This book review appeared in Canadian Philosophical Reviews (September 1988), Vol. 8, no. 9, pp. 346-348.

DENIS J. HILTON, editor.
Contemporary Science and Natural Explanation:
     Commonsense Conceptions of Causality.
New York: New York University Press 1988.
Pp. xii + 244.
ISBN 0-8147-3443-X.

This anthology grew out of a symposium, Knowledge-based strategies in causal attribution, at Clare College, Cambridge, 1985. It would appear, however, that most of the ten papers finally published were written by invitation subsequent to that symposium. Perhaps because of these origins, unlike the Proceedings of many other symposia, there are here neither replies to, nor discussions of, any of the contributions.

There seems to have been little thought given to the question of the intended audience for this book. All but one of the contributors, Rom Harré, are psychologists, yet philosophers figure prominently, although certainly not predominantly, in every one of the bibliographies accompanying the papers. If the book is intended to be interdisciplinary, there is remarkably little attempt, either by the editor or by the contributors, to integrate, or to synthesize anything from, the diverse articles.

For a philosopher picking up this volume, the most informative and provocative part will, surprisingly, surely be Part I, the four papers devoted to empirical studies concerning how persons go about selecting what they take to be causes. Part II, which is by design 'more philosophical', will, very likely, strike the professional philosopher as being, if not familiar in its entirety, at least familiar in many places. It can be read with some profit and interest, but will not prove nearly as revelatory as the first half.

The thrust of this book is significantly different from the usual philosophical fare about causality. Historically, in the philosophical literature, questions about causality have principally focused on the issue of the metaphysical nature of 'the' causal relation and on the issue of how it is possible to know causes (along with the associated problem of projecting such knowledge to unexamined cases). These issues are scarcely touched upon in this book.

Among these writers, the problems to be tackled in examining causality include [my reconstruction]: 'What sorts of conditions are regarded as normal against whose background the explanandum is seen as anomalous?'; 'What sorts of linguistic or conversational cues prompt certain kinds of causal explanations and eliminate others?'; 'What sorts of causal paradigms do persons invoke and modify in their particulars to explain concrete events?'; 'To what extent can a regularly repeated causal factor "mask" the recognition of the contribution of another?'; and 'To what extent are causal attributions rule-governed or the result of associations?'

These sorts of questions, it should be obvious, are a far cry from the normative or ideal reconstructions offered by certain formalist philosophers of the 1940s and 50s. Indeed, among these present writers, the earlier theories of Hempel, Braithwaite, et al, are so thoroughly discounted as to merit only an occasional swipe. The philosophers who are cited approvingly, and whose writings remain fertile pastures, are the critics of the formalists, e.g. Dray, Scriven, and especially Hart and Honoré, whose Causation in the Law is cited by three of the four experimentalists whose papers comprise Part I.

In little more than one generation, the formalists' theory of explanation has – at least in the writings of these psychologists – been relegated to the scrap heap of intellectual history. In retrospect, this development may be seen to have been inevitable. For the trouble with the logical (normative) theory of explanation was that it provided no grounds for a research program. Having argued that explanation is a logical relationship between propositions, and having explicitly and caustically eschewed the relevance of any and all psychological factors, the formalists barred the areas of explanation and causality to empirical researchers. But when one today reads 'an explanation of something is given by someone to someone' (57), one immediately sees in it the complete repudiation of the normative model. Indeed virtually each and every aspect of the formalists' model – that explanation involves universal or statistical laws, that explanation and prediction are 'symmetric', etc. – are explicitly or implicitly dismissed, in one paper or another in this book, as simply inconsistent with empirical data about how persons actually generate causal explanations.

Apart from giving the lie to the formalists' theory, what has empirical investigation revealed to date about how human beings make causal attributions? On the evidence of this book: not much. But really, one ought not to expect otherwise. Attribution theory is still in its nascency. Its subject matter is exceedingly complex, and powerful theories in the field would seem to be years away. Even so, future discussions of causality can never again proceed in the kind of high-handed a priori fashion of the period 1739-1960. Like it or not, future discussions of causality will have to take more and more cognizance of the experimental research of psychology and psycho-linguistics.

Philosophers unused to reading the professional writings of psychologists may find some of the prose in this book offputting. There is, for example, an overabundance of acronyms. To be sure, much of the writing is workmanlike. But some of it is, no doubt unintentionally, hilarious: 'unusual events, such as sudden deaths, often violate a desired aspect of the status quo. They are in that sense goal failures' (180). And, regrettably, some of it is of the sort which has provoked (obviously ineffectual) perennial lampooning by persons who admire the English language: 'conditioning results from increments in the strength of the association between event representations resulting from contiguous pairings of the events with the level of the conditioning being related monotonically to associative length' (95). 'Event representations' is, unfortunately, just one of numerous instances of nouns being used as adjectives ('covariation information' (35), 'abnormal conditions focus model' (39), etc.) Indeed, there are even instances of the resurrecting of obsolete nouns just to be misused in compound grotesqueries: 'enablement conditions' (179). Farewell, gerunds!

Overall, the conclusions drawn in this book are too few, too controversial, and too underdetermined to provide any secure grounds on which to base plausible theories of causality. If anything, this particular collection of papers warrants the verdict that psychological theories of causality are currently in disarray. Nonetheless, even if you, quite reasonably, decline to purchase this book for your own use, you might well want to borrow a library copy to catch the flavor and direction of contemporary empirical research in this field.


Simon Fraser University

Return/transfer to Norman Swartz's Philosophical Notes

Return/transfer to Norman Swartz's Home Page