320 H. M. JOHNSON 



the sense organ. If the study is regarded as one in behavior, 

 these data are considered as " explained " when they are satis- 

 factorily referred to the characters of the stimuli and the struc- 

 ture, condition and physiological processes of the organism. If 

 the study is to be stretched into one of (hypothetical) " mental 

 content," which some consider the subject-matter of " animal 

 psychology," it is necessary for the student to make certain 

 arbitrary assumptions regarding the animal's possession of 

 ' ' mental states ' ' and regarding the degree of similarity obtaining 

 between these hypothetical mental states and his own. Postu- 

 lating the validity of these assumptions, and basing his inferences 

 on them as much as on the observed facts, the student may now 

 attempt to "translate" the animal's behavior into "terms of 

 his own conscious experiences." This sort of treatment of the 

 data has been severely criticized on the ground of unwarranta- 

 bility. The criticisms have not been altogether effective, how- 

 ever. But it would seem advisable in any event for students to 

 report all the stimulus-conditions and all the facts of behavior 

 before indulging in interpretations in terms of mental content, 

 in order that questions of fact and questions of interpretation 

 may not become confused by the readers. Whatever one's 

 standard of interpretation may be, the animal's responses remain 

 the same. 



The responses chosen for study may of course be either muscular 

 or glandular. PavlofI and members of his school have devised 

 the well-known method of studying the effect of sensory stimuli 

 on the saliva-reflex. Yerkes and Morgulis 1 have discussed 

 certain shortcomings in the technique. Some of these the 

 experimenters themselves had considered. Could the number of 

 variables be reduced, a very interesting and valuable set of data 

 are possible. The reason why some experimenters have chosen 

 muscular responses is merely a greater interest in them. It 

 should be remarked also that we are not yet warranted in predict- 

 ing the effectiveness of stimuli in occasioning muscular responses, 

 from their effectiveness in occasioning glandular responses. A 

 stimulus-value or stimulus-difference sufficient to produce a 

 change in the rate of saliva-secretion or of the quality of the 

 saliva might be quite inadequate to cause an animal to choose 



1 Yerkes, R. M., and Morgulis, S.: The Method of Pavloff in Comparative Psy- 

 chology. Psychological Bulletin, 1909, pp. 257 ff . 



