THE PRIMITIVE METHOD OF REASONING. G3 



directly to the sum of systemic knowledge, and its ])ro(luct is always in- 

 definite and uncertain. 



When the processes of reasoning by analogy and homology respectively 

 are compared they are found to be antithetic — to stand in the relation of 

 the face and the obverse of the same shield. Analogic inference (or 

 reasoning from what is recognized to be like in only one or a few attri- 

 butes) is discursive and suggestive, leading on the one hand to discrimi- 

 nation and thus to refinement of observation, and on the other hand to 

 the invention of hypothesis ; while homologic inference is incursive and 

 constructive, leading on the one hand to identification and thus to re- 

 finement of observation, and on the other hand to the elimination of 

 aberrant observation and incongruous hypothesis, or to generalization. 

 Thus analogic inference is analytic, homologic inference synthetic ; the 

 one makes for the accumulation, the other for the assimilation of observa- 

 tions ; the one leads to speculation, the other to generalization ; the first 

 adds to the quantity, the second to the quality of knowledge. 



Empiric knowledge, like knowledge of a higher order, is possession and 

 arrangement of facts. When the facts are few retention is easy, and the 

 facts are accessible for comparison with little arrangement; when the 

 facts are many retention is more difficult, and the facts can be kept acces- 

 sible only by definite arrangement; and in this way classification grows 

 up with the accumulation of knowledge. Again, when the facts are few 

 and the classes small the arrangement may rest on one or, at most, a few 

 attributes without inconvenience to the user, but when the facts are many 

 and the classes large it is necessary to subdivide the classes by additional 

 attributes, else the user is swamped beneath the burden of his own pos- 

 sessions ; and in this way classification becomes more and more refined 

 with the accumulation of knowledge. Now, in the beginning of the 

 acquisition of knowledge through observation there is little need for 

 arrangement with a view to retention, and accordingly nearly all effort 

 is directed toward discrimination, and contrasts are sought and differ- 

 ences magnified through the process of reasoning from the unlike, and 

 thus analogic inference is develoi)e(l. Later, when the fiicts of observa- 

 tion are too numerous for retention witliout arrangement, a larger and 

 ever-increasing share of effort is withdrawn from acquisition and devoted 

 to assimilation, and comparisons are made and resemblances sought, to 

 the end that like may he grouped witli like, and tlius homologic inference 

 is developed to the partial exclusion and legitimate regulation of analogic 

 inference. Accordingly, analogic reasoning is the more primitive, homo- 

 logic reasoning the more developed. 



The body of empiric knowledge, like that of higher order, reacts on the 

 method of acquisition in every stage, and thus there is a normal progres- 



X-BuLL. Geol. Hoc. Am., Vol. 6, 1894. 



