56 Françoise Longy
However, such traits also deserve, it seems, to be called “artifactual” because without
voluntary human intervention the features with the desired effects would not have appeared,
would not have been selected long ago, or would not have been recently maintained. Here
we need to be a little more precise about the meaning of the expression “artifact function”.
Vagueness does not pose much of a problem when considering inanimate objects, since
they can have only artifact functions (they may have functions only if they are used or
produced by intentional agents). But, with biological items the situation is more complex
since there is more than one option.
First and foremost, what does artifact mean? Sperber convincingly defends the claim
that artifact is a family resemblance notion rather than one that “could be defi ned precisely
enough to serve a genuine theoretical purpose” (2007: 124). In fact classical defi nitions
of artifacts present a choice between two possible conditions that are not equivalent: 1) to
have been “intentionally made or produced for a certain purpose,” and 2) to be “the product
of human actions” (Hilpinen 1999). The second condition is the less restrictive of the two,
since human action does not necessarily presuppose clear purposes. However, the fi rst
condition is not very restrictive either. It just supposes some purposefulness in making or
producing something. According to that condition, an entity, a feature, or a function will
be artifactual if it results from purposeful human action, even if a clear anticipation of the
result is missing. Moreover, as we know, a series of limited short-term aims may produce
unforeseen long-term effects.
So unpredictability can go hand in hand with artifactuality. Unforeseen new features
and new functions may indeed appear thanks to what Darwin considered an unconscious
form of artifi cial selection, the form that “results from everyone trying to possess and
breed from the best individual animals” (1859: 34). Besides, there is a continuum of
intermediate stages from unconscious artifi cial selection to completely planned artifi cial
selection, from the unconscious domestication of some species in prehistoric times to the
consciously pursued aim of nineteenth-century breeders or of present GMO producers. All
this supports a broad application of the term artifact in the biological domain, one that
covers the whole range of human interventions. So in the biological realm, just as in the
nonbiological realm, artifactual functions should include intended functions (effects that
have been clearly anticipated and have been obtained as the result of consciously planned
interventions) and nonintended functions (unanticipated effects that have been obtained at
the end of a series of short-term oriented interventions).
Let us continue to clarify the issue by taking a closer look at nonintended functions.
When we considered the nonintended functions of inanimate objects, we spoke of socio-
culturally selected functions. It is true that many of these functions can probably be
accounted for by SEL with some form of sociocultural selection replacing natural selec-
tion. For instance, in the story of manufactured artifacts, sociocultural selection through
economic competition plays an important part. Is this, however, the case with all non-
intended functions of inanimate objects? Sperber claims that cultural functions do not