Efforts to bare the psychonomic nature of the semantic reference (representation) relation have been remarkably scanty; in fact, the only contemporary account developed with any care is the one proposed by Osgood. However, not even Osgood has looked deeply at the difficulties that beset any attempt to analyze reference in terms of common effects appropriately shared by a symbol and its significate.

We summarize the evidence and arguments usually employed against the use of the profit maximization assumption in microeconomic theory, and then pass directly to the methodological arguments. Two arguments are considered. The first summarizes positions which have been taken by various defenders of the “people who think a theory should be given up just because it is false are naive and confused” view. To rebut this view, we develop a scheme for classifying theoretical assumptions and show that the specious plausibility of the argument in question derives from ignoring the distinctions we have drawn. The second argument attempts to show that a better theory is more likely to be attained by continuing the development of the orthodox theory. We show that this view is mistaken and then discuss steps which have been taken to develop a better theory.

Probability measures can be constructed using the measure-theoretic techniques of Caratheodory and Hausdorff. Under these constructions one obtains first an outer measure over “events” or “propositions.” Then, if one restricts this outer measure to the measurable propositions, one finally obtains a classical probability theory. What I argue is that outer measures can also be used to yield the structures of probability theories in quantum mechanics, provided we permit them to range over at least some unmeasurable propositions. I thereby show that nonclassical probability theories can be seen to arise naturally within the framework of possible worlds semantics.

Following an account of the incommensurability argument, an objection, based on assumptions concerning rival theories, is examined and rejected. This rejection leads to an alternative direction of criticism of incommensurability, a direction that involves the articulation of comparative standards of theory evaluation that are independent of meaning invariance.

The concept of “fitness” is a notion of central importance to evolutionary theory. Yet the interpretation of this concept and its role in explanations of evolutionary phenomena have remained obscure. We provide a propensity interpretation of fitness, which we argue captures the intended reference of this term as it is used by evolutionary theorists. Using the propensity interpretation of fitness, we provide a Hempelian reconstruction of explanations of evolutionary phenomena, and we show why charges of circularity which have been levelled against explanations in evolutionary theory are mistaken. Finally, we provide a definition of natural selection which follows from the propensity interpretation of fitness, and which handles all the types of selection discussed by biologists, thus improving on extant definitions.

We examine critically the interdependence between science and philosophy which Sklar asserts in Space, Time, and Spacetime. We find that such a view makes it difficult to criticize the ideas of science, like that of absolute space, on their own merits, without importing extraneous philosophical associations. It also impedes appreciation of the importance, and subtlety, of explanation in scientific theory. As a result, particular explanations, such as the one Newton offered of his bucket experiment, are dismissed facilely—indeed, all geometric explanation appears illegitimate—, and such general attitudes as conventionalism appear to lack rationale.

In a previous paper (Simon and Groen 1973) it was proposed that theories of empirical phenomena should satisfy conditions of finite and irrevocable testability (FITness conditions). Roughly speaking, a theory is finitely testable if, for every set of observations that falsifies the theory, there exists a finite subset of observations that falsifies it. A theory is irrevocably testable if, for any finite set of observations that falsifies the theory, any superset of observations that includes that set also falsifies it—a theory falsified by observations cannot be resuscitated by adding new observations to the original set.

Frank Jackson and Robert Pargetter (1977) propose a method for synchronizing clocks at rest at distant points of an inertial system in Euclidean space, which, they claim, (i) does not depend on Einstein's signalling method (Einstein 1923, pp. 27–29) and (ii) provides a basis for denying the conventionality of distant simultaneity. I am afraid, however, that the new method presupposes that the simultaneity of distant events relatively to the chosen inertial system has been already determined by Einstein's or some other method. Jackson and Pargetter describe their method as follows:

In their recent paper, Jackson and Pargetter (1977) have attempted to prove that simultaneity is not a matter of convention. They attempt to show that it is an empirical fact whether or not a moving rod A′B′ is perpendicular to a plane XY. If it is an empirical fact, then such a perpendicular rod can be used to synchronize clocks. We shall show that it is not an empirical fact, but it is in fact synchronization dependent.

Robert Torretti's objection is verificationism writ large. We reply that verificationism is to be rejected. Carlo Giannoni's objection is that our test for tilting fails because the rod might tilt and yet no current flow through its mid-point. We reply that nevertheless we can test for tilting because there would still be differences detectable with a glavanometer.

