The Inherent Normativity of Functions in Biology and Technology 121
organs fi ts this picture, as we apparently see them as supporting the purposes of organisms,
similar to the way external objects support our lives.
The justifi cation for speaking of a “good” heart seems additionally to be inspired by the
vision of being able to “improve” upon existing hearts. And in fact, thanks to medical
technology, the dividing line between biological organs and artifacts is blurring, now at
an increasing rate. Open-heart surgery and the implementation of pacemakers are two ways
to “improve” the performance of a poorly functioning or “bad” heart. Our medical practice
invites us to incorporate our organs into the artifactual realm.^24
However, there are major obstacles in extending the analysis of artifacts in this direc-
tion, along the lines followed in this chapter, in that it would require us to make sense of
the idea that we use our bodily organs to realize certain goals. On one hand, it may be all
right to say that I use my hand to pick up a pencil from my desk, since the “handling” of
my hand answers to much the same features as my handling of the pencil. On the other
hand, it makes no sense to say that I use my heart to circulate my blood through the arter-
ies and veins of my body. I cannot refrain from doing so, hesitate in doing so, do it slowly
or carelessly, nor any of the other things that characterize the intentional action of using
something. Linguistic practice is fairly accurate in this; the only borderline case is perhaps
“use your head.”
Talk of better and worse livers, hearts, and so forth, can be justifi ed only insofar as we
regard these items as falling within the engineering domain, as being quasi-artifacts, so to
speak, being amenable to improvement and redesign. When we abandon this perspective,
there is no normativity to be found in nature and no normativity inherent in function
talk.
Consequently I conclude that the advantage that the PF and SE theories have always
claimed over their rival theories—of uniquely being able to account for malfunction—is
illusory. The normativity of artifacts is not what it has seemed to most people. We have
functions as causal contributions to a system capacity, functions as causal contributions
to surviving and reproducing behavior, functions as causal contributions to designs, func-
tions as causal contributions to a selection process, functions as intended behavior, and
functions as behavior made use of, but none of these functions uniquely supports norma-
tive “ought to” statements—applied to functional items rather than persons—to the extent
that we see a point in uttering such statements. This should not be seen as an invitation
to add a new function concept to this collection, functions as normativity-supporting
behavior. No defi nition in naturalistic terms of such a concept seems to be possible.
Besides, what work would this concept do, apart from covering our normative talk in
certain situations? I argue that we can account for our use of this normative talk in other
ways.
The etiological theories remain strong contestants, however, in the battle for the “true
account” of function, as far as the explanatory role of function attribution is concerned. I
do not discuss this role in this chapter, nor, therefore, whether there is such a “true