Functions 529sure, but not because they make compensating internal adjustments. Whales are
also self-repairing; if it is injured, a whale can grow new tissue, for example. But
watches cannot repair themselves in this way. This way of thinking of organisms
views them not so much as bags of good tools, but instead as analogous to people
who are striving towards some goal. The organism is an object whose goal is to
attain and then to maintain its integrity (or perhaps its capacity to reproduce) in
the face of a changing environment. A person, too, will change their behaviour
to get what they aim at if the environment is changing. If you are trying to lose
weight, you will find a new gym when the one you have been going to closes down.
And if a person’s plans are blown off course (their crash diet repays them with a
swift weight gain) then, so long as they still have their goal, they will pick them-
selves up and set off after their goal again. If we stress the differences between
whales and watches, and instead focus on the similarities between whales and
weight watchers, we are likely to arrive at anagent-based modelof heavy function
talk.
It is no coincidence, then, that if one is fed on a diet of examples from modern
Anglo-American evolutionary biology one is likely to construct an artefact-based
model of function, while if one is instead fed on a diet of developmental biology
one is more likely to construct an agent-based model of function. The study of
development makes goal-directedness the most tempting focus for an account of
function. The study of evolution instead encourages us to conceptualise organisms
as sets of tools for solving environmental problems. Thus George Williams, an
influential figure among contemporary adaptationists, makes clear allusion to the
artefact model when he begins his celebrated workAdaptation and Natural Selec-
tionwith the words, ‘Evolutionary adaptation is a special and onerous concept
that should not be used unnecessarily, and an effect should not be called a function
unless it is clearly produced by design and not by chance’ [Williams, 1966, vii]. It
is interesting to contrast this with some introductory remarks in C. H. Wadding-
ton’sThe Strategy of the Genes, a work that focuses strongly on development.
Here Waddington does not focus on adaptation to environment as characterising
the organic realm, but rather he mentions both the dynamic manner in which the
form of an individual organism is maintained, and:
...the second major peculiarity of living things, their ‘directiveness’
...Thisreferstothefactthatmostoftheactivities of a living organism
are of such a kind that they tend to produce a certain characteristic
end result. The most inclusive type of end-result, it is often stated, is
the continued life of the organism; actually its reproduction, and the
passing of its hereditary qualities to the next generation, should be
regarded as a still more general goal. [Waddington 1957, 2]More recently, Stuart Kauffman, known for work on the ‘self-organisation’
of complex organic systems, makes direct reference to a Kantian view of goal-
directedness when describing ‘autocatalytic sets’. These are sets of molecules that