Peter Richard Forrest

In this paper I propose an indubitability analysis of knowledge. The motivation for this analysis is a conviction I have that the Cartesian analysis of knowledge as indubitability is not completely mistaken, although it requires considerable weakening if it is to be satisfactory. My analysis may be contrasted to those which treat knowledge as a species of the genus justified true belief. For although on my analysis, ‘S knows that p’ entails ‘S has a justified true belief that p’, I am not proposing indubitability as a differentia distinguishing knowledge from other members of the genus justified true belief. The situation here is analogous to that of the conceptual relation between vermillion and red: ‘X is vermillion’ entails ‘X is red’, but there is no straightforward differentia which would entitle us to think of vermillion as a species of the genus red. Another point of contrast is that species-of-justified-true-belief analyses start with an analysis which is too weak and strengthen it. My starting point is an analysis which is too strong which I then weaken.

David Armstrong in his A Combinatorial Theory of Possibility proposes that non-actual possibilities may be treated as fictions grounded in instantiated universals. In this paper, I first provide some objections to his theory. Then I make the case for Foundational Nominalism, the Armstrong inspired thesis that the whole of ontology supervenes on particulars described much as in the Quinean Nominalism that Armstrong rejected as an ontological ostrich. Finally, I argue that Foundational Nominalism permits a fictional theory of possibilities similar to Brian Skyrms’ Tractarian Nominalism.

Conceivability is, I say, prima facie evidence for possibility. Hence, we may count the cost of theories about possibility by listing the ways in which, according to the theory in question, something conceivable is said nonetheless to be impossible. More succinctly we may state a principle, Hume's razor to put alongside Ockham's. Hume's razor says that necessities are not to be multiplied more than necessary. In this paper I count the cost of David Lewis's modal realism, showing that many of the objections are replied to by Lewis only at the cost of multiplying necessities.

Consider the things that exist—the entities—and let us suppose they are mereologically structured, that is, some are parts of others. The project of ontology within the bounds of bare mereology use this structure to say which of these entities belong to various ontological kinds, such as properties and particulars. My purpose in this paper is to defend the most radical section of the project, the mereological theory of the exemplification of universals. Along the way I help myself to several hypotheses: the existence of merely possible worlds; that particulars have thisnesses; and that mereology is far from classical. Moreover, the way I characterize instantiation might be judged too complicated to be plausible. At the end of the paper, I reply to these objections based on complexity.