INTERFERENCE OF HABITS IN THE WHITE RAT 61 



the former. The amount of the interference will probably 

 depend much upon the ease of discrimination between the 

 stimuli for the two habits. We are not prepared to contribute 

 upon this point. (Because the rats ranked the buzzer as the 

 same as handclaps we have felt justified in assuming that un- 

 trained rats would learn " buzzer to the left ' as readily as 

 " hand claps to the right.") 



Mrs. Binnie Pearce, in research from this laboratory as yet 

 unpublished, found even more striking interference in visual 

 habits. Using the same T-shaped box, she trained rats to 

 run one way for light and the other way for darkness. When 

 she then attempted to train them to reverse this behavior, the 

 task was found all but impossible. 



We are not familiar with any other work where an animal 

 has had to learn the opposite of a previously acquired habit. 

 There are many cases where different habits have been set up 

 in succession and where interference has been more or less ex- 

 plicit. However, in order to secure comparable data, it is neces- 

 sary that the stimuli be known and the responses simple. The 

 study of interference in mazes, latch boxes, etc., suffers for this 

 reason. Not only must the stimulus be known in the case of 

 the first habit, but the second stimulus must be known physi- 

 cally and also physiologically in terms of the first one. Thus 

 one can know whether or not the stimulus for the second habit 

 is for the subject in that situation the same as the first stimulus 

 (positive transfer). Where the type of habit set up is kinaes- 

 thetic as opposed to auditory or visual, the control of the stim- 

 ulus is very difficult because the stimulus lies in the animal's 

 movements. The most feasible procedure is to reduce the 

 problem to such an extent that only one or two prominent kin- 

 aesthetic experiences are presented. The senior author is work- 

 ing upon this problem at the present time, although interference 

 is but one phase of the study. 



VIII 



Relative rates of error elimination in interfering habits. — With 

 particular reference to the 30-day rats and rats 25-29, it is of 

 interest to raise the following question: In what parts of the 

 learning curves does the interference, as measured by the rela- 

 tive rates of error elimination, occur? 



