[Elster, Jon (1986), Reply to Comments on Making Sense of Marx , Inquiry 29 (1):65-77]

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

VI. Reply to Comments

Jon Elster

[start of page 65]

The main theme in most of the contributions to the symposium on Making Sense of Marx is methodological individualism. In the first part of my reply I consider the objections raised to this, in my opinion, trivially true doctrine. Against Taylor I argue that social relations, seen in abstraction from their relata, have no causal efficacy. Against Wood I argue that my defence of methodological individualism and my criticism of functional explanation are less closely related than he makes them out to be. Against Slaughter I argue that he holds two inconsistent views on the importance of individual desires and beliefs in social explanation. Against Meikle I argue that his view that entities are 'real natures' with a normal path of development needs to be restated in terms of dynamically stable processes. In the second part of the reply I deal with the individual contributions one by one. The replies to the 'fundamentalist Marxists' Slaughter and Meikle are relatively brief, because of the dismissive, unscholarly nature of their comments. Similarly I do not have much to say to North and Taylor, whose brief comments do not contain much with which to disagree. I reply at greater length to Wood, conceding the point he makes in the last section of his comment but rejecting his argument concerning functional explanation.

Underlying most of the contributions to this symposium, or so it appears to me, is a feeling of puzzlement. Why would anyone want to write a big book to sort out the chaff from the wheat in Marx when, at the end, only a few straws of nourishing grain remain? Appropriate (or at least natural) reactions might include:
- with friends like this, who needs enemies?;
- a mountain of critical exegesis has given birth to a mouse-sized insight;
- if the author thinks of himself as subscribing to some kind of Marxism, it can only be 'sentimental Marxism' (Michael Walzer);
- if the goal of the book is to formulate substantial propositions about society Marxian exegesis is a singularly ill-chosen vehicle; if the goal is exegetical, references to later and current discussions seem out of place.

[end of page 65, start of page 66]

I sympathize with these reactions, since I lived through them all in the process of writing Making Sense of Marx. I persisted, nevertheless, because I had a strong feeling that the Marxist wood remains even when every single tree has been chopped down. The book, inevitably, is largely about details; and on almost every point of detail Marx can be shown to have been inaccurate, incoherent, careless or misguided. Yet, for reasons well stated by North in his comment, Marx's contribution should be seen in a broader perspective. In his historical and political writings he laid down the terms of problems that later writers have found increasingly important, and provided rough examples of what a possible solution might look like. My attempt in the concluding chapter to convey a feeling for the wood may have been too feeble to offset the impression left after the relentless chopping down of trees in the preceding chapters.

In particular, I believed and believe that Marx's normative views were sufficiently important to justify the effort. Here I may also have made a mistake, in not stating more simply and plainly my basic assumptions. I believe that in the real world workers are by and large exploited by capitalists and that this is unjust, although one can construct examples, with some real-world relevance, in which workers are either unexploited or, if exploited, are not treated unjustly. I believe that the good life is one of self-realization through work, and especially self-realization jointly with other people, although facts about industrial society and human psychology make it difficult to implement this ideal fully.1 In the book, what follows 'although' in each of the two preceding sente'ices may have acquired undue importance and distracted attention from what precedes the qualifying clauses.

The contributors have of course more to offer than their feeling of puzzlement (if I have identified it correctly). They have caught me out in some mistakes, for which I am grateful. They have forced me to reconsider some of the arguments in which I believed deeply, and made me see that matters are less simple than I thought. In other cases I remain unconvinced. Some of the objections offered I believe to be plainly wrong; others rest on an insufficiently close reading of my book. Since the central point in all but one (North) of the comments concerns my views on methodological individualism, I begin by discussing that issue. I then go on to deal with the contributions separately.

I. Methodological Individualism
I am, frankly, disappointed that this issue should have such a central place in the contributions. I regard methodological individualism as trivially true,

[end of page 66, start of page 67]

worth stating only because triviality notwithstanding it was regularly violated by Marx. Clearly the symposiasts think otherwise. In discussing their objections I begin with the less radical (Taylor) and end with the most extreme (Meikle).

