Unification as a Regulative Ideal

I The Unity of Science Movement is dead. If philosophers ever believed that science could be organized as a hierarchy of theories founded on general principles with the basic generalizations of “higher level” theories derivable from those of more “fundamental” theories, then they do so no more. The doctrine that chemistry is reducible to physics, biology to physics and chemistry, psychology to biology, and the social sciences to psychology has suffered from scrutiny of crucial junctions—particularly those between biology and the physical sciences, and between psychology and biology. The following points are, I hope, relatively uncontroversial. (1) Some sciences, particularly parts of biology, psychology, and the social sciences, are not happily viewed as collections of laws that can be organized in an axiomatic system. (2) Some sciences legitimately employ functional concepts (like the concept of a gene) and historical concepts (like many taxonomic concepts) that resist identiacation in structural terms. (3) Even to the extent that some “higher-level” sciences contain generalizations whose component concepts can be speciaed in terms of the concepts of “lower-level” sciences, the derivation of these generalizations from the speciacations and the principles of the “lower-level” sciences would not be explanatory. We wouldn’t deepen our understanding of Mendelian rules of inheritance by deriving them within molecular cell biology, even if we could give the derivation. The questions raised by this symposium center on how far we want to depart from the picture of one science, hierarchically organized. Some philosophers, Ian Hacking, John Dupre, and Nancy Cartwright, have