What are the main errors of Bertrand Russell's A History of Western Philosophy?

Question

This discussion of A History of Western Philosophy complains about its errors and omissions, but doesn't give examples of the former. A previous question asked "how inaccurate" Russell's book was, and the answers indicated the primary areas where it errs are in early Greek and Mediaeval philosophy. But what are the main specific errors, on these topics or any others? For instance, did he exaggerate the extent to which Mediaeval theodicies were panglossian? (That's just an example I invented to illustrate why this question isn't a duplicate.)

Answer 1

I can offer one example, which I take from Isaiah Berlin concerning Russell's treatment of Kant's doctrine of space and time. This treatment, Berlin suggests, profoundly miscontrues the central doctrine of the Critique of Pure Reason:

Kant is treated in greater detail [than Hume], and once more Russell follows his preferred and somewhat Napoleonic method of concentrating his fire against the position on which he regards the enemy as strongest, leaving the rest to collapse and vanish of itself. In this case the doctrines of space and time are selected as the principal target, and after complaining that Kant gives no adequate explanation to account for the particular order or characteristics in time and space of particular material objects (which Kant might have regarded as a metaphysical, i.e., in some sense illegitimate question), Russell seems to assume that in Kant's system space and time must either be subjective, i.e., in some sense be empirically given, or, if not, that they must in some sense derive from, or belong to, Things in Themselves. This does not so much refute as ignore the central doctrine of the Critique of Pure Reason, according to which material objects (in Space and Time) are neither Things in Themselves (which according to Russell in the end would amount to a metaphysical transcendentalism of a neo-Platonist type), nor yet a set of (or logical constructions out of) sense-data, at any rate not in the ordinary phenomenalist sense. (Isaiah Berlin, 'A History of Western Philosophy by Bertrand Russell', Mind, Vol. 56, No. 222 (Apr., 1947), pp. 151-166 : 163.)

'Sense-data' is not the best term, reflecting as it does theories of perception prevalent at the time of writing (1947) rather than the intellectual world of the Critique. But if we replace it with 'sensory perceptions', Berlin does appear to convict Russell of a significant misunderstanding of Kant.

Answer 2

Something I happened to read today in Russell's "The Scientific Outlook", may be relevant to the above comment about Russell having no sympathy for philosophers like Kant:

Hume, nearly two hundred years ago, threw doubt upon induction, as, indeed, upon most other things. The philosophers were indignant, and invented refutations of Hume which passed muster on account of their extreme obscurity. Indeed, for a long time philosophers took care to be unintelligible, since otherwise everybody would have perceived that they had been unsuccessful in answering Hume. (part 1, chapter 3)