Chemistry Reference
In-Depth Information
commitment to physicalism and a deflationary, externalist idea of the
role of idealization in explanation appears to be highly restrictive. It rules out much
scientific practice, at least so far as explanation in chemistry goes. Those of us
interested in defending a more inflationary view of explanation in chemistry need
an alternative strategy. Even if one is not committed to physicalism, one must admit
that there are significant problems confronting an account of explanatory idealiza-
tion given that idealizations will misrepresent the causal production of phenomena.
But while Strevens account is restrictive, in another sense it is surprisingly permis-
sive. A result of this permissiveness is that even if idealization in chemistry is the
distortion of causal difference-makers - a central claim of this chapter - this does
not entail giving up on the kairetic criterion of explanatory relevance.
Strevens
Strevens
'
adopts a “two-factor” approach to causal explanation. His aim is to
distinguish the metaphysics of causation and a criterion of explanatory relevance.
He is ecumenical in his attitude to the metaphysics of causation and defends the
idea that identifying the causal relation - causal influence - is a necessary but not a
sufficient condition for explanation. A
one
-factor view is the idea that the causal
influence is sufficient for explanation. According to Strevens, the disadvantages
with the one-factor view are many. For one thing, it lets in too much. It seems to
commit us to a view of explanation in which
all
casually salient factors - all the
actual causal influences - are explanatorily relevant. But explanation is selective.
So, while one factor of explanation is the causal relation itself, the other factor is a
non-causal criterion of explanatory relevance. And that is what the kairetic criterion
provides. It is a selection rule driven by the kairetic procedure of optimization.
A high-level relation of causal influence - “
c
is a cause of
e
” - is not an assertion of
causal influence, but rather an assertion of an explanatory relation between two
events, namely
c
and
e
. The explanatory relation is both a “low-level causal
influence relation and a high-level explanatory relevance relation” (ibid, p. 51).
Combining a high-level explanatory relation with a low-level (i.e. physical) causal
influence relation, “gives you both the truth conditions for causal claims and an
account of the high-level
is a cause of
relation, now understood not as a purely
causal relation but as the causal-explanatory relation, the relation that an event must
bear to another event in order to participate as a cause in its explanation” (ibid).
According to the two-factor approach, in explaining a phenomenon we select
just those and only those causal facts that make a difference to the causal production
of the explanatory target, underwritten by a metaphysical dependency relation of
causal influence while remaining silent as the metaphysics of causation itself.
But note that Strevens account is also a
modular
theory of explanation. One doesn
'
t
just select for explanatory relevance; one also selects a “domain of dependence
relations that must be appreciated in order to understand the phenomenon to be
explained” (ibid, p. 5). If the domain of dependence is a causal, then we have a
causal explanation. Strevens himself points out that the domain of dependence need
not be causal. Mathematical dependence is one candidate of non-causal dependence
of interest to philosophers of scientific explanation (ibid, p. 180). It sounds like
causal influence isn
'
t even a necessary condition for explanation, and Strevens
himself suggests a “partial revision” of his account such that “the difference-
making criterion takes as its raw material any dependence relation of the
'
making
'
