530 Tim Lewens
act as mutual catalysts and reagents, which Kauffman views as primitive living
systems:
In a collective autocatalytic set, there is no central directing agency.
There is no separate genome, no DNA. There is a collective molecular
autopoetic system that Kant might have been heartened to behold.
The parts exist for and by means of the whole; the whole exists for
and by means of the parts. [Kauffman, 1995, 275]3 THE SELECTED EFFECTS ACCOUNTIn this section I want to evaluate the most influential modern theory of func-
tions. This is the Selected Effects, or SE account. The SE account has its roots
in Wright’s [1973; 1976] account of functions, and has been developed and de-
fended in most detail by Ruth Millikan (e.g. [Millikan, 1984; 1989]) and Karen
Neander (e.g. [Neander, 1991a, b]). SE analyses of one sort or another dominate
two recent collections on functions ([Buller, 1999b] and [Amundsonet al., 1998]).
The SE account gets much of its appeal from the way it seeks to supply a bio-
logical surrogate for the natural theologians’ intelligent design. As the quotations
from Williams, and from Krebs and Davies indicate, many modern evolutionary
biologists continue to see the traits of organisms as useful tools produced by de-
sign in response to problems laid down by the environment. Elsewhere, Williams
also licenses an updated version of Paley’s argument from design: he explicitly
recommends that biologists take note of Paley’s criteria for determining when a
trait should be attributed to design rather than chance — when, that is, it is an
adaptation [Williams, 1992, 190]. Of course such biologists do not think that an
intelligent designer is responsible for these functional traits; however, if we begin
by supposing that in the realm of artefacts, functions are established through the
intentions of users or designers, then one who is liable to describe organisms in
terms of function and design needs to find a non-intelligent natural process that
can play the role of intelligent design in the organic realm. Because the whole
range of artefact talk is found in biology, and because it tends to be used in just
the same way that it is used on its home turf, we want a way of making sense of
biological functions that respects the connotations of heavy function statements
when they are applied to artefacts.
Philosophers typically find three such connotations when heavy function talk is
used, regardless of whether we are dealing with organisms or artefacts:
Function statements are explanatory: Suppose we are looking at the
radiator at the front end of a car. The questions ‘What is that for?’,
‘What is that thing’s function?’ and ‘Why is that thing there?’ all
seem equivalent in this context. When we give an answer that states
a function — ‘It’s for keeping the engine cool’ — we therefore answer
the question ‘Why is it there?’, and we thereby explain the presence