[OPE-L:5554] Re: William of Ockam's Razor and Political Economy

From: Gerald_A_Levy (Gerald_A_Levy@email.msn.com)
Date: Sat May 12 2001 - 08:42:02 EDT

Re Allin's [5550]:

> There's a problem here with the term
> prediction".
> Jerry, you're taking it to mean "telling ahead of
> time", which indeed  it literally means.  Yet in the > sense in which Paul
is using the term
> -- which is fairly widespread in the philosophy of > science -- no
particular temporal order is
> implied.  Etymologically, "prediction"
> and "forecasting"  ought to be pretty much
> synonymous, but in the
> philosophy of science the act of telling ahead of > time is labeled
"forecasting", while "prediction"
> is given a more generalized meaning.

OK, I agree that the word 'prediction' has different
meanings in different subjects. I won't say that
your understanding of 'prediction' is wrong:
rather I question whether the meaning that is
widely given to the term in the philosophy of
science should also be used in political economy
(and 'social science' generally).

> This may seem like semantic legerdemain, but
> there is a basis for it.
> Consider Newtonian mechanics, as applied to
> the celestial sphere.
> Did Newton really latch onto something, with his > account of how
> physical bodies interact via gravitation?  An
> affirmative answer seems
> to be validated by the ability of the Newtonian
> theory to predict the
> existence of unknown planets, such as Pluto.
> Given the existing
> observations on planetary motions, the
> Newtonian theory suggests that
> those observations are best explained by
> postulating the existence of
> a new planet at (roughly)  such and such a
> location.  We look
> carefully, and -- Wow! -- there is a planet there.  > Since the theory
> was initially constructed with no such data and
> questions in mind,
> this is very impressive.
> Yet the question of temporal order is secondary.  Consider another
> well known celestial example: the "predictions" of Einstein's theory
> of Relativity with respect to the perihelion of Mercury.  The problem
> was that the orbit of Mercury around the sun exhibited some peculiar
> features that were not readily explicable on the Newtonian theory.
> To account for the observed orbit, a Newtonian would have to postulate
> some unobserved planet that was mucking things up.  In itself that
> hypothesis was not unreasonable.  After all, Mercury (being so close
> to the sun) is hard enough to see, and another planet even closer to
> the sun would be very tricky to spot.  But it was never spotted.  The
> sole reason for believing in the existence of such a planet was the
> Newtonian theory.
> Along comes Einstein and says, "Hey, this orbit for Mercury is just
> what my theory predicts!" (with no unobserved planet).  Of course, he
> wasn't actually _predicting_, in the sense of forecasting, since the
> facts about Mercury's orbit were well known already.  But his theory
> "produced the phenomena" in a way that Newton's did not.  In
> principle, one might worry that Einstein had "cooked up" his theory
> precisely to "account for"  the previously known facts about Mercury's
> orbit.  Yet any physicist who examined Einstein's theory would
> immediately see this worry was unfounded.  Einstein's theory was an
> very general one, lacking "free parameters"  that could be
> surreptitiously tuned to match the perihelion of Mercury.  The
> "prediction" was highly impressive, even if it was after the event; no
> less striking a confirmation of the fact that the theorist was "onto
> something" than the Newtonian prediction of unknown planets.

I won't take issue with any of the above. I
question, though, whether methods and techniques
used to verify or dis-prove or test a theorem
in the 'natural sciences' can also be used in
political economy.

The examples that you give above largely
concern astronomy and physics.  In astronomy,
some theorems can be proved though observation.
Yet, in political economy we don't have unchanging
regularities in nature that can be observed which
then can be used to validate/invalidate a political-
economic theorem. In physics, a theorem
concerning a causal relationship between two
variables can be tested by placing them in a
vacuum. Yet, we can't do this in political economy
(the ceteris paribus assumption is a very poor
substitute for a vacuum!). In the other natural
sciences, as well, there are established and
accepted ways of proving a proposition, e.g. a
litmus test in chemistry. Yet, there are no litmus
tests in political economy.

Thus, I don't think that we can apply the same
way of thinking about the 'social sciences' as we
do in the natural sciences. Of course, there are
points of contact and the social and natural
sciences are not _completely_ separate: one might
conceive of the social world as existing within the
natural world but because of humans ability to
comprehend and transform their world (within
limits), i.e. because they have subjectivity,  human
conduct can not be predicted in the same
manner as it can in the natural sciences. [Nicky
in 5548 seems to have similar reservations as I do
on this score].

[An interesting history of thought question is how
Marx viewed these issues.  I would suggest that
he had contradictory positions on this subject,
e.g. when he takes the position in the "Preface
to the First Edition" of Volume 1 that the
tendencies of capitalist production work "with iron
necessity towards inevitable results", he is
suggesting a meaning to tendencies under
capitalism that is very different from that explained
elsewhere, e.g. in Volume 3.  Similarly, when he
wrote in the same "Preface"  that from his
standpoint "the evolution of the economic formation
of society is viewed as a process of natural
history", he seems to suggest that the Darwinian
perspective on evolution can be applied to
understanding the history of class struggle. This
(teleological) perspective on 'necessity' and 'inevitability', does not seem
to be in keeping with
propositions advanced elsewhere about his
philosophy (e.g. in the "Theses on Fuerbach")].

> What's the relevance to Marxism?  Paul is arguing (and I agree) that
> the labour theory of value is a predictive theory in the extended
> sense.  It can "produce the phenomena" (not necessarily ahead of time)
> in a way that (e.g.) Value Form theory cannot.

VFT might counter, though, by claiming that
embodied labor theories of value shave too closely
with Occam's razor by over-simplifying the subject
matter (I took this to be part of what Geert was
saying in 5527 ).  Andy seems to agree in 5545
when he wrote: "If a theory tells me that money
doesn't exist, or is purely 'inessential', I reject
that too."   Do you agree that this is an issue
in this debate?  How would you (and e.g. Gary
and Ajit) answer Andy's rejection of a theory
using this criteria?

> Paul also argues (and
> again I agree)  that this sort of predictive theory > is needed to
> validate (or even make sense of) key points of
> Marxist theory, such as
> the theory of relative surplus value.
> The idea of relative surplus value is that
> capitalists pursue an
> increase in surplus value by reducing the labour
> content of the
> workers' means of subsistence.  (snip, JL)

Suppose that output/working hour increases in
that part of Department II that produces
commodities that are solely consumed by
capitalists while output/working hour remains
constant in Dept I and the rest of Dept II.  Is
that an example of relative surplus value where
the labor content of the workers' means of
subsistence can remain unchanged?

In solidarity, Jerry

This archive was generated by hypermail 2b30 : Sat Jun 02 2001 - 00:00:07 EDT