How Biological, Cultural, and Intended Functions Combine 57
necessarily suppose something similar to natural selection, and I agree with him (2007:
128). He insists on not needing replicators such as genes. One can also dispense with a
true mechanism of selection that supposes competition among variants.^10 Consequently,
in his defi nition of a teleofunction, Sperber supplants the notion of selection with the
broader one of propagation: “an effect of type F is a teleofunction of items of type A just
in case the fact that A items have produced F effects helps explain the fact that A items
propagate.”^11 Natural selection is then simply one of the mechanisms that can explain
propagation and evolution.
This new defi nition, which is meant to apply indiscriminately to biological, artifactual,
and cultural teleofunctions, needs some clarifying remarks. Above all, a new notion has
appeared—that of a “cultural function.” From what Sperber says, one can extract the fol-
lowing characterization: a function is cultural if the propagation on which it relies involves
at some stage some mental representations. For instance, domesticated wheat has the cul-
tural function of nourishing humans because the cultivation of wheat for this end propa-
gated thanks to mental representations. Indeed, it propagated because some farming
practices were consciously imitated, because people thought and talked about how to grow
wheat, because books about growing wheat were written and read and so on and so forth.
First, it should be noted that the replacement of a selectionist characterization of functions
by a propagationist characterization has consequences only for cultural functions. As a
matter of fact, the only biological mechanism that may explain a propagation that is
“helped by the fact that Xs do F” is natural selection. Second, artifactual and cultural
functions, which overlap to a large extent, remain somewhat distinct. In fact neither sort,
apparently, includes the other. Behaviors typically have cultural functions, but behaviors
are not artifacts.^12 On the other hand, purely intended functions of artifacts are not cultural
functions, at least not of the propagated sort. In fact it is generally admitted that an artifact
has a purely intended function when it leaves the hands of its creator and has not yet been
reproduced and diffused.
Now that these clarifi cations have been made, let us resume our investigation of the
functions of a cultivated or domesticated species. A good example, analyzed by Sperber,
is that of wheat and barley whose cultivation began thirteen thousand years ago by sowing
part of the collected seeds in chosen locations instead of simply eating all there was.
Sperber explains:
It can be quite advantageous for a plant to have a large proportion of its seeds used by humans as
food, provided that the remainder of the seeds serves the goal of reproduction and dispersal in a
particularly effi cient way. When this became the case for various species of cereals, feeding humans
became a biological teleofunction of the seeds, that is, an effect that contributed to the greater
reproductive success of varieties of cereal providing better food. Both the feeding function and the
reproduction function of seeds are simultaneously biological and cultural/artifactual functions of
cultivated cereal. The plants take biological advantage of their cultural functions and humans exploit
culturally, and more specifi cally economically, some of the biological functions of the plants. There