Russellian Propositions JUDY PELHAM ALASDAIR URQUHART ∗ Department of Philosophy, University of Toronto, Toronto, Ontario, Canada Bertrand Russell, in the first decade of this century, held an unconven- tional view of propositions. He took them to be complex abstract entities resembling logical formulas in their basic structure, but differing from formu- las in that they may contain physical objects as constituents. The aim of this paper is to give an account of Russell’s notion of a proposition during the period 1903-06, and to explore the extent to which the logic which coexisted with that account of propositions is feasible. The period 1903-06 lies between Russell’s completion of The Principles of Mathematics (Russell 1903) and the beginning of the writing of Prin- cipia Mathematica (Whitehead and Russell, 1910). During this time Russell worked on a form of type-free logic, which he called the “no-classes theory” or “substitutional theory” (Russell 1905b,1906b), as a resolution to his paradox (or “the contradiction”, as he called it). Russell’s view of propositions given in the Principles changed and developed with his work on the paradoxes. Our elaboration of Russell’s notion of proposition is based on the formal evidence of the substitutional theory, as it was worked out in unpublished manuscripts of 1905-06. This theory is based on Russell’s conception of propositions as structured non-linguistic entities, and their fundamental logical properties. Russell’s ideas about propositions have come back into favour of late in connection with theories of direct reference (Kaplan 1986,1989) and the situation theory developed by Barwise, Perry, Etchemendy and others (Bar- wise and Etchemendy 1987). Recent authors, although they have found their * Research partially supported by the Social Sciences and Humanities Research Council of Canada. 1