In lieu of an abstract, here is a brief excerpt of the content:

Hume Studies Volume XX, Number 1, April 1994 pp. 138-139 ALAN MUSGRAVE. Common Sense, Science and Scepticism: A Historical Introduction to the Theory of Knowledge. Cambridge University Press, 1993. xiii+310. $59.95 cloth; $18.95 paper. Musgrave's book tells the story of a sceptic's progress and conversion. The sceptic begins with an attack on the very possibility of knowledge (chap. 1) and then fights back against the defences of empiricists (chaps. 4-9) and rationalists (chaps. 10-13). Overcoming the last-ditch response that truth is after all subjective (and thus easily secured against sceptical attack), she emerges victorious at the start of the final chapter. She there undergoes a conversion —to a form of Karl Popper's "fallibilist realism": infallibility is standardly beyond our grasp, but beliefs about the unobserved, when they have "withstood serious criticism," may be classed as reasonable and hence (if also true) as knowledge. The drama is played out with historical figures—like Locke and Berkeley as empiricist defenders of the possibility of knowledge, and Descartes as the leading rationalist. (Hume contributes more to the cause of scepticism.) But though these—with Kant—are the main protagonists, the historical scope is not the standard one of histories of early modern philosophy: there appear also Sextus Empiricus, Mill, Russell, the founders of non-Euclidean geometry, and even Tarski and Gödel. The book is not an introduction to recent debates in epistemology. Causal and reliabilist conceptions of knowledge are never discussed; externalism and intemalism never appear; the picture of knowledge as justified true belief, despite a Gettier-style challenge (5), reappears later essentially unscathed (280ff). The book's greater concern is with the defence of Popperian views on perception and scientific knowledge. Musgrave's defense of " fallibilist realism" is hard to assess. According to the fallibilist, "a belief is reasonable iff it has withstood serious criticism" (281), and a reasonable true belief counts as knowledge. Whether this is correct depends on what "withstanding serious criticism" involves. If it is merely a matter of surviving attempts at falsification, then it is surely insufficient to make a true belief knowledge. (What if the belief is in a domain where neither truth nor falsehood are ascertainable by us, for example, "the number of stars is odd"?) If "criticism" involves more than attempts at falsification, then we need to hear more about it: can it include criticism that the belief is based on insufficient evidence? Can it include criticism from a different epistemic perspective (for example, looking from outside at the brain-in-the-vat)? The Fallibilist case needs more careful exposition if it is to establish itself as both Hume Studies Book Reviews 139 plausible and distinct from the rivals that explicitly place both causal and rational conditions on knowledge. The book is not an introduction to the history of epistemology. There are chapters that will give undergraduates a good sketch of classic lines of argument and a brisk exposé of some weak points. (The chapters on Berkeley and Descartes struck me as the best.) Hume—"the Scotsman of our famous trio of British empiricists" (145)—appears mostly for his views on induction, but Musgrave quickly leaves the text behind to discuss three recent responses. One might try adding a premise to turn an inductive argument into a deductive one; one might try saying that inductive argument has its own sui generis validity; rejecting these, Musgrave prefers to say that our reasoning about the unobserved is in fact not inductive, but a matter of Popperian conjecture and refutation. And that (as the end of the book will argue) saves it from being irrational. There is a problem with some of Musgrave's historical expositions. It is not in the simplification and sketchiness, what are to some degree inevitable. It is in the inaccuracies that sometimes suggest recklessness. It is hardly important that the nine words of the empiricist Latin tag "Nihil in intellectu..." contain three mistakes (15-16). But any student who hears Berkeley described as thinking "that we cannot contemplate tables or trees or stars, but only ideas of tables or trees or stars" (97) is in danger of forming a seriously misleading impression...

pdf

Share