Mechanisms and Laws: Clarifying the Debate - Mechanism and Causality in Biology and Economics - page 125

Before we address Leuridan's arguments, it is necessary first to set the record
straight. When Leuridan asks “Can cs-mechanisms really replace p-laws?” the word
“really” suggests that somebody has claimed that they can. Is this true? Do
mechanists really insist that scientists can discover, explain, predict, and control
the action potential, heredity, long-term potentiation, natural selection, and neuro-
transmitter release (to name a just few of the lengthy examples that mechanists have
discussed) without forming generalizations about them? Do mechanists think that
the Hodgkin-Huxley model of the action potential, the theory of evolution by
natural selection, and the current models of long-term potentiation and neurotrans-
mitter release make no appeal to regular occurrences? In a word, no. They are quite
explicit on this matter (see, for instance, Bechtel and Richardson 2010 , p. 232;
Glennan 1996 , p. 52, 2002 , p. 345; Machamer et al. 2000 , p. 3, p. 7; Bechtel and
Abrahamsen 2005 , fn. 1, 437; Craver 2007 , Ch. 3, pp. 233-34). James Bogen
( 2005 ), the mechanist most critical to the role of generalizations and regularities
in our thinking about causation, stresses at great length the importance of Mitchell's
treatment of p-law statements (and the regularities they describe). He also
emphasizes the many epistemic roles that generalizations play in the search for
mechanisms (cf. Bogen 2005 , p. 401):
(a) to describe the phenomenon to be explained;
(b) to suggest and sharpen questions about causal mechanisms;
(c) to describe constraints on acceptable mechanistic models;
(d) to measure or calculate quantities relevant to the mechanism;
(e) to support inductive inferences without which mechanisms could not success-
fully be studied;
(f) to support extrapolation of mechanistic knowledge to new cases;
(g) to design effective experiments to test mechanisms;
(h) to simulate the behavior of mechanisms.
The list could no doubt go on. In short, no mechanist denies that biologists
search for regularities and routinely formulate generalizations (p-law statements)
that can be used for prediction, explanation, and control of phenomena. Indeed, it is
hard to see how any significant human activity could be pursued without discover-
ing and representing (in some sense) such regularities. The mechanist claims simply
that it is useful to ask further about the material structures those generalizations
describe and about how this affects the various tasks scientists perform. In many
areas of science, scientists seek to describe mechanisms in order to explain, predict,
and control phenomena. If one places the idea of mechanism at the center of one's
thinking about those sciences, one suddenly sees p-laws in a new light, with new
roles to play (compare Bogen's list to Leuridan's emphasis on prediction, explana-
tion, and control). The question is not whether biological phenomena operate in
accordance with p-laws or exhibit p-regularities but rather how the search for those
regularities fits into the central aim of describing mechanisms.