Re: [Fis] CONSILIENCE & interdisciplinarity

Dear Pedro, and All,

Pedro stated that Wilsonian consilience involves reductionism as one of its
central motifs. I worry that the term 'reductionism' may be understood
differently by different people. In the philosophy of science, the
prevailing view is that a theory, such as thermodynamics, reduces to
statistical mechanics if statistical mechanics is able to explain the
empirical success of thermodynamics, and explain its limitations. For
example, statistical mechanics is able to explain Brownian motion, which
lies outside the scope of thermodynamics. In fact, statistical mechanics
provides a Whewellian consilience of chemical, thermodynamics, and Brownian
inductions that is highlighted by the agreement of many independent
measurements of Avogadro's number (as argued by Perrin in a book called
Atoms around 1916). The reductionist's view is that we should dispense with
thermodynamics altogether. In the case of statistical mechanics, it is
supported by the Whewellian consilience of inductions achieved by
statistical mechanics.

On the other hand, for a long time, neither statistical mechanics nor
thermodynamics (van der Waals theory) could account for the measured values
of critical exponents in liquid to gas phase transitions, or phase
transitions in ferromagnetism. But there were some relatively simple
phenomenological models used by physicists that modeled the phenomena
successfully using power laws. These laws were not reduced to statistical
mechanics until very recently, when Wilson in 1971 used the method of
renormalization borrowed from quantum field theory to derive an
approximately correct set of critical exponents from a precise assumptions
about the intermolecular forces. It's safe to say that theories that manage
to explain *some* disparate phenomena rarely, if ever, explain all the
phenomena. Neuroscience has not explained all psychological phenomena, and
molecular genetics has not explained all biological phenomena.

Whewell made a distinction between laws of phenomena, and laws of causes.
Thermodynamic laws count as laws of phenomena, whereas the laws of
statistical mechanics are laws of causes, in some liberal sense of 'cause'.

This brings me to the other issue that Pedro has raised--whether we should
view the sciences in terms of a strict hierarchy, in which physics informs
all the special sciences, but is never informed by them, or whether a circle
of sciences is a better metaphor because it allows that every science can be
informed by any other science. It seems to me that if the circle metaphor
implies that connections between sciences are entirely symmetrical, then it
says, incorrectly in my view, that statistical mechanics and thermodynamics
have an equal status. I agree that the strict reductionist view that falls
out of a strict hierarchical view of science is too rigid. A better view
seems to be a mixture of both. After all, theories, such as statistical
mechanics, have achieved a bone fide kind of consilience that has not, and
will not, be achieved by any theory at the phenomenological level. On the
other hand, higher level theories, no matter how consilient, are in a
constant state of development, which is driven by the development of less
consilient phenomenological theories below them in the hierarchy. There is
a kind of coevolution between the levels.

Whewell is someone who pushed for a hierarchical view of science, in which
laws of phenomena are at the bottom and laws of causes are at top. This is
a hierarchy of theories, not disciplines. He thought that laws of causes
supersede laws of phenomena. If this view implies that there is a point at
which phenomenological research contribute nothing to the progress of
science, then it is wrong. The successful consilience of some inductions
does not mark the end of science--there is always the consilience of new
inductions.

Malcolm

_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis
Received on Wed Oct 6 23:03:11 2004