Russell’s distinction between knowledge by acquaintance and knowledge by description has been recently re-examined in frequently controversial epistemo-logical contributions. The present essay reflects upon the pertinent papers by D. F. Pears, J. Hintikka, R. Chisholm, W. Sellars, A. J. Ayer, and others, but is pri-marily founded on Russell’s significant formulations from his writings published between 1910 and 1918. By employing an auxiliary device of a late-Wittgen-steinian language game, I explore at first the situation in which human subject is “experiencing” and naming particular objects (Russell’s sense-data and sensibi-lia) and later the subject’s acquaintance with universals. The reconstruction of such situations shows that, contrary to Russell’s assumptions, even the “purest” acquaintance cannot function without knowledge by description, i.e. without sta-ting propositions about the object of acquaintance (whatever its nature). Then the only “descriptionless” alternative would be a kind of intuitive knowledge of such objects which is difficult to reconcile with the position held by Russell in the 1910s. Whatever the consequences, this topic retains its fundamental epistemological significance.