Technical Artifacts, Engineering Practice, and Emergence 281
The fi rst interpretation is rather weak. Taken by itself, it would make almost any mac-
roscopic property (including functional properties) an emergent property. So in that case
the ontological interpretation of the second condition (reduction) for emergence has to
discriminate between ontologically emergent and nonemergent properties. The problem is
how to interpret the reduction relation in an ontological sense. Van Gulick (2001: 4) dis-
cusses at least fi ve different ways this ontological link has been interpreted in the context
of the mind-body problem: elimination, identity, composition, supervenience, and realiza-
tion. Many macroscopic properties will lose their ontologically emergent status under
suitable interpretations of, for instance, the composition or realization relation.
We do not enter into a discussion of these various ontological reduction relations here.
We are interested in the ontological relation between the function of a technical artifact
and its underlying physical structure and, since it is quite common to claim that technical
functions are realized by physical structures, we focus on the realization-reduction relation.
Given that technical functions satisfy the fi rst condition (in general it makes no sense to
attribute the function of an artifact to its components) but fail on the second condition
(they are claimed to be realized by physical structures), the conclusion appears to be
straightforward that they are not ontologically emergent properties relative to their physi-
cal emergence base. Take a relatively simple example of a material technical artifact, for
example, a mechanical clock. Its function is to measure time and this function is realized
through its physical structure, which consists of many parts.^8 The physical construction as
a whole has the functional property of being a clock, a property that cannot be attributed
to any of its physical parts separately. These parts, however, in their specifi c confi guration
are considered to realize the function of measuring time. This conclusion therefore seems
warranted: the function of measuring time is not ontologically emergent on the physical
structure of the clock, because it is ontologically reducible to that physical structure
through the relation of realization.
This line of reasoning, however, is highly problematic, since it would inevitably lead
to the conclusion that the functional property of, for example, being a clock is a mind-
independent, intrinsic feature of the physical construction. But technical functions are
generally considered to be mind-dependent features of the world; physical objects can
acquire a technical function only by being embedded in a context of intentional action.
On the one hand, there is a close relationship between the function of a technical artifact
and its physical structure (not any physical structure will realize the function); on the other
hand, this function is intimately tied to intentional features of the world (practices of
human action).^9
The problem with this line of reasoning concerns the interpretation of the phrase “this
function is realized through its physical structure.” This phrase is misleading in that it
confl ates a function with the physical behavior (or capacity) corresponding to that function.
What is realized through the physical construction is the physical behavior (capacity) that
is necessary for physical constructions of this type^10 to have the function of measuring