Date: Thu, 13 Mar 1997 22:03:47 -0500 (EST)
From: Justin Schwartz <jschwart-AT-freenet.columbus.oh.us>
Subject: Re: M-I: game theory



I'm a pretty severe critic of Roemer on lots of counts, but I think Louis
misses the boat and Barkley catches it on this exchange. Game theory is a
useful tool. It can't be confused with reality, but it can be usefully
applied to it. As Elster argues, it does capture in very precise terms at
least part of the notion of contradiction. In particular the idea of the
Prisoner's Dilemma and its analogue, the core of which is that
individually rational (i.e., selfish) action will have perversely
suboptimal consequences, is key to an important aspect of contradiction.
It is in fact just this notion that drives Marx's theory of the tendency
of the rate of profit to fall. I borrow this idea from Daniel Little,
andother analytical MArxist, who in his The Scientific Marx conclusively
shows that much od the apparatus now formalized in AM is not only not
alien to Marx's own thought, but integral to it.

This, and the living and immediate relevance of game theory and rational
choice theory to real world problems, should not surprise us. Capitalism
is a system that systematically reproduces narrow economic rationality. So
it manifests a lot of the effects of such narrow economic rationality. The
science that studies these is rational choice theory, of which game theory
is a species; neoclassical economics is another species of it. As Marx
said of the theorist of the bourgeois economists of his own day, these
contain a lot of truth if they are understood in their historical context
and not reified and universalized.

Louis really hates economic rationality because he thinks, correctly, taht
it cannot explain class solidarity, which is not, on the individual level,
economically rational action. I note that this is the central problem for
Lukacs' History and Class Consciousness, a work that no one can accuse of
insufficient Hegelianism or attention to contradiction; but then, as
Lukacs himself demonstrated in The Young Hegel, Hegel was much preoccupied
by economic rationality in the version presented by Adam Smith. 

However, while ER and its working out in game theory cannot explain class
solidarity, it can, as Barklay notes, go a long way to explaining the
dynamics of class conflict, which is a zero-sum game in game-theortic
terms--in the long run, what one class wins, the other loses; and each
ideally seeks to maximize its expected utility. It can also explain, on
the individual level, why in part class solidarity is so hard to achieve.
Here the locus classicus is Alan Buchanan's Revolution and Rationaity
(reprinted in his Marx and Justice), which discusses revolution as a
public good (in ecpnomic terms) of a familiar sort, one which it is in the
intereests of each who will benefit from it for others, but hot himself,
to bring about. It's because revolutionary action is individually
irrational, at least in part, that we have such a hard row to hoe. There
is a concrete appplication of game theory in a context that matters
practically. This goes a long way towards making sense of the class struggle.

Louis cpmplains that Roemer doesn't explain a lot of stuff with his models
that he doesn't try to explain. If Louis is interested, Roemer does indeed
have an explanation of historical materialism--I can't lay hands on it
right now but willprovide the reference. But Roemer is trying to do other
things with his models. In his analysis of exploitation he is trying to
show its essential logical structure and dynamics. I think, and have
argued at lenth, that his particular approcah is defective on a number of
key counts, including areas where I think Louis and I would agree. Roemer
is insufficiently realistic in his models. But the problem is not that he
is doing formal modeling, but that his models leave out key aspects of
capitalism that are historically important and causally relevant. 

Louis complains that Roemer's math makes his eyes glaze. I would be more
charitable than Andrew W. A., who thinks that Louis wants everything to be
intelligible to someone completely illiterate and uninformed. Sometimes I
do think Louis wants us all to commit ourselves to writing nothing but
propaganda, but then I reflect that his own writing is unintelligible to
most people, and indeed most workers--I don;t mean because it is unclear,
but because of the vocabulary and background of knowledge it
presupposes--so I think it must be something else. With the math, I
suspect that it is the C.P. Snow Two Cultures issue, the abhorrence of
anything formal. "Grey is theory, but Green is the tree of life," as
Goethe puts it. Well, that's as may be. I wouldn't do Roemerian analysis
if I could, which I can't, although I can follow it, but I'm glad we have
people who can prove theorems on important issues. Their significance,
however, requires interpretation. 

