# Statistical Explanation and Statistical Relevance

Wesley C. Salmon
Richard C. Jeffrey
James G. Greeno
Pages: 128
https://www.jstor.org/stable/j.ctt6wrd9p

1. Front Matter
(pp. i-vi)
(pp. vii-viii)
3. Preface
(pp. ix-2)
W.C.S.
4. Introduction
(pp. 3-18)
Wesley C. Salmon

Suppose that a small plane crashed upon takeoff from an airport near Denver on July 15, 1971. We ask why the crash occurred. Our interest in the event might be highly practical; the FAA, for instance, investigates such accidents in order to improve flying safety. When we know why occurrences of various types happen, we can often do something about controlling them. At the same time, our interests might be largely theoretical. To someone concerned with aerodynamics, the search for an explanation of the crash might be the result of sheer intellectual curiosity. In either case—these two motives are...

5. Statistical Explanation vs. Statistical Inference
(pp. 19-28)
RICHARD C. JEFFREY

Hempel is not the first philosopher to have held that causal explanations are deductive inferences of a special sort: in thePosterior Analytics¹ Aristotle distinguishes a special sort of deductive inference – the demonstrative syllogism – in these terms:

By demonstration I mean a syllogism productive of scientific knowledge, a syllogism, that is, the grasp of which iseo ipsosuch knowledge.

He then lays down defining conditions for this special sort of inference:

… the premisses of demonstrated knowledge must be true, primary, immediate, better known than and prior to the conclusion, which is further related to them as...

6. Statistical Explanation
(pp. 29-88)
WESLEY C. SALMON

Ever since his classic paper with Paul Oppenheim, “Studies in the Logic of Explanation,” first published in 1948,¹ Carl G. Hempel has maintained that an “explanatory account [of a particular event] may be regarded as an argument to the effect that the event to be explained …was to be expectedby reason of certain explanatory facts” (my italics).² It seems fair to say that this basic principle has guided Hempel’s work oninductiveas well asdeductiveexplanation ever since.³ In spite of its enormous intuitive appeal, I believe that this precept is incorrect and that it has led...

7. Explanation and Information
(pp. 89-104)
JAMES G. GREENO

One purpose of this paper is to suggest a way around an apparent paradox in the current theory of statistical explanation. The paradox has been mentioned in earlier analyses [2] and can be illustrated by the following situation. Suppose that a boy, Albert, is convicted for stealing a car. Attempting to give an explanation, a social worker points out that Albert lives in San Francisco, where there is a high delinquency rate. However, it is also noted that Albert’s father earns \$40,000 per year, and sons of men with high incomes have a low delinquency rate.

One view which agrees...

8. Postscript 1971
(pp. 105-110)
Wesley C. Salmon

The essay “Statistical Explanation” is reprinted here with only one significant change aside from the correction of minor typographical errors; namely, the inequality has been added to the formula on page 55 which defines the screening-off relation. Its omission from the original printing was a foolish oversight, and the need for the correction was called to my attention by J. A. Coffa. In order forDto screen offCfromB, Cmust be irrelevant toBin the presence ofD, butDmust not be irrelevant toBin the presence ofC. Coffa has also pointed...

9. Bibliography
(pp. 111-112)
10. Index
(pp. 113-118)