[bowman] fundamental PROCESSES IN HISTORICAL SCIENCE 577 



(2) If the contradiction be by a trustworthy record, and the at- 

 tendant circumstances of the discrepancy are known, then, (a) if the 

 circumstances constitute a legitimate explanation of the conflicting 

 statements, each statement is to be accepted as correct in the sense 

 which the circumstances justify; but (6) if the circumstances show 

 that one of the statements must be an error, without showing which, 

 silence should be observed concerning the point in contradiction, un- 

 less either record is confirmed bya second, or more, independent trust- 

 worthy records; in which case the statement of the records in agree- 

 ment is to be accepted. 



(3) If the contradiction be by a trustworthy record, and the 

 attendant circumstances are unknown, the statement in neither record 

 should be accepted, whether confirmed by other independent trust- 

 worthy records or not; the scientific requirement in this case is silence 

 concerning the point in contradiction. 



The 3rd rule involves the chief departure from the present treat- 

 ment of discrepancy. The scientific necessity for this rule is proved 

 experimentally in the Origin and Treatment of Discrepancy pre- 

 viously mentioned (Trans. Roy. Soc. Can., 1911, v, 127-178; on point 

 concerning confirmation by other records, p. 173). 



V. Narration. 



Scientific results, including historical narratives, should be pre- 

 sented ordinarily (cf. Cases 4 and 9 of the experimental test) in posi- 

 tive, and not in negative, form, because persons receiving information 

 wish, as a rule, to know what a thing is, not what it is not. 



Historical narrative should be categorical in form, because, by 

 the fundamental principles of science, all doubtful and hypothetical 

 conclusions are debarred from the accepted final results designed for 

 general publicity, and especially for printed publication ; in the words 

 of Goethe, "When a man writes a book, let him tell me what he knows. 

 I have guesses enough of my own." 



The treatment of records above outlined constitutes a scientific 

 historical method because it meets the general and ordinary require- 

 ments of science and because it conforms to the general and ordinary 

 principles and practice of actual trustworthy intercourse. The pre- 

 vailing method is constrained to its present incorrect and probable 

 processes by a fundamental error in its attitude towards records 

 as such. With a consideration of this point the present discussion 

 will close. 



Sec. I and II 1916—38 



