Notes on: Collier, A.  (1978) 'In Defence of Epistemology'.  Radical Philosophy 20.

Dave Harris

Bhaskar's rejection of epistemology evades the problems.  What we need is a new epistemology, which should be materialist, rather than a new idealism which is what you get in Bhaskar.

After the intervention of Althusser, the split between theory and practice became both the political and then epistemological issue.  For Althusser, seeing theoretical production as practice, and seeing the unity of theory and practice as a political rather than a general task offered a solution.  However, this need not be seen as a rejection of epistemology.  Althusser's position moved towards internal criteria to establish validity rather than external, and this is acceptable because the criteria clearly are different in different sciences, because their objects differ—but they're all doing the same task, trying to make a theory correspond to its object.  In this sense, we need to establish general ways for science to proceed [as an empirical issue] rather than relativizing.

Althusser saw philosophy as a matter of class struggle rather than some master discourse to guide science, but philosophy cannot be seen as the same as any social practice.  It must be partisan to defend objectivity avoid errors and illusions.  Althusser's Essays in Self Criticism retracts the emphasis on epistemology, but still argues that it is possible to disentangle the ideological bits from the scientific interests in, say, Darwin and Marx.  Science is not seen as relatively autonomous in terms of other practices, but what makes it genuinely different is its claim to appropriate the real in objective knowledge.  If it is successful, it can escape its social genesis [sounds very much like Popper], and it is this that makes science universal [I prefer the Habermas line that it is capitalist technology and the spread of bourgeois interest that seems to make science universal]. Science has to exist somewhere beyond the social and the materialist for Althusser, if we are not to equate it with sociology or psychology.

Cutler et al develop a kantian/platonic line, where discourse theory is used to launch a radical critique of Marx.  All that survives from Marxism is the notion of classes as agents, and a general anti humanism.  This is an abandonment rather than a revision, and it is based on the view of Marx as essentialist.  Criticizing his rationalist epistemology leads them to reject all epistemology.  The argument is seen best in criticisms of the labour theory of value which are used to explain circulation and equilibrium.  This is economist and functionalist for Cutler Et Al and leads either to an orthodox crisis theory or some teleological and functionalist notion of survival of capitalism as a matter of reestablishing equilibrium.  However, for Collier, this approach is too polarised and has no concept of the dialectic.  Further, the discussion of reproduction is not teleological, and nor can contradictions be equated with crisis.  Althusser will suffice to argue this.

Any notion of an expressive totality or essentialism in Marx is based on the conception of a 'trans historical subject', but Marx also says that social structures have their own dynamic.  Cutler Et Al see no break between young and mature Marx.  Marx's notions of politics as class struggle must be rejected, because politics do not just simply represent classes, but other connections are still possible, for example when alliances are forged as a result of class struggle, as in the analysis of the 18th Brumaire and the state.  The state is also never simply a direct representation of class interests, but some parties are closer to classes at least than others [this touches on the old discussions between Marx and Weber about the connections between class and party].  Classical Marxism was never reductionist, since economic classes were seen as not the same as political ones, and making them so was a practical matter rather than an effect of structure.  It is also possible to argue that social structures produce ideology without being functionalist about it—so that dysfunctional ideologies are equally possible, including bourgeois socialism.  Overall, this reading is too assertive.  It is possible to have a materialist analysis of subjectivity, and the connections between Marxism and psychoanalysis look promising.

Cutler Et Al assume that expressive totalities are the only kind.  They argue that there are only two alternatives instead—rationalism or holism, and empiricism, or atomism, and neither require a particular interest in epistemology.  Marx never really introduced new ideas of totality, but we can deny both atomism and holism and opt for 'decentred structures in dominance' as in Althusser.  There are other alternatives as well outlined in Marx.  For example  the 1857 Introduction describes a 'rich totality'  and refers to the 'concentration of many determinations', and argues that science needs to be grounded in reality as a point of departure, says that it is an illusion to regard the real as a product of thought [or discourse] and says that the real lies outside concepts.  The approach is not empiricist because thought is needed, nor is it rationalist because the real lies outside thought.  It is a kind of correspondence theory, suggesting that the relation between concepts is adequate if it corresponds to the complexity determinations found in the real.  The method of presentation makes Marx look essentialist, perhaps, but it is really are the result of a complex method of enquiry, as in all science.  There is never simple link between subject and object.

