626 THE AM E NIC AX XA TUEALIST [Vol. XLIII 



3. He is led to suppose that my account, being wrong in prin- 

 ciple, must also be wrong in details, so that he attempts to cor- 

 rect details, by giving usually the same explanation that 1 had 

 already given ; or at times some other causal explanation that he 

 devises, without experiment, to account for the facts; this often 

 leads him into curious errors. 



4. He is inevitably driven to suppose that I reject determinism 

 in behavior. 



5. He is naturally led (though here without real basis as it 

 seems to me, save by additional error of the same sort) to attrib- 

 ute to me psychic, anthropomorphic and finalistic explanations 

 in place of causal ones. 



Let us look briefly at certain of these points. 



1. My concept " trial and error" the author is forced by his 

 mistake to suppose designed to take the place of an experimental 

 analysis based on a real knowledge of the causes of reaction. His 

 account of its origin is : 



As he has not taken into account the phenomena of sensibility to 

 differences, he has been led to speak of trial and error (p. 191). 



The fact is that the idea of trial and error was based precisely 

 on reactions due to ''sensibility to differences"; to changes in 

 the conditions, and I have again and again stated that fact. 

 For example: 



tion by "trial and error" in Paramecium? An examination of the 

 facts shows that as a general rule the effective stimuli consist of some 

 change in the conditions. 15 



The purpose of the concept was to bring under a unified point 

 of view a complex of stimulations and reactions which had the 

 interesting and important result of bringing the organism into 

 conditions adapted to it, yet not including "anything differing 

 in essential principle from such methods of action as we find 

 in the inorganic world." 16 The concept is useful only in at- 

 tacking the problem of how adaptation is brought about without 

 the operation of "final causes." To speak of "trial and error' 

 of course does not get rid of the necessity of finding a deter- 

 mining cause for every phase of the reaction, any more than 

 to call a reaction "tropism" has this result. It has been pre- 

 *" The Behavior of Paramecium, ' ' Journ. Comp. Neurol, and Psychol, 

 16 " Behavior of the Lower Organisms," 1906, p. 343. 



