Chemistry Reference
In-Depth Information
simpler than canonical models because while the latter are optimized by providing
a range of bounded values for a specific parameter, idealized models provide
them with a definite value (an extreme default value) and this can make it easier
to deduce the explanatory target. This certainly seems true of frontier orbital
and orbital symmetry models. Third, idealized models are good predictors because
only non-difference-makers are distorted. Frontier orbital and orbital symmetry
models are indeed prized as good predictors. In fact, Woodward and Hoffmann
s
approach is a much prized means to predict the course of chemical reactions
(Brush
1999
, p. 286).
But what is the
intrinsic
explanatory function of idealized models? Strevens
'
“ontology-first” approach implies that while an extra-ontological sense of explana-
tion performs some constructive role in scientific practice, there is nothing
“internal” to the kairetic account to motivate the use of idealized models. They
perform an extrinsic explanatory function in the communicative or pragmatic sense
of explanation, but that is not a part of Strevens
'
account of explanation as causal
difference-making because idealization is the distortion or omission of
non
-differ-
ence-makers. As Strevens points out, it is only by drawing on an extrinsic idea of
“explanation” that one can deliver a positive role for idealized explanations: we
need “a novel proposal about the meaning and purpose of idealization that is
independent of the kairetic account” (op. cit., p. 315). The novel proposal is
presumably that unlike other accounts of explanation, in particular pragmatist
accounts that (Strevens argues) provide little in terms of a positive cognitive
function for idealization, idealizations certainly raise awareness of causal irrele-
vances. But idealized models are still explanatorily second best because the ideal-
izations themselves do not contribute to the casual entailment of the explanandum,
nor do they tell us why certain causally salient factors are explanatorily irrelevant.
As for the other proposed virtues of idealized models, emphasising their simplicity
and predictive virtues doesn
'
t seem to get us very far. Pragmatists would presum-
ably accept this. If idealizations merely make non-difference-makers explicit in the
communicative sense of explanation, then they are explanatorily redundant in the
ontological sense of explanation. The result is a deflationary account of idealized
explanation because idealization is not connected to causal difference-making.
It is entirely legitimate for Strevens to regard idealizations as explanatory in
an extrinsic sense given his commitment to the ontological sense of explanation.
This sense of explanation would seem to pose problems from significant parts of the
explanatory content of chemistry. By assigning a zero value to the contribution of
sigma bonds and thus to regard them as irrelevant to the causal entailment of the
explanatory target is one thing. But it should be readily apparent that a considerable
difficulty remains when one attempts to cash out the explanatory function of
any
orbitals (atomic or molecular; and concerning the latter, “frontier” or otherwise).
The orbital concept is predicated on the semi-classical electron configuration
model. The orbital idealization has lost it physical significance except in the case
of one-electron systems (like the hydrogen atom) because individual electrons in
many electron atoms are not in stationary states; only the atom as a whole could be
said to be in such a state (Scerri
2001
, S79). Atomic orbitals are essentially
'
