The Inherent Normativity of Functions in Biology and Technology 113
designed,” by which components come to be seen to exist “for” their functions, that
grounds such expectations for Davies. It seems to me, however, that Davies is overly
restrictive in this. Not only is there no limitation to complex systems built into any of the
theories of function, CR-type or PF-type, but my examples show that we are inclined to
use the “ought to” language expressing justifi ed expectations much more widely than we
attribute functions.
7.5 The Arbitrariness of the Attribution of Normative Functions in Biology
In direct relation to function attributions, on the other hand, the corresponding normative
statements occur less widely than might be expected. If we witness the birth of a kitten,
we expect the kitten’s heart to beat and circulate the kitten’s blood just as much as we
expect the kitten itself to live. We can say “The kitten ought to go on living” just as well
as we can say “The kitten’s heart ought to go on beating.” If we watch some fi shes in a
pond that is frozen over, or if we put a butterfl y in the refrigerator in order to be able to
take photographs of it more easily afterward, we can say, for instance, to a protesting child,
“It ought to survive.” This is diffi cult to understand if such normative talk is linked to the
attribution of function. Biologists are extremely reluctant to attribute functions to organ-
isms, and even more reluctant to come up with the normative judgments that would
accompany such function attributions.
In fact normative judgments are already withheld from, or at least only sparingly applied
to, the other biological items apart from organs that are eligible for function attributions.
Functions are also applied to traits, like the forkedness of the snake’s tongue or the dif-
ferent maturation times of the pistil and the stamina of the fl owers of monoecious mono-
clinous plants, and also of forms of behavior, like the mating dance of the stickleback or
the mantling by which birds of prey hide their catch from the hungry eyes of other birds.
Insofar as the normativity of traits or behaviors is concerned, we are far more reticent in
pronouncing evaluative statements. Do we call the fl owers of a white deadnettle (Lamium
album) in which the maturation times of pistil and stamina coincide malfunctioning? It
sounds odd to do so.
In the social sciences we see a distinction in the way function attributions are subject
to malfunction and evaluative statements that refl ects the differences between artifact
functions and biological functions. Many social institutions are the result of design and
subsequent implementation, monitoring, and adjustment, and can accordingly be seen as
social artifacts, for example, the legal system of a country. Of such institutions it is entirely
in order to say that they function poorly or malfunction. In the case of social practices that
have evolved “organically,” this is no longer true.^15 If the Hopi rain dance does not con-
tribute to social cohesion, does the dance malfunction? Such issues make clear how enor-
mously questionable such claims are, even apart from the more general problem discussed