While it may be, as Fodor has said, that "it wasn’t God that
Darwin killed, it was Mother Nature," there continues to
be a need to ensure that we don’t fall back into outmoded and
mythological ways of thinking. Davies’ book is a sustained attack
on what he sees as one such rearguard action, ironically one that
places Darwin’s own theory of natural selection at its core.

Biological thought has always revolved around the notion of function.
Recent philosophizing about biology has often attempted to show
how this notion can be regarded as naturalistically legitimate,
how its apparent appeal to purposes or the intentions of some
sort of designer can be explained away. Since the work of Cummins
and Wright in the mid-1970s, it has been common to contrast two
general approaches to the analysis of function in biology, associated
with these two authors. A Cummins view (a systemic account)
emphasizes the idea that for something to have a function is for
it to make a contribution to a wider system – so, to take a standard
example, the fact that one function of the heart is to pump blood
is seen as telling us that the heart’s pumping blood is its contribution
to a wider system for the distribution of nutrients around the
body. A Wright view (a selective account) on the other
hand emphasizes the history that has led to us having hearts:
ancestors with hearts that pumped blood well dominated those that
lacked hearts altogether or whose hearts did not pump so efficiently,
so today we have hearts that pump blood.

It appears that one problem with systemic accounts is their "promiscuity"
– hearts do lots of other things in other systems (e.g. make noises
that doctors use for diagnosis) but we are not inclined to think
of these as functions of the heart. And here a selective account
has an obvious answer: hearts were selected because of their role
in pumping blood but were not selected for their capacities to
make noise. Selective accounts also appear to be able to ground
our idea that organs can malfunction, that there are norms of
performance in the nature of things: hearts are the outcome of
a selective history focused on pumping blood, so pumping blood
is what they are supposed to do, so we can objectively judge that
something has gone wrong when a heart is unable to pump blood.

Davies sets out to show that these appearances are false. One
line of argument is that functions understood systemically are
unavoidable. Before a biologist starts telling a story about selective
history (and Davies reminds us that getting evidence to support
such stories is far from straightforward), he or she must have
identified something that at least has a Cummins-style systemic
function here and now. Although it is possible that every specimen
of a species should begin to malfunction, if we imagine coming
upon a totally alien life-form for the first time we would not
in such a case have any ground for identifying the proper function.
(In not so alien cases, structural similarities might well guide
our thinking – a cave-dwelling species of animal has things that
are in the right place to have been eyes.) Davies further argues
that the typical explanatory focus of selective function accounts
– the persistence of a feature in a population – can be handled
systemically by treating the population itself as the relevant
system and taking selection as a causal process within such a
system, so the systemicist can give us all the historical explanations
promised by the selective account.

To counter the promiscuity objection, Davies claims that to invoke
a function is to think in terms of a hierarchical system: a function
is a contribution to a higher-level capacity of a system. He seeks
to show that some of what critics have claimed must be taken as
systemic functions are not functions at all because we do not
have the right kind of hierarchical system – the water cycle is
a mere cycle, like the seasons, not a hierarchical system like
the menstrual cycle, so we need not think that clouds have a function
of producing rain to encourage vegetative growth; the solar system
lacks the right sort of hierarchy, so we need not think that one
of Uranus’ functions is to perturb the orbit of Neptune. While
this move may allow Davies to make a distinction in some cases
between functions and mere effects, he has to allow that there
are a lot of functions to be found since so much of the
world is hierarchically organized. So he is prepared to be somewhat
more promiscuous with respect to function-attributions than the
man in the street, or at least in this philosopher’s armchair.

The negative lines of argument point to weaknesses in the selective
history account – which history is relevant? does it have to be
a matter of natural selection? In general it seems that our judgment
about whether we have a function determines the answers here,
whereas it ought to be the other way round on a selective account
– and in particular to the rejection of its central idea of natural
norms. For this Davies has two main strategies. One is to ask
what possible naturalistic sense we can give to the idea that
something is "supposed" to do something, when the only
facts are that ancestral counterparts were selected because they
did it. We may create expectations, but a history of selection
cannot. Davies offers a Humean story in which our epistemic expectations
(most of the hearts we have examined have pumped blood and we
have begun to see the systemic structure involved, so we expect
this one to as well) are the basic cause of our regression to
pre-Darwinian notions. While it may be odd to think of Millikan,
Dretske or Kitcher as hankering after non-anthropocentric norms,
Davies has at least offered them the challenge to spell out what
kind of normativity a history of differential reproductive success
is meant to create, and when Plantinga has argued that proper
functioning can only be understood within a supernatural context
we are surely due an account that unambiguously dispenses with
such excesses.

The second line of attack is intended to show that the supporters
of selective accounts cannot in fact deliver on their claimed
derivation of the intelligibility of malfunctions. If natural
selection is going to operate on some entity, we need three categories:
a generic type containing variations, a narrow type not selected
for, and a narrow type selected for. (The narrow types will no
doubt contain some variation, but not as much as the generic type.)
Taking the heart again, Davies asks us what is to count as a heart
in a selective account of heart function. Present-day hearts have
been selected for pumping, so they are tokens of the third category,
a category defined in terms of its success in pumping. But when
an object that cannot pump is said to be malfunctioning it is
being identified, not as a token of the third category, but as
one of the first, generic category. And the story that supposedly
gives rise to a natural norm doesn’t apply to things in that category.
(On Davies’ own showing, such an incapacitated object may well
belong to the second narrow category. What would that tell us
to do? I think this shows how any norm we might associate with
it derives, not from the nature or history of things, but from
our own interests.)

Davies has many detailed discussions of issues relating to function
in the philosophy of biology and of what the invocation of function
by biologists actually achieves, and he well shows the resources
available if one wishes to eschew any appeal to history in one’s
analysis of function-attributions. It is perhaps unfortunate that
he has focused so narrowly on issues in biology – in a couple
of places he explicitly and perhaps too hurriedly rejects attempts
to unify biological functions with intentional design, but he
nowhere refers to discussions of the interpretation and legitimacy
of functions in sociological theorizing, for instance, which may
point to a greater role for some type of historical appeal than
he is prepared to concede.

© 2001 Ed Brandon

Ed Brandon is,
by training, a philosopher, and now is working in a policy position
in the University of the West Indies at its Cave Hill Campus in
Barbados.

Categories: Philosophical

Tags: Behavioral and Cognitive Sciences, Science (General)