Carnap's attempt to develop an inductive logic has been criticized on a variety of grounds, and while there may be some philosophers who believe that difficulties with Camap's approach can be overcome by further elaborations and modifications of his system, I think it is fair to say that the consensus is that the approach as a whole cannot succeed. In writing a paper on problems with inductive logic (and with Camap's approach in particular), I might therefore be accused of beating a dead horse. However, there are still some (e.g., Spirtes, Glymour and Scheines 1993) who seem to believe that purely formal methods for scientific inference can be developed. It may still then be useful to perfonn an autopsy on a dead horse when establishing the cause of death can shed light on issues of current concern.

Relativistic quantum theory poses a number of conceptual problems over and above the non-relativistic mechanics. From a purely mathematical point of view it is also much more sophisticated. Even the kinematic theory poses considerable difficulties: locality, antimatter, negative energy, and charge are already systematically linked in ways that cannot simply be paraphrased. A further complication is that in the midst of this structure we have what is usually called the Newton-Wigner representation (or, in the case of the Dirac theory, the Foldy-Wouthuysen representation), in which fields and states are no longer covariantly described and where the “local” self-adjoint quantities (NW-local observables) do not obey microcausality, but only satisfy equal-time commutators with respect to a particular inertial frame. If we pass to this represenation, we obtain at a stroke the basic structure of non-relativistic quantum field theory; it is not too hard to descend from that to a many-particle mechanics, and to recover the usual definitions of localization.

This essay explores one of the fundamental ways in which feminist methodology in philosophy of science differs from most standard philosophy of science. Methodology in general arises from goals which the methods are designed to reach. Feminist goals in philosophy have been scandalous precisely because they are avowed political interventions into what has been considered a cool, apolitical academic activity. Another way to put this point is that feminism self-consciously inserts certain moral and political values into an activity considered to be neutral among moral and political values; it advocates a moral and political point of view in an area presumed to be objective in the sense that it is without any point of view. And precisely because the goals of feminist efforts in philosophy of science are moral and political ones, they appear inimical to the central goals of philosophy of science itself—the morally neutral and apolitical production of theories about how science works.

In Bayes or Bust?, John Earman (1992, 104-113) attempts to express in Bayesian terms a sense of “projectibility” in which it is logically impossible for “All emeralds are green” and “All emeralds are grue” simultaneously to be projectible. I argue that Earman overlooks an important sense in which these two hypotheses cannot both be projectible. This sense is important because it connects projectibility to lawlikeness, as Goodman intended. Whether this suggests a solution to Goodman's riddle remains unsettled, awaiting an account of lawlikeness. I explore one avenue that might prove illuminating.

Suppose we discover that object a is G, where our background beliefs K include that Ea. Earman notes that by Bayes's theorem, hypothesis H (“All E's are G”) is confirmed—i.e., .

According to general wisdom, quasianalysis belongs to the large family of Carnap's ingenious, but finally failed contributions to epistemology and philosophy of science. In this paper I want to show that this is not the case. Rather, Carnapian quasianalysis is to be considered as a promising theory of a representational constitution of scientific objects. That is to say, I intend to embed Carnap's approach of quasianalytical constitution in the framework of a general theory of meaningful representation (cf. Mundy (1986)).

The outline of this paper is as follows: In section 2 I recall the basics of the quasianalytical approach, taking into consideration not only the well-known account in “Der Logische Aufbau der Welt” (Aufbau) but also a rather unknown first version of quasianalysis (“Quasizerlegung”) which Carnap developed in an unpublished manuscript written in 1923. This paper deserves attention not only for philosophicohistorical reasons, rather it contains quite a lot of interesting features of the quasianalytical approach which do not appear in the Aufbau account.

There are many reasons why my “special science,” psychology—and more specifically cognitive psychology—lends itself to particularly fruitful feminist philosophical analysis. Many central questions in feminist philosophy of science and feminist science studies are well illuminated there, especially those that are reflected in specific methodological issues. For example, in feminist science critiques tensions have arisen about the role of empiricism: many stress the importance of paying careful attention to certain strictures of empiricist methodology (close attention to accuracy in data collection, hypothesis formation, replication of “findings” about sex differences, and so on), yet concerns have also been raised about empiricism as an overall feminist epistemological framework—specifically, concerns about whether such a framework can adequately entertain the most transformative feminist political projects. I think such tensions in feminist science criticism are best addressed by paying close attention to their deployment in particular working contexts of feminist science.

Commonly-cited criteria for the assessment of hypotheses, research projects, and theories include empirical adequacy, simplicity, conservatism, explanatory power, generality of scope, fecundity, reproducibility, and interconnectedness with other going theories. The list of criteria for good scientific practice typically includes objectivity (though what objectivity comes to is much disputed), respect for evidence, open mindedness, and tentativeness. Feminist scientists and science scholars have of late proposed additional and, in some cases, alternative criteria (e.g., Biology and Gender Study Group 1989, Bleier 1984, Fausto-Sterling 1985, Haraway 1978, Keller 1983 and 1985, Longino 1994, and Nelson 1990). We here consider the nature and rationale of some of these proposed criteria, focusing on several articulated by Helen Longino. A research program in neuroendocrinology investigating a hormonal basis for alleged sex-differentiated lateralization and feminist critiques of this research serve as our case study.

