Paul Humphreys

(? - 2022)

This chapter provides six arguments in favor of scientifically based ontology and against speculative ontology, a branch of analytic metaphysics. Parts of contemporary speculative ontology are untenable because they are factually false; intuitions are not domain-invariant; conceptual analysis is too closely tied to everyday experience; what counts as an acceptable philosophical idealization is left unarticulated; the world is not scale-invariant; and anthropocentric epistemology does not always minimize epistemic risk. Nevertheless, specifically philosophical arguments are necessary when making ontological claims and complete deference to scientific consensus is unreasonable. The overall strategy is to recognize that different domains of reality require different methods of discovery and justification, and that a significant part of contemporary metaphysics is employing methods that are inappropriate to its goals.

Rather than distinguishing between the natural and social sciences on the basis of their subject matter, it is often philosophically fruitful to start with the fact that most social phenomena are not amenable to investigation by laboratory experimentation. One can, in some cases, use randomized experimental designs to investigate causal relationships between social phenomena, but ethical and practical difficulties often preclude the use of such techniques. In the absence of these two empirical approaches, two principal alternatives are often used: causal models and quasi-experimentation. I shall focus on the former here, because the techniques involved offer an interesting example of how the context of investigation can affect the structural form of mathematical models used to describe phenomena. I have an ulterior motive, however. Little attempt has been made to link the substantive work on causal modeling with philosophical theories of probabilistic causality.

The elements of structural models used in the social sciences are built up from four fundamental assumptions. It is then shown how the central idea of qualitative probabilistic causality follows as a special case of this covariational account. The relationships of both instrumentalism and common cause arguments for scientific realism to these structures is demonstrated. It is concluded that a predictivist argument against a thoroughgoing instrumentalism can be given, and hence why the difference between experimental and non-experimental contexts is important to arguments for scientific realism.

A great deal of attention has been paid by philosophers to the use of computers in the modelling of human cognitive capacities and in the construction of intelligent artifacts. This emphasis has tended to obscure the fact that most of the high-level computing power in science is deployed in what appears to be a much less exciting activity: solving equations. This apparently mundane set of applications reflects the historical origins of modem computing, in the sense that most of the early computers in Britain and the U.S. were devices built to numerically attack mathematical problems that were hard, if not impossible, to solve non-numerically, especially in the areas of ballistics and fluid dynamics. The latter area was especially important for the development of atomic weapons at Los Alamos, and it is still true that a large portion of the supercomputing capacity of the United States is concentrated at weapons development laboratories such as Los Alamos and Lawrence Livermore.

The topic of this session is “physical randomness”. It might be doubted whether such a subject exists, for definitions of randomness have hitherto almost all been mathematical in nature. The only exceptions of which I am aware are the preceding paper by Benioff and a paper by Wesley Salmon. These attempts to inject some empirical content into randomness are highly desirable. But anyone attempting to formulate a physically based definition of randomness should at some point make clear what the connection is (if any) with a more traditional notion of disorder - that of indeterminism. Repeated reference to quantum mechanical examples whenever physical randomness is discussed indicates that a primary motivation for considering physical randomness to be important is because of the current belief that data sequences associated with quantum mechanical experiments are irreducibly random. (As an indication of this, in any situation in which physical randomness is discussed, translate the remarks about quantum phenomena into remarks about coin tossing, and they will lose much of their interest.)

This handbook provides both an overview of state-of-the-art scholarship in philosophy of science, as well as a guide to new directions in the discipline. Section I contains broad overviews of the main lines of research and the state of established knowledge in six principal areas of the discipline, including computational, physical, biological, psychological and social sciences, as well as general philosophy of science. Section II covers what are considered to be the traditional topics in the philosophy of science, such as causation, probability, models, ethics and values, and explanation. Section III identifies new areas of investigation that show promise of becoming important areas of research, including the philosophy of astronomy and astrophysics, data, complexity theory, neuroscience, simulations, post-Kuhnian philosophy, post-empiricist epistemology, and emergence. Most chapters are accessible to scientifically educated non-philosophers as well as to professional philosophers, and the contributors - all leading researchers in their field -- bring diverse perspectives from the North American, European, and Australasian research communities. This volume is an essential resource for scholars and students.

This book provides a post-positivist theory of deterministic and probabilistic causality that supports both quantitative and qualitative explanations. Features of particular interest include the ability to provide true explanations in contexts where our knowledge is incomplete, a systematic interpretation of causal modeling techniques in the social sciences, and a direct realist view of causal relations that is compatible with a liberal empiricism. The book should be of wide interest to both philosophers and scientists. Originally published in 1989. The Princeton Legacy Library uses the latest print-on-demand technology to again make available previously out-of-print books from the distinguished backlist of Princeton University Press. These editions preserve the original texts of these important books while presenting them in durable paperback and hardcover editions. The goal of the Princeton Legacy Library is to vastly increase access to the rich scholarly heritage found in the thousands of books published by Princeton University Press since its founding in 1905.

This collection of essays is the definitive version of a widely discussed debate over the origins of the New Theory of Reference. In new articles, written especially for this volume, Quentin Smith and Scott Soames, the original participants in the debate, elaborate their positions on who was responsible for the ideas that Saul Kripke presented in his Naming and Necessity. They are joined by John Burgess, who weighs in on the side of Soames, while Smith adds a further dimension in discussing the contributions of philosophers such as Follesdal, Geach, Hintikka, and Plantinga. Also included are lengthy excerpts from Follesdal's 1961 Harvard dissertation and a careful examination by Sten Lindstrom of the respective contributions of Kripke and Stig Kanger to the development of modal semantics. The collection will be essential reading for anyone acquainted with these influential ideas.

It is argued in this paper that although much attention has been paid to causal chains and common causes within the literature on probabilistic causality, a primary virtue of that approach is its ability to deal with cases of multiple causation. In doing so some ways are indicated in which contemporary sine qua non analyses of causation are too narrow (and ways in which probabilistic causality is not) and an argument by Reichenbach designed to provide a basis for the asymmetry of causation is refined. The importance of referring causal claims to an abstract model is also emphasized.