THE INSTRUMENTALIST VIEW ON THE CHOICE OF QUESTIONS
What, then, determines which question will be pursued by scientists? Not surprisingly, perhaps, the only widespread answer offered to this question is technological. I often ask scientists orally, or try to divine from their written work the answer they would give to the question, how do you decide which question to pursue next? If I elicit any answer at all it is, I choose any question which I have the tools to pursue.
The tools may be experimental, such as high energy accelerators, mathematical, such as recent solutions to non-linear equations of a given type, or scientific, such as a general theory of superconductivity. The answer that techniques determine agenda has been expounded at length only by the celebrated instrumentalist philosopher Pierre Duhem, in his magnum opus The Aim and Structure of Physical Theory. (See also Millikan’s autobiography.)Duhem’s answer has truth in it, but it is a false answer, and one which is empirically refuted. Undeniably, there exist in any given field of investigation some standard techniques. No doubt new techniques invite new investigations. The discovery of radioisotopic organically assimilable elements has opened a vast field of research by enabling us, with the use of tracing techniques, to study diverse questions of assimilation and of metabolism and more. In some cases the rendering of a new technique useful for research is most ingenious, for instance Fraunhofer’s discovery of the method of comparing the diffraction indices of diverse materials by the use of solar absorption spectral lines (which, subsequently, were named after him). A more obvious instance is Einstein’s use of the absolute differential calculus in his development of his theory of gravity.
True, then, as Duhem’s view is of many cases, it is most unsatisfactory even in those cases. For, the response these cases should evoke is not, thank goodness techniques evolve which enable us to tackle new problems! Rather, it should provoke us to ask, how come techniques evolve and how come they are applicable beyond their original intended domain of application? It is no accident, as I have tried to show elsewhere (Journal of the History of Ideas, 1969), that the absolute differential caluclus was there when Einstein needed it; rather, the problems which led to the development of that calculus, as well as to Einstein’s theory of gravity, have a common ancestry - in the philosophy of Leibniz. It is not always that the questions asked by the inventor of a technique relate directly to the questions asked by the one who applies it.
The questions answered by radiophysics differ greatly, for example, from the questions asked by the biochemist. Attempts to relate the two, however, constitute very interesting parts of physics and of physical chemistry.Sometimes, of course, behind a technical question there is a theoretical question meant to be solved by it. When J. J. Thomson invented the primeval television tube he was interested not in the tube and in what it could bring about, whether television or mass-spectroscopy, but merely a theoretical question concerning the mass and charge of the smallest particle. This example, and its like - and there are many of them - utterly refute Duhem’s theory of choice of questions. In these cases it is not that new techniques allow the study of new questions, but that the interest in new questions provide incentives for the development of new techniques. New techniques do not evolve, as Duhem’s theory tends to suggest, they are developed with given interests in mind. How do these interests get chosen?
No doubt, existing techniques do constitute a factor determining choice of problems, only not the sole factor. Sometimes they do constitute the sole factor, and that is when the new technique is so powerful that its employment is satisfactory for almost everyone. Lakatos has observed that the techniques of modern formal logic, of the combined sources of the Frege-Russell-Whitehead stock, and of the more formalist Hilbertian stock, has offered so much challenge that all problems to which these techniques are not relevant were forgotten for a generation or two. This is understandable, but also it is a certain loss - not merely of leaving certain avenues unexplored, but also of leaving the lone explorers of these avenues isolated and forgotten.
VI.