THORNDIKE'S ANIMAL INTELLIGENCE 443 



dency to read severe advice to cognate disciplines. However, 

 the contemporaneous attitude toward ideational interpretation 

 of isolated cases of behavior was settled "once and for all" 

 by these papers. The tendency to seek in experimentation an 

 excuse to express views on the whole field of human and com- 

 parative psychology characterized other papers of this period, 

 and in fact such a tendency is not without value in the develop- 

 ment of any new field. Some interpretative positions were un- 

 founded; the ideational significance of retentive permanence and 

 of a sudden drop in the learning curve are doubtful propositions. 



Two chapters are new. The first concerns the province of 

 the science. It maintains the thesis that comparative psychol- 

 ogy may deal with behavior for its own sake. Criticism is here 

 futile, for we enter the realm of personal opinion. A better 

 exposition of the validity of the objective method is difficult to 

 find, but perhaps this statement is too much colored by the 

 reviewer's bias. The chapter on the laws and hypotheses of 

 behavior is one of the most important in the book, inasmuch 

 as it expresses the author's mature opinion formed in the light 

 of a decade of subsequent experimentation. The reviewer 

 finds, however, more opportunity for dissent and criticism than 

 in the earlier work. A few points may be noted. 



(i) The law of causation holds -sway; the behavior of any 

 organism is wholly determined by the stimulus and the nature 

 of the organism. This conception is emphasized because, it is 

 asserted, two popular doctrines (in the sense of widely accepted, 

 I presume) exist in defiance of it. The first is the doctrine that 

 behavior is random. I deny that the term necessarily implies 

 a meaning antagonistic to the above principle. A similar objec- 

 tion might also be made on just as good grounds to the phrase 

 "accidental success" which is so liberally sprinkled throughout 

 the book. That the term "random" can be employed without 

 doing violence to the principle is evident from the fact that 

 the author himself uses it twice (p. 255). If, as implied, any 

 great number of writers actually do employ the term in the 

 objectionable sense, I venture to suggest that a few names be 

 specified. The second is the doctrine "that the need for a cer- 

 tain behavior helps to create it." This term can also be used 

 without violating the principle. A "need" may represent the 

 presence of a persistent irritating stimulus or the absence of a 



