Realism and Artifact Kinds 195
over the nature and main procedure of designing. I take this rough-and-ready characteriza-
tion to be suffi cient for my purposes. What is important is that the whole process of pro-
duction must be, at least to some extent, successful. The success is evaluated with respect
to two distinct factors. Concerning the conceptual design phase, the physical constraints
posited must be adequate for the object to be able to perform the intended function,
whereas the physical realization is successful when the realized object meets the structural
requirements specifi ed by the design. It is very diffi cult to individuate general constraints
of success for both the design phase and the phase of physical realization, but there are
no doubts that in many circumstances we actually do succeed in realizing successful
designs and in producing objects according to such designs. I take it that this constitutes
suffi cient detail for the time being.
11.5.1 Two Related Problems
The phases of design and physical realization^7 can be intertwined in such a complicated
manner that it can be very diffi cult to distinguish one from the other. On the one hand,
the physical realization phase may not be a mere execution of intentions previously stated
but it may also have creative aspects. On the other hand, the phase of designing may
involve automatic procedures thanks to past experiences, either personal or collective, that
do not require intentionality. These considerations have led to the belief that these two
phases are nothing but mere abstractions. The conclusion may be drawn that it is not pos-
sible to isolate designers’ intentions for the attribution of functions, which is why it is
meaningless to attribute function according to design intentions.
A further problem concerns the attribution of functions to the component parts of arti-
facts. A possible way of producing an artifact o for a function F is by means of trial and
error, and in such circumstances the designer may ignore how o performs the intended
function F—that is, how the different parts of o contribute to the overall activity of o. The
strategy of adopting an intentional criterion for the attribution of function to the parts of
artifacts is therefore not a satisfactory strategy.
I reply to both criticisms as follows. Here I am not trying to formulate a general theory
of function attribution to artifacts, so I do not think it is a real problem if the proposed
principle does not capture some function attribution. My aim is to fi nd out which criterion
for function attribution may allow for a classifi cation of artifacts that is compatible with
a realist position. I agree that according to the present proposal if an object is produced
without a full awareness of the goal, this object will not be an item of an artifact kind.
Maybe it can still be considered an artifact according to some notions of “artifact” but
such notions are not of interest here. I do not presuppose that all the objects we consider
to be artifacts are objects of the same nature and hence objects that belong to the same
type of kinds. What is of interest here is those objects that populate our everyday lives,
objects that we judge to be—and that actually are—intentionally produced for a certain