In a recent paper, William K. Goosens (1978) objects to the arguments I set out some time ago attacking the logical empiricist analysis of reduction as applied to genetics (Hull 1972, 1973, 1974, 1976a). In these works I did not argue against the claim that Mendelian genetics was being reduced to molecular biology. Nor did I conclude, as Goosens asserts, that in the case of genetics, “reduction is insignificant” (p. 93). To the contrary, I repeatedly stated that, “given our pre-analytic intuitions about reduction,” the reduction of Mendelian to molecular genetics “is a case of reduction, a paradigm case” (Hull 1974 p. 44). And in agreement with Kenneth Schaffner (1967) I argued that reduction “in some very important, pre-analytic sense has taken place and is taking place in genetics” (Hull 1973, p. 634). The target of my objections to reduction in genetics was not “reduction” in some pre-analytic sense. Nor was it the various explications which have appeared since. Rather it was the notion of theory reduction set out a generation ago by such logical empiricists as Ernest Nagel (1961), The particular version of this analysis which I chose to attack was that presented by Schaffner (1967, 1969); for Schaffner's later views, see his (1974, 1976, 1977). I chose Schaffner's explication to attack because I thought it was the best of its kind and because it was the only defense of the logical empiricist analysis at the time which used genetics as one of its chief examples. Contrary to Goosens' assumption, in criticizing Schaffner's explication, I did not thereby “endorse” it (p. 82).

Marjorie Grene has argued in a number of contexts (1971, 1974, 1976) that any attempt to reduce the explanation of human actions to “mechanistic” explanations is doomed to failure in advance because it is absurd. “This argument examines the status of the reductivist thesis in its own terms and reduces it to absurdity …” (1971, p. 55); “the attempt to reduce human purposive, or ‘intentional,‘ action to physiology and ultimately to physics and chemistry is an absurdity rather than simply a confusion” (1976, p. 183). I wish to show that this argument is fallacious because it involves a petitio principii; and if Plato can be credited with introducing the same argument in the Theaetetus, as she suggests, so much the worse for Plato.

Desmond Clarke's remarks on “my” absurdity argument are puzzling. i) Although I do indeed still believe it to be a valid argument, I certainly would not claim credit for it. I believe that “Reducibility: Another Side Issue?” put the general problem of the reducibility of mind into a somewhat unorthodox context, but the particular claim Clarke is attacking forms only one very unoriginal step in the general argument of that essay. ii) Some points that Clarke makes I would certainly agree to, and believe I have agreed to in some of the texts he cites, while others are far from anything I meant to convey. Under Clarke's (a), 1) of course the “antinomies” of the paper above referred to trade on ambiguities. That's one of the points of the paper. 2) in both “Reducibility … (Grene 1971)” and “To Have A Mind …” (Grene 1976), I have tried to specify the sense in which “mechanism” was being used. That I do not take such terms as self-explanatory should be clear, for instance, from the remarks in the latter paper that immediately follow the passage Clarke criticizes. 3) I have no idea how physics is to develop in the future, nor did I wish to suggest that I did. With the rest of Clarke's (a) I entirely agree. As to his (b), I fail to see its relevance to my argument, unless it simply leads up to (c) and (d), with which, again, I wholeheartedly concur. Nancy Maull's paper, which Clarke refers to, for example, seems to me to shed great light on what used to be thought the problem of reduction; I don't see how her argument conflicts with “mine.” iii) Clarke's two “interpretations” of “my” argument are even more puzzling. The first paragraph is clear, as far as it goes. But the second is entirely murky. Perhaps I may try to reconstruct it. “The question for the scientist of a reductivist disposition in this context,” Clarke writes, “is: would it be possible to construct an alternative theory (which is not intentional, teleological, etc.) which adequately explains the same phenomena as the human sciences?” Fair enough. But apparently one is not allowed to reflect about an answer to this question. One has to wait and see if such a theory does develop. And meantime to notice that any theory by its very nature is intentional: this is to use some part of the human sciences to establish the explanatory power of all the human sciences. Is that what is meant? If so, I would reply: that the claim that theories are inherently intentional is not a thesis of any “human science” but a philosophical thesis about the nature of intelligible discourse. If a wholly non-intentional theory of human action were stated it would invalidate itself as theory. Thus, to respond to the major accusation of (ii), we would have an answer to Clarke's question such that the question could not have been asked. What we have, then, is not a petitio but a valid reductio ad absurdum.

It has been suggested by several philosophers (Lewis 1976, Putnam 1962) that many of the so-called paradoxes of backward time travel can be resolved if we conceive of the backward time traveller as having a zig-zag or N-shaped world line in spacetime. In this I am in general agreement. But there is still a problem in conceiving of backward time travel this way. In this note I will show how we can solve this problem by conceiving of backward time travel in terms of the closed time-like world lines in certain general relativistic space-times. Indeed, it has often been claimed (cf. Dwyer 1978, Godei 1949, Tipler 1974) that such world models as Godei spacetime show (or contribute to showing) that backward time travel is conceivable. Our discussion will help to make clear just why this claim is correct.