It is often said that singular causal statements express a relationship between one event and another or between a fact and an event. This is a very strong view, which has the following simple corollary: singular causal statements whose cause-term purports to refer to an event (or a fact) and whose effect-term purports to refer to an event express a relationship between an event (or a fact) and an event.

Thus, both Davidson and Kim would claim that the singular causal Statement

(1) Socrates’ drinking hemlock at dusk caused his death expresses a relationship between two events, referred to, respectively, by the expressions “Socrates’ drinking hemlock at dusk” and “his death.” For Kim, but not for Davidson, an event is analyzable as a thing's having a property at or during a time. The event of Socrates’ drinking hemlock at dusk consists of Socrates’ at a certain time having the property of drinking hemlock at dusk. I shall not here try to choose between their respective theories of events but will only note that both theorists would say that (1) expresses a relationship between events however the latter are to be construed.

This paper is concerned with the genesis of mathematical knowledge. While some philosophers might argue that mathematics has no real subject matter and thus is not a body of knowledge, I will not try to dissuade them directly. (One might do so by developing a theory of meaning and truth, which together with observations from the sociology of mathematics would imply that mathematical knowledge exists. Mathematicians do seem to make knowledge claims, so all one needs is a theory which shows that here at least appearances are real.) I shall not attempt such a refutation because it seems clear to me that mathematicians do know such things as the Mean Value Theorem, The Fundamental Theorem of Arithmetic, Godel's Theorems, etc. Moreover, this is much more evident to me than any philosophical view of mathematics I know of — including my own. So I am going to take mathematics as my starting point.

In this paper I wish to propose and defend a form of the Representative Theory of Perception (RTP). According to this version of the theory, when a subject perceives some object x to be in a state P1 he does so by being aware of some modfication M1 (from a range M1,M2, … Mn) of some object E. The subject's way of perceiving any one of a range of objects x,y,z, … is that of being aware of some modification of E (leaving it open whether E is the subject himself, a person, an ego or a sense-field, e.g., a visual field). It will be a necessary condition of someone's perceiving x to have property P1 that E be modified in some way. And in order for the modifications of E to be representations, a given property P1 will need to be causally related, characteristically, to one or more specific modifications, M1,M2, … Mn, of E. (For simplicity, instead of taking the physical states or modifications P1,P2, … Pn to be states of individual physical objects, I shall assume them to be modifications of the physical world (in toto) symbolized by ‘X'.)

The notion of possible worlds — once an abstruse offspring of Leibnizian theology — seems to enjoy a new lease on life in the hands of contemporary modal logicians and semanticists. The phrase “possible world” crops up with increasing frequency, and, as it is the case with many philosophical catchwords, its very familarity creates a presumption of understanding. Yet, although some related problems, particularly the one concerning cross-world identification of individuals, have received some critical scrutiny, the very idea of possible worlds has been borrowed from Leibniz, and its intelligibility taken for granted even without its theological roots.

My misgivings about this notion are not due to a prejudice against the concept of the possible itself: far from “passing” away, it remains a fundamental and irreplaceable element of human thought.

A convention is a regularity in behavior, sustained by a system of preferences and expectations, that has a possible alternative: so claims David Lewis. That it is not, is what I hope to show.

In Convention; A Philosophical Study, Lewis gives an analysis of what he hopes is “our common, established concept of convention.” In developing his analysis, Lewis extensively employs the technical vocabulary of game theory. According to Lewis conventions are established by “populations” in order to solve recurring “coordination problems.” Lewis gives eleven examples of “coordination problems” which he claims are solved by convention. One example is the following:

Quine's arguments against the attribution of essential properties de re to individuals have been the motivation for attempts at reinstating essentialism as a respectable metaphysical thesis and at defending the coherence of modal logic in general.

I shall argue here along somewhat different lines, that the particular version of essentialism Quine objects to is in fact untenable but that this conclusion is far from entailing a commitment to some version of conventionalism, and in particular that it does not entail the view that the only kind of necessity that is coherent is de dicto necessity.

In what follows, I shall assume, without arguing for it, that de re essentialism and subjunctive conditionals are intimately related, and in particular, that any version of de re essentialism which conflicts with our basic intuitions about subjunctive conditionals is untenable.

