The type of reasoning which has gone into Kepler's thinking, Hanson characterizes as retroductive. The form of the inference is – (1) Some surprising phenomenon P is observed: (2) P would be explicable as a matter of course if a hypothesis H is true: and (3) And hence, there is a reason to think that H is true. H does not emanate from some unaccountable creation as hypothetico-deductivists think nor from simple repetitions of observations as inductivists think. It emanates from a mode of thinking which seeks to find out a plausible pattern into which what are observed are fitted. A hypothesis provides such as a plausible pattern. Before we test a hypothesis, it must at least be plausible and not just a conjecture. Of course, apart from its plausibility, the hypothesis must satisfy further conditions such as if a hypothesis H is meant to explain a phenomenon P, then H cannot itself rest upon the features in P which required explanation. ‘That is why the peculiar colour and odour of Chlorine (P) are not explained by reference to atoms in a volume of Chlorine, each one having the colour and odour in question (H). Grasping this point is essential for any understanding of the fundamental concepts of modern particle physics'⁶ .

Of course, the current work on discovery has gone much ahead of Hanson in terms of sharpness of articulation and rigour of analysis. But, the credit of putting on defensive the Popperian position on discovery goes to Hanson's path-breaking work.