The Open Border 237
between artifacts and organisms and to import selectionist concepts and mechanisms on
the basis of these similarities—if this suits their explanatory, descriptive, or constructive
purposes.
Moreover, in transferring concepts, researchers seek to establish a particular mode of
coexistence of the intentionalist and selectionist frameworks. Neither in EA nor in ED
does this coexistence amount to a facile eclecticism or a tangle of ambiguities: researchers
seem aware that the success of their fi eld depends on realizing the coexistence purpose,
and their research is shaped by this awareness. Summing up, the goal of ED is to supple-
ment traditional design methods with evolutionary algorithms that explore uncharted por-
tions of design space; I have not found any claims that ED might replace traditional design,
that is, that it might lead to more effi cient or effective designs in the well-traveled portions
of design space. That evolutionary processes should supplement traditional, intentionalist
methods is a driving force behind the development of ED: In section 13.2.2, I have
described how various procedures are evaluated and improved with regard to their capacity
for yielding innovative designs. In EA, coexistence is constructed along different lines.
Researchers admit that in principle the intentionalist framework can be used to describe
and explain the archaeological record, but they claim that lack of information prevents
them from using it in practice. Therefore evolutionary concepts and models are chosen as
substitutes. Again this mode of coexistence shapes research: if it turns out that evolutionary
models can do their explanatory work only by reintroducing intentionalist elements,
researchers admit that there are diffi culties, and additional transfer is seen as advisable. In
practice these diffi culties may turn out to be insurmountable—but that would be a success
criterion derived from the explanatory-substitution mode of coexistence. Thus the two
modes of coexistence both lead to constraints on satisfactory results within the program
and act as heuristics for modifying the program.
This leads to my argument against the confl ict image presented in section 13.1—specifi -
cally, against the separatist and unionist image of the relation between organisms and
artifacts. On these images, a problem-oriented, nonmetaphorical, and open-ended transfer
of concepts between the domain of organisms and that of artifacts is either ruled out or
vastly lacking in ambition. Neither option is, in my opinion, attractive.
On the fi rst, separatist judgment, researchers that indulge in the type of transfer described
earlier should be charged with smuggling by the philosophical border patrol. This should
not lead to symbolic sanctions: if the classifi catory, explanatory, and constructive resources
employed in EA and ED belong in the realm of organisms, transferring them to a qualita-
tively different realm can only yield results by accident. Both programs should fail. It may
be possible to suspend this judgment as long as the techniques employed and the concepts
transferred are suffi ciently general to avoid ambiguities. Insofar as cladistics involves only
a statistical analysis of the resemblances within a set of objects^20 and embryogenesis
involves no more than some translation mechanism, this may be correct. However,
ED and EA are bound to exhaust this tolerance rather quickly, given the researchers’