New Left Review has surely made a major contribution to the development of Marxism by publishing a thought-provoking interview with Lucio Colletti, and an important article by Bob Rowthorn, which both appeared in nlr 86 July /August 1974. These two texts together cover a variety of important issues. This letter is only concerned with the area on which both of the two articles overlap; i.e. the question of the relation between Marxian political economy and the works of Sraffa and his followers. Although I am far from hostile to the main lines of arguments made by both Colletti and Rowthorn in this area, I feel that there is some scope for misinterpretation and the derivation of mistaken conclusions from what they argue.

The first point I wish to comment upon is the assertion, and it is no more than a mere assertion, that Colletti makes when he says that Sraffa’s ideas are incompatible ‘with the entire foundation of Marx’s analysis’ (pp. 21–22). Whatever the truth of this assertion, and whatever the truth of Colletti’s criticisms of Baran, Dobb, and Sweezy, Colletti must be criticised for making it without any explanation. It is far from clear that such a deep incompatibility exists. As the assertion stands it is likely to provide another weak excuse for some Marxists to dogmatically throw overboard the entire edifice of Sraffa’s work. At the moment Marxism is cautiously entering a period of renewed theoretical development, of escaping from its isolation and learning from the healthy and scientific aspects of other schools of social science. Although Marxism is a distinct theoretical system from Classical economics this does not mean that Marx did not learn from Classical economists such as Ricardo or Smith. In fact he adopted some of their categories and propositions, such as, for example, the distinction between exchange value and use value. The latter was utilized and endowed with a distinctive importance in Marx’s works.

On the question of the Sraffa school it is necessary to draw a distinction between the formal propositions of Sraffa, Garegnani, and others, and the theoretical interpretation of these results by Neo-Ricardian writers. Rowthorn provides an excellent critique of the Neo-Ricardian school in his essay, which I entirely accept, but this should not be taken as an attack on Sraffa’s formal results. In fact Rowthorn nowhere makes such an attack in his essay, he explicitly recognizes the importance of the formal Sraffian results (p. 73), and in other works footnote1 he has accepted a solution to the transformation problem on the same lines as Bortkiewicz and Sraffa.

I do not assert that interpretations of the Sraffa system can be entirely disconnected from the formal results. But these two aspects of the work of the Sraffa school must be distinguished. This point is reinforced by the fact that the Neo-Classical school now accepts the formal results of Sraffa’s Production of Commodities by Means of Commodities but they give them an entirely different interpretation. Unlike both Neo-Ricardians and Marxists they argue that Sraffa’s formal work can be integrated with marginalist utility theory and their own general equilibrium analysis.

The formal results of the Sraffa system are derived by logical deduction based on stated assumptions. Sraffa shows certain relations between prices, profits, wages, embodied labour inputs, and physical inputs and outputs, in a commodity-producing economy. His Production of Commodities by Means of Commodities contains very little other than a derivation of such relations. He does not provide a systematic conceptual analysis of the capitalist mode of production, or a theory of exploitation, alienation and fetishism. That is the major weakness of his work. But its strength lies in the power it has to expose logical contradictions in the Neo-Classical system, and even in the writings of Marx. I know of no reason why we should reject the logical force of these formal results, as long as the stated assumptions are regarded as satisfactory. Neither do Colletti or Rowthorn state such a reason, nor even assert that such a reason exists.

It must be emphasized, however, that some of Sraffa’s results are inconsistent with some major parts of Marx’s work, e.g. Marx’s solution to the transformation problem in the third volume of Capital. It is almost certain that other important inconsistencies exist. These cannot be glossed over or regarded as mere ‘corrections’ of Marx. Of course their character must be made clear. Insofar as they exist they are logical inconsistencies in Marx’s work: nothing more, nothing less. They may provide logical ammunition for an attack on a significant part of Marx’s work, but that attack cannot be made on the grounds of logic alone, the logical results have to be interpreted in one way or another. The Neo-Ricardians have provided one such interpretation—a faulty one. As far as I am aware, a complete ‘Neo-Marxian’ critique of Marx, which is consistent with Sraffa’s formal results, has not been constructed. I am certain that such a critique must be developed. But at this stage we cannot predict how destructive such a critique will be. At this stage of theoretical development there is simply no basis for Colletti’s bald assertion that Sraffa’s work is incompatible with the ‘entire foundation’ of Capital.

The strength of Capital lies in the scientific and materialist character of its conceptual system, and its consequent revolutionary relevance in the context of capitalist society. Sraffa’s work does not develop such a conceptual system, and there is no trace of a revolutionary critique of capitalist society. This is why the editors of New Left Review are wrong when they describe the Sraffa system as a whole ‘corpus of concepts’ (nlr 86, p. 2). Sraffa’s formal work is nothing more than pure logic. We reject it at the cost of logical consistency. Opponents of empircism are aware that facts do not speak for themselves, they require a system of concepts to endow them with theoretical meaning. In a similar way, logical syllogisms do not speak for themselves. As long as they are complete and consistent they can only be criticised for a lack of realism in their assumptions. That is the only ground upon which Sraffa’s formal results can be criticized (interpretations of Sraffa, are, of course, far more vulnerable). For these reasons the character of Sraffa’s formal system has hardly the same weight and significance as Marx’s Capital, whatever the veracity of the latter work.