Re: individualism vs holism.

[Alan G Isaac <aisaac@xxxxxxxxxxxx>]

I think Ted and Larry's posts show that defining terms
is going to be crucial and that there is no generally
accepted terminology that is adequate to the distinctions
that need to be drawn.

Many people confuse methodological individualism (MI) with a peculiar
kind of essentialism that equates individuals to their desires and
treats desires as exogenous to social context.  Atomism seems as
reasonable a term as any, although atomism is generally applied to more
than desires.  I do not believe I have ever read anyone who actually
defended atomism, although it is easy to find people ready to attack it
in the belief that they are attacking MI.

To understand what might be meant by MI, it might be worth considering
what Hayek meant by it, since Hayek clearly and explicitly recognized
the social embeddedness of individuals.  What was crucial was that the
individual is the actor and the locus of beliefs/knowledge and desire,
so that any "explanation" of "social outcomes" is incomplete unless it
accounts for the individual actors.  Obviously the crude functionalism
associated with popular Marxist arguments (e.g., there is a reserve army
of labor because that serves the interests of the ruling class) are
incomplete, for the kinds of reasons that have been explored in detail
by Jon Elster.  But even statements like "the firm maximizes profits"
become suspect, since there is no reference to the beliefs/knowledge and
desires of the individual actors who determine the firm-level outcomes.

Once we have such an understanding of MI, one may reasonably wonder if
it doesn't sound a lot like methodlogical holism (MH).  After all, if
individuals are socially embedded, a society has an obvious ontological
precedence over any individual.  Yet while the distinction may be
difficult to draw in a persuasive fashion, the flavor of the distinction
is well communicated by sociological literatures in which explanation is
allowed in terms of group actions---under MI, only individuals act, and
group outcomes are explained by the individual actions---or in political
economy literatures that use a concept of the social good that is not
short-hand for the good of the constituent individuals.

In that context, the passion surrounding the debates over MI can be
understood.  One might easily claim that the worst excesses of humanity
and the greatest threats to human freedom have had the explicit
rhetorical protection of a holistic concept of the social good (either
in good faith or as cynical rhetoric).  I think this is a true claim,
but it is really irrelevant:  the *methodology* under discussion is the
methodology of *explanation* in social science.  If we can find useful
explanations of social phenomena that do not require reference to the
constituent individuals, then this is a challenge to MI as generally
understood.

As an analogy, consider Boyle's law.  It is a useful and exploitable
property of gasses that makes no reference to the constituent molecules.
Even more important, the entire literature on emergence suggests
situations where aggregate level explanation supplants explanation at
the level of the constituents.  Finally, to turn more concretely to
social science, consider the literature on zero-intelligence traders:
the point of this literature is that efficient outcomes can be assured
by certain institutional structures indepedently of the characterization
of the participants.

Alan Isaac