Biology Reference
In-Depth Information
not at all the issue. The point is, rather, to explain some particular phenomenon
or to provide a mechanism which underlies some phenomenon. From a scientific
perspective, the point is to provide an explanation of some robust phenomenon.
From a philosophical perspective, we aim to explain how such modeling differs
from eliminative reduction.
Let us begin with mechanism and its rebirth as a philosophical ideal. Most
generally, as Carl Craver (in press) says, mechanistic explanations are 'con-
stitutive explanations.' They are not narrowly 'causal' explanations, moving
from cause to effect, but explain how some phenomenon or behavior is 'con-
stituted'. Usually, we begin with some set of phenomena, or some behavior, to
be explained. The explanation involves the articulation of parts, or components,
typically in functional terms. Having detailed the capacities of the parts, and the
character of their interactions, mechanistic models are sufficient to predict sys-
temic behavior. The most recent articulations of mechanistic explanation within
philosophy have taken two rather different forms. One, initially due to Peter
Machamer, Lindley Darden, and Carl Craver (2000), describes the development
of interlevel explanatory models. A second, due largely to Stuart Glennan (1996,
2002), fits a more conventional emphasis on theories and laws. Both approaches
assume that mechanistic explanation has a particular structure. Let us begin
with the structure from Machamer, Darden, and Craver, often referred to as
the 'MDC' model. We assume we have, say, an organism with a characteristic
behavior we want to explain. These behaviors, on the MDC model, can differ
depending on the entities, characteristic activities, and the spatio-temporal orga-
nization of the containing system. Spatial and temporal structure on this picture
are crucial. The result is a schema for understanding mechanistic explanations.
Craver (2002, ยง 5.3) says, at one point, 'In such schemata, higher level activities
() of mechanisms as a whole (S) are realized by the organized activities ()
of lower level components (Xs), and these are, in turn, realized by the activities
() of still lower level components (Ps)'. On Glennan's approach, a mechanism
is a complex system, whose parts are so organized that they interact to yield
a characteristic behavior (see Glennan, 1996). In a recent version of his view,
'A mechanism for a behavior is a complex system that produces that behavior
by the interaction of a number of parts, where the interactions between parts
can be characterized by direct, invariant, change-relating generalizations' (2000,
p. S344). We do not intend to explore the differences in emphasis between the
views, so much as the alternative they offer.
So we begin with a behavior. We might want to explain, for example, the
behavior of the bacterium,
E. coli
, and its culinary 'preferences' on various
media. For its maintenance
E. coli
needs a reliable source of energy. Sugar solu-
tions are natural choices for the experimentalist and favorites for the bacterium.
To explain how
E. coli
uses sugars, we need to see what happens as the various
sugars are consumed and processed. This is the study of metabolism. We look
