In a few short paragraphs of The Problems of Philosophy Bertrand Russell presents his theory of introspective knowledge based on the concept of knowledge by acquaintance. In this article, I critically analyze these comments by Russell and their proposed application by contemporary authors, including Brie Gertler (2001; 2011) and Laurence BonJour (2003). I show how these theories differ from the competing ‘inner‑sense theories’ and try to show that they are inconsistent with Gareth Evans’s ‘transparency’ observation. Then I compare acquaintance theories with Alex Byrne’s (2005; 2012) ‘transparency’ theory and show that Byrne’s theory offers a simpler account of mechanisms governing introspection and attribution of mental states to other agents than the observer.

The article analyzes Bertrand Russell’s distinction between knowledge by acquaintance and knowledge by description contained in his writings from 1910 to 1912. I point to some problems that arise from Russell’s concept of direct knowledge if they are observed in the light of modern psychological and neuroscientific research. In some cases knowledge by acquaintance may be mediated by certain unconscious operations and by mental representations of conceptual nature. I point out however that some of the examples given in The Problems of Philosophy can be characterized by a different degree of indirectness. Consequently, I propose to view Russell’s distinction as a typology. Because Russell narrows knowledge by description to specific descriptions, Russell’s division is not exhaustive and therefore is not a logical one. The article also shows that there are reasons for both narrowing his category of direct knowledge (e.g. by excluding universals) and for expanding it (e.g. by unconscious experiences).

The aim of the paper is to provide an answer to the following question: How to interpret the principle of acquaintance and what role does it play in Bertrand Russell’s epistemology? It seems that the principle itself should be so adjusted as to take into account two concepts: (1) an act of judgment as a multiple relation and (2) the division of reality into what is directly given (sense‑data) and physical objects. I show, contrary to Russell’s assurances, that the content and the role of the principle is not clear, and its acceptance leads to a paradox. Having discovered that consequence, Russell abandoned the dualistic division of reality and with the help of the method of logical constructions, sought a position that embraced phenomena