Until the late 20th century, it was widely agreed that scientific discovery was or at least partly involves an unanalyzable creative act of (Romantic) genius. This view, which came to dominate the discussion of discovery in the first half of the 20th century, has been replaced by a wide range of more sophisticated approaches. The central issue that has divided these is whether or not the process of formulating new hypotheses – what Whewell called “colligation” – is part of discovery proper. Some argue that it is not. They argue that there is no logical way to have new ideas and that a philosophy of discovery can only be an empirical study of the processes of thought that generate such ideas.
Others, influenced by contemporary research in cognitive science and in particular studies of creativity, suggest that there are rules that guide the generation of novel thoughts. They maintain that the context of discovery should be distinguished from the context of justification and that there are heuristics (rather than inductive or deductive) that can be used to guide the search for truth.
In addition to these rich accounts of knowledge generation, some philosophers have developed more specialized theories of the logic of discovery. In particular, Norwood R. Hanson argues that the process of formulating a new hypothesis follows a special logical pattern that differs from both inductive logic and hypothetico-deductive reasoning. Other philosophical authors have elaborated on this theme, drawing on the work of Charles Sanders Peirce.