Theory & Evidence

by Clark N. Glymour

The description for this book, Theory and Evidence, will be forthcoming.

Genres: Philosophy

396 pages, Paperback

First published January 1, 1980

About the author

Clark Glymour is Alumni University Professor in the Department of Philosophy at Carnegie Mellon University and Senior Research Scientist at Florida Institute for Human and Machine Cognition. He is the author of The Mind's Arrows: Bayes Nets and Graphical Causal Models in Psychology (MIT Press), Galileo in Pittsburgh, and other books.

Reviews

[Joshua Stein · Rating 3 out of 5]

There's no doubt here that Glymour's approach to understanding the philosophy of science and the ways in which we come to prefer one theory to another represents a significant contribution to the rationalist literature. His criticisms of Lakatos (or, at least, Feyerabend's reading of Lakatos) and Feyerabend, among others, are significant and useful, and could easily be updated to relevant criticisms of modern arationalist views of scientific methodology and argumentation. However, the book also suffers from all of the historical problems of writing in the philosophy of science during the period. It is excessively technical, wandering off into the mathematics of Bayesian probability theory when a straightforward commentary on the underlying ideological contents would be sufficient. (Glymour's chapter on Bayesianism actually offers a rather straightforward case, but in illustrating his proficiency in probabilities, Glymour dramatically increases the page count and makes the argument far more convoluted.)

His notes on the history of physics, in particular, are very instructive, and in those cases his use of a more technical approach seems much more relevant to the argument he is advancing. Because he takes theories, in those cases, to be principally mathematical entities (or, at least, best expressed as mathematical entities) he gives us a look at the role of adjudicating between theories based on the contents of their mathematical expressions. It is illustrative of his whole approach such that, if one can take the time and energy to work through the commentary on the mathematical representations, his approach is easy enough to understand.

Very little of the history of physics, apparent from his analysis of the mathematical contents of various physical theories, is going to be unfamiliar to modern philosophers of science, however his use of the mathematics makes for a rather different analysis than what those in the social studies of science are going to be familiar with. Reading much of the mainstream literature can leave the impression that either abstract ideological entities (like Kuhn's ephemeral paradigms) or concrete and political entities. (like those discussed by Latour or Shapin and Schafffer) Glymour offers one of the better accounts of the role of mathematical representations of theories I've come across, (though perhaps not quite as accessible and engaging as McAllister) and so offers a worthwhile contribution.

I won't recommend the book to very many folks, because without some proficiency in mathematical physics, it is going to prove more frustrating than it is worth. I suspect this is why the book has not been widely read in the contemporary social studies of science, which it likely should be on account of its phenomenal section of historical notes. However, for those who are so inclined, it is a worthwhile and challenging read, with some very useful analysis of the recent philosophy of science, and the history of science.

[Chris Lawrence · Rating 2 out of 5]

Very glad to have finished this, although (or perhaps because) I understood very little of it. I had the uneasy feeling that there were parts of it which I might possibly have been able to understand had the author resisted the temptation to be as opaque as possible... :(