"Marxism, Functionalism and Game Theory" by Jon Elster

The scope of Elster's paper is quite large, and targets multiple ideas to establish its conclusions. Elster argues (a) in favour of methodological individualism, (b) for a discontinuation of the employment of functionalist and teleological conceptions of various types of social relation in Marxist theory, and (c) so as to advocate the application of game-theoretic methods of analysis of these types. The focus of the first half of the paper is an attack on the functionalist idea of socio-political analysis in Marxism. It is a commonplace, not only from the examples advanced by Elster (Marx's posthumous "Theories of Surplus Value" and multiple works in contemporary and recent sociological literature) but also from conversations with those committed to Marxist ideas, that social and political phenomena ought to be understood as existing both for and because of  purposes that are defined by their role in whatever system is currently dominant contemporaneously with the institutions themselves. To shed light on this idea we will take a closer look at the particular terms used by Elster in his analysis:

(i) Weak Functional Paradigm (WFP): derived from Bernard Mandeville, this refers to the thesis that

"an institution or behavioral [sic] pattern often has consequences that are (a) beneficial for some dominant economic or political structure; (b) unintended by the actors; and (c) not recognized by the beneficiaries as owing to that behavior. [sic]" (my emphasis). The description as "weak" clearly registers that this particular thesis is by no means contentious, as it is bound to be the case that at least one institution or behavioural pattern already has exhibited such features.
(ii) Main Functional Paradigm (MFP): this thesis is generated by suggesting that the "latent function" (that is, a role satisfying (a), (b) and (c) of the terms of the WFP) of a behavioural pattern or institution explains why the latter exists.
(iii) Strong Functional Paradigm (SFP): this thesis is rather radical, generated by universalising these concepts such that all existent institutions and behavioural patterns possess such functions.

Elster is critical of (and hostile to) these explanations, and comes to the conclusion that of the three different modes of explanation in the sciences, "causal", "functional" and "intentional", the last of these three is to be chiefly employed in sociological analysis. Now he says something here which is puzzling, and merits far more discussion than I can give it here. He says, "The social sciences make extensive use of intentional analysis, at the level of individual actions. Functional analysis, however, has no place in the social sciences, because there is no sociological analogy to the theory of natural selection" (My emphasis).

Now, I'm not necessarily a physicalist, but that phrase seems to ring alarm bells. In the first place, there actually have been models proposed whereby cultural and social changes can be understood in terms analogous to those of biological natural selection, namely those of Richard Dawkins and Daniel Dennett in their expositions of the existence and nature of "memes". If institutions and behavioural patterns (a) are reducible to patterns of observal behaviour, (b) reproduced over time and (c) face selection pressures on the basis of factors which portend an objective idea of "fitness", and behavioral patterns/institutions are not more than a collection of these patterns, then there does indeed exist such an analogy. In the second place - and this position evidently carries a hefty theoretical burden - the existence of a division of "intentional" and "causal" explanations is possibly non-existent. If intentions are the results of regular physical changes occurring in the brain of an organism exhibiting  behavioural patterns that are, in principle, causally predictable, then rather than it being the case that "the proper paradigm for the social sciences is a mixed causal-intentional explanation" as Elster says, it may simply be the case that "the proper paradigm for the social sciences" is just an extraordinarily complicated causal explanation. This strikes me as highly amenable to evolutionary accounts of institutions, and one that will be grist for the Marxist mill if a physical and scientific position on the relationship between the abstract character of Capital and the (no less abstract) idea of "fitness" be provided.

Elster goes on to provide a possible game-theoretic analysis for certain situations in sociological analysis, particularly for the striking worker. To use Elster's terminology, "E" represents the egoistic option, and "S" the solidary option. In the striking worker case, the worker stands to lose much if he opts for S (strike) and every other worker opts for E (work and earn), stands to gain much if he opts for E and every other worker opts for S. Now let us assume that the workers can be divided into two groups: one particular individual worker in one group, and all the other workers in the other group. If

"1" represents the outcome of both worker and all-others opting for E,
"2" represents worker and all-others opting for S,
"3" represents worker opting for E and all-others opting for S, and
"4" represents worker opting for S and all-others opting for E,

then analysed at a group level, if we assume that the more workers strike, the higher the possibility for success, then an order of options arranged in order of preference, the first the most and the last the least, would look something like this:

2 - 3 - 4 - 1

at an individual level, from the point of the worker, the preference looks like this:

3 - 2 - 1 - 4

The first prioritisation is the "Assurance Game" and the second is the "Prisoner's Dilemma". Elster believes that the use of game-theory can grant Marxism the analytical resources to properly consider the individual subjects as members of groups while simultaneously employing the benefits of methodological individualism. But if it results in all the emotionally-charged metaphysical and theological language being replaced by statistical and mathematical analysis, then count me out.