In the philosophy of science, we are to assess critically and on their intrinsic merits various proposals for a consistent interpretation of quantum mechanics, including resolutions of the measurement problem and accounts of the long-range Bell correlations. In this paper I suggest that the terms of debate may have been so severely and unduly constrained by the reigning orthodoxy that we labor unproductively with an unhelpful vocabulary and set of definitions and distinctions. I first review this situation and how we arrived there. Then I present an alternative conceptual framework, free of many of the standard conundrums, and ask why we seem unwilling to pursue it as a serious possibility.

I begin by sampling how some of the central tenets of the standard, or “Copenhagen”, theory of quantum mechanics came to be formulated. As is typical for his writing on broader philosophical issues, Niels Bohr's pronouncements on the interpretation of quantum mechanics are often difficult to understand and at times just plain opaque.

Why do scientists accept or reject theories? More specifically: why do they change from one theory to another? What is the role of empirical tests in the evaluation of theories?

This paper focuses on a narrowly-defined question: in judging theories, do scientists give greater weight (other things being equal) to successful novel predictions than to successful deductions of previously-known facts? The affirmative answer is called the “predictivist thesis” (Maher 1988).

It is primarily philosophers who are interested in this question, and they have treated it mostly as a normative or logical problem. Can the writings of historians of science tell us how scientists have treated novel predictions in the past? Until recently historians have rarely addressed this point.

Here I describe and defend a version of space-time substantivalism. My account is meant partly as a reply to an argument against a particular substantivalist position that can be found both in John Earman and John Norton's “What price spacetime substantivalism? The hole story”, and Earman's World Enough and Spacetime. Their argument, the Hole Argument, purports to show that substantivalism leads to a radical form of indeterminism within a class of theories that includes our best theory of spacetime, namely, General Relativity (GTR). The hole argument contends that the substantivalist must view diffeomorphic models of a spacetime theory as representing genuinely distinct physical situations. It is this contention I argue against.

The version of substantivalism I present, although not a traditional account, incorporates the important features of traditional substantivalist views.

Building on an idea by Pitowsky (1990), David Malament (private communications), Hogarth (1992) and Earman and Norton (1993) have shown how it is possible to perform computational supertasks—that is, an infinite number of computational steps in a finite span of time—in a kind of relativistic spacetime that Earman and Norton (1993) have dubbed a Malament-Hogarth spacetime.

Definition 1 A spacetime (M,g) is Malament-Hogarth just when there is a future endless curve with past endpoint and a point q∊M such that .

(Hereafter, the symbols “q” and “λ” are assumed to have the properties they have in Definition 1.1 shall also speak of a “λ-curve”.)

Various examples of Malament-Hogarth (hereafter, M-H) spacetimes are given in Hogarth (1992), but the following artificial example from Earman and Norton (1993) is perhaps the simplest.

A common device in popular presentations of science is a sequence of views from cosmic to terrestrial to local to microscopic, thereby placing the subject to which the program is devoted in a proper perspective. I wish to use an adaptation of this device to place the announced topic of our panel — “Do Explanations or Predictions Provide More Evidential Support for Scientific Theories?” — in perspective. My four steps, from the largest to the smallest scale, are the following:

1. 1. A brief summary of the world view suggested by the discoveries of the natural sciences and by philosophical reflections on them;

2. 2. a consideration of the methodology for scientific investigation, upon assumption that this world view is approximately correct;

3. 3. the formulation of a version of Bayesian scientific inference satisfying the desiderata of step 2;

4. 4. the topic of our panel — explanation vs. prediction.

According to a standard view, predictions of new phenomena provide stronger evidence for a theory than explanations of old ones. More guardedly, a theory that predicts phenomena that did not prompt the initial formulation of that theory is better supported by those phenomena than is a theory by known phenomena that generated the theory in the first place. So say various philosophers of science, including William Whewell (1847) in the 19th century and Karl Popper (1959) in the 20th, to mention just two.

Stephen Brush takes issue with this on historical grounds. In a series of fascinating papers he argues that generally speaking scientists do not regard the fact that a theory predicts new phenomena, even ones of a kind totally different from those that prompted the theory in the first place, as providing better evidential support for that theory than is provided by already known facts explained by the theory.

Traditional empiricist epistemologies have raised major philosophical obstacles to a clear understanding of instruments. Empiricist philosophers have given scant attention to instruments as a separate topic of inquiry because the reliability of instruments is presumably reducible to the epistemology of common sense experience. Instruments function presumably to magnify our physiologically limited sensory capacities by linking the specimen's sensory properties to accessible empirical data; such data are then validated by the same empiricist standards used to access ordinary (middle-sized) phenomena. Thus, no epistemic insight is revealed by studying instrumentation per se, according to empiricists.