Taylor's main point is that my view does not allow structures as causes. By a structure he means a set of relations, e.g. relations of production defined in abstraction from the specific relata. In his opinion structures thus defined, have causal efficacy. I disagree. I think beliefs (and more generally, attitudes) about structures, thus defined, have causal efficacy but as I say in the book, methodological individualism does not hold within intensional contexts. In extensional contexts, what has causal efficacy is a relation with its relata or as I put it individuals together with their relational properties. Consider the causal efficacy of relations of production in, say, a capitalist society. Any given individual is affected by these relations in two ways. On the one hand his belief that capitalist relations obtain in his society enters among the determinants of his motivations emotions, and ultimately actions. On the other hand his behaviour is heavily constrained by his own relations to property and to other property owning individuals with whom he interacts. My question to Taylor then is. how can relations of production have causal efficacy over and above what they have in these two ways? It appears as if he might want to say: by affecting the beliefs that the individual entertains about the relations. To this I have a predictably trivial reply: individual beliefs are shaped by the experiences of the individual, which in this case include (i) the property relations in which he stands personally and (ii) those he learns about indirectly, from other individuals whose beliefs have been formed in either of these two ways.

Wood addresses the issue of methodological individualism somewhat obliquely, by arguing that in my book it 'seems to serve as a (misleading) name for certain objections he has to the use of functional explanations in social theory'. I shall attempt to reply to his criticism of my views on functional explanation later. Here I first want to record my surprise that he chooses to ignore my explicit statement (M, pp.4-5) that 'there is no logical connection [between methodological individualism and functional explanation], since the collectivist methodology may also be wedded to a causal mode of explanation. Conversely, functional explanation may be compatible with methodological individualism if one insists on the nec essary existence of some underlying mechanism. In chapter 6.2.2 of my book, for instance, I discuss in some detail 'macro-macro' causality in the explanation of collective action Conversely biology rests on both methodological individualism and functional explanation. Against Wood who states that methodological individualism as a principle of biological science 'makes no sense', I believe that one cannot overstate the importance

[end of page 67, start of page 68]

in modern biology of the principle that the structure and behaviour of or~anisms must he explained in terms of their fitness-enhancing effect for the individual organism rather than for the group.

I had problems understanding exactly what Slaughter found objectionable in my espousal of methodological individualism. Let me focus therefore on one specific issue. the role of desires and beliefs in explaining aggregate economic phenomena such as an economic crisis. Slaughter here seems to argue for two conclusions which, if I understand him correctly, cannot both be true. First. he says that '[Marx] does not concern himself with such entities', and then that Marx, unlike Keynes, has as a startingpoint that desires and beliefs 'are changeable and, furthermore, that they come into being and change not arbitrarily but as a reflection of changes in the life of society ...' (p. 47). I think (but I am not quite certain) that both of these statements are meant to constitute objections to my reading of Marx. However, I subscribe entirely to the second statement, as will he clear from chapter 1.3.1 of my book. I obviously disagree with the first, as must Slaughter if he wants to stick to the second. I also agree completely with the view that Marxist analysis 'shows that there are mechanisms of historical change of society which operate without first passing through the consciousness of any of the individuals involved ... - "behind their backs", so to speak' (p. 47). Indeed, since chapters 1.3.2 and 1.5.3 of my book are largely devoted to a discussion of such mechanisms. I cannot see how Slaughter can use their existence as an argument against my interpretation of Marx.

Meikle's attack against the doctrine of methodological individualism is based on a defence of essentialism, defined as 'a metaphysics of whole entities with real natures . . . [which] develop along lines of potential that are part of their nature, unless their development is frustrated or terminated by some internal or external condition' (p.30). I would like to propose a different characterization - entirely consistent with methodological individualism - of the kind of processes Meikle has in mind. I consider the transformation of, say, an embryo into a mature human being as a dynamically stable process of interaction among lower-level entities. Such processes have the property of homeorhesis (a dynamic extension of homeostasis) which can be illustrated by a ball rolling in a valley sloping towards the sea. If the ball is diverted from its course on the floor of the valley and set rolling up the hillside, then sooner or later it will return to the floor and approach the course it would have taken in the absence of the diversion - unless the shock was so strong as to send it over into the adjoining valley and make it pursue a different course altogether. This concept of dynamic stability seems indispensable for understanding how an entity can have a normal development' and under what conditions it can be interrupted. If we do not have some idea of the inner workings of the entity, the idea of

[end of page 68, start of page 69]

a 'normal' development and the conditions under which deviations occur are left dangling in the air.

