interpretations of function ascriptions (the intentional, etiological, and causal-role theories
of functions) leads him to the conclusion that each of them has its own problems when
accounting for the normativity of functions: neither facts about intentions, nor selection
histories, nor facts about causal roles can ground normative claims with regard to func-
tions. Next he analyses in detail three kinds of relations that might form the origin of the
normativity associated with functions, namely means-ends relations, part-whole relations,
and type-token relations. Means-ends relations differ from nonnormative causal relations
in that means ought to or are supposed to contribute to bringing about the ends. Part-whole
relations may have a normative dimension if the whole is a hierarchically organized system
with a good of its own. Finally, tokens may instantiate a type better or worse, so the type-
token relation is of a normative nature. Insofar as these relations play a role in function
theories, the normativity of functions may be grounded in these relations.
Franssen discusses the role that the idea that functions are inherently normative
plays in the debate on adequate theories of functions. One of the chief touchstones when
assessing function theories is how they deal with the alleged normativity of functions.
Etiological theories are generally taken to be able to account for normativity, whereas
causal-role theories deny that there is anything normative about functions. So a question
arises about the list of adequacy criteria for function theories: should it include the criterion
that a theory of functions has to account for the inherent normativity of functions? To
answer this question Franssen analyzes in detail the nature of normative statements with
regard to functions. He argues that the normativity traditionally associated with functions
derives from human intentionality; it is either related to the justifi cation of beliefs about
functional items or to reasons for specifi c actions with regard to such items. So this nor-
mativity is not inherent to functions. This means that the advantage of etiological theories
over causal-role theories with regard to accounting for the normativity of functions is
illusory.
Davies argues that although most of us have the intuition that the parts of living things
are supposed to fulfi ll certain functional tasks, we should give up this intuition and stop
talking about functions. He asks why we are moved to theorize about the concept of nor-
mative functions and concludes it is because we are conceptual conservatives regarding
the concept “purpose.” Conceptual conservatives are committed to preserving or otherwise
“saving” concepts that strike us as especially important, including our concept “purpose”
as it applies to organisms. And yet insofar as the genealogy of our concept of “normative
functions” traces back to a largely theological worldview, we now regard it as false or
unpromising, and insofar as we are psychologically constituted to apply this concept with
undue generosity, we ought to relinquish the orientation of the conceptual conservative
with respect to normative functions on the grounds that it diminishes rather than facilitates
the growth of human knowledge. This casts doubt upon the main theories of functions in
the philosophical literature, except for the theory of systemic functions that eschews the
alleged normative dimension of biological purposes.
90 Part III