Natural Selection and Causal Productivity - Mechanism and Causality in Biology and Economics - page 147

a level: the death of the French literary critic, Roland Barthes, who was struck by a
laundry truck while crossing a Paris street on the way home from meeting with
then-President Fran¸ois Mitterrand (Glennan 2010a ). Other defenses of the new
mechanistic philosophy, such as Bechtel and Abrahamsen ( 2005 ), Craver ( 2007 ),
and Craver and Bechtel ( 2007 ), emphasize both the multilevel nature of mechanis-
tic explanation and the importance of situating of a mechanism in its context (see
especially Craver 2001 on this latter point). So, how can I claim that accounts under
the new mechanistic philosophy are decompositional?
A distinction made by Salmon between etiological explanations and constitutive
explanations is useful in answering this question. Salmon states that both types
of explanation are “thoroughly causal.” However, according to Salmon, etiological
explanations “explain a given fact by showing how it came to be as a result of
antecedent events, processes, and conditions” ( 1984 , p. 269). Constitutive explana-
tions, on the other hand, show “that the fact-to-be-explained is constituted by underly-
ing causal mechanisms”; they exhibit “the internal causal structure of the
explanandum” ( 1984 , p. 270). I would suggest that etiological explanations are “at” a
level, whereas constitutive explanations cite lower levels by citing the parts that make
up the whole (i.e., they are decompositional). According to Salmon, we can expect that
most explanations will have both etiological aspects and constitutive aspects, but we
should also recognize that there are some cases of pure etiological explanation and
some cases of pure constitutive explanation. Salmon gives the explanation of “the
presence of a worked bone that is thirty thousand years old in an Alaskan archaeological
site” as an example of a pure case of etiological explanation, noting that “to explain this
fact, it is not essential to look for the causal constituents of the bone” ( 1984 ,p.270). 5
In general, the newmechanists seem to agree with Salmon that most explanations
include both etiological and constitutive aspects; however, whereas Salmon's
account emphasizes etiological explanations, the new mechanist philosophy
emphasizes constitutive ones. Indeed, Craver explicitly distinguishes his project
from Salmon's in exactly this way, stating, “The variety of explanation that I am
interested in is constitutive (or componential) causal-mechanical explanation: the
explanation of a phenomenon, such as the opening of a Ca 2+ channel, by the
organization of component entities and activities” ( 2007 , p. 8). Similarly, Bechtel
acknowledges that “mechanistic explanations are inherently reductionistic insofar
as they require specifying the parts of a mechanism and the operations the parts
perform” ( 2011 , p. 538). Thus, Darden's example of the mechanism for the segrega-
tion of genes seems to be the exception rather than the rule, and Glennan distinguishes
ephemeral mechanisms from his primary account of systems mechanisms, which
do involve the decomposition of a system into parts (Glennan 2010a ,p.258). 6
5 He also states, “Microphysics is invoked to ascertain the age of the bone, but not explain its
presence in the site where it was discovered” ( 1984 , p. 268).
6 Illari and Williamson ( 2010 ) also seem to understand MDC mechanisms as being decompositional.
Kuorikoski ( 2009 ) usefully distinguishes between mechanisms that involve decomposition and those
that do not; he agrees with Skipper and Millstein ( 2005 ) that natural selection falls into the latter
category. (Thanks to Till Gruene-Yanoff for the pointer to the paper by Kuorikoski).