In an enormous (about 550 pages), ponderous, repetitive and (to me) ultimately quite boring book, David Hull uses in-depth examples from numerical taxonomy and cladistics to show that science evolves, like organisms, by natural selection in a competitive environment.
David Hull says that successful science is done mostly by small groups of collaborators working within a research program and that their reward is prestige or credit (rather than money or power). Inter-group rivalry can be vitriolic but stimulates progress through competition.
Accounts of vitriolic arguments (often for trivial reasons, such as the meaning of a word) occasionally enliven Hull's otherwise turgid compendium. For example, Paul Ehrlich predicted that, by 1970, traditional nomenclature and taxonomy would be replaced by computerized systematics. At a meeting in St Louis, 'when one taxonomist asked indignantly, "You mean to tell me that taxonomists can be replaced by computers?" Ehrlich replied, "No, some of you can be replaced by an abacus." Thereafter, Ehrlich did not consider the give-and-take after a paper truly successful unless he brought at least one taxonomist to the point of tears' (page 121).
David Hull takes the sociology of knowledge seriously (too seriously for me, I admit) and denies that science advances only by 'reason, argument and evidence', but he rejects the extreme view that changes in science are caused entirely by changes to its social conditions. Instead, science advances by argument and evidence as these principles are understood within an academic context; hence scientists often argue about what counts as science or about what constitutes a good argument and accuse one another of irrationality, missing the point and arguing at cross-purposes.
Altogether, David Hull has compiled a valuable set of arguments and examples in support of evolutionary epistemology (very close, in fact, to Karl Popper's epistemological Darwinism) that would have been pleasant to read at half its length.