Hard going. Very tediously written & a joy to finish. Shame because there's some good stuff in there which the author tries hard to hide.
Improved a bit on second reading (after a few months) - I promoted it from 2 to 3 stars. But still harder going than it needs to be. I felt there was a leaner, clearer, more engaging book struggling to get out.

the book begins with the "dean's challenge" -- can a philosopher shed any light on scientific enquiry (especially the relationship between evidence and hypotheses) that is useful to actual scientists. then, the rest is an attempt to make good on this challenge. there are at lest two fundamental problems: (1) this presumes that philosophers haven't had much to say that should interest scientists, which is false (see  Scientific Reasoning by  Howson and Urbach); (2) the author puts forward an "ordinary language" account of evidence that seems like it is not interesting to scientists or philosophers. this project seems like a step backwards in understanding evidence.