There is a certain view abroad in the land concerning the philosophical problems raised by Tarskian semantics. This view has it that a Tarskian theory of truth in a language accomplishes nothing of interest beyond the definition of truth in terms of satisfaction, and, further, that what is missing — the only thing that would yield a solution to the philosophical problem of truth when added to Tarskian semantics — is a reduction of satisfaction to a non-semantic (and ultimately physical) relation. It seems to me that this view either misidentifies the philosophical problem altogether, or encourages a seriously misleading picture of the nature of the problem.

The view I have in mind is nowhere more persuasively at work than in a recent paper by Hartry Field.1 In this paper Field argues that a Tarskian theory of truth for a natural language is impossible if we insist on Tarski's case-by-case elimination of ‘satisfies'. More fundamentally, however, he argues that a Tarskian theory could provide nothing of philosophical interest beyond the admittedly interesting reduction of truth to satisfaction, and his ground for this claim is, roughly, that a Tarskian theory does not reduce its primitive semantic relation — satisfaction — to a non-semantic relation.

Can ‘disposition’ be understood as a contrast term, the contrast being ‘occurrence'? Put it another way: do ‘dispositional predicate’ and ‘occurrent predicate’ form a contrast pair? I shall show that if ‘occurrent’ is taken as simply meaning ‘non-dispositional', then ‘occurrent’ has no applications (i.e., no predicates are occurrent (Sec. 2). However, if ‘occurrent’ is given an independent meaning so that predicates like ‘break', ‘bend', ‘disintegrate’ etc. are occurrent predicates, then it is not the contrast of ‘dispositional'. Its contrast is rather what I shall call ‘remainant’ (Sec. 3).

In the by now well known talks he gave at Princeton, Saul Kripke claimed that “[t]heoretical identities … are generally identities involving two rigid designators and therefore are examples of the necessary a posteriori.” (Published as “Naming and Necessity,” in G. Harmon and D. Davidson, eds., Semantics of Natural Language (Dordrecht, 1972) 253-355; (hereafter referred to as “NN”; this quote p. 331.) A rigid designator is an expression that designates the same object in all possible worlds when it is used. So Kripke is claiming that ‘Water is H20’ and ‘Heat is the motion of molecules’ are generally identities involving expressions like ‘water’ and ‘the motion of molecules’ which designate the same objects in all possible worlds. If the identity statement is true, both sides designate the same object rigidly, i.e., in all possible worlds, and therefore the statement is necessarily true. On the other hand, whether it is true is determined ultimately by appeal to experience. It follows that if true, the identity is necessary a posteriori.

Proper recognition came belatedly to the work of Karl Popper. The novelty and power of his comprehensive philosophy went largely unnoticed for decades, his views being misapprehended, to the extent they were apprehended at all, as an uncompelling variation on the dominant Positivist theme. In the past two decades, however, recognition has become widespread. He can lay claim to being the initial figure in a vital and flourishing tradition in the contemporary literature; his views find expression even in undergraduate curricula; and he finds himself the welcome subject of a two-volume issue in the Schilpp series, The Library of Living Philosophers. The appearance of this collection is the occasion of this notice.

Unfortunately, the spectre of misapprehension surrounds Popper still, and haunts what should have been some of the most interesting pages in this collection. At issue here is Popper's conception of rational scientific methodology, a conception which has received substantial criticism from inductivist and anti-inductivist quarters alike. Partly from conviction, and partly for the sake of argument (Popper is an anti-inductivist), I shall adopt the latter viewpoint in this essay. My sympathies, however, will be with his critics.

In these days of inflation, perhaps we should not be surprised that the fourteenth and latest addition to the Library of Living Philosophers, should require two volumes. Previous subjects, including Bertrand Russell, Albert Einstein and G.E. Moore, were adequately accomodated within the covers of one volume. This expansion is hardly justified by the contents of the volumes. The most interesting and useful material is to be found in Popper's opening autobiographical section, but the other contributors and critics for the most part are disappointing in their lack of critical bite.

For this reason perhaps Popper's reply to his critics adds little to our understanding of his philosophical positions, although even when his critic is as acute as A.J. Ayer, Popper's reply has an air of defensive obfuscation.

Another recent book giving a brief but clear exposition of Popper's philosophy by a former student, Bryan Magee also fails to give his views the critical examination they deserve.