ABSTRACT
The contemporary Russellian identifies a singular proposition with an ordered pair of an n-ary sequence of entities and an n-ary relation.1 But since such pairs correspond to propositions many-one, this raises the question, which pair is the proposition that (say) Brutus stabbed Caesar? The question points to a more general worry, what I shall call ‘the representation problem’: How can it be that a sequence – in effect, a mere set – represents the world, and, in particular, represents it as being thus-and-so? The Russellian has no answer, and would seem forced to acknowledge that, although propositions are like sequences in being structured, only in the case of sequences is it clear how the attribution of structure is to be understood; in the case of propositions, the attribution of structure must remain unanalyzed. Thus, although she recognizes on intuitive grounds that the proposition that Brutus stabbed Caesar differs from the proposition that Caesar stabbed Brutus, when pressed to explain how they differ, the Russellian is forced either to remain silent, or traffic in uncashed metaphors.2
