Abstract

In one of its latest papers Timothy Williamson has drawn a distinction between two readings of the phrase "possible F", where "F" is a predicate variable: the predicative and the attributive. In what follows, on the one hand I will hold that the first reading naturally applies to the phrase "possible object", thereby supporting a moderata conception of possibilia as entities that possibly exist. Moreover, I will maintain that one such conception provides the best possible account of Tractarian objects. On the other hand, I will claim that the second reading of the phrase "possible state of affairs" supports a parsimonious ontology of states of affairs, according to which a possible state of affairs is nothing over and above the de re possibility its constituents have of being related in a certain way to each other. As a result facts, i.e. possible states of affairs actually obtaining, are the only genuine states of affairs that there are, or, to put it differently, for any given state of affairs it is redundant that it is a fact. Besides, I will also claim that, appearances notwithstanding, this ontology of Sachverhalten underlies the Tractatus itself. Finally, I will try to show how, again appearances notwithstanding, this thrifty ontology of Sachverhalten squares with the semantic theses of the Tractatus according to which an elementary proposition presents a possible state of affairs as its sense and it is true if ‘that’ possible states of affairs obtains.