Functions 545towards some normal end state, and only by reference to this end state can we
reckon some developmental trajectories truncated, and others of the proper length.
By hypothesis we cannot understand ‘normal’ here to mean ‘statistically typical’;
instead, normal development means proper development. Proper development
itself is then understood to mean development that proceeds as though it were
guided by some successful plan.
If we think that goal-directedness is merely a projection onto nature that enables
further inquiry into the mechanism of development, it is now less pressing for us
to ground talk of goal-directedness by giving an account of which biological facts
determine that one kind of development is proper, another improper. For the
appeal to the developmental plan no longer explains development at all; rather,
it tells us only which patterns of interactions between genetic and other resources
we should seek to focus on as the primary basis for a subsequent attempt at
mechanical explanation. It gives us a way of setting the subject matter for what
will then be a causal explanation for how a fertilised egg comes to acquire new
capacities over time. It would not be a threat to biology if it turns out, for
example, that our perception of organic nature as uniquely goal directed (compared
with, say, inorganic systems in physics) derives from a projection of our own
personal striving for survival onto the organic world. That projection can still
serve an essential function in providing what Kant calls a ‘guiding thread’ for the
subsequent articulation of mechanical explanations of development. Without some
way, however it is grounded, of picking on a form of organisation as the proper
one, we could not pick out patterns of change as development, let alone explain
them mechanically. Yet once the mechanical explanation is given the need for
teleological concepts disappears, at least until we wish to draw back from a focus
on the causal interactions of matter and re-characterise some collections of this
matter as constituting organisms. Goal-directedness, on this view, is the ladder
we kick away.
8 NATURALISMAlthough Kant scholars are beginning to pay a great deal of attention to under-
standing Kant’s views about biology, philosophers of biology very rarely defend
accounts of function that take their inspiration from Kant. Part of the reason
surely lies in the difficulty of Kant’s work, and in the fact that it is hard to un-
derstand any part of it without already understanding all the rest. Yet I suspect
another reason may lie in a perception that an account of functions should be
naturalistic, and that naturalism entails that an account should give truth con-
ditions for function statements that use only biological terms. This is a mistake.
Roughly speaking, a naturalistic account of some phenomenon should not rely on
the existence of any events or objects whose existence is denied by our best science.
Projectivist accounts in ethics (e.g. [Blackburn, 1998]), for example, pass this test
because although they do not identify goodness with some property endorsed by
science, neither do they rely on the existence of properties denied by science in the