Mechanisms and Laws: Clarifying the Debate - Mechanism and Causality in Biology and Economics - page 124

question is not whether mechanisms can replace laws, simpliciter. Rather it is
whether mechanisms can replace p-laws and, correlatively, whether descriptions
of mechanisms can replace p-law statements in our thinking about science.
Yet Leuridan's thesis is narrower still. He distinguishes two “kinds” of
mechanism: complex system mechanisms (cs-mechanisms) and Salmon/Railton
mechanisms. Leuridan characterizes cs-mechanisms as stable configurations of
robust objects that produce stable behaviors ( 2010 , p. 319; see also Glennan
2002 , pp. 344-46). 1 Leuridan does not define Salmon/Railton mechanisms, except
to say that they involve causal processes and causal interactions ( 2010 ,p.319).
However, if we follow Glennan, they might be understood as “sequences of
interconnected events” or “a chain or web of events leading to a particular event”
such as “a boy hit a baseball; the baseball ricocheted off the tree and crashed into the
window” (Glennan 2002 , p. 345). Salmon/Railton mechanisms are singular causal
chains. Substituting into Leuridan's title question yields something closer to the
question he in fact addresses: “Can cs-mechanisms really replace p-laws?” Leuridan
is not always clear to distinguish this ontological question from its epistemological
twin: “Can models of cs-mechanisms replace p-law statements?” 2 But in either case,
he concludes they cannot.
More precisely, Leuridan presents four theses, two of which he describes as
ontological, and two of which he describes as epistemological:
First, [cs-] mechanisms are ontologically dependent on stable regularities. There are no
[cs-]mechanisms without both macrolevel and microlevel stable regularities. [L1]
Second, there may be stable regularities without any underlying [cs-]mechanism. [L2]
Third, models of [cs-]mechanisms are epistemologically dependent on pragmatic laws.
To adequately model a [cs-]mechanism, one has to incorporate pragmatic laws. [L3].
Finally, pragmatic laws are themselves not epistemologically dependent on mechanistic
models. They need not always refer to a mechanism underlying the regularity at hand. [L4]
(Leuridan 2010 , pp. 318-19)
We have inserted the qualification to cs-mechanisms specifically, given that
Leuridan offers no argument even purporting to show that non-cs-mechanisms are
dependent upon p-laws (see Bogen 2005 , 2008 ), a gap to which we return below.
Leuridan argues that cs-mechanisms cannot replace p-laws in our thinking about the
ontology of science because cs-mechanisms are ontologically dependent on p-laws
(L1), but the opposite is not the case (L2). Furthermore, he claims that cs-mechanisms
cannot replace p-laws in our thinking about epistemology because models of
cs-mechanisms are epistemically dependent on p-laws (L3) and not vice versa (L4).
1 The full passage is “Contrary to Salmon/Dowe mechanisms, complex systems mechanisms
(cs-mechanisms) are robust and stable. They form stable configurations of robust objects, and as
a whole they have stable dispositions: the overall behaviors of these mechanisms” (Leuridan 2010 ,
p. 319).
2 On closer inspection, Leuridan's question is still too imprecise since it does not specify the
purpose for which cs-mechanisms are intended to replace p-laws (or vice versa). Mechanisms and
stable generalizations serve many functions in our thinking about science. Perhaps mechanisms are
useful for some philosophical purposes and laws are useful for others.