In addition to this conceptual point I want to make a substantial claim: I do not believe societies are dynamically (or statically, for that matter) stable. Stability can arise by accident, by intentional design, or by selection and evolution. Of these mechanisms, the first can never be excluded, but is unlikely on a priori grounds. The second is inappropriate since societies are much too complex to allow planning for stability. The third rests on a superficial analogy since the notions of reproduction and fitness of societies are not well defined. 2

I now turn to the individual contributions. Some are generally positive, in others critical remarks dominate. Some are in sympathy with my analytic framework, while others are squarely within a different tradition ('fundamentalist Marxism'). I devote most space to Wood's contribution, since it is both critical and sympathetic. To the others I have less to say - either because there are fewer objections to respond to, or because communication appears less feasible.

II. Meikle
Meikle. with Slaughter, represents fundamentalist Marxism in this symposium. Since the criticism of this strand in Marxist thought was a tertiary task of my book - after the interpretation and assessment of Marx's own writings - it is not surprising that they do not like it. Expressions of dislike are not, however, a substitute for argument. Thus I object strongly to Meikle's unscholarly method of discussion. Let me give an example. Meikle cites a passage from page 513 of my book ('If [Marx's theory of revolution] fails to persuade us, it is no doubt because he himself was so persuaded of the necessity of communism that he did not feel an argument was needed' [p.33]), and then goes on to discuss it as if it were all I had written on the topic. The cited sentence is, however, taken from a half-page preview of the 20-page concluding chapter. The sentence itself is a summary of a four-page discussion, which in turn is a résumé of a 13-page treatment in an earlier chapter. These analyses contain a number of arguments which Meikle could have taken up, instead of picking on an isolated sentence. Unfortunately, much of his comment has this character. Here is another example. Meikle and I disagree on the possibility of creating abundance in communism. I believe this is 'cloud-cuckoo land' - and Meikle adds that 'no argument is offered' (p.34). Since here he only cites from the concluding chapter, I must conclude that he did not read the argument I offer on page 231 of my book.

Meikle asserts that my treatment of Marxian economics is shallow and uninformed. but does not substantiate this view. If he disagrees with my

[end of page 69, start of page 70]

(admittedly) brief comment on Morishima's treatment of the iteration problem, why doesn't he try to argue for the exegetical or substantive significance of the iterative procedure? If he wants to invoke Morishima's highly mathematical treatment as an authority, why the heavy sarcasm about arithmetic and matrices'? If he disagrees with my treatment of the 'qualitative' transformation problem, why doesn't he tell us what is wrong with it instead of just citing the names of Aristotle, Böhm-Bawerk, and Jevons? Here as elsewhere in Meikie's contribution the ratio of argument to invective is close to zero,

These practices discharge me from any obligation to respond to all of Meikle's other comments. I must draw attention, however, to his view that my interpretation of historical materialism is 'an anachronistic howler' (p. 35), since it imputes to Marx 'a theory that "explains" history as a sequence of modes of production, each with its characteristic forces of production and "corresponding" relations of production' (p. 35). In his view, this interpretation was canonized by Stalin in 1938, and has as little to do with Marx as the views of the medieval and early modern Catholic Church have to do with Aristotle. This is a very strange position. For one thing, it represents itself an anachronistic howler, since the interpretation he questions was developed by Plekhanov and others long before 1938. For another, it is hard to see how it can be sustained in the face of the many statements by Marx which affirm as explicitly as one could wish the theory he is denied to have held. Moreover, since he believes that the notion of correspondence' is vague and indefinite, why doesn't he discuss my fairly extensive attempt (M, pp.258-67) to render it somewhat more precise and definite?

III. Slaughter
Besides the comments on methodological individualism, Slaughter's contribution centres on problems related to class theory and the labour theory of value. I begin with responding to three of his objections concerning class theory.

