On revolutions and progress in economic knowledge: definitions and conclusions
I. 1 In the foregoing chapters, including their titles, we have been making much use of the term `revolution'. It may seem reasonable to hope (and believe) that the omission so far to define and analyse this term precisely or searchingly need not have, and has not, been a source of confusion. As we noted in the Foreword, the term `revolution' has been very much used by economists, especially since Keynes, and for some time before the work of T. S. Kuhn appeared. It can therefore be used and understood without any of the implications regarding the history and philosophy of science attached to it in Kuhn's work.
I. 2 Presumably a `revolution' is a process of comparatively fundamental or comprehensive, and/or comparatively rapid, change, affecting either a part or the whole of the subject. We are concentrating here on `revolutions' in, or appreciably affecting, the subject as a whole. For example, the Keynesian `revolution' was only directly concerned with one major branch of the subject, but it might be considered sufficiently important to have reshaped the subject as a whole. As we noted earlier, we are not concerned here with partial `revolutions' in branches, or subdivisions too small to affect significantly the subject as a whole. We shall take up subsequently the concept of methodological `revolutions'.
I. 3 It is impossible to be at all precise regarding the rapidity of the processes of fundamental change, if these are to be described as 'revolutionary'. The idea of `a long revolution' does not seem to be contradictory (and has been, in fact, the title of a history of recent cultural changes in England). Also, complete conversions, or processes of change, may be, and usually have been spread out over a considerable period, although it may be possible to identify a decisive phase or episode with the publication of a particular book.