[Levine, Andrew (1986), Review of J. Elster (1985): Making Sense of Marx, The Journal of Philosophy, p. 721-728]

[ Books | Articles | Reviews | Index | The main menu ]

Review of Making Sense of Marx

[start of page 721]

Making Sense of Marx. JON ELSTER. New York: Cambridge University Press; Paris: Editions de la Maison des Sciences de l'Homme, 1985. xv, 556 p. Cloth $49.50, paper $15.95.

The principle objective of this book is, its title announces, to make sense of Marx - that is, of Marx's substantive, theoretical positions. It is accordingly encyclopedic in scope: addressing, as Marx himself addressed, many central questions of philosophical anthropology, economics, social philosophy, the philosophy of history, sociology, political theory, and the analysis of ideology (where elements of all these disciplines and social psychology intersect). A second objective is to defend "methodological individualism" as the proper explanatory strategy for the social sciences. In Elster's view, methodological individualism is indispensable too for making sense of Marx. Marx, of course, seldom advanced individualist explanations. But, if Elster is right, all that is worth taking seriously in Marx's thought can he reconstructed in methodological-individualist fashion. Doing so, Elster thinks, saves the "rational kernel" (as Marx might have put it) of Marx's thought from the indefensibility that threatens many of Marx's own formulations and from the obscurantism that afflicts so much of what has come to be identified as Marxism. A third objective of this book is to show how that rational kernel, suitably developed and elaborated, is of continuing importance for understanding the range of topics Marx addressed, and still provides our best purchase on many of these issues.

Like Elster's earlier books, Making Sense of Marx displays an astonishing erudition and mastery of technical material in a variety of fields, joined with the analytical philosopher's gift for supplying distinctions and structure - even in domains philosophers seldom venture.1 If all thinkers are, as Isaiah Berlin maintained,2 hedgehogs or foxes, then Brian Barry was right, some years ago, to describe Elster as "the prince of foxes."3 It would require many more pages than I am allotted just to describe some significant fraction of Elster's re-

[end of page 721, start of page 722]

construction and assessment of Marx. Anyone interested, even remotely, in any of the many issues discussed is well advised to turn to the text itself. The effort is bound to be rewarding.

It is feasible, however, briefly to characterize this book's general tenor and place in recent writing on Marx. Over the past two decades, a number of economists (Elster lists Nobuo Okishio, Michio Morishima, Christian von Weizäcker, Paul Samuelson, Ian Steedman, and John Roemer) have successfully elaborated the analytical foundations of Marxian economic theory; and a host of philosophers have joined Marxian theoretical concerns with more traditional treatments of central issues in social philosophy, philosophy of history, and the philosophy of social science. Elster's book stands very much on the shoulders of these analytical reconstructions. To a degree, Making Sense of Marx is a synthesis of the state of the art circa 1985. However, Elster passes on very little without significant criticism. Particularly in the chapters on economics and political theory, Elster's own contribution is evident primarily in the insights he provides on the work of others. Much the same is true in more strictly philosophical contexts where two seminal texts - G. A. Cohen's Karl Marx's Theory of History: A Defense 4 and Roemer's A General Theory of Exploitation and Class5 - dominate his concerns. In "analytical Marxism's" still underdeveloped areas - for instance, in discussion of class alliances, revolutionary motivations, and ideology - Elster's contribution is more positive and substantial. These topics are, moreover, among those where traditional Marxian theorizing appears to have deviated most from the methodological individualism Elster endorses.

In promoting methodological individualism throughout the range of topics discussed, the prince of foxes shows, perhaps for the first time, intimations of joining the hedgehogs.6 But if as a fox Elster is nothing less than spectacular, scattering insights at breakneck pace over a territory so broad that "following up his references would be a liberal education in itself,"7 his reiterated, but always implicit, de-

[end of page 722, start of page 723]

fense of methodological individualism is less successful. To mount an adequate challenge to Elster's stance on methodological individualism, it would be necessary to do to Making Sense of Marx - or some important fragment of it - what Elster does to Marx: to tease out his substantive position from comments made largely in passing, reconstruct it, and submit it to critical assessment. Since space considerations prohibit attempting that here, I shall instead venture some general comments on Elster's stance in what is, after all, an old debate.

