Abstract
: Suppose a scientist discovers a new, unpredicted phenomenon. How can one ascertain the causes, properties and laws of the phenomenon? How can one plan the investigation of the circumstances that affect the phenomenon, and of the effects that the new phenomenon could produce? If the phenomenon is completely unexpected and does not fit any previous theory, it is impossible to provide a theoretical prediction of its likely properties. In the empiricist tradition, therefore, the recommended method was to investigate all possibilities, because in such cases it is impossible to exclude a priori anything. William Stanley Jevons provided a clear criticism of this method. It is impossible to investigate all possibilities, because they are boundless. Is it then impossible to plan the research of unexpected new phenomena? No. Jevons pointed out an alternative. According to Jevons, a scientist confronting a new, unexpected phenomenon, should compare it to other known phenomena to establish analogies. This comparison should allow the researcher to find out one or several known phenomena similar to the new one. This paper will present and discuss Jevons’ proposal in the context of late 19th century methodology of science. Keywords: scientific method; empirical research; analogy; Jevons, William Stanley.