142 K. S. LASHLEY 



A comparison of the pairs of trials made during the same 

 period of practice with those made in successive periods shows 

 a greater diversity of errors in the latter. In them there were 

 247 pairs of like errors and 493 diverse errors; that is, 49.05% 

 of the errors made in successive trials were diverse. In the 

 successive trials made on the same day there were 274 pairs of 

 like errors and 439 diverse errors; 44.48% of the errors made 

 in these successive trials were diverse. The trials separated by 

 a 24 hour interval are thus seen to include 10% fewer duplicate 

 errors than those without intervening time: a greater diversity 

 of activity occurs where successive trials are separated by a 

 considerable interval of time than where they follow each other 

 immediately. 



This is the result to be expected on the hypothesis considered 

 above and it seems to place the latter on a firm basis. Two 

 questions, however, are unanswered: first, is this the only factor 

 involved in producing the results noted when different distribu- 

 tions of practice are used; second, is the hypothesis applicable 

 to other types of activity than motor forms such as archery or 

 typewriting. It is not possible at present to say how great would 

 be the effect of the conflicting habits in retarding the rate of 

 learning in any particular case. If the diversity of activity is 

 reduced 10% by the concentration of practice from one to two 

 trials per day, a 10% increase in the amount of practice neces- 

 sary for learning is the least which could be expected, but the 

 fact that the disadvantageous activities are repeated more fre- 

 quently with increasing concentration of practice would help 

 to fix them as habits and thus produce an even greater retarda- 

 tion of learning. At a rough estimate, the persistence of the 

 same errors through prolonged periods of practice seems ade- 

 quate to account for such retardation of learning as has been 

 found to result from concentrated practice during the formation 

 of motor habits. 



If the same explanation is not applicable to language habits, its 

 validity for motor habits is questionable. In view of our ignorance 

 of the mechanism of association in such activities generalizations 

 here must be made with caution. There can be little doubt, 

 however, that false associations occur in the formation of langu- 

 age habits and it seems probable that these may interfere with 

 learning in the same way that habits of making wrong turns in 

 the maze interfere with the learning of the correct path. 