For philosophers, if not social scientists,8 the debate between methodological individualists and holists has passed into desuetude once it became evident that the rival positions could not be formulated in a way that would make for a genuine and interesting disagreement, at the level of generality at which the issue traditionally had been posed.9 Particular explanations can sometimes be classified, informatively, as individualist or holist.10 But it is unclear whether in general there is anything about which individualists and holists disagree. Ostensibly, holists regard social wholes as somehow more than the sum of their parts, and individualists do not. But, on closer examination, plausible accounts of how wholes are more - or no more - than their parts seem, finally, to converge - rendering the contrast illusory. What must be asked of Elster, therefore, is whether his version of methodological individualism can he formulated so that it has genuine, plausible rivals - and whether, if so, he can provide good reasons for defending this individualism against those rivals. It will help in getting to the heart of the issue, to consider briefly how the second question might be addressed.

What reasons might there be in general for favoring one explanatory strategy or another? One possibility is that we take our lead from science itself, generalizing on the basis of successful explanatory practices. Thus one might base support for methodological individualism upon explanatory strategies in economics or other social

[end of page 723, start of page 724]

sciences which explain successfully in paradigmatically individualist fashion. But, however matters may look for philosophers of physics, it seems overly credulous, at best, for philosophers of social science to draw firm conclusions on the basis of actual explanatory practices. It therefore seems more promising to defend methodological individualism by supplying expressly philosophical reasons. This is what Elster does; and I will go on to suggest what I think his reasons are. 11 These reasons do not, however, sustain all that Elster implies they do.

Societies are collections of individuals, just as individuals are collections of cells; and social phenomena are effects of individuals' actions in much the way that individuals' actions are effects of the behaviors of the cells that compose individuals. It is clear, however, at least for individuals' actions, that ontological reducibility (decomposability without remainder) does not entail explanatory reducibility. The best explanation for individuals' actions need not make essential reference to behavior at the cellular level. If, for instance, reinforcement schedules affect the rate of performance of some behavior, these reinforcement schedules will be represented in specifiable changes at the cellular level - they will be describable neurophysiologically - but the best explanation for changes in behavior frequencies is almost certainly not at the cellular level. Reinforcement schedules have neurophysiological "microfoundations." But unless current views are widely off target, the microfoundations, in this instance, are irrelevant for psychology's explanatory aims. Similar considerations apply to social phenomena. Individuals compose societies. But it does not therefore follow that the best explanation of social phenomena need appeal to the behaviors of individuals. In all likelihood, supra-individual, relational properties - population density, kinship relations, social norms, and so on - will sometimes be explanatory. But these properties are not properties of individuals, except in the irrelevant sense that societies are decomposable ontologically into individuals. Very generally, with social phenomena, as with individual behavior, what is explanatory cannot be specified a priori. 12

[end of page 724, start of page 725]

What does seem fair to require of any purported explanation, though, is that it be susceptible to representation at a microfoundational level. If, for instance, reinforcement schedules could not even in principle be reconstructed microfoundationally, then we would be justified - because we believe that ontologically wholes are just the sum of their parts - to view the purported explanans in the way we might an occult cause; as unconnected to its explanandum by any plausible, causal mechanism. A lack of microfoundations is therefore undesirable; and it is fair, in thinking philosophically about social-scientific explanations, to proscribe purported explanations that dispense in principle with microfoundations. It is not clear from what Elster writes whether he intends more than this; whether he wants to insist, in addition, that the microfoundational account actually be known 13 or even, as he appears sometimes to suggest, that the best explanations, at least in principle, will always be microfoundational. The former claim states a desideratum of a good explanation, but hardly a necessary condition. The latter claim is almost certainly wrong. If Elster intends, as he should, only that in principle there be a microfoundational account - but not that it be the best explanation or even that it be known - it is unclear whether very much social-scientific explanatory practice actually controverts his strictures.

