42 Beth Preston
rare exception rather than the rule in material culture, as in biology. And as Cummins
(2002: 166) remarks, if neo-teleology is applicable only to the spread of rare, radical
novelties, then it is not a signifi cant theory.
Now what about building items of material culture? Cummins calls this a Paley ques-
tion, with reference to the watchmaker analogy made famous by William Paley. Paley’s
point was that if you fi nd a complex artifact like a watch on a deserted beach, you naturally
assume it must have had an intelligent designer in order to exist at all; likewise, if you
fi nd an eye or a stomach, you should assume an intelligent designer of these complex
biological items. But what evolutionary theory shows is that you do not have to answer
Paley questions about biological traits by appeal to intelligent design, because the observed
results can be achieved by long-term, mechanical, incremental processes—including
natural selection—that are insensitive to the eventual complex structural “design” as well
as the ultimate proper function. It is widely assumed that Paley was right about material
culture, though, and that cultural selection is necessarily sensitive to function and design
while natural selection is not. But Cummins’s argument will go through for material culture
only if Paley was wrong—and wrong in the same way—about material culture. In short,
the existence of a watch on a deserted beach, like the existence of a stomach, must be
accounted for by a long history of incremental variations that was not from the beginning
aimed at the creation of watches.
We have already been oriented in this direction by the preceding point about the
rarity of radical novelty. In fact, the nature of human inventiveness is overwhelmingly
a matter of making small changes in existing material culture rather than producing
radical novelty out of nowhere. As Henry Petroski (1992, especially ch. 3) demonstrates
at length and in detail, inventors are in the fi rst instance critics of current technology, but
constructive critics with ideas for incremental improvements. The resulting variations
in artifact traits provide ongoing incremental changes on which cultural selection acts,
just as mutation and various other evolutionary mechanisms provide incremental changes
on which natural selection acts. In light of this observation, let us consider the history
of mechanical watches of the sort Paley had in mind, the early forms of which appeared
in the sixteenth century. First, such watches depend on the development of two basic
technologies, glassmaking and metallurgy, both of which have histories stretching back
many millennia. Second, they depend on the development of machining techniques
for producing very small parts capable of precision operations, which also predate the
advent of watches. Finally, watches depend on the prior history of mechanical clocks,
which fi rst appeared in the fourteenth century in the form of large, weight-driven
tower clocks in public buildings. It is certain that early glassmakers and metallurgists
were not aiming at watches. Neither were early machinists. And arguably, neither were
the early clockmakers, who were working on a vertical mechanism driven by large weights
that was not even conceivably portable. It was only with the invention of a spring-driven
mechanism and early portable clocks (e.g., for use onboard ships) that the sort of