SCIENCE (and common SENSE ) 33 



trajectories "too diflFerent" from those already observed. The physi- 

 cist accepts data on a cosmic ray event he cannot reproduce because 

 he feels it classifiable with other "events of the same kind." The 

 palaeontologist accepts as datum a single jawbone or cranium, and 

 seeks to reconstruct the entire animal involved. The fragment is 

 unique, irreproducible, but he finds in it some filiation or likeness in 

 type with other analogous remains. The sense of analogy renders 

 the datum acceptable and the reconstruction possible. 



Facts historical and scientific. The distinction drawn between the 

 facts of history and the facts of science (and common sense) is far 

 less sharp and significant than is supposed by one general school of 

 thought— here represented by Langer's statement: 



. . . nor is historical truth judged by the same criteria as the truth of 

 scientific propositions. For to science, as Lord Russell once remarked 

 in an academic seminar, "A miracle would not be important if it hap- 

 pened only once, or even very rarely"; but in history the point is to 

 find out what did happen just once, what were the specific facts about 

 a specific occasion. 



As we have just seen, scientists do accept irreproducible data, "spe- 

 cific facts about a specific occasion," if they conceive the occasion 

 to be recurrent. Russell and Langer notwithstanding, the acceptabil- 

 ity of historical data seems to be judged by much the same criterion. 

 In the operations of historians and scientists there is a major dif- 

 ference in temporal reference— is this diflFerence in itself decisive? 

 Surely not! Seeking to reconstruct what happened in the past, the 

 historian must and does take as data only presently available written 

 reports, and the presently surviving physical remnants of antiquity; 

 and part of the interest of his work ( as of the scientist's ) lies in its 

 relevance to present or future situations somehow analogous to those 

 of the past. When as scientist I consider my data, I work in part as 

 historian. I, too, refer to events of the past: I consider my own ex- 

 periments not in their present immediacy but in my memory and 

 (more reliably) in my written laboratory records; I consult the pres- 

 ently a\^ailable reports of my predecessors and contemporaries who 

 have investigated analogous phenomena. My data may be experi- 

 mentally reproducible, as the historian's are not; and perhaps this 

 loould be a decisive diflFerence save that often as scientist I accept the 

 weaker criterion of recurrence. 



