The meaning of this argument is clear. 'Abstract labour' is an abstraction, in the sense that it is a mental generalization of the multiplicity of useful, concrete kinds of labour: it is the general, common element of all these kinds of labour. This generalization, moreover, as Sweezy goes on to point out, corresponds to capitalist reality, in that in this kind of society labour is shifted or diverted according to the direction of capital investments; hence a determinate portion of human labour is, in accordance with variations of demand, at one time supplied in one form, at another time in another form. This proves the secondary importance in this regime of the various specific kinds of labour, as against labour in general or in and for itself. In spite of Sweezy's plea that 'the reduction of all labour to a common denominator . . . is not an arbitrary abstraction, dictated in some way by the whim of the investigator' but 'rather, as Lukács correctly observes, an abstraction "which belongs to the essence of Capitalism",' despite this, in the absence of what seems to me the decisive point, 'abstract labour' remains, in the last analysis, essentially a mental generalization.
The defect of this interpretation of 'abstract labour' lies not only in the fact that -- if abstract labour is a mental generalization -- it is not clear why what this labour is supposed to produce is something real -- value ; but also in the fact that this opens the door to the transformation of value itself into an abstract generality or idea as well. For, in the sense that here only useful and concrete kinds of labour are regarded as real, whereas 'abstract' labour is seen as a merely mental fact, so too only the products of useful kinds of labour or use-values are real, whereas value, the merely general element common to them, is abstract.
The interpretation that Bernstein adopted was precisely this one. 'Value' is ein Gedankenbild, a mere thought-construct: it is in Marx's work a formal principle which serves to bring system and order to the complexity of the analysis, but itself has no real existence. 'Insofar as we take into consideration the individual commodity', Bernstein comments, 'value loses any concrete content and becomes a mere mental construction'. Hence it is clear that 'the moment that labour-value is only valid as a mental formula (gedankliche Formel ) or scientific hypothesis, surplus value also becomes a pure formula, a formula based on a hypothesis'.
This interpretation had, of course, already been advanced before Bernstein by Werner Sombart and Conrad Schmidt, in time for Engels to confront it in his Supplement to Volume III of Capital. Value, according to Sombart, is 'not an empirical, but a mental, a logical fact' while for Schmidt the law of value within the capitalist mode of production is a 'pure, although theoretically necessary fiction'.
It is striking that even at this point, decisive for the genesis of 'revisionism', Engels's response is both uncertain and substantially erroneous. Even if he makes some reservations towards Sombart and Schmidt, he ends up by accepting their essential thesis (that is, the unreal nature of the law of value when commodities are produced under capitalist conditions ), and hence falls back to the position of Smith (already criticized in its time by Marx) which had relegated the action of the law of value to precapitalist historical conditions.
In other words, 'abstract labour' and 'value' -- the point on which everything hangs -- are understood simply as mental generalizations introduced by the scientist, in this case by Marx; ignoring the fact that, if this were effectively so, in introducing these generalizations Marx would have been committing a 'clumsy error' and the whole of Böhm-Bawerk's critique would indeed be correct. The central argument of Böhm-Bawerk's critique -- already present in Geschichte und Kritik der Kapitalzinstheorien (pp. 435ff.) and restated in 1896 in Zum Abschluss des Marxchen Systems (a text which may have influenced Bernstein) -- was that if 'value' is the generalization of 'use-values', it is then use-value 'in general' and not, as Marx had argued, a qualitatively distinct entity. Marx's error, according to Böhm-Bawerk, was the error of those who 'confuse abstraction from the circumstance in general (von einem Umstande überhaupt ), and abstraction from the specific forms in which this circumstance manifests itself'; the error of those who believe that to abstract from the differences between one use-value and another is to abstract from use-values in general ; for the real value is use-value, the true theory of value a theory of value-utility. According to Böhm-Bawerk, this 'wrong idea' he attributes to Marx means that instead of seeing in 'exchange value' a relation or a mere quantitative proportion between use-values, and hence, like any
relation, an unreal value outside the entities related together, Marx invoked the existence behind exchange-value of an objective being 'value', without seeing that this 'entity' was only a 'scholastic-theological' product, a hypostasis arising from his defective logic.
The response that has traditionally been given to these objections by Marxists is well known. It consists, at most, in an appeal to the original conception of Ricardo who had, as can be seen from his last incomplete memoir, already before Marx distinguished between Absolute Value and Exchangeable Value. However, apart from Marx's remarks on the tendency of Ricardo's analysis to dwell more on 'exchange-value' than on 'value' itself, this response is further weakened by the fact that, confronted by the non-coincidence of 'values' and 'costs of production', this interpretation has continuously been forced to fall back on to Sombart-Schmidt positions or even Bernstein positions. For once it is accepted that value is not identified with the concrete exchange-values or competitive prices at which the capitalistically produced commodities are in fact sold, this interpretation retreats to a position of attributing to 'value' the significance, essentially, of an abstraction. Dobb's case is typical After stating that 'value [is] only an abstract approximation to concrete exchange-values', that this 'has generally been held to be fatal to the theory, and was the onus of Böhm-Bawerk's criticism of Marx', he limits himself to concluding that 'all abstractions remain only approximations to reality . . . it is no criticism of a theory of value merely to say that this is so.
E Q U I V A L E N C E A N D S U R P L U S V A L
U E >>
84 For this critique of Smith by Marx, see Theories of Surplus Value, Part I, London, n.d., pp. 71-2.
85 E. Böhm-Bawerk, Zum Abschluss des Marxschen Systems (in a volume of writings in honour of Karl Knies), Vienna, 1896; English translation by Paul Sweezy: Karl Marx and the Close of his System, New York, 1949, pp. 73-4. Hilferding's reply to Böhm-Bawerk, which is the best Marxist critique of the theory of marginal utility, is nonetheless deficient on this question -- cf. Hilferding, op. cit., p. 127: 'We have in fact nothing more than a disregard by Marx of the specific forms in which use-value manifests itself.'