However, some Marxian explanations do - insofar as they appeal to the unfolding of an immanent teleology or to extracausal functional interrelations.14 It is far from clear, though, whether Marx himself could be charged with this sin of his neo-Hegelian followers. Marx admitted to "coquetting with . . . modes of expression peculiar to [Hegel]",15 and he certainly advanced functional explana-

[end of page 725, start of page 726]

tions. On occasion, he was insensitive to the need to supply causal mechanisms. These theoretical indiscretions are documented incisively in this book. But nowhere does Elster show Marx committed to views that in principle deny microfoundational accounts. Methodological individualism is less of a sword to wield against any part of the Marxian corpus, including those parts Elster finds indefensible, than first appears.

The "rational kernel" of Elster's methodological individualism, then, is just the relatively unportentous insistence that, in social-scientific explanations, microfoundational accounts - accounts at the level of individual actors - be available in principle. But what of the nature of these accounts? Here too, it might appear, Elster's methodological individualism implies a substantive position of some interest; but, as before, more is suggested than is actually delivered.

Unlike other Marxian "methodological individualisms," 16 Elster's does not require that individuals be regarded as rational economic agents - bent on maximizing their distributive shares or minimizing their efforts. Individuals do maximize in the accounts Elster provides, but no assumptions are made about the values they aim to realize. Altruism (conditional and unconditional), general benevolence, even self-abasement, are possible values, along with the egoism assumed by economists. It is a weak notion of individual rationality as maximizing behavior, not the much richer concept of individual interest widely supposed throughout economics, that is the basis for Elster's individualist explanations. It is not clear, though, what, if anything, can be concluded from this way of depicting behavior.

Even on the "thinnest" account of rational agency, individuals are frequently irrational. 17 In what sense, then, is individual rationality explanatory? If the claim is just that individuals maximize but are also subject to interferences which for analytical purposes can be abstracted out, there is no particular problem. Rational behavior would then be an idealization of actual behavior and would be explanatory to the extent that actual behavior approximates the ideal.18 I suspect that, throughout most of Making Sense of Marx, Elster implicitly regards rationality this way. But his more direct treatments

[end of page 726, start of page 727]

of the question suggest a different understanding,19 where rationality is a template against which existing behaviors may be compared and categorized. On this understanding, though, the idea of rationality may be helpful in organizing descriptions and focusing reflection, but it would not explain the phenomena it helps to describe.20

Rationality must be explanatory, however, if it is to provide a basis for a microfoundational account. Counterfactual reconstructions of phenomena - achieved by "instantiating extreme values of one or more variables"21 - are not enough. It can be worth while to work with reconstructions, even to the point of advancing substantive positions on the basis of claims that are known to be false.22 Elster's success in deriving Marxian positions from individual maximizing behavior is sufficiently impressive to dispel doubts about the merits of the project. The point is just that much less follows about explanation than Elster leads his readers to suppose.23

Indeed, it may be best, pace Elster, to construe at least some of his claims in behalf of methodological individualism as speculations about the explanations for social phenomena. The claim, then, would be that individuals' (maximizing) choices will always - or frequently - matter explanatorily. This claim actually is controversial. It might turn out, for instance, that, for the phenomena social scientists aim to explain, the opportunity sets that agents confront are so restricted that choosing among alternatives can be effectively discounted. Or, more germane to current social theory, it might be that individuals are so profoundly socialized that individuals' choices, though phenomenologically real, will be of little or no explanatory import. Some of Marx's followers-for example, those who talk of "structures" causing "structures" - would then be in substantive disagreement with Elster's speculation. And so would ethnomethod-

[end of page 727, start of page 728]

ologists, symbolic interactionists, and practitioners of a host of other sociological fashions. In this dispute, Elster may be right, and he likely does have Marx on his side. But this is a question for the progress of science to settle, and not at all an issue in the philosophy of explanation.

