ABSTRACT
In The Structure of Scientific Revolutions, Thomas Kuhn proposed that normal science is a 'puzzle-solving' activity. Kuhn's idea is that mechanical puzzles have definite solutions; the challenge of working on them is to find such solutions. Kuhn's proposal appears to be an analogy - the everyday practice of normal science is like the work of solving puzzles - but nowhere, so it seems, does he consider the work of solving any particular type of puzzle. Work on a puzzle is the continual search for ways of finding that relevance. Most puzzle-solvers initially sort the pieces to find the border components, frequently sorting for general colours of sky, buildings, grass, and the like at the same time. In some ways, comparing chemistry and jigsaw puzzles is absurd: reasoning in chemistry is not the reasoning of jigsaw puzzles, and chemical practice is hardly the search to fit 'pieces' of a puzzle together.
