236 JOHN D. DODSON 



experimenters divided their subjects into three groups of two 

 rats each and trained two with punishment, which was a " light 

 electric shock/ ' two with reward, which was " milk-soaked 

 bread/' and two with a combination of reward and punishment. 

 Of the two rats trained with reward and punishment one 

 finished in 490 trials, or the 49th day, the other in 550 trials, 

 or on the 55th day; but the first of these made a perfect record on 

 the 30th, 31st, 34th, 37th, 40th, 43rd and 44th days; the latter 

 made a perfect record on the 48th day. Of those trained with 

 punishment one finished in 550 trials; the other had not finished 

 when the experiment was discontinued on the 60th day; but the 

 latter made a perfect record on the 37th, 50th and 51st days. 

 Neither of the animals used with reward finished or made a per- 

 fect record during the entire series. From these results the 

 experimenters came to the following conclusions : 



It seems evident from this experiment that a combination of punish- 

 ment and reward-motives is more effective in bringing about visual 

 discrimination in the rat than is either punishment or reward used 

 alone. It seems evident, also that punishment is more effective than 

 reward, at least in so far as the rate of learning is concerned. 



Whatever else the above experiments may or may not have 

 proved, they demonstrate beyond a question of doubt that no 

 experimenter can arbitrarily choose a single strength of stimulus 

 or degree of hunger and say he has the most favorable condition 

 for training any living organism. The experiment on the relative 

 value of punishment and reward as motives shows almost noth- 

 ing of the relative value of these two motives. The strength of 

 electrical shock used may have been the most unfavorable to the 

 learning process while ;the degree of hunger was the most fav- 

 orable, or the strength of shock may have been the most favorable 

 while the degree of hunger was the most unfavorable. It seems 

 likely from the results that both the electrical shock and the 

 degree of hunger were too weak to keep the animals up to their 

 greatest capacity. 