(i) He makes fun of my definition of class - 'a group of people who by virtue of what they possess are compelled to engage in the same activities if they want to make the best use of their endowments' (M, p.331) - by asking rhetorical questions such as 'Are burglars a class? Are procurers a class?' (p.54). In the sentence immediately preceding the cited definition, however, I explicitly list the activities which can enter into the constitution of a class: 'working vs. not working, selling vs. buying labour-power, lending vs. borrowing capital, renting vs. hiring land, giving vs. receiving commands in the management of corporate property. These enumerations

[end of page 70, start of page 71]

are intended as exhaustive.' Surely this ought to have told Slaughter that burglars and procurers do not on my view form separate classes. (ii) His sarcastic remarks on the notion of 'non-tangible endowments' (p.55) in my definition of class also reflect a superficial reading. It takes no account of the structure of my argument, which goes as follows. On grounds of consistency with Marx's texts and conformity with his theoretical intuitions, managers must be admitted as a class. Since managers typically do not move into that class by virtue of tangible endowments (property in the means of production), we must consider whether intangible endowments (skills, cultural capital, and the like) can be the foundation for their class position. Correct or incorrect, this is an argument that cannot be demolished simply by saying that 'there is nothing "non-tangible" about' (p.55) Marx's specifications of property and labour in the historical modes of production. (iii) Slaughter seems to think that l disagree with the view that 'Marx's theory of class . . . fisi inseparable from the theory of class struggle' (p. 54). I do say, however (M, p.319), that '[i]n deciding between various interpretations [of class], I try to choose the one that is most plausible in the light of Marx's theory of the class struggle' - a principle which is then implemented at several occasions (M, pp.323, 324, 328, and 329).

Slaughter claims that 'Marx did not argue that commodity exchange arises on the basis of a common factor shared by all commodities, but that in the process of exchange it is labour which regulates the terms of this exchange' (p.52). His fundamentalist companion, Meikle, holds a different view: 'Marx's profoundest criticism of Ricardo is that he is concerned only with the magnitude of value, and not at all with explaining the fact that only a certain sort of labour produces values' (p.36). If I have understood them correctly, Slaughter believes that the labour theory of value is purely quantitative, and Meikle that it is both quantitative and (above all) qualitative. On this point, I side with Meikle. The locus classicus for the qualitative interpretation is a passage in Capital I (cited on page 139 of my book) in which Marx says that if two commodities exchange in certain proportions, then, 'whatever those proportions may be', they must be reducible to 'a third, which in itself is neither the one nor the other'. It appears to me that the phrase which I have italicized shows that Marx was not only concerned with deducing the quantitative terms of the exchange, but also with understanding its very possibility.

IV. Wood
With Wood's contribution we move from the mists of fundamentalism into the daylight of scholarly discourse. Let me begin with section VII of his contribution, concerning Marx's views on justice. Wood is well known as

[end of page 71, start of page 72]

the best exponent of the view that Marx did not subscribe to anv non-relativistic theory of justice. In my book I agree that there is much to be said for this view, but I then go on to argue that there are also elements in Marx's thought that point in a different direction. The most important textual argunients for this alternative reading are (i) Marx's frequent reference tc capitalist exploitation as robbery. theft, embezzlement, etc.. (ii) the self-defeating critique of rights in Critique of the Gotha Programme. and (iii) the Passage from the 1861-63 manuscripts discussed by Wood in his contribution above.

Of these, Wood discusses the concept of robbery on pages 137-8 of his Karl Marx, also referred to in his contribution here. I do not think. however, he has answered the objection G. A. Cohen made in his review of Wood's book (Mind XCII [1983], pp.440 ff.): 'Now since . . . Marx did not think that by capitalist criteria the capitalist steals, and since he did think he steals, he must have meant that he steals in some appropriately non-relativist sense. And since to steal is, in general, wrongly to take what rightly belongs to another, to steal is to commit an injustice . . .' (op. cit.. p 443),3 (Could it be that Cohen has 'come out on top'?) Wood does not discuss the problem posed by the Critique of the Gotha Programme: that Marx here dismisses the principle 'To each according to his contribution' because of the 'defects' (Misstände) to which it gives rise, such as that of allocating income to workers independently of the size of their family, without noticing that he is thereby committed to some superior principle of justice (most plausibly equality of welfare) in the light of which the contribution principle can be seen as defective.

