THE PROCESS OF MEMORIZING. 127 



finally adopted followed as far as possible the original complete memo- 

 rization method of Ebbinghaus, except that in our arrangement memo- 

 rization need not be completed. Ebbinghaus measured the total 

 number of repetitions that it took to enable a subject to learn a series 

 of nonsense syllables, that is, to speak the words of the series without 

 prompting. We sought to measure the value of each repetition by its 

 saving in the reaction-time of successive members of the series as they 

 were exposed seriatim. We measured the reaction-time of each word, 

 instead of that of the group process. An analogous short cut in the 

 psychology of reading is to measure the reaction to individual words 

 instead of the total time it takes to read a page. In certain respects our 

 method resembles the Mtiller and Pilzecker's 1 Treffemethode, but 

 instead of "paired" associates, our associates were continuous, and 

 the decrease of the reaction-time in successive repetitions of the series 

 showed the increase of perseveration. The technique is the direct 

 outgrowth of Professor G. E. Miiller's lectures on memory measure- 

 ments, which one of us was fortunate enough to attend at the University 

 of Gottingen in the winter of 1910-11. We would hereby express our 

 obligation to Professor Miiller, while expressly disclaiming for him any 

 faults in our technique. According to the theory of our method, any 

 saving of time between the reaction-time in responding to the first 

 exposure of a series of words and the reaction-time in responding to the 

 second exposure must be due to the influence of memory. The memory 

 process itself would be complete when the reaction-time for each 

 member of the series is zero or less; that is, when each member of the 

 word series could be pronounced before it appeared, as in the complete 

 memorization of Ebbinghaus. 



This method seemed to satisfy our demand for a practiced process. 

 To be sure, words are not commonly read during a gradual exposure; 

 certainly not during the kind of exposure that was used in our experi- 

 ments. But, after all, something not so very different appears to be 

 involved in all rapid reading. Only a part of the words of a page are 

 actually fixated by the reader. Most of them are read chiefly or partly 

 in indirect vision where the visual cues to the identity of the words are 

 few. The more familiar the text the fewer the words that are actually 

 fixated, and the more significant become the imperfectly seen cues of 

 extra-foveal vision, and the more important are the central or memory 

 factors in the process. Fundamentally, then, the method is as practiced 

 and natural as the process of reading itself. 



A special device to procure a slow, gradual increase in the visual 

 exposure of the words, so that reaction differences might be exaggerated, 

 was to expose the words backwards one letter at a time. This insured 

 a gradually increasing number of cues to the identity of the word, until, 



Stiller and Pilzecker, Experimentelle Beitrage zur Lehre vom Gedachtnis, Zeitschr. f. Psychol., 

 Erganzungsband, 1, 1900. 



