Mechanisms and Laws: Clarifying the Debate - Mechanism and Causality in Biology and Economics - page 132

why would one ever suppose that it would be impossible for them to obtain among
causally connected random events (whatever sense we can make of that notion)? 7 It
seems there is no interpretation of the idea of irregularly behaving components that
sustains even the negative existential thesis entailed by Leuridan's modal claim that
no (c)P-regularity can be produced by irregularly behaving components.
Perhaps what Leuridan means to claim is that very few of the mechanisms
described in biology textbooks explain higher-level (c)P-regularities without
appealing to regularly behaving components. If one wants to discuss the kinds of
mechanism that biologists typically study, then one must acknowledge that there
are true p-generalizations about the components of mechanisms. True enough. But
this claim is entirely independent of the ontological thesis that cs-mechanisms
depend on lower-level regularities. And no mechanist denies that there are true
p-generalizations about the components of mechanisms.
If the argument does not work for cs-mechanisms, it certainly will not work
for mechanisms in general (as his title and introduction suggest). That is, it
cannot establish, as Leuridan claims, that there can be no mechanisms without
microlevel stable regularities. It seems one-off mechanisms (the “Salmon/Railton
mechanisms” discussed above) might well work without microlevel stable
regularities. Such mechanisms probably would not be so scientifically interesting,
and we might never know about them, but they might well exist.
Leuridan might, at this point, have entered a long debate about the regular
character of causality. Perhaps he could endorse the view that the components in
a mechanism can properly be said to causally interact with one another only if there
exists a p-regularity relating events of one type to events of another type. If all
mechanisms have interacting parts, and if there can be no interactions among parts
without p-regularities, then there can be no mechanisms without p-regularities.
That's certainly an ontological thesis, and it's one with a grand tradition. It's also a
view that some mechanists (such as Bogen 2005 , 2008 ; Machamer 2004 ; Darden
2006 ) explicitly challenge.
As we mentioned above, Bogen ( 2005 , 2008 ) argues that causation and regular-
ity are conceptually distinct. One set of Bogen's arguments turns on the implicit
thesis that causation is local (or, in other words, intrinsic): that whether A causes B
depends on facts about A, B, and their relation to one another and does not depend
on how other A-type things and B-type things behave when they interact. What
matters instead is whether A and B are connected by some determinate sort of
activity. One need not buy the metaphysics of activities to appreciate the intuitive
pull of locality. Imagine a world composed only of two billiard balls traveling
through space-time toward one another until one day they clack together and fly
back in the directions whence they came. Whether they interacted would seem not
to depend on whether any other billiard balls ever meet or on whether the same
billiard balls ever meet again; neither is true in the world we are considering. The
causal interaction is a fact about them and them alone (i.e., an intrinsic feature of
7 Again, note that Leuridan is operating with a most unorthodox notion of “irregularity.”