I resist the teleological interpretation of Marx – that all of Marx is there in every text, and if a text seems to say something that contradicts all-of-Marx, then we just have to either categorize Marx’s works to shunt it to the side – it was polemical! – or decide that it was an unfortunate collateral gesture. On the other hand, I’m not sure that my idea of Marx as constructing his all-of-Marx-ness in his text really purges the teleological impulse completely. Take the issue of the notebook, or the draft. We have these things. They were preserved. But the facile notion that Marx, too, having these things, goes back over them suffers both from lack of proof and automatic assumptions about research and writing that I have found, both in my personal experience and as an editor of others, to be false. I have found, instead, that one’s vital discoveries tend to fade and change and be renewed – that old intentions get submerged by new ones. Yet characteristic themes and inclinations will assert themselves, and the repressed will return.
This is why I favor the problem-based approach to reading monumental texts. For any theme or thesis carries with it both the problems it responds to and the new problems it creates. A problem is as much a token of memory as a thesis. Stripping a writer of his problems – translating his text into something like a list of answers such as you can find in the back of the math textbook - trivializes him.