In reading Wittgenstein one can, and for the most part perhaps should, treat the expression ‘language-game’ as a term of art, a more or less arbitrarily chosen item of terminology meaning something like ‘an actual or possible way of using words’. It would then be a fairly routine task to work out answers to such questions as what features of the ways a word is used are emphasized by this term of art, what philosophical purposes are served by the description of primitive language-games or of variations on actual language-games, or in exactly what way those purposes are supposed to be served.

It is persons who are ontologically committed. But a person is not ontologically committed by virtue of his character, his height, his social standing or whatever, but by virtue of the sentences he assents to. Hence we should look to sentences for a criterion of ontological commitment. This is precisely what is done by advocates of what I will call the Referential theory. In this paper I argue that the Referential theory faces serious objections related to the role paraphrase must play in it. I then present a modified Referential theory and go on to discuss certain implications of the modified theory. More precisely, the paper divides up as follows. In §1 I present the Referential theory. In §2 I argue that it is in trouble over paraphrase. In §3 I consider a general objection to the arguments of §2. In §4 I present the modified theory. In §5 I consider the implications of the modified theory for polemic over what there is and for the existence of properties or universals.

Prichard held, like some others before and since, that there is a categorial difference between knowing and believing:

To know is not to have a belief of a special kind, differing from beliefs of other kinds; and no improvement in a belief and no increase in the feeling of conviction which it implies will convert it into knowledge. Nor is their difference that of being two species of a common genus. It is not that there is a general kind of activity, for which the name would have to be thinking, which admits of two kinds, the better of which is knowing and the worse believing, nor is knowing something called thinking at its best, thinking not at its best being believing.

Jonathan Glover and I, while not in such deep disagreement about the ethics of killing as to make all communication impossible, still disagree enough to make sustained confrontation worthwhile. (Very briefly, Glover is a utilitarian, while I am not.) At minimum, such confrontation should make it clear what are the most fundamental issues at stake in ethical arguments about various kinds of killing.

Happiness requires something in its own nature, or in ours, to give it influence, and to determine our desire of it and approbation of pursuing it (Richard Price, Review of the Principal Questions in Morals, Ch. 1, Sec. 1).

Thales believed that everything is water:

A far from foolish thesis when you think

That your best vintage claret, Indian ink,

The knees of politicians, Pharaoh's daughter

The brains of all the nincompoops who court her,

The mouse, the tiger and the bobolink

(One must perhaps except the kitchen sink)

Are all, white-coated analysts report, a

Good seventy-five per centum aqua pura.

In any case, Creation's primal stuff,

The mixture for the Eternal Cook's plum duff,

Is likelier to be something that can cure a

Thirst, or the smell that emanates from barges,

Than a barn dance of low electric charges.