On Roy Bhaskar's A Realist Theory of Science.It is quite a long journey from philosophy of science to railway history, but Roy Bhaskar's A Realist Theory of Science is still worth packing, even if these days it is usually more a matter of luggage in advance than reading for the trip.
Initially my enthusiasm stemmed from the answers that Bhaskar offered for some troubling but fairly narrowly defined questions concerning the cognitive status of the natural sciences. I had become interested in these as an undergraduate studying physics and philosophy at Bristol in the late 1970s, partly as a reaction against Paul Feyerabend's lively and iconoclastic Against Method (1975) and Science in A Free Society (1978), two highly readable books that combined an attack on the orthodox claim that science can progressively tell us more about nature with a critical re-evaluation of the place of science and scientists in contemporary society. I felt far from happy about Feyerabend's more sceptical implications. Scientific reasoning in some shape or form seemed potentially a better force for social good than the epistemological free-for-all that Feyerabend seemed to promise. So I turned to a study of theoretical explanation in the natural sciences and came to the not-very-profound conclusion that scientific theories were best understood as provisional statements about the characteristics of entities that exist in the natural world; that is, as ontological claims.
But a move to Manchester and an introduction to Bhaskar's book in the early 1980s took me much further down the path of philosophical realism. A Realist Theory certainly confirmed my feelings about the importance and nature of theoretical reasoning. But it also made me realise that I had absorbed too much of the orthodox philosophical thinking about the ontology of nature. Bhaskar argues that to understand the practical achievements of scientists in grounding their theoretical arguments through experimentation, we must see the world as an open, complex set of natural ''mechanisms".
If A Realist Theory of Science were simply a technical treatise on these philosophical aspects of the natural sciences, I doubt whether it would have had quite such a lasting impact on me. But Bhaskar's interest stems from a concern with the wider political uses to which models of scientific reasoning are put. The central ideas of A Realist Theory were developed in The Possibility of Naturalism (1979), which sought to show that social structures exist and that it is possible to study them in the same way as natural ones. Both impressed partly because they drew upon Continental as much as Anglo-American schools of thought. For a time I thought seriously about focusing my doctoral research on philosophical realism. But ironically, a lesson I had learned partly from Bhaskar eventually dampened my enthusiasm.
Strong on ontology, A Realist Theory is less convincing on epistemology, yet it does emphasise that scientific knowledge is historically conditioned. Increasingly my interests turned to understanding that historical dimension, a road that led me to the social history of engineering and railways.
Yet in a way A Realist Theory and its sequel still keep me on the rails. For over two decades, historians and sociologists have creatively explored the socially constructed aspects of science and technology. Bhaskar has helped me retain my confidence that such analysis does indeed have a proper subject. But some historians now seem to think that a little warmed-up linguistic philosophy can send all historiography crashing into the buffers. Perhaps it is time to unpack that luggage again.
Colin Divall is professor of railway studies, University of York and head of research, National Railway Museum.