This article records tradition as it has been passed down and reported. Its sources are not yet part of the atlas's verified catalogue.
Bertrand Russell's distinction between knowledge by acquaintance and knowledge by description, worked out most fully in a 1910 to 1911 paper of that title and repeated a year later in The Problems of Philosophy, was meant to answer a question that had troubled him since his break with idealism, namely what a mind can know without the mediation of inference. Acquaintance, in Russell's technical sense, is a direct cognitive relation to an object, a case in which the mind is directly aware of the object itself with nothing standing between the knower and the thing known. He restricted genuine acquaintance to a narrow class of items, chiefly the sense data of the present moment, such as the particular patch of color or the sound one is hearing right now, along with universals like whiteness or diversity, and, more controversially, oneself. What one is acquainted with, Russell stressed, is not a truth about the object but the object itself, so acquaintance is knowledge of things rather than knowledge of truths.
Almost everything else we claim to know, however, we know only by description. Description based knowledge concerns objects that are not themselves given to the mind but are picked out indirectly, through a definite description built from concepts and relations the mind is acquainted with. Russell's stock example contrasts Scott the man, whom his contemporaries might know directly, with the phrase the author of Waverley, which specifies an individual uniquely by a relation to a book without putting the mind in direct contact with the person. One can know a great many truths about the author of Waverley, including that he wrote a particular novel, without ever having met him, so long as one grasps the descriptive phrase and knows, or can verify, that exactly one thing satisfies it.
This distinction dovetailed with the theory of definite descriptions Russell had already developed in logic, in the earlier paper On Denoting, where he had shown that a phrase like the present King of France need not name an entity in order to be meaningful. Analyzed logically, the sentence the present King of France is bald asserts that there exists an x such that x is a present king of France, that x is the only such thing, and that x is bald. Since no such x exists, the sentence comes out false rather than meaningless or committed to a nonexistent king. Knowledge by description exploits exactly this structure, to know an object by description is to know that some unique thing satisfies a given descriptive condition, built entirely from concepts one is acquainted with, without the object itself ever being presented to the mind.
Russell treated this apparatus as the foundation of a genuinely empiricist epistemology, one meant to reduce every legitimate belief, in principle, to a chain terminating in items of direct acquaintance. He put the guiding principle bluntly, that every proposition we can understand must be composed wholly of constituents with which we are acquainted, since otherwise we could not grasp what the proposition asserted at all. Bismarck himself, as a historical agent known to us only through testimony and record, becomes an object known by description, built out of sense data, memories, and other people's acquaintance, rather than an object anyone alive today could claim direct acquaintance with.
This fed directly into the logical atomism of Russell's middle period, the view that the world ultimately consists of atomic facts, simple particulars standing in simple relations, mirrored by atomic propositions whose truth or falsity depends on how those facts stand. The acquaintance relation supplied the epistemological anchor for that ontology, since only what could be encountered directly, sense data and universals given in experience, counted as a genuine simple, while everything more complex, other minds, enduring physical objects, historical figures, had to be reconstructed as a logical description built out of those simples. Russell's own regressive method acknowledged a complication in this picture, that we often trust premises because we already trust their consequences rather than the reverse, yet he never abandoned the conviction that understanding has to bottom out somewhere in direct contact with what is known, and acquaintance was his name for that bottom.