Abstract

This essay suggests to understand scientific competition from a developmental viewpoint which is relative to two things: a changing problem domain of competing programs, and each program's developmental strategies. A metaphor of "Treasure Hunt" Race is introduced to highlight my point, and to differentiate my view from those of Lakatos and Laudan. Specifically in this essay, I focuses on various aspects of a changing "problem domain" and its "structure" , aspects that play important roles in the logic of program competition and development. Similar to heuristics or models which are suggestive to program development, domain problems, structures, and strategies are also suggestive in theory construction and modification within a program. A program's choices of domain problems and of development paths therefore strongly influence the fate of a program in its development and competition. Furthermore, solutions of domain problems offered by one program sometimes can be "incorporated" by other competing programs. I thus introduce a related aspect of program development and competition: an interactive process among competing programs by "borrowing" or "learning" each other's concrete results is always going on. In short, program development is not isolated from other programs, and each program's need for developments affects the outcome of program competition. Therefore, a program succeeds because of it has a fruitful empirical advances in its chosen developmental path, and its empirical results are sufficiently "incorporatable" or "learnable" to other programs, and thus it attracts great attention. Finally, many historical examples are used to illustrate my points, and in particular, the history of modern optics Newton, Huygens, Fresnel is my major historical reference point. Two short historical studies from my viewpoint are conducted: the differences in developmental strategies between Huygens and Fresnel explain the failure of the former and the success of the latter, and similarly between Huygens and Newton for the similar reasons. Newton's and Huygens' optical programs also illustrate the two previous reasons for the success of a program