Changing the Mission of Theories of Teleology 25
tinct kinds? These “natural” objects have been modifi ed by human activity longer than we
can remember.
Further blurring the natural-artifi cial distinction, some artifacts have biological charac-
teristics. Consider computer viruses—they engage in self-replication, supposedly an exclu-
sive biological trait, so much so that we use biological terms like virus and spread to
describe them. Yet they are clearly designed by human beings, and have teleological func-
tions, though not very admirable ones. Add to that the instances of animals creating tools:
apes modifying branches, ants building anthills, bees making beehives and honeycombs,
otters smashing shells against rocks they put on their bellies as they fl oat, and many others
(see fi gure 2.6). Then there are the various cases of cats, apes, and elephants putting brush
to canvas (or walls) and producing paintings—artworks (allegedly) that are “artifacts,” but
not human-made. The old view of man as the only animal that uses tools has been deci-
sively debunked by more than thirty years of new discoveries of animal behavior. Of
course animal tools are not as sophisticated as human artifacts, but that is a matter of
degree, not kind. So the hard-line distinction between natural objects and human artifacts
has become blurred, and the hard distinction between natural and artifact functions blurs
with it.^2
2.2 DON’T Let Teleofunctional Theories Neglect Contexts
One aspect of teleological functions that seems often to be neglected is context. Objects
develop or are given functions in a context. Cummins (1975) very clearly makes his notion
of function dependent on context. It was also a central feature of Millikan’s LTOBC—she
even gives us technical terms—“Normal conditions” and “Normal explanations.” But
other theories spend too much time focused on trying to defi ne function, and pay too little
Figure 2.6
A chimpanzee using a rock as a tool.