This brilliant and controversial work investigates the relationship between 'individual' and 'scientific' knowledge.

The philosophical theory of scientific explanation proposed here involves a radically new treatment of causality that accords with the pervasively statistical character of contemporary science. Wesley C. Salmon describes three fundamental conceptions of scientific explanation--the epistemic, modal, and ontic. He argues that the prevailing view is untenable and that the modal conception is scientifically out-dated. Significantly revising aspects of his earlier work, he defends a causal/mechanical theory that is a version of the ontic conception. Professor Salmon's theory furnishes a robust (...) argument for scientific realism akin to the argument that convinced twentieth-century physical scientists of the existence of atoms and molecules. To do justice to such notions as irreducibly statistical laws and statistical explanation, he offers a novel account of physical randomness. The transition from the "reviewed view" of scientific explanation to the causal/mechanical model requires fundamental rethinking of basic explanatory concepts. (shrink)

The final work of a distinguished physicist, this remarkable volume examines the emotive significance of time, the time order of mechanics, the time direction of thermodynamics and microstatistics, the time direction of macrostatistics, and the time of quantum physics. Coherent discussions include accounts of analytic methods of scientific philosophy in the investigation of probability, quantum mechanics, the theory of relativity, and causality. "[Reichenbach’s] best by a good deal."—Physics Today. 1971 ed.

This thesis addresses the problem of causal asymmetry. This problem may be characterized as follows: what is the relation R such that if an event c causes an event e c bears relation R to e but e does not bear relation R to e. The traditional Humean account of causal asymmetry is that "R" may be replaced by "temporally prior." Difficulties with this account based on consideration of cases of simultaneous causation and backward causation have given rise to non-Humean (...) accounts of causal asymmetry. This dissertation analyzes some of these difficulties along with various non-Humean approaches. I offer a new theory of causal asymmetry which does not depend upon the temporal priority of causes to their effects and which differs from any of the other theories available. ;In chapter one the problem of causal asymmetry is clarified. This problem is distinguished from related issues. At the same time, I articulate the various philosophical considerations which bear on the scope and importance of this problem. I also state what I take to be the conditions an adequate account of causal asymmetry must satisfy. In the second chapter I discuss the classical Humean analysis of causal priority in terms of temporal priority. It is argued that arguments which have been put forward to show that backwards causation is logically impossible are unconvincing and that there are cases of simultaneous causation. In the third chapter the manipulability theory of causation is treated. Various formulations of this theory, from the most anthropomorphic to the least anthropomorphic are treated as they bear on the problem of causal asymmetry. Manipulability theory is found to be inadequate. ;In chapter four counterfactual theories of causation are considered. A number of counterexamples to these theories are presented and developed. In chapter five, two views which are grouped under "regularity theory" are analyzed and criticized: those of Bromberger and Berofsky. Mackie's earlier and later accounts of causal asymmetry along with Sanford's view are discussed in chapter six. In chapter seven, Erik Brown's recent view is considered alongside "transference theory." In the final chapter I develop a new theory of causal priority which utilizes the "circumstantial" character of the causal relation in accounting for causal asymmetry. (shrink)

This paper examines Wesley Salmon's "process" theory of causality, arguing in particular that there are four areas of inadequacy. These are that the theory is circular, that it is too vague at a crucial point, that statistical forks do not serve their intended purpose, and that Salmon has not adequately demonstrated that the theory avoids Hume's strictures about "hidden powers". A new theory is suggested, based on "conserved quantities", which fulfills Salmon's broad objectives, and which avoids the problems discussed.

The physics of time asymmetry has never been a single well-defined subject, but more a collection of consistency problems which arise in almost all branches ...

Causation has traditionally been analyzed either as a relation of nomic dependence or as a relation of counterfactual dependence. I argue for a third program, a physicalistic reduction of the causal relation to one of energy-momentum transference in the technical sense of physics. This physicalistic analysis is argued to have the virtues of easily handling the standard counterexamples to the nomic and counterfactual analyses, offering a plausible epistemology for our knowledge of causes, and elucidating the nature of the relation between (...) causation and physical science. (shrink)

Two proposals for a physicalistic analysis of causation — the so-called transference model and an account given by J. L. Mackie — are examined and found wanting on the score of physical objectivity. This shortcoming can be remedied, but it is further argued that both proposals embody a too restricted conception of what a physicalistic analysis of causation should be. A more general program is proposed.

The contemporary view of the fundamental role of time in physics generally ignores its most obvious characteric, namely its flow. Studies in the foundations of relativistic mechanics during the past decade have shown that the dynamical evolution of a system can be treated in a manifestly covariant way, in terms of the solution of a system of canonical Hamilton type equations, by considering the space-time coordinates and momenta ofevents as its fundamental description. The evolution of the events, as functions of (...) a universal invariant world, or historical, time, traces out the world lines that represent the phenomena (e.g., particles) which are observed in the laboratory. The positions in time of each of the events, i.e., the time of their potential detection, are, in this framework, controlled by this universal parameter τ, the time at which they are generated (and may proceed in the positive or negative sense). We find that the notion of thestate of a system requires generalization; at any given τ, it involves information about the system at timest(τ) ≠ τ. The correlation of what may be measured att(τ) with what is generated at τ is necessarily quite rigid, and is related covariantly to the spacelike correlations found in interference experiments. We find, furthermore, that interaction with Maxwell electromagnetism leads back to a static picture of the world, with no real evolution. As a consequence of this result, and the requirement of gauge invariance for the quantum mechanical evolution equation, we conclude that electromagnetism is described by a pre-Maxwell field, whose τ-integral (or asymptotic behavior as τ → ∞) may be identified with the Maxwell field. We therefore consider the world of events in space time, interacting through τ-dependent pre-Maxwell fields, as far as electrodynamics is concerned, as the objective dynamical reality. Our perception of the world, through laboratory detectors and our eyes, are based onintegration over τ over intervals sufficiently large to obtain an aposteriori description of the phenomena which coincides with the Maxwell theory. Fundamental notions, such as the conservation of charge, rest on this construction. The decomposition of the common notion of time into two essentially different aspects, one associated with an unvarying flow, and the second with direct observation subject to dynamical modification, has profound philosophical consequences, of which we are able to explore here only a few. (shrink)

Agendanken experiment is proposed for distinguishing between two models accounting for the macroscopic arrow of time. The experiment involves the veloeity revesal of components of an isolated system, and the two models give contrasting predictions as to its behavior.