Wood does discuss the third textual support for the view that Marx had a non-relativistic conception of justice. and claims that it is based on a misreading of the text. I agree with him. I made a mistake in not noticing that 'ein Unrecht' is not within the scope of 'Erkennung' in the cited text, but is rather governed by 'Beurteilung', which, as Wood points out, is not similarly a 'success-word'. Although I could make two or three moves to weaken the strength of his objection, they would at most make the cited text more compatible with the non-relativistic conception, assuming that it could be defended on other grounds. They could not turn it into an argument for that view, as I wrongly thought it to be. Yet I believe that the interpretation can be defended on other grounds. viz. arguments (i) and (ii) above.

I now turn to Wood's discussion of functional explanation. Section II embodies the unfounded charge, discussed above, that I have confused the issue of methodological individualism with that of functional explanation. In section III Wood briefly sets out the main explanatory claims of historical materialism, asserting that the relations of production 'tend to adapt to the growth of productive powers, facilitating both the efficient employment of

[end of page 72, start of page 73]

the existing [productive] powers and their further growth' (p.14). A central argument in the relevant chapter of my book is that one cannot expect the relations of production to enhance both the employment and the growth of the productive forces, just as in general one cannot expect two different functions of one variable to be maximized by the same value of the latter. Wood chooses not to address this problem.

In section IV Wood goes on to argue that in any functional explanation, the explanans is a disposition or a tendency to produce the explanandum, and that any explanation which does not have this form cannot be a functional explanation, not even a crude one. Let me first set aside a verbal issue. I can't see that it matters whether we call attempts to 'explain a piece of behaviour merely by citing its beneficial consequences for someone or something' (p.16) a crude functional explanation or no explanation at all. I use the term 'explanation' at the level of explanatory intent: when someone offers an argument that purports to explain something, I call it an explanation even if it falls hopelessly short of my criteria for a good explanation. Hence my definition of a functional explanation is simply what all such explanatory attempts - good and bad ones - have in common: the citing of beneficial consequences as the explanans.

Good functional explanations must, in my view, have certain additional features: a demonstrated feedback mechanism or a lawlike connection between the tendency of the explanandum to produce the explanans and the occurrence of the explanandum. (Wood would say that the tendency to produce what I call the explanans is the explanans, but, to repeat, this is a merely verbal issue.) I note with some interest that in Wood's book he agrees that 'it may be reasonable to ask for a causal explanation of the existence and workings of the persistent tendency which grounds a teleological explanation', while conversely we 'are seldom inclined to ask for teleological explanations for causal laws'.4 I would strengthen both of these observations. Any scientist worth his salt would regard the causal explanation of the tendency as the ultimate goal. Thus I disagree with Wood's statement5 that 'Darwin's theory shed light on the teleological explanations biologists had been using for centuries, but did little to legitimate and nothing to discredit - these explanations'. It certainly did discredit earlier explanations in terms of species-benefits. as distinct from benefits to individual organisms; more subtly, it discredited any explanation based on 'ecological benefit' (measured, say, in life-span), as distinct from reproductive benefit (measured in number of reproducing offspring). Conversely, I would argue that scientists never ask for teleological explanations of causal laws.

In section V Wood traces the development of my 'murky. indecisive, confused, and inconsistent' (p.19) views on functional explanation, and concludes that 'Elster seems still to be seeking for a way of pressing

[end of page 73, start of page 75]

philosophical objections to functional explanation, but has yet to find a way to do so successfully' (pp.19-20). This is not an inaccurate description. My intuition is that functional explanation (when not backed by demonstration of feedback from explanans to explanandum) is inferior to causal explanation. Although I cannot ask anyone to take that intuition seriously until it has been successfully turned into an argument, I do not think it is mere prejudice either. It is based, among other things, on the following convictions. (i) Everything can in principle be explained causally, but not everything can in principle be explained functionally. (ii) Every functional explanation contains a causal subargument (to the effect that the explanandum causes the explanans), but not every causal explanation contains a functional subargument. (iii) While a functional explanation can be improved by producing another functional explanation,6 'the limiting case of an improved functional explanation is a causal explanation'.7 (iv) Functional explanations invite arbitrary practices, by manipulation of the time perspective or of the interests which are being served. (v) In actual practice, scientists seek for causal explanations or for functional explanations backed by (causal) feedback mechanisms; they never use functional explanations backed by a lawlike connection between the tendency of the explanandum to produce the explanans and the occurrence of the explanandum. For me, the cumulative impact of these and other convictions is so strong that I cannot take Wood's advice and 'give up this futile quest' (p.20) for a refutation of functional explanation. But I agree that I have yet to get my ideas fully into focus. I should add, however, that these self-confessed doubts do not impinge upon anything I say in Making Sense of Marx, where I maintain a deliberately cautious stance on these issues.

