¹ Social Science Research Council, Theory and Practice in Historical Study, New York, 1946, p. 136n.

² Needless to say, the “unique events” whose occurrence is asserted may range from complex institutional structures such as European feudalism through styles in art such as the Gothic to individual personalities such as Napoleon.

³ Carl Hempel, “The Function of General Laws in History”, The Journal of Philosophy, XXXIX (1942). It should be made clear, of course, that Professor Hempel introduces the concept of “explanation sketches” as part of an argument showing the essential continuity, rather than the discontinuity, of historical explanation with explanation in other sciences.

⁴ Five meanings of “primary” or “most important” have been analyzed by Ernest Nagel, “The Logic of Historical Analysis”, The Scientific Monthly, 74, 1952.

⁵ Patrick Gardiner, The Nature of Historical Explanation, Oxford, 1952, pp. 53, 60–61.

⁶ Of course, if we ever did achieve a body of solid theory adequate for explaining human behavior, this theory might be formulated in a special language remote from ordinary language and from the language that is now regarded as the language of historiography. Accordingly, we might not be inclined to regard the explanations that such a theory provided as cases of what is now called “historical explanation.” In this sense, it is clearly true to say that it may be impossible to write history in terms of rigorous and exact statements of laws, if by “history” is meant a discipline whose results are expressed in what is now ordinary language. However, this is at least in part a terminological issue. Whether or not we used the term “historical” to characterize the explanations derived from such high-level theories, they would in fact provide explanations of the same events that are now putatively explained by historians. And since we are usually willing to call an explanation a “better” explanation when the generalizations it invokes are more exact, systematic, and embracing, such explanations would be “better” explanations, in at least one sense of the term, than those which historians can now provide. It should be added immediately, however, that a history written in such a specialized language as is here envisaged, if “history” it be called, could hardly be expected to replace history as it is now understood. We would still wish also to talk about what has happened in human affairs in the everyday language in which those affairs are actually conducted. To keep only a high-level theoretical explanation of human affairs and to dispense with history written in ordinary language would be like having only a micro-physics and no macro-physics. (The specific analogy has been suggested to me by Ernest Nagel.)

⁷ This point has been put, of course, in an intentionally simplified way. The case is strengthened when we consider that usually the historian is confronted by more than two alternatives, and that these alternatives are not precisely defined legal conditions (as in the example above) but elements in complex empirical states of affairs.

⁸ Reinhold Niebuhr, Faith and History, New York, 1949, pp. 117–18.

⁹ Reinhold Niebuhr, Faith and History, New York, 1949, pp. 117–18.

¹⁰ Patrick Gardiner, The Nature of Historical Explanation, London: Oxford, 1952, p. 105.

¹¹ Cf. Nagel, op. cit.

¹² Of course, it is possible to write individual biography institutionally, and to describe institutions in terms of the careers of individuals. Frequently the “difference” between one approach and the other may be only a difference between two mutually inter-translatable languages, both of which are saying the same thing. But if this is the case, then the accounts of events that are given by the two languages are not simply not in disagreement; they are not even really different, and cannot be described as being on different “levels” any more than two identical descriptions of a dog, one in English and one in German, can be described as being on different “levels.” One language or the other may be more convenient for the job in question, but this is the only issue. At least sometimes, in other words, there seems to be no objective reason for choosing between “two” approaches to history only because, in fact, there is no real difference between them.