[bowman] fundamental PROCESSES IN HISTORICAL SCIENCE 543 



the correctness, even in the case of a most probable conclusion, is not 

 a necessary result, but only accidental. In history, as in other sciences, 

 therefore, these fundamental principles disallow probable conclusions, 

 and their formulation for acceptance, except in so far as they represent 

 averaged results. 



h. Probability tested by the methodic principles of science. These 

 principles, which are necessary for examining the scientific status of 

 any process or method of investigating and solving problems, theoreti- 

 cal and practical, are defined with illustrations on pp. 502-506 above. 



i. An incorrect process is one in which the operator deviates from 

 a requirement of a correct process, believing that such deviation is proper. 

 If an operator makes such a deviation unintentionally and only by 

 oversight, he has made only an incidental error; but if he makes the 

 deviation deliberately and habitually and in the belief that this 

 course is correct, this practice constitutes a process which, because it 

 differs from the correct process, is incorrect. According to the 

 fundamental principles of science, confirmed in the foregoing experi- 

 mental test, conclusions should be formulated for acceptance on the 

 basis of probability only in the case of averaged results. Under the 

 prevailing method, however, probable conclusions which do not 

 represent averaged results are systematically included with the accep- 

 ted results of historical investigation. This deviation from a correct 

 process is manifestly intentional and is made in the belief that the 

 course is proper; hence the practice does not merely involve incidental 

 error, but it constitutes a process, which, because it deviates from a 

 correct process, is incorrect. 



ii. Where an operator deviates from a correct process, the result, 

 if correct, is nevertheless unscientific because accidentally obtained. 

 Where an operator deviates from a correct process, one is disposed to 

 infer that his result will necessarily be incorrect. Ordinarily it will 

 be incorrect, but not necessarily so. In the illustration of this prin- 

 ciple on p. 502, it was found that if an operator devised, e.g., a partial 

 deviation from a correct mathematical process and applied it, he 

 might still get a correct result, but only if he happened to make a 

 further mistake which would exactly counterbalance the first error. 

 The occurrence of such an exactly counterbalancing error, however, 

 could be only accidental. This ultimate analysis of any correct 

 result obtained, after a partial deviation from a correct process, to an 

 accident, corresponds exactly with the result of the complete depar- 

 ture from all correct processes and all scientific principles which is 

 involved in taking probability instead of necessity as the criterion of 

 conclusions; for with probability as the criterion, all correct results, 

 except the averaged, will be only accidental because, with this excep- 