Cutler Et Al have not dispensed with epistemology, but rather have reverted to an earlier one of idealism, when discourses become necessary for knowledge, and this slides towards discourses as the only source or type of knowledge.  It is important to look at language rather than ideas, however, and Cutler does not deny the existence of real objects—it is just that they are irrelevant as a test.  But this makes it kantian, says Collier, operating with the old concept of things in themselves, that we can know.  They deny that thought is merely a matter of a subject grasping an object, but they also deny the materialist principle that thought should correspond to its object.  For them, there is no necessity in knowing the real world.

[Then a confused bit, which might even be similar to Bhaskar] the real world is given is a possibility to experience , So therefore it must be knowable at least as a contingent possibility.  It is possible to see the real world as the conception to produces an effect on our imagination, but it is equally reasonable to assume that the correspondence between experience and the real world shows that there really is a world like that.  If we reject this notion, the only criterion left is internal consistency of discourses, and this will lead to problems are relativism and choice, or perhaps anarchy [as in Feyerabend].  The usefulness of scientific discourse is particularly dubious if there is no correspondence to the real world.  We are left only with establishing logical connections, for example between concepts [the concept of base and the concept of superstructure], with only an accidental relation to reality.  Or we can assume idealism or essentialism.

Classical Marxism offers a much better approach.  Althusser spells it out in the notion of the decentred structure in dominance.  This structure is not reducible to its elements, and it is not teleological.  It does not constitute the elements, but rather limits them.  It does depend on the real world if it is to be reproduced according to 'geological and biological laws'. There can be accidents and floors, for example in the form of class antagonisms.  It is neither anatomist nor a holist conception, neither the sum of individual wills nor a single will.  Marx explains how the reproduction of the economic system reproduces the whole structure, and how the ideological level does as well, but not in terms of functionalism: if there are no such mechanisms, the 'mode of production could not be instantiated'(18), and the ideology that assists reproduction is selectively promoted [looks awfully functionalist to me!].  To reject such materialism for sociology means rejecting in biology too.  Modern biology shows how reproduction is possible without teleology.

Epistemology in the human sciences.  Realism claims a privilege here, so is this dogmatic?  Possibly, but every science must be realist.  Science makes progress if it 'lets things speak for themselves' rather than trying to bend them to our own needs.  New practices then become possible.  Natural science did appear as a 'mutation' and this 'had something to do with the bourgeoisie' [!].  Can we see the emergence of human sciences in this way?  These sciences became concrete through classification rather than developing law like regularities.  Human sciences remain probabilistic.  Marxism is both a concrete and an abstract science for Collier, but these relations needs clarification.

Most abstract sciences are tested independently of their applications, for example in experiments.  Human sciences can't measure [so they can still do abstract experiments?  Thought experiments?] so they are necessarily imprecise, but this is still better than learning by experience alone.  There is a need to risk concepts against practice, through systematic testing.  This led, for example, to ruling out the parliamentary road to socialism [strange way of thinking of this, as if it were just some scientific experiment], since it was obvious that the bourgeois state could not be captured, and this led to Marxists testing theories whether the state is neutral or has to be smashed.  The question is answered by observing the daily runnings of bourgeois states, gathering 'all sorts of empirical facts'.  Liberals would see some of these as only abuses.

Are Hindess and Hirst idealists?  They deny it, and agree that material objects are still out there, but the relation to them is the central question for Collier.  Hindess and Hirst might have no idealist ontology, but they persist with an idealist epistemology, and dangerous affects results, such as having to eliminate large bits of Capital, and developing dubious politics.  If no relation to reality is possible at all, why gain knowledge at all?
back to marxist controversies