Wood largely accepts my criticism of 'particular Marxian attempts at functional explanation for their empirical failings' (p.20). with the exception of Marx's philosophy of history where, he argues, teleology is being used with a narrative rather than an explanatory purpose. He attempts to come to grips with some of the texts I cite in favour of my view, but most of them (enumerated in the index of my book under 'teleological conception of history') he ignores. Let me briefly summarize what I believe to be the most important of these textual arguments. (i) Given Marx's penchant (accepted by Wood) for unsupported functional explanation in other contexts, it is not 'gratuitous to ascribe explanatory intent to' (p. 20) his teleological statements about history. (ii) Wood agrees that various statements from the Paris manuscripts embody teleological explanations of history, but he does not discuss the numerous texts from the mature economic manuscripts (cited on pages 13-15 of my book) which are very close in spirit to the early texts. (iii) He does not discuss my documentation (M, pp.430-2) of Marx's teleological attitude to the bourgeois revolutions. nor my argument (M, p.437) that Marx's own policies in the German

[end of page 74, start of page 75]

Revolution of 1848 were shaped by this outlook. (iv) Wood does not discuss my argument (M, pp. 367-71) that Marx's views on working-class consciousness and trade unions have strong overtones of teleological explanation. (v) Nor does he discuss my argument (M, p.124) that Marx's theory of the falling rate of profit, with the unsubstantiated assertion that the main tendency will dominate the various countertendencies, only makes sense if we impute to him the a priori assumption that capitalism is doomed. (vi) Finally, he does not discuss my criticism (ch. 5) of the view that the productive forces call into being relations of production optimally suited for their further development. I believe that my case rests on more solid foundations than what the reader can glean from Wood's account of it.

V. North
There is not much in North's comment I can respond to. At several points he expresses his dissatisfaction with the generality and vagueness of my analysis, arguing for the superiority of the transaction-cost approach to property rights. Now it is certainly true that I do not make transaction costs into the cornerstone of my analysis, although I do refer to them at various places (M, pp.192-3, 198-9). I suspect that North may be right that my treatment would have been improved had I taken more account of the transaction-costs revolution. I think, however, that North matches the vagueness and generality he imputes to me when he attempts to explain the relevance of transaction costs to my analysis. Let me try, nevertheless, to respond to some of his challenges.

One puzzle must be stated at the outset. In my book I cite and praise North's Structure and Change in Economic History for pointing the way to an explanation of property rights which, although inconsistent with Marx's own theory of property rights, is both more intrinsically plausible and more consistent with Marx's theory of the class struggle. This is the view that the property rights that obtain in a society tend to be those which maximize the surplus accruing to the rulers. These rights may often be (statically or dynamically) inefficient, but then why should we expect efficiency if those with the power to implement it have nothing to gain thereby? In the present contribution North seems to be much more concerned with the conditions for efficient property rights. This concern seems to have led him away from predictive definiteness and towards the truism that a potential for reduced transaction costs will be realized only if property rights change so as to allow entrepreneurs to cash in on the gain, which in turn depends on what happens at the level of the state. I am not charging North with a contradiction, but I want to record my puzzlement.

North argues that my interpretation of Marx in terms of 'development-

[end of page 75, start of page 76]

fettering' (p. 59) is remarkably consistent with Mancur Olson's recent theory. I disagree entirely. On Olson's view. what makes a property-rights regime increasingly ill-suited to the deyelopment of the productive forces is internal institutional change, such as the rise of organized interest groups and monopolistic elements. On Marx's view, the cause of the dynamic inefficiency is the development of the productive forces brought about by the regime itself. A property-rights system which is optimal for the rate of change of the productive forces at one level of their development may cease to be optimal at a higher level. On Olson's theory. the dynamic efficiency of, say, capitalism could be restored simply by having a revolution that cleared the ground by restoring the original competitive form of the svstem. This would be unthinkable on Marx's theory.8

