[bowman] fundamental PROCESSES IN HISTORICAL SCIENCE 575 



III. Is the record trustworthy? 



The trustworthiness of a record as a whole, or in its respective 

 parts, or in its respective statements, depends upon its exempHfica- 

 tion as a whole, or in the respective parts, or statements, of the 5 

 requisites of trustworthiness located experimentally in Part I of this 

 paper. These requisites are the same qualities by which we judge, 

 with essential correctness, of the trustworthiness of our fellow men in 

 actual intercourse; and the test, as applied to records, should conform 

 to the test as applied to men. In actual intercourse, when brought into 

 contact with a man with whom we expect to be more or less closely 

 associated or to have business dealings, we do not make a preliminary 

 test of his trustworthiness by selecting a particular body of his state- 

 ments and faring forth after verification or disproof. We judge of 

 his trustworthiness by his language, bearing and interests, and ordin- 

 arily in an association of any length, we do not judge amiss. Only a 

 few do we accept as essentially trustworthy in the fullest sense; a 

 few we reject as essentially untrustworthy; and between these two 

 extremes, there is every variety of intervening grades. The state- 

 ments of an essentially untrustworthy person we reject as a basis of 

 action, save in points where we have other grounds for believing 

 that he is correct. The statements of an essentially trustworthy 

 person we accept as a basis of action, save in points where we have 

 grounds for believing that he is in error. Moreover even in a person 

 accepted as essentially trustworthy, exceptions in his trustworthiness, 

 and not mere unintentional errors on his part, will in most cases be 

 detected eventually. Thus most men will shield members of their 

 own family, at least by silence, up to a certain point; and few men 

 are wholly without honest prejudices or personal predilections. We 

 discover these points gradually in the case of a trusted associate; 

 we differentiate them in our reliance upon his trustworthiness ; but 

 apart from these exceptional points, if their number is not 

 excessive (and ordinarily the number is not excessive), we still 

 give him our general confidence, subject to these differentia- 

 tions. Differentiations of trustworthy points in untrustworthy per- 

 sons may be similarly located; and these differentiations of both 

 kinds, with their respective grounds, made in actual intercourse, and 

 assembled and digested, form a valuable amplification of the 5 requi- 

 sites in judging of the trustworthiness of records as a whole or in their 

 respective parts or statements. Records are but the recorded utter- 

 ances of men; and just as men in actual intercourse have a language, 

 and bearing, and interests, which may be studied and located by their 

 fellow men, so a record has a language, and bearing, and interests, 

 which may be studied and located by a historian. The only difference. 