It is ironic that the analytical tools Elster employs so deftly arose in the non and even anti-Marxian confines of neoclassical economics and its extensions in public-policy studies, and that these tools have been used, in the main, for ideological and political purposes deeply at odds with Elster's. However, on this count, even orthodox Marxists ought not object. It is a case of turning the enemies' weapons back against the enemy.24 Where orthodox Marxists have reason for disquiet is in the devastation Elster visits on Marxian orthodoxy. Part of the original motivation for joining analytical philosophy (and contemporary economics) with Marxism was to defend Marxian positions. But, as ought to have been expected, clarification has revealed vulnerabilities - prompting development, revision, and in some cases abandonment of traditional Marxian claims. Over the past decades it has become clear that relatively little that Marx wrote can be accepted as is. Still, as Elster's book amply attests, Marxian theory, regarded as a collection of core insights and substantive positions,25 remains a timely and fruitful basis for theoretical work. Making Sense of Marx is therefore not like the "refutations" of Marxism supplied by disillusioned ex-Marxists of earlier generations. Elster remains radical and sympathetic to the Marxian project, and he defends many of its fundamental components. The burden, after Elster, is on more orthodox Marxists to justify how one can be more orthodox than Elster's analyses warrant, as it is on everyone else to justify adherence to views at odds with those Marxian views which, Elster demonstrates, do make sense.

University of Wisconsin /Madison

[end of page 728]

[Collected from their respective pages]

1 Among Elster's earlier writings, see especially Leibniz et la formation de l'esprit capitaliste (Paris: Aubier-Montaigne, 1975), Logic and Society: Contradictions and Possible Worlds (New York: Wiley, 1978), Ulysses and the Sirens: Studies in Rationality and Irrationality (New York: Cambridge, 1979), Explaining Technical Change (New York: Cambridge, 1983), Sour Grapes (New York: Cambridge, 1983).

2 The Hedgehog and the Fox (London: Weidenfeld & Nicolson, 1953).

3 "Superfox," Political Studies, XXVIII, 1 (January 1980): 136-143.

4 Oxford: University Press; Princeton: University Press, 1978.

5 Cambridge, Mass.: Harvard, 1982.

6 Declarations in favor of methodological individualism are already evident in Elster's earlier writings. See, for instance, "Marxism, Functionalism and Game Theory: The Case for Methodological Individualism," Theory and Society, XI (1982): 453-483. But before this book, Elster's writings, considered as a whole, do not, in proper hedgehog fashion, center on any particular theme so much as they cluster around a constellation of topics-roughly on rationality (and irrationality) and their ramifications for understanding human behavior in societal and psychological contexts.

7 Brian Barry, op. cit., p. 137.

8 Now including even Marxian social scientists. For a survey of the revival of this debate from that unlikely perspective, see Scott Lash and John Urry, "The New Marxism of Collective Action: A Critical Analysis," Sociology, XVIII (1984): 33-50.

9 The classical statement of this view is Ernest Nagel's discussion in The Structure of Science (New York: Harcourt, Brace & World, 1961), pp. 535-546. See May Brodbeck, ed., Readings in Philosophy of the Social Sciences (New York: Macmillan, 1968), for a useful collection of articles on the debate in its earlier incarnations.

10 For a perspicuous and generally deflationary discussion of the issue from the perspective both of social science and of evolutionary biology, see Elliott Sober, "Holism, Individualism and the Units of Selection," in P. Asquith and R. Giere, eds., PSA 1980, Proceedings of the Philosophy of Science Association Meetings (East Lansing, Mich., 1981), reprinted in Sober, ed., Conceptual Issues in Evolutionary Biology (Cambridge, Mass.: MIT Press, 1984).