North has me implying that 'capitalist society, when compared with communist or socialist societies. will produce suboptimal development' (p. 62) of [presumably] the productive forces. I don't know where he got that idea, which is never stated in my book and in which I do not believe. I think socialism could be superior to capitalism on several counts, such as distributive justice, economic democracy, and the opportunity for self-realization in work, possibly even with respect to static efficiency, but I do not think it will be superior on grounds of dynamic efficiency. Indeed, what little understanding we have of the conditions for technical change would lead us to expect that the spur to innovation would be reduced with the scope of the entrepreneurial role. An important, open question is whether the trade-off between dynamic efficiency and the other values is such that, all things considered, socialism would be preferable.

VI. Taylor
Besides his comments on methodological individualism, Taylor's observations concern three topics: the peasantry, international relations, and values. On the first point, I think he misses the distinction between intracommunity and inter-community solidarity which underlies much Marxist writing on the topic. The reason why peasant rebellions have been endemic in history is solidarity among village members; the reason why they have rarely given rise to large-scale movements is the isolation of villages from one another. I agree, however, that Marx also and wrongly denied intracommunity solidarity, and that I culpably may have left the impression that I agreed with him. I also accept Taylor's argument that I erred in thinking that motivation by informal social sanction falls outside selfishly rational behaviour.

Taylor thinks that Marx and I neglect the role of international relations in the domestic development of societies. In my book. however, I document

[end of page 76, start of page 77]

and discuss two such theories in Marx: the view that the successful communist revolution will involve a diffusion of revolution from the industrially backward East to the advanced Western nations, and the view that it will come about by the diffusion of technology in the opposite direction. I agree with 'I'aylor. however, in that neither theory is as central in his writings as the endogenous explanation traditionally associated with his doctrine. In the book I alsso suggest that interaction between advanced and backward countries will, if anything, make a successful revolution less likely than if they had developed in isolation from each other. Whether correct or not, this observation seems to be the kind of argument Taylor claims to be missing from my account.

A final remark about values: I agree with Taylor that Marx placed an unfortunate emphasis on mankind as a whole, with a concomitant disregard of local communities, and I say as much in the final paragraph of chapter 6.


1 In Making Sense of Marx this view is merely sketched in a way that probably does not carry much convincing power. I have tried to offer more of an argument in my 'Self realization in Work and Politics'. Social Philosophy & Policy 3 (1986) forthcoming.

2 For a further discussion, see my Explaining Technical Change (Cambridge Cambridge University Press. 1983). pp 61-64.

3 In his Karl Marx (London Routledge & Kegan Paul 1981) pp. 137-8, Wood argues that Marx's analogy between military plunder and capitalist exploitation helps us understand why the latter is not unjust (in a non relativistic sense) 'For Marx the relation between plunderers or conquerors and their victims or tributaries is not something economically accidental, but must constitute a regular production relation as determined by the stage of development of the victims' productive powers. Hence there is good reason to think that the transactions (ranging from military incursions to tax collection) between plunderers and plundered correspond to the prevailing mode of production, and are just according to Marx's concept of justice.' I fail to see the relevance of this observation, which shows only (or at most) that plunder is not unjust in a relativistic sense.

4 A. Wood. Karl Marx, op. cit.. pp. 107 and 252.

5 Ibid., p. 105.

6 Hence Wood is wrong in interpreting my statments about nuts and bolts as restricted exclusively to causal explanations For an example of how one can open the black box of a functional explanation, while still remaining within the functionalist framework see my 'Further Thoughts on Marxism Functionalism and Game Theory', in B Chavance (ed.), Marx en Perspective (Paris: Éditions de I'École des Hautes Études 1985), pp. 629 ff.

7 Ibid., p.630.

8 See also my review of Olson, 'The Contradictions of Modern Societies', Government and Opposition 19 (1984), in which I present Olson as a latter day Veblen - and North as a latter-day Marx!

Received 28 February 1986

Jon Elster, Department of Political Science, University of Chicago, Chicago, Illinois 60637, U.S.A.

[end of page 77]

[Elster, Jon (1986), Reply to Comments on Making Sense of Marx , Inquiry 29 (1):65-77]

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