1 How to Identify Scientific Revolutions? (in: Torres J. M. (ed.) (2010), On Kuhn’s Philosophy and its Legacy, CFCUL, Lisbon, pp. 251-282.) Erik Weber and Dunja Šešelja Centre for Logic and Philosophy of Science, Ghent University {Erik.Weber, Dunja.Seselja}@UGent.be Abstract Conceptualizing scientific revolutions by means of explicating their causes, their underlying structure and implications has been an important part of Kuhn's philosophy of science and belongs to its legacy. In this paper we show that such “explanatory concepts” of revolutions should be distinguished from a concept based on the identification criteria of scientific revolutions. The aim of this paper is to offer such a concept, and to show that it can be fruitfully used for a further elaboration of the explanatory conceptions of revolutions. On the one hand, our concept can be used to test the preciseness and accuracy of these conceptions, by examining to what extent their criteria fit revolutions as they are defined by our concept. On the other hand, our concept can serve as the basis on which these conceptions can be further specified. We will present four different explanatory concepts of revolutions – Kuhn's, Thagard's, Chen's and Barker's, and Laudan's – and point to the ways in which each of them can be further specified in view of our concept. Key words Research traditions, scientific revolutions, taxonomic changes. 1. Introduction Conceptualising scientific revolutions has been one of the central issues in the explication of scientific development in general. This topic was especially emphasised in Thomas Kuhn's The Structure of Scientific Revolutions. What Kuhn, as well as his successors (e.g. Larry Laudan, Paul Thagard etc.) tried to do is to offer an account of revolutions which explicates the structure underlying this process: the ways in which revolutions emerge and the changes they provoke on epistemological, methodological and ontological level. In contrast to such approaches, revolutions can also be analysed in terms of identification criteria. Let us clarify this difference by making an analogy with a medical diagnosis. A disease can be specified in two ways. One way is to give a full explanation of the physiological processes constituting the illness. This is an account that would be useful for a biomedical researcher who wants to produces knowledge that is useful for prevention and cure. Knowing how the disease emerges and develops is useful background knowledge in such research. On the other hand, a doctor diagnosing a patient will rather refer to symptoms as the identification criteria of a disease. The full account of the disease process is irrelevant in the process of diagnosing as such. The biomedical researcher will need both approaches: if she runs clinical experiments, she will also need symptoms as identification criteria, in order to design and interpret her experiments. 1 Going back to the case of scientific revolutions, the above mentioned attempts at characterizing revolutions are comparable to a full account of a disease process. A concept based on a set of identification criteria of revolutions – similar to a set of symptoms used for identification of a disease – is missing in the literature. 2 The aim of this paper is to offer such a concept. What we are after is a set of identification criteria, which captures the meaning of the term “scientific revolution” as it is used by scientists, historians and laypersons. In other words, there should be a substantial overlap with an intuitive usage of this term. In a broader context, our concept should also serve as a