The Inherent Normativity of Functions in Biology and Technology 119
for cutting, because in the absence of alternatives, using this knife still may be the best
option if one is pressed hard enough.
As a second-order fact, the fact that “This is a malfunctioning knife” can be subsumed
under the second-order fact “This is useless as a knife,” which expresses the fact that
because of its physical characteristics, any person has a reason not to use this object for
cutting. It is the fi rst-order fact of the particular features the object has that distinguishes
between an object’s being a malfunctioning knife and its not being a knife at all and lacking
any property that would give it knifelike capacities. Among the features that a malfunc-
tioning knife has and an object that is not a knife at all lacks is the historic feature of
having been designed as a knife.
Similarly, the second-order fact that “This is a good knife” can be subsumed under the
broader second-order fact “This is useful as a knife.” Something that is not a knife, in the
sense of not having been designed as a knife, can still have features that make it the case
that someone who has a need for cutting has a reason to use it for cutting. For most arti-
facts, if something that was designed for the particular task at hand is available, provided
it is in working order, there will usually be a conclusive reason to use this item, rather
than an arbitrary object that happens to be able to do the job though it was not designed
for it. Other circumstances where we must make do with what is available are nevertheless
far from rare.
The mere attribution of function, “This is a knife,” cannot be subsumed under the fact
that “This is useful for cutting,” because of the existence of malfunction: a particular knife
may be a broken or otherwise useless knife. It may be thought that “This is a knife”
expresses the second-order fact that because this object has certain features (including its
design and manufacture history), someone who has a reason for cutting has a reason to
use a token of the narrow, design-historical type to which it belongs.^21 It is questionable,
however, whether this is indeed true. If most tokens of the type are malfunctioning, I would
say this is false. My suggestion is to read the statement “This is a knife” as expressing a
different sort of normative fact, which adds reasons for believing something to the reasons
for doing that dominate the previous cases. On this reading, “This is a knife” expresses
the normative fact that because this object has certain features (including its design
history), if one has a reason for cutting, one has a reason to believe that one has a reason
to use this object for cutting. This is weaker than saying that one is justifi ed in believing
that one has a reason to use it for cutting, because there might be other reasons that speak
against this belief, such that on the balance of reasons one should not believe that one has
a reason to use it for cutting, and believing so would then not be justifi ed.^22
So a statement of the form “This object is of functional kind F” expresses both a fi rst-
order fact—the object was designed for the purpose of F-ing—and a second-order fact—
the object’s being designed for the purpose of F-ing gives one a reason to believe that it
will be useful for F-ing. The fact that it was designed for the purpose of F-ing makes a
difference to the question what to believe and, subsequently, how to act for people who