[OPE-L] "Levels of abstraction"

From: Jurriaan Bendien (adsl675281@TISCALI.NL)
Date: Wed Jan 23 2008 - 17:19:38 EST


I have no objection to the notion of "levels of abstraction" per se - science is replete with statements about "ceteris paribus" clauses (assumptions about "other things being equal"). What I object to is the abuse of the notion, in order to disregard problems which are in reality foundational. If we "assume away" certain problems, we can be left with the question of "what then validates the abstractions in the first place". This problem occurs in modelling all the time, leads to such confusions as that we assume what we have to explain. In particular, we model (create an analogy, isomorphism or likeness) in advance of a comprehensive theory precisely because we do not yet know exactly how to explain something, and therefore the starting assumptions are decisive ("What would be a non-arbitrary starting point?").

One thing I learnt from Marx is that we not just always abstract "from" something, but that it is decisively important what exactly we abstract from, and to what extent we are aware of that. And we can abstract from a real condition existing mind-independently, or we can abstract from ideas, i.e. we can abstract from abstractions.

If we abstract from abstractions, the question then arises what regulates or disciplines those abstractions. If the abstractive procedure is not regulated or disciplined by facts and logic, it could evolve in any way whatever. This is arbitrary or speculative. There is of course nothing per se wrong with arbitrary speculative or fanciful abstractions, but to call that "science"" would be wrong. 

I think generally Marx aimed to abstract from a real phenomenon in order to fix the essence of the object, and then provide an ordered story successively integrating the specifics, the proof being that you have really integrated the specifics in a rational manner. I provide a very brief indication of my interpretation in an article to be published in David Laibman's review. 

The main point is that I think that the dialectics of a subjectmatter have to be "discovered" through a systematic study of the subjectmatter. This implies (1) that you cannot specify an a apriori "dialectical method" which is good for all times and places and all contexts, and (2) that there is no "one" correct dialectical method, it requires your own ingenuity. Once you have discovered the dialectics of the subjectmatter, you can write a dialectical story about it, revealing the internal necessity of the subjectmatter, and integrating all salient aspects. But that is your own story.

I think dialectical logic is a logic which moves between different logical levels, but which is nevertheless non-arbitrary, insofar as in the end all salient elements are integrated in a self-explanatory story which is logically sound. The logical solidity, however, is only proved by the logical "coherence" of the story about the totality which you describe and explain.

It is just that a lot of the ordinary human reasoning we do (mediated by language), although it is non-arbitrary, does not conform to the rules of formal logic. In that sense, dialectical thinking is part of ordinary life. I cannot devise a measure of an empirical phenomenon, without moving between logical levels in a language to arrive at an adequate categorization. But in order to show it is non-arbitrary, we also have to be able to give good reasons for our specifications, by referring to relevant circumstances. Something may not be formally-logically compelling, but we can nevertheless explicate good reasons for our abstractive procedure. You cannot have abstraction without specification.

The problem with dialectical reasoning is that if it becomes speculative or vague, then, whereas it may be creative, it becomes anyone's guess as to what it really means, it could mean all sorts of things and this becomes arbitrary. In this sense, Rosa Luxemburg opined that she would have preferred Das Kapital without the "Hegelian roccoco", it was more a hindrance than a help. Marx strove for a dialectical depiction of the totality of capitalism, but in a sense he overreached himself, to the extent that there are lacunae, in his bold vision it is not unambiguously verifiable what the rationality of particular assumptions is. In addition of course, he did not prepare many manuscripts to the point of publication. In defence of Marx, however, is the fact that capitalism contains a series of mutually opposed but mutually presupposing forces which constantly have to be mediated, so it is not unreasonable to want to tell a dialectical story about it. But it would be wrong to say that you could only tell a story about it in a dialectical way.

Fred Moseley and others have put a lot of effort into proving that Marx's method was not unreasonable. This is true I think. But it does not mean that Marx is necessarily correct, or that it is the only method to follow. He could be partly correct partly wrong, and so on. Marx's concepts do not justify themselves, because they were Marx's concepts, and he could have told the same story in many diferent and equally valid ways. If we insist on only one method as the correct one, we get a ""tyranny of concepts".

The very nature of dialectical reasoning means that, although it is non-arbitrary, it does not in and of itself constitute a formal proof or an empirical proof. The only proof in dialectics is that of meaningful coherence, i.e. these propositions are mutually at least compatible (they are not inconsistent). If I can adduce good reasons for why I abstract in this way or that way, that is rational thinking, but that is not the same as saying I am necessarily correct in my inferences or assumptions. Whether I am correct about something depends on formal, empirical and practical proofs, not on the quality of my dialectical reasoning. You might for example impress somebody with your dialectical insight, connecting diverse phenomena and making sense of them, but it might only be of heuristic value. Heuristic insights are valuable, but they are not proofs.

If we are constantly debating about what would be the correct starting point, or what would be the correct method, then we never get to the point of doing any research looking at data and suchlike. All we do is that we just generate more and more heuristics, that is all. This might prove we are creative and attentive to the existence of problems, but it does not actually solve any problem. All it says is that you "might solve a problem" in certain ways. But in the end, frankly, I think the question is whether we do solve it. It is nice to be in the position of being an authority who tells others how things ought to be explained, but the real authority to my way of thinking is the one who really explains them, who really solves the problems.

Anyway, what my father taught me when I was young (a lesson which I often forgot) is that if you are in situation of chaos, or you are caught up in a "mire of contradictions" (maybe not of your own making), the way out is to start solving some problems, even if they are only very simple ones, since you get nowhere at all just by showing your awareness that there are many problems to be solved. And if you solve some simple problems, it inspires confidence to tackle some more. In this sense, Marx is not necessarily a good guide, since he diagnoses an enormous range of problems far beyond what he could solve - eager for knowledge, you may become ensnared in a predicament of biting off more than you can chew. 


This archive was generated by hypermail 2.1.5 : Thu Jan 31 2008 - 00:00:06 EST