It can not be our purpose to give here a complete account of the phenomenological history of mathematical doctrine. It will be enough to refer to the battle of opinions in mathematical theory which was waged within the last eighty-five years, since Riemann's inaugural lecture on Non-Euclidean Geometry. Furthermore, the revolution which Einstein's theory of general relativity created is indicative of the complete absence of any general awareness that mathematics as a science has any relation to social reality. If we are going to establish such relations, we are conscious of the fact that they would be hotly contended by many mathematicians.

Adaptationism is often taken to be the thesis that most traits are adaptations. To assess this thesis, it seems we must be able to establish either an exhaustive set of all traits or a representative sample of this set. Either task requires a more systematic and principled way of individuating traits than is currently available. Moreover, different trait individuation criteria can make adaptationism turn out true or false. For instance, individuation based on natural selection may render adaptationism true, but may do so by presupposing adaptationism. In this article, we show how adaptationism depends on trait individuation and that the latter is an open and unsolved problem.

Building on two nonproblematic claims, I argue for a qualified endorsement of Hume's intuition that there must be a time-difference between cause and effect. Those claims are: (i) that the statement ‘A caused B’ is meaningful only if we have a criterion for saying ‘A’ and ‘B’ refer to distinct events; and (ii) that an adequate view of what it is to be an event must illuminate the enterprise of seeking to establish a singular causal statement. Specifically, I argue there must be a time-difference when cause and effect are modifications of the same physical object.

The semantic confusion in Europe about the term “anthropology” has of late been considerable. On the one hand there is meant by it, and quite justifiably, human biology and medical anthropology. On the other hand, the work of some contemporary thinkers, under the name of “philosophical anthropology,” has recently gone beyond the narrower compass. This has been noticeable at both the German and the international European philosophical conventions since the last war. In addition to this, there appeared the term “Kulturanthropologie,”—for the first time probably as the title of E. Rothacker's contribution to Systematische Philosophie, a collection of essays edited by Nicolai Hartmann and published in 1942. But “Kulturanthropologie,” in Rothacker's sense, is not a translation of “cultural anthropology,” which has in recent years grown to such eminence among scholarly fields in England and America. In the United States, (1) “anthropology” designates a group of scholarly fields ranging on the one hand from biological anthropology to archaeology and the study of the beginnings of history on the other, often including history, linguistics, and other realms of research; (2) it is neither one of the natural sciences nor of the humanities but one of the social sciences; and (3) the actually new results of this young discipline, especially at Harvard, Yale, and the University of Chicago, are arrived at by means of team work by anthropologists, psychologists, and sociologists, which in Germany has not been the case so far.

Feyerabend's “Classical Empiricism” (1970) draws on a 17th century Jesuit argument against Protestant fundamentalism. The argument is very general, and applies to any simple foundationalist epistemology. Feyerabend uses it against Classical Empiricism—roughly, the view that what is to be believed is exactly what experience establishes, and no more—which he identifies as among other things Newton's “dogmatic ideology.”

In this paper, I examine the claim that any physical theory will have an extremely limited domain of application because 1) we have to use distinct theories to model different situations in the world and 2) no theory has enough textbook models to handle anything beyond a highly simplified situation. Against the first claim, I show that many examples used to bolster it are actually instances of application of the very same classical theory rather than disjoint theories. Thus, there is a hidden unity to the world of classical physics that is usually overlooked (by, for example, Nancy Cartwright who argues for the claims above). Against the second claim, I show that the practice of classical physics involves an enormous (infinite) number of models the use of which cannot be written off as merely ad hoc. Thus, although classical physics cannot, of course, model every situation in nature, it has a much larger domain than some would have us believe.

In this article, I discuss calls for access to empirical data within controversies about climate science, as revealed and highlighted by the publication of the e-mail correspondence involving scientists at the Climatic Research Unit at the University of East Anglia in 2009. I identify several arguments advanced for and against the sharing of scientific data. My conclusions are that, whereas transparency in science is to be valued, appeals to an unproblematic category of ‘empirical data’ in climate science do not reflect the complexities of scientific practice in this field.

The logical empiricists knew that scientific theories sometimes arise out of the attempt to reconcile or unify two existing theories. They also thought that, at best, old theories would be retained as approximations to their successors. Kuhn lost both insights when he rejected the logical empiricists' formal approach in favor of an exclusively historical and psychological one. But when Putnam tried to restore such ideas he failed to provide them with the historical support they require. An account of revolutionary unifications is defended as reconciling Putnam's realist ideas about methodology with Kuhn's historical approach.

