Biological and Cultural Proper Functions in Comparative Perspective 49
causal-role, accidental, Cummins, or (my preference) system function—that does not
require a history but only a role in an embedding system of some kind. For example, a
pen of entirely novel design that works as intended by its designer has the system function
of writing in virtue of fi lling that role in the system of artifacts used in that activity (paper,
pencils, erasers, etc.). The only question is whether the prototype pen works well enough
to be substituted for a regular pen in writing notes or signing documents; not whether it
has been used for such activities in the past and reproduced on account of this use. On the
other hand, if the prototype pen does not work well enough to be substituted for a regular
pen—if it is a failure as far as its intended function goes—then it has neither a system
function nor a proper function.
A common intuition among function theorists is that novel prototypes must have
proper functions, and a common solution is to argue that the intentions of designers
establish them (Millikan 1999; Vermaas and Houkes 2003). But as I have argued (Preston
2003), this seemingly plausible and innocuous move has serious repercussions. In particu-
lar it threatens the widely accepted and important distinction between historically
conditioned function (what something is supposed to do, i.e., its proper function) and
current function (what something in fact does on a given occasion, i.e., its system func-
tion). I argued further that without the distinction between proper function and system
function it would be impossible to appropriately describe and account for the social pro-
cesses involved in the use, production, and reproduction of artifacts. I will not rehearse
these arguments here, but I stand by them. So in appealing to a history of use and repro-
duction to pick out the proper functions of artifacts, I am not carelessly ignoring the alleged
proper functions of novel prototypes. Rather I wish to assert that such artifacts have no
proper functions.
However counterintuitive this conclusion may seem, there is a bright side to it from the
perspective of comparing biological and cultural functions. I have argued here that cultural
selection is like natural selection in crucial respects, but that precisely these similarities
mean that cultural selection does not pick out the proper functions of artifacts any more
than natural selection picks out the proper functions of biological traits. Furthermore, I
have argued that although contributions to fi tness leading to reproduction may well pick
out biological proper functions, fi tness cannot be used to pick out the proper functions of
artifacts. Cultural fi tness is at best only vaguely analogous to biological fi tness; and, more
importantly, cultural reproduction is often independent of fi tness in ways that biological
reproduction is not. Finally, I have recommended an approach that instead looks to patterns
of use leading to reproduction to pick out the proper functions of artifacts. And this recom-
mendation preserves a signifi cant analogy between biology and culture—the centrality of
processes of reproduction that ensure the continuing production of tokens of standardized
types, while allowing for variation leading to new types of things.