Let’s begin with a story—not the full strange story of my title but an early part of it.
One day some philosophers decided to take a walk down through history. They had passed through ancient Greece and Rome, continued
on through the Middle Ages and the Renaissance, and had come to the 16th and 17th centuries, where they stumbled upon the
Scientific Revolution. “What intricate design these experimental instruments and practices display!” they exclaimed. The philosophers
marveled at the intelligent order manifested in the theories and explanations. Where could all this design have come from,
of a sudden? They agreed that it was virtually a priori true that there can be no design without an intelligent designer,
but who or what could the intelligent designer be in this case? Since the preceding generations of inquirers had done nothing
comparable, they concluded that these 17th-century natural philosophers had hit upon some intelligent method of discovery,
a tool that amplified their intellect. For how else, short of appealing to a direct revelation from God to the innovators,
could they possibly explain this explosion of successful problem-solving activity? How else could they account for the production
of so much epistemically interesting design, following upon centuries of sterility?
Now this is a true3 story, or at least an archetype with actual instances! A variant of the story is already true of Descartes, as we shall see,
who, along with Bacon, is considered a founder of modern scientific method. Bacon died in 1627 and Descartes in 1650, so neither
was in a position to reflect on the later work of Newton, Leibniz, and company.
This paper is an updated version of ideas that I presented at the Ghent Congress on Discovery and Creativity
in spring 1998. I have tried to retain the original flavor of the paper. An ancient ancestor with the same title
was presented at the University of California, Davis, and a more recent version to the British Society for
Philosophy of Science. I received valuable comments at these venues. Thanks to Gaye McCollum Nickles
for much help and to Joke Meheus, both for organizing the congress and for her criticisms, which I have not
yet fully addressed. Thanks also to Yoichi Ishida. The paper draws upon previous work supported by U.S.
National Science Foundation grants. For companion essays that provide more detail and more justification at
some points, see Nickles, 2003a, 2003b. The reader should keep in mind that, today, the ideas of evolutionary
computation are more familiar to, and are taken more seriously by, philosophers than in 1998.