While pleased to be associated with Ellen Meiksins Wood’s position against Alan Carling and ‘Rational Choice’ or ‘Analytical’ Marxism, we believe that Carling’s caricature of Wood’s Marxism and ours as ‘everythingism’ requires a rebuttal.footnote1 Wood’s own reply to Carling does not confront all that is at stake here, especially the refusal to deal seriously with the basic philosophical and methodological challenge posed for Marxism by Marxists working with this ‘everythingism’.footnote2 But first, the term ‘everythingism’ itself needs to be set aside in favour of what we actually argue. Calling it an ‘unfortunate strain of Marxian thought’, Carling defines everythingism (p. 98) as the view that ‘you need a complete explanation of something before you can have any explanation of something’. He rejects this caricature in favour of a practical approach which, not ‘aiming for an utterly exhaustive explanation’, gets ‘along as best we can—one bit of explanation at a time’.

For reasons explained elsewhere, we follow Lukács and Althusser in using the term ‘overdetermination’ to (1) criticize the many different kinds of determinist arguments within Marxist theories, and (2) offer an alternative, non-determinist kind of Marxism. Notwithstanding the problems it raises (as, indeed, basic theoretical terms always do), we find ‘overdetermination’ to be far more precise, useful and suggestive than the ‘everythingism’ it displaces. In a sense, our reply to Carling involves explaining how and why overdetermination is not subject to the dismissive critique he offers of everythingism.

Because everything is related to everything else, because the conditions of the existence of any event are infinite in number and variety, and because each of those conditions have in turn their conditions of existence, it is, as Carling sees, quite impossible to produce complete or exhaustive explanations of anything. There are, broadly speaking, three ways to handle what is an epistemological as well as a practical problem.

Some people—rarely Marxists—decide to throw up their hands in frustration, presume the effort at explanation to be hopelessly misguided, and prefer meditations on undecidability, the futility of partisanship in politics or much else, and so forth. A second group does produce explanations by (1) presuming that some conditions of existence (causes or determinants) are more important than others, (2) identifying these essential conditions, and then (3) building explanations around them. Carling proceeds in this way—‘one bit at a time’. Like many others, Carling places great confidence in such explanations as the most adequate or best, since they are superstructures built upon, and from what are taken to be the essential determinants of social life. Sometimes only those explanations are endorsed that proceed from one essentialized determinant (this was perhaps Carling’s view in his 1986 New Left Review article essentializing individual rational choices in social theory).footnote3 At other times, endorsed explanations build upon more than just one essence, as in Carling’s current view that rational-choice logic needs to be ‘supplemented’ by ‘other explanations—especially functionalist explanation’ (another essentialism).

Ours is an example of a third and completely different response to the interrelatedness of all things. We proceed from a presumption that no theorists, ourselves of course included, can ever know which one or several of an always infinite set of explanatory causes or elements are ‘the most important or influential’. That presupposes the complete explanation, which is precisely what is out of reach. We presume, rather, that no essences or essential causes exist. Thus the inevitably partial explanations that any of us has everr constructed, or could ever construct, are never more or less adequate, better or worse in any general or absolute sense. Rather, explanations are different in accordance with (a) which factors we select to emphasize and (b) whether or not we claim that the factors we select are the essential causes of what is to be explained.