’s achievements is that he has shown how fruitfully what would remain of the doctrine we have inherited can be enriched and extended.
1 For reasons given in my “Functional Explanation, Consequence Explanation, and Marxism” I am not certain that explanations of causes by consequences should be considered functional explanations, but that issue is irrelevant to Elster’s article (“Marxism, Functionalism, and Game Theory: The Case for Methodological Individualism”), so I shall here fall in with the standard practice of regarding what I would call consequence explanations as functional explanations. Much of this reply has already appeared in the article mentioned above, published in Inquiry, and I am grateful to the editor of that journal for allowing it to be reproduced here.
2 Cohen, Karl Marx’s Theory of History, henceforth referred to as KMTH.
3 Quoted in KMTH, p. vii.
4 Marx, The Poverty of Philosophy (Lawrence and Wishart edition), p. 166.
5 As opposed, for example, to their socially developed needs, reference to which would be inappropriate here (though not, of course, everywhere).
6 For a set of correspondences of relations to forces of production, see KMTH, p. 198.
7 I call such power “economic” in virtue of what it is power over, and irrespective of the means of gaining, sustaining, or exercising the power, which need not be economic. See KMTH, pp. 223–24.
8 The common practice of overpopulating the superstructure is criticized in my review of Melvin Rader’s Marx’s Interpretation of History.
9 Small letters represent phrases denoting particular events, and capital letters represent phrases denoting types of event. Where the letters are the same, the particular event belongs to the type in virtue of the meanings of the phrases denoting them.
10 For a detailed account of the nature of the primacy of the forces, see section 5 of chapter 6 of KMTH, which also discusses the transitional case.
11 In a widely favored idiom, he may not know the mechanism linking cause and effect, or, as I prefer to say, he may be unable to elaborate the explanation. I use both forms of expression in the sequel.
12 Elster, Ulysses and the Sirens, p. 34.
13 See KMTH, pp. 148–49.
14 Elster, “Marxism, Functionalism, and Game Theory,” p. 453.
15 Cited in KMTH, p. vii.
16 Hence to say, as some Marxists do, that “class struggle is the motor of history” is to abandon historical materialism.
17 That is the simplest way of confirming a functional explanation without establishing a mechanism. For more complicated ways, see KMTH, chap. 9, secs. 5 and 7.
18 See the exchange between Elster and myself in Political Studies: Elster, “Cohen on Marx’s Theory of History”; Cohen, “Functional Explanation: Reply to Elster.” See especially Elster p. 126, and Cohen, p. 133–34, and “Functional Explanation, Consequence Explanation, and Marxism.” One result reached in the latter article bears mention here. I show that if Elster is right about what functional explanation is (he says what it is in Ulysses and the Sirens), then he is wrong that natural selection is necessary to sustain functional explanations in biology. It follows that he is also wrong in the corresponding claims about sociological functional explanation at p. 455 and p. 463 of “Marxism, Functionalism, and Game Theory: The Case for Methodological Individualism.”
19 Elster does not always distinguish this criticism from the one I make in the next paragraph; see, for example, his comments (p. 458) on the passage from The Eighteenth Brumaire. If he is right, both criticisms apply, but he does not properly separate them.
20 And sometimes it is unclear that a step has been taken from a statement of functionality to a functional explanation, and, therefore, it is correspondingly unclear that a fallacy has been committed. Thus, for example, I do not share Elster’s confidence that Marx’s use of the word “means” in the quotation from vol. 3 of Capital on p. 457 proves that Marx is offering a functional explanation, and I am sure that he is wrong when he claims (p. 456) that Marx subscribed to “the main functional paradigm.”
[This is the passage from Marx:
The circumstance that a man without fortune but possessing energy, solidity, ability and business acumen may become a capitalist in this manner [i.e., by receiving credit]—and the commercial value of each individual is pretty accurately estimated under the capitalist mode of production—is greatly admired by the apologists of the capitalist system. Although this circumstance continually brings an unwelcome number of new soldiers of fortune into the field and into competition with the already existing individual capitalists, it also reinforces the supremacy of capital itself, expands its base and enables it to recruit ever new forces for itself out of the substratum of society. In a similar way, the circumstance that the Catholic Church in the Middle Ages formed its hierarchy out of the best brains in the land, regardless of their estate, birth or fortune, was one of the principal means of consolidating ecclesiastical rule and suppressing the laity. The more a ruling class is able to assimilate the foremost minds of a ruled class, the more stable and dangerous becomes its rule. (Marx, Capital, vol. 3 [International Publishers], p. 600–601 [Penguin Edition, pp. 735–36].)—Ed.]
21 It is, indeed, a point I made myself: see KMTH, pp. 267ff.
22 See, further, my “Functional Explanation.”
23 Elster, “Marxism, Functionalism, and Game Theory: The Case for Methodological Individualism,” p. 472.
24 Ibid., p. 474.
25 Ibid., p. 472.
26 Ibid., p. 476 (emphasis added).
27 See “Exploring Exploitation,” p. 12, where he concludes that “in modern capitalist economies the notion of exploitation should be linked to the lack of power over investment decisions rather than to the fact (or to the possibility) of capitalists having a high level of consumption at the expense of workers.”