I’ve just finished reading Bertrand Russell’s “The Problems of Philosophy”. I loved the book even though I disagree with almost everything he has to say.

Here are some modest reflections about it:

1. This book is a perfect hinge between modernity and the twentieth century. On the one hand, Russell sums up the contributions of modern philosophers like Descartes, Spinoza, Berkeley, Hume and Kant to the theory of knowledge; on the other, he sets out the foundations of the new epistemology that developed in the first half of the Twentieth Century and gave birth to Wittgenstein’s Tractatus, the Vienna Circle, logical positivism and Popperian falsificationism.
2. I was surprised to learn that Russell, for all his empiricism and positivism, was also a Platonist. That is, he thought that universals such as the principle of induction, mathematical entities (numbers, geometrical figures), the law of causality, etc. are all real. Moreover, Russell claims that they don’t exist as material entities but as perfect, immutable, immaterial forms. Russell is not willing to follow Plato in his mystic moments, and he’s not willing to embrace mystic readings of Plato’s dialogs. In addition, he doesn’t want to trace a very sharp division between doxa and episteme; as a good British empiricist, Russell sees a continuity between common sense and philosophical-scientific knowledge. He trusts our human instincts and claims that sense-data (the “sensible world”) play a very important role in providing raw material to our knowledge. But, other than that, he’s a full fledged Platonist.
3. Concerning the last point: this is perhaps where we feel more alien to Russell now. After Sellars’ denunciation of the “myth of the given” and Quine’s holistic understanding of knowledge processes as always based on ontological commitments, we are reluctant to accept Russell’s lineal view of knowledge acquisition, which starts from value-neutral atomic data and uses it to build more complex and theoretical knowledge forms, such as knowledge by description (or inference). We now believe that sense-data are already penetrated by theory, and that our knowledge processes are circular (but not necessarily visciously circular).