While realists had a great deal of sympathy for the pragmatists' empiricism and their rejection of idealism and its coherence theory, they still kept (in the beginning) allegiance to the correspondence theory suitably updated to be acceptable in post-Kantian epistemology. From their point of view the pragmatists gave a very nice analysis of how we humans go about deciding on whether or not to accept any given statement as true, but the question of how we go about deciding whether or not to believe (accept as true) a statement is quite different from the question of what makes that statement true; and pragmatists have confused the two. What does make a statement true is not human inquiry (the evidence we've turned up), but the real facts of the case, what really is so.

Through Russell's efforts (relying on the logical work of G. Frege and his young student L. Wittgenstein) realists could show that any complex statement can be analyzed into logical truth functions of simple logically basic statements. These logically unanalyzable basic statements provide the foundation on which knowledge is erected; hence this outlook embodied a return to classical foundationalistic epistemology; the justification of a belief lies in the fact that it stands on this bedrock, incorrigible foundation. If such basic statements were somehow knowable, then the apparatus of symbolic logic could allow us to demonstrate how all more complex statements could be built upon this foundation. The "atomic" building block propositions led Russell to call his version of realism "logical atomism."

The problem for realists was saying what these atomic statements were about. Russell hoped to remain true to the British tradition of empiricism and yet avoid the elephant trap of subjectivity by which the skeptic had caught Hume by making the "objects" of atomic propositions objectively real elements of experience, so-called sense-data. Unlike the subjective "ideas" or "impressions" of classical British empiricism, "sense data" are objectively real entities, not the possessions of private subjective minds. "Physical objects" which we experience as populating the world are in effect constructions built up by the knower out of these objectively real elements, the sense data. The problems of sense data theory were legion, and dealing with them occupied a great part of English philosophy in the period between the wars. Today, most epistemologists regard this as a failed experiment.