244 Tim Lewens
sense, then, natural selection was already copied from the realm of artifacts. According to
many standard presentations, selection explanations account for adaptation by telling us
that a variety of forms are produced; the ones that fi t local demands well are retained for
further modifi cation, and gradually a well-adapted system is built up. If this is how selec-
tion explanations work, then the selectionist view seems once again to be true for techno-
logical innovation, but rather too obviously true to provide much insight to the student of
technical change. (For an elaboration of this argument, see Lewens 2002; Lewens 2004:
ch. 8.)
14.2 Two Responses to the Basic Problem
There are many responses we could give to the mean-spirited antievolutionary argument
sketched in the previous section. One response consists of pointing out that we should not
expect too much from evolutionary theories of technology change. For example, an evo-
lutionary view of technical innovation should neglect neither variation nor selection. This
means that evolutionary views of innovation will credit a wide range of factors with
explanatory value. On the variation side, these factors will include the current state of
technical know-how, the nature of dominant design heuristics, and the material basis of
artifact manufacture, all of which help to determine what alternatives are available for
selection to act upon. On the selection side, these factors will need to include the competi-
tive environment a given technology fi nds itself in, as well as consumers’ conscious and
unconscious desiderata for technical artifacts. Since the latter can, in principle at least,
take many forms, evolutionary views are likely to lead to skepticism of any grand theory
that sticks its neck out regarding the general determinants of artifact success. Certainly
we should not expect the most useful, or the best-designed, or the cheapest, artifacts always
to be the ones that succeed.
Depending on one’s standpoint, then, it may be a strength rather than a weakness of the
evolutionary view that it offers few clues in itself regarding which factors are most impor-
tant in explaining technological change. Evolutionary theories provide useful standpoints
from which to articulate the rashness of monistic, or “deterministic,” theories of technol-
ogy change, regardless of whether such monistic theories locate determining power in
technologies themselves or in the societies that produce and make use of them.
Note that what is distinctive about evolutionary views here is precisely their lack
of distinctiveness: they tend to partially endorse many aspects of competing theories of
technological change. Consonant with this, the culmination of a series of discussions of
evolutionary theories of technical innovation by John Ziman (2000) and collaborators is
the modest claim that evolution provides a unifying “paradigm of rationality” for describ-
ing technology change (Ziman 2000: 313). I take this to mean that evolutionary theories
provide a framework in which contentious issues familiar to historians and economists