--Justin

> Barkley:
> 
> >     1)  Although a lot of Roemer's presentations presume 
> >methodological individualism, they don't have to.  Let the 
> >"individuals" be groups or classes.  Then, let the "games" 
> >begin.  The analysis works out.  Are not the capitalists 
> >and the workers each trying to get as much for themselves 
> >as possible?
> 
> Louis: "Let the individuals be classes?" "Aren't the workers trying to get
> as much for themselves as possible?" These formulations are alien to
> everything that Marx was about. And what if "the analysis works out" on its
> own terms. This is not germane since it is highly likely that what Roemer is
> modeling in his corn economy simply has no correspondence in history.
> Marxists are focused on history, not laboratory conditions where
> game-experiments can be conducted. Roemer's methodology doesn't explain the
> rise of capitalism, nor does it identify the role of the class struggle in
> history. Unless an analysis can do this, it isn't worth a dime for the
> purpose of understanding or changing the world. 
> 
> >     2)  I see no reason not to view the conflicts in a 
> >non-cooperative game as being awfully similar to a 
> >"contradiction."  Why isn't a contradiction?
> 
> Louis: I have no idea what your use of the word "contradiction" means. I use
> contradiction in the sense that the USSR was a contradictory society, or
> that black nationalism has contradictory aspects, or that Bonapartism is a
> contradictory phenomenon. What do you mean when you say "contradiction"? A
> "conflict" between the bourgeoisie and the working-class in some kind of
> Roemerian game is not an example of a contradiction. 
> 
> A much better example would be the Irish working-class in the US in the
> 1860s. This most oppressed layer of the working-class was coopted by the
> ruling class in Northern cities in order to act as a bastion of racism. In
> England, the Irish were treated like "niggers". Even Engels himself was
> capable of the most loathsome racist epithets when he wrote about the Irish.
> Then the Irish turned around and became like Ulstermen to the blacks in the
> northern cities after they migrated to the US. This can not be modeled in
> games. It requires class analysis rooted in the totality of social
> relations, including race and gender. AM is the perfect place to hide from
> all these sorts of messy problems, since everybody acts "rationally".
> 
> 
> >     3)  It is certainly true that Roemer does not present 
> >a Hegelian dialectical view.  But, non-cooperative game 
> >theory has a dynamic version in its "repeated game" form in 
> >which a sort of abstracted historical process can be played 
> >out.  There are people out there running simulation models 
> >of this sort of thing.
> 
> Louis: Good luck to them. This seems like a perfectly harmless activity.
> 
> >     4)  Of course a very basic aspect of this is the 
> >question of the use of mathematics.  Many argue that 
> >dialectical analysis is inherently unmathematical and thus 
> >would rule out game theory.  Now, on the one hand, Marx 
> >himself did not hesitate to use mathematics.  Indeed, Phil 
> >Mirowski has argued that he was the first true mathematical 
> >economist.  On the other hand, it is possible to interpret 
> >dialectical analysis in a mathematical manner, the quantity 
> >becomes quality being equivalent to a phase transition in 
> >nonlinear dynamics.  I have even written a paper on this, 
> >but will not name it or where it was presented or where 
> >it is currently under review or who was in the audience for 
> >fear of being dinged for self-aggrandizement and awful 
> >academic elitism, blah blah blah.  But in that paper I 
> >recognize that if one insists that dialectics cannot be 
> >mathematical, then the whole analogy is out the window.  
> >But why can it not be mathematical?  Too complicated for 
> >Third World peasants to understand?  So is much of Marx.
> 
> 
> Louis: I am sure that the list would appreciate a presentation on your
> article just as it would one from Zarembka on Venezuela. There really is
> nothing to be afraid of when you favor us with such presentations except the
> possibility that laypeople will tell you that you are full of shit. On the
> up side, you might receive the adulation of the untutored mob.
> 
> You must remember, however, that many of us--myself included--haven't done
> anything with math in many years and won't know what you are talking about
> if you use mathematical notation. That is one of the reasons I get
> frustrated with Roemer. Much of his text is riddled with mathematical
> notation that I am convinced proves nothing. I am spending the next year or
> so doing remedial work in math just so I can confront the sort of
> mystification represented by Roemer's mathematics and satisfy myself that it
> is grandstanding.
> 
> What would impress me more than anything would be an example of AM making
> sense of the class struggle. I have yet to read a word in Elster, Cohen or
> Roemer that explains anything going on the world. When Roemer does descend
> from the very rarefied world of the Oxford commons room and tries to speak
> about the class struggle, he sticks his foot in his mouth nearly every time.
> For example, he says that recent wars between socialist nations like China
> and the USSR have "provoked a crisis in Marxist theory". Really? I have no
> idea what makes him say this. A classical Marxist analysis of the relations
> between the United States and China on one side, and the USSR and Vietnam on
> the other, during the mid-1970s would explain a lot more than Roemer's
> lifeless and mind-numbing excursions into corn economy and his calculus
> cartwheels.
> 
> 
> 
> Louis Proyect
> (www.columbia.edu/~lnp3)
> 
> 
> 
>      --- from list marxism-international-AT-lists.village.virginia.edu ---





     --- from list marxism-international-AT-lists.village.virginia.edu ---