Critics of empiricism have championed the cause that all sensory data are theoryladen. Yet even this dictum has the effect of minimizing the philosophical significance of instruments. This is because many critics of empiricism are working within the empiricist's distinction between the subjectivity of theory and the apparent objectivity of data.

Scientists must often solve analytically or computationally intractable problems. Philosophers of science have problematized the nature and role of laws, theories and evidence by examining how scientists use abstractions (Cartwright 1990), idealizations (Laymon 1985, Brezinski et. al. 1990), approximations (Cartwright 1983; Laymon 1989; Ramsey 1990, 1992), and visual images (Wimsatt 1991, Ruse 1991) in such situations. Here, I problematize the nature and role of concepts by examining ideal types as a response to intractability. I focus on the work of Carl Wagner, a former director of the Max Planck Institute for Physical Chemistry in Goettingen. Wagner used ideal types in his research on the oxidation of metal alloys during the middle third of this century.

I argue that: natural scientists do use ideal typical concepts; these concepts are distinct from ‘standard’ concepts; and ideal typical concepts provide one means of connecting a tractable theoretical model with the experimental evidence.

The primary focus of philosophy of science in this century has been on theories and theoretical knowledge. Except for concern about whether phenomenal reports or reports of objects and events should count as data reports and whether data might be theory-laden and what consequences that might have for objectivity of science, data have been taken to be relatively unproblematic. But for scientists, questions about what are the data are often of paramount concern. Such controversies often overshadow controversies over theories or explanatory models. Scientists’ questions about data, moreover, do not turn on the philosophical questions just identified. Instead, for scientists, the overriding question is whether purported data is actually informative about the phenomenon in nature under investigation or whether it represents an artifact; that is, whether it was so much the product of the procedures used to generate the data that it could not be interpreted as providing information about the underlying phenomenon.

In the spring of 1994, Lindley Darden spent five months as a visitor in Joshua Lederberg's Laboratory for Molecular Genetics and Informatics (MGI) at Rockefeller University. Michael Cook is a Research Associate in the MGI Lab. This paper discusses reasoning strategies at use in that lab. In Part I, Lindley Darden recounts her experiences in the lab and discusses reasoning strategies that she observed in use there. In Part II, Michael Cook presents a technique of methodical hypothesis generation in the light of an anomaly for a model.

The reasoning strategies to be discussed in this paper are reasoning in hypothesis formation, reasoning in experimental design, and reasoning to generate hypotheses in the light of an anomaly. More specifically, the reasoning strategies may be characterized as (1) abstraction-instantiation (Darden, 1987; 1991; Darden and Cain, 1989), (2) the systematic scan (Lederberg, 1965), and (3) modular anomaly resolution (Darden, 1991; 1992; forthcoming).

In order to discuss the question of the reduction of chemistry it will be necessary to begin with a brief review of what philosophers mean by reduction in science. I follow most authors on this subject by starting with the writings of Nagel (Nagel 1961). As is well known, Nagel stipulates that two formal conditions, namely connectability and derivability should be fulfilled in order to say that reduction of theory T2 to theory T1 has occurred. In addition he stipulates a non formal condition, that the primary or reducing science should be supported by experimental evidence.

Furthermore Nagel considers that reductions occur in two main varieties which he calls homogeneous and heterogeneous respectively. In homogeneous reduction the terms used by the reducing theory are also common to the theory to be reduced. For example Galileo's science of freely falling bodies was conceived of as separate from the mechanics of celestial bodies.

Hacking (1983) introduces an attempt to defend scientific realism on the basis of the reality of theoretical entities. This position, which is called entity realism, is based on disconnecting the reality of theoretical entities from the truth and explanatory power of theories that account for them. In this way, two problems can be avoided. First, if theories about theoretical entities are rejected, the entities themselves do not have to go with them and the realist thesis that we can have knowledge of what exists in the world can be sustained. Second, theoretical entities, which will replace theories as the grounds for the realist position, would be protected from attacks on the validity of the inference to the best explanation which underlies classical or “theory” realism. In other words, theoretical entities would be able to survive the collapse of the inference to the best explanation.

In this paper, I revisit a problem in immunology and molecular genetics that I had first tried to understand some philosophical implications of over twenty years ago. At that time, some immunologists such as Mel Cohn at the Salk Institute, referred to it as the GOD problem, which was the acronym for Generator Of Diversity (also see Cohn's more recent discussion in his 1994, 41-48). In the early 1970s there were three or four different theories that had been proposed to account for the way in which antibody diversity is generated and considerable argument among the proponents of the different approaches to the problem (see Cohn, 1994). In 1971-72 I attempted to assess the strengths and weaknesses of the competing theories of antibody diversity from the perspective of a “logic of comparative theory evaluation” that had seemed to work reasonably well in physics (Schaffner, 1970), and I drafted an essay that ultimately sided with one of those theories (somatic mutation).