246 THE SOCIAL LIFE OF ANIMALS 



by rats. The exact determination of the place in this 

 line of increasing complexity at which an action 

 ceases to be a simple reflex and becomes a more 

 elaborate tropism, or the point at which the tropism 

 gives way to an instinct, has never been made. That 

 is, we do not know just how far down in develop- 

 ing patterns instinctive behavior extends. 



There is the added complication that the word 

 "instinct" has been loosely used. The most workable 

 definition that I have arrived at is a modification of 

 an older one of Wheeler's: An instinct is a com- 

 plicated reaction which an animal gives when it re- 

 acts as a whole and as a representative of a species 

 rather than as an individual, which is not improved 

 by experience, and which has an end or purpose of 

 which the animal cannot be aware. Too frequently 

 the word has been applied to any bit of behavior 

 whose origin and motivation the observer did not 

 understand, with the unfortunate paradoxical im- 

 plications that thereby the action was explained and 

 at the same time could not be further explained. 

 As a result of this uncritical usage many careful 

 workers disapprove employing the word under any 

 conditions, and particularly in the field of social 

 activities. 



In recent years some students of social life have 

 attempted to avoid the term "social instinct," while 