11 Perhaps the wisest strategy for thinking constructively about explanation would be to adopt a procedure roughly analogous to the method of "reflective equilibrium," correcting philosophical considerations against scientific practice and vice versa, according weight to explanatory practice to the degree we are confident in our explanations. Philosophical considerations would therefore count for more in the philosophy of social science than in the philosophy of natural science; but they would not count exclusively.

12 It is sometimes held that, if we knew more, we could dispense with explanations at the level of aggregations of parts, and so, if we knew everything, all explanations would be at the level of the most elementary parts. It should be obvious that this view is false. World War II was, in the sense in question, just an aggregation of subatomic particles in motion. But knowing all there is to know about these subatomic particles would not help us, in all likelihood, in knowing, say, the causes of World War II. It is fair to speculate that physical descriptions of wars are not even in principle the best explanations of these phenomena, even if wars could be entirely described in physical terms.

13 This apears to have been a basis for Elster's well-known challenge to G. A. Cohen's defense of functional explanations in historical materialism; cf. his "Review of G. A. Cohen: Karl Marx's Theory of History: A Defense, " Political Studies, XXVIII, 1 (January 1980): 121-128. See also Cohen, "Functional Explanation: Reply to Elster," op. cit., pp. 129-135. See also chapter 5 of Making Sense of Marx, for more plausible objections to historical materialist explanations.

14 So too would social scientific explanations that treat society as an entity irreducible ontologically to its constituent individuals. Emil Durkheim's "conscience collective," for instance, could be construed this way.

15 See Afterword to the second German edition of Capital, vol. 2 (New York: International Publishers, 1967).

16 Compare Adam Przeworski, "The Challenge of Methodological Individualism to Marxist Analysis" in Pierre Birnbaum and Jean Leca, eds., Sur l'Individualisme (Paris: Presses de la Fondation Nationale des Sciences Politiques, 1985).

17 Elster's most recent book before Making Sense of Marx is devoted precisely to the "subversion of rationality"; see Sour Grapes, op. cit.

18 A textbook parallel in the philosophy of natural sciences is Boyle's ideal-gas laws, which are explanatory because the behavior of actual gases approximates the behavior of ideal gases.

19 See Sour Grapes, op. cit., chapter 1.

20 The two interpretations just sketched parallel difficulties familiar in the interpretation of Weberian "ideal-type" concepts. As before, an old debate in the philosophy of social science is here revived in new guise, but not advanced.

21 See W. G. Runciman, A Critique of Max Weber's Philosophy of Social Science (New York: Cambridge, 1972), p. 33.

22 It has been suggested that this procedure is pervasive - and useful - throughout economics. See Daniel Hausman, Capital, Profits and Prices: An Essay in the Philosopy of Economics (New York: Columbia, 1981). The most successful ventures in Marxian methodological individualism too have relied essentially on similar assumptios; see Roemer, op. cit.

23 The heuristic value of Elster's (minimalist) individualist assumptions and of the stronger (egoistic) assumptions of others itself calls for explanation. Making Sense of Marx provides valuable data for such an investigation, but little help in making sense of it. Similar questions could be raised of some Weberian explanatory agendas.

24 Another striking instance of the same phenomenon is evident in the recent work of G. A. Cohen, where the Lockean (and Nozickian) idea of "self-ownership" is put to pro-socialist purposes. See, "Nozick on Appropriation," New Left Review, 150 (March/April 1985): 89-105.

25 For Elster, as for "analytical Marxists" generally, Marxism is emphatically not a distinctive "method" (as Georg Lukacs, for instance, insisted), but a collection of substantive positions. A tentative conclusion reached by the careful study of Marx undertaken in the past decades by investigators with analytical training is that, despite what is so often proclaimed, there is, finally, nothing distinctive about Marxian methodology - at least insofar as it yields defensible, substantive positions.

[Levine, Andrew (1986), Review of J. Elster (1985): Making Sense of Marx, The Journal of Philosophy, p. 721-728]

[ Books | Articles | Reviews | Index | The main menu ]

1 1