It has been alleged that Bayesian usage of prior probabilities allows one to obtain empirical statements on the basis of no evidence whatever. We examine this charge with reference to several examples from the literature, arguing, first, that the difference between probabilities based on weighty evidence and those based on little evidence can be drawn in terms of the variance of a distribution. Moreover, qua summaries of vague prior knowledge, prior distributions only transmit the empirical information therein contained and, therefore, their consequences for long-run frequency behavior are “a priori” in at best a Pickwickian sense.

One of the most serious theoretical obstacles to contemporary spacetime substantivalism is Earman and Norton's hole argument. We argue that applying the bundle theory of substance to spacetime points allows spacetime substantivalists to escape the conclusion of this argument. Some philosophers have claimed that the bundle theory cannot be applied to substantival spacetime in this way due to problems in individuating spacetime points in symmetrical spacetimes. We demonstrate that it is possible to overcome these difficulties if spatiotemporal properties are viewed as tropes rather than universals.

I am grateful to Professor Wheatley for his note, [3], on my analysis of interrogatives, [1]. His comments bring out very clearly a number of considerations that deserve our closest attention. For example, he shows that if we can classify interrogatives as true and false—as I proposed to do—then we can properly inquire about what sentences contradict them, and what sentences are contingently or logically equivalent to them. Furthermore, he shows that, on my analysis, no indirect question can contradict any other indirect question and he accordingly, and correctly, looks for such contradictories among declarative sentences. (Different considerations than any that Professor Wheatley brought out will equally well show that no direct question will contradict any other direct question: “Is Freeman Governor of Minnesota?” and “Is not Freeman Governor of Minnesota?” express different concerns with one and the same topic of concern, hence, are equivalent.) Finally, he notes, [3], p. 54, that in my article, I have treated only inquisitive questions and have said nothing about deliberative or practical questions. While practical questions (as well as their practical answers) demand a different kind of treatment than that which Wheatley speculatively, albeit very tentatively, extrapolates for me, still these questions do form a kind of interrogative that did not get any attention in my article. (But the problems are so complex that I shall not try to say anything about them in this Reply.)

There is a sense in which a philosophic theory can be confirmed. We may ask what its effects were on the development of scientific theory,—did it clarify ideas and help open up new areas of research, or did it constrain the work of science? In this essay, we shall try to judge the significance of dialectical materialism from this standpoint. We shall be concerned with the bearing of this philosophy on scientific work, especially in the Soviet Union.

Now dialectical materialism is not a “philosophy” in the same sense which the word has when applied to, let us say, logical empiricism. The difference is evident when a writer like Haldane asks, for instance, how far scientific discoveries have verified the principles of dialectical materialism.' We could not ask in a similar way whether scientific discoveries have confirmed the principles of logical empiricism. For the latter affirms no factual propositions concerning the world; it enunciates solely the principle of verifiability, and there is clearly no scientific discovery which could possibly refute the requirement of verification. If it makes sense to ask how far scientific discoveries verify dialectical materialism, then the latter must obviously contain factual propositions concerning nature which should be stated in a way to permit of confirmation of rejection.

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.

Evolutionary models can explain the dynamics of populations, how genetic, genotypic, or phenotypic frequencies change with time. Models incorporating chance, or drift, predict specific patterns of change. These are illustrated using classic work on blood types by Cavalli-Sforza and his collaborators in the Parma Valley of Italy, in which the theoretically predicted patterns are exhibited in human populations. These data and the models display properties of ensembles of populations. The explanatory problem needs to be understood in terms of how likely an observed change, in either a population or an ensemble, would be under drift alone; this is fundamentally a matter of chance. Understood in this way, issues of drift and chance undercut most recent philosophical, but not biological, discussions of the role of “genetic drift.”

John Greenwood (this issue) claims that neglect of an important methodological distinction has contributed directly to the “epistemic impoverishment” of empirical studies of all forms of professional psychotherapy. I challenge this claim, as well as other important claims he makes about the efficacy of psychoanalysis and other forms of psychotherapy.

A recent article (Leeds and Healey 1996) argues that the modal interpretation (Copenhagen variant) of quantum mechanics does not do justice to immediately repeated non-disturbing measurements. This objection has been raised before, but the article presents it in a new, detailed, precise form. I show that the objection is mistaken.

Consider a red circle, and suppose it is a paradigmatic thing. Some philosophers maintain that a thing is ontologically analyzable into a particular exemplifying properties, those properties truly ascribed to the thing by the customary words. Our red circle, then, consists of a particular, say a; two properties, red and circle; and exemplification, a tie tying a, red, and circle into “the red circle.” Upon this analysis, a is bare, i.e., not re-recognizable as such, whereas red and circle are natured, i.e., not bare, and, further, are universal, i.e., they may be exemplified by several particulars (or, somewhat differently, they may be in several things).