*The Cognitive Structure of Scientific Revolutions

Hanne Andersen, Peter Barker, Xiang Chen

Cambridge, 2006, 199 pp., £45 h/b Ð ISBN 0 521 85575 6

*Beyond Kuhn

Edwin H.-C. Hung

Ashgate, 2006, 160 pp, £40 h/b Ð ISBN 0 7546 5347 1

Reviewed by David Lorimer

Kuhn Reconsidered
Many readers will be familiar with the work of Thomas Kuhn on the structure of scientific revolutions. His initial work came out in 1961, and was subject to considerable criticism. He refined his theories over the years, but many people are unfamiliar with the later modifications he made to his work. These two books extend Kuhn's legacy, providing some useful new concepts as well as a detailed analysis of his notion of incommensurability between competing theories. Some of Kuhn's categories have entered into popular usage, such as paradigms, scientific revolution and the distinction between normal and revolutionary science. Both books try to reconcile aspects of traditional philosophy of science (including Popper) with constructivist accounts favoured by social scientists.

Andersen, Barker and Chen aim to legitimise the study of the cognitive content of science while restoring the ideas of conceptual revolutions and incommensurability to a central position. They build on the work of Wittgenstein and Eleanor Rosch on the idea of graded structure in human concepts, whereby concepts lack a single common feature connecting all category members, but nevertheless have a family resemblance. This is useful when applied to the central idea of a paradigm. The authors extend this work to present their own idea of 'frame theory' as a means of modelling conceptual structures through relations of similarity and dissimilarity. Frame theory accommodates both family resemblance and graded structure. They illustrate this with reference to core concepts of nuclear physics in the 1930s.

In this context they define normal science as research that conserves conceptual structures and revolutionary science as research that revises conceptual structures in radical ways. This normally requires the reclassification of existing entities in ways that were formerly impossible. It is interesting to note the influence of Gestalt theory on Kuhn's ideas of incommensurability, as the Gestalt switch between perception of the old lady and the young lady (see illustration in Janine Edge's conference report) does not apply so simply to scientific revolutions. Kuhn favours the idea of a taxonomic shift in his later work. Indeed, the authors point out that careful consideration of the Copernican revolution shows that Ptolemaic and Copernican astronomy had highly similar conceptual structures, and that the real revolution takes place with Kepler, whose concept of orbit is truly incommensurable with the earlier concept of circular motion.

Hung's ambitious book has the distinction of separate forewords by Rom Harré from Oxford and Peter Lipton from Cambridge. In it he tackles and claims to solve the central problems in the philosophy of science, namely scientific explanation, the structure of scientific theories, incommensurability, scientific change and physical necessity. His central concept is that of representational spaces (Kuhn's paradigm theories) for describing generic theories. He argues that the representation of reality has two steps: the construction of a representational space, followed by the modelling of reality with configurations of 'objects' in this space. Likewise, the search for truth in science progresses through an empirical and a theoretic stage. A conceptual shift takes place when the current conceptual framework is replaced by a new one. These shifts can occur either through extension, reduction, restructuring or replacement. Hung then moves on to language, pointing out that every conceptual theory has its own characteristic vocabulary. He argues that conceptual theory determines a conceptual language while 'what corresponds to statements are the models which representational spaces enable scientists to construct.' In this sense, languages are not so much statements as vehicles that carry statements. His rigorous discussion of incommensurability and empirical data enables him to reject the relativism favoured by social constructivists. He distinguishes empirical data as statements about mental states produced in observers, while the word phenomenon denotes the observation or phenomenal content of the empirical datum.

Both of these volumes are technical and aimed largely at specialists in philosophy of science. However, their broad ideas are important in refining and extending the most significant concepts in Thomas Kuhn's work.