Febbuaet 11, 1910] 



SCIENCE 



225 



complex for experimental control, and it 

 is of interest to discover whether this early 

 judgment is now superseded, and whether, 

 quite apart from results, a method has been 

 devised for experimenting on thought. The 

 method now suggested is certainly direct and 

 obvious. The person whose mental processes 

 are to be observed is given a problem to solve ; 

 in some experiments the problem has been of 

 the easiest, in others it has demanded careful 

 attention; but in all cases it has been such 

 that a solution could be reached in a few sec- 

 onds, at the end of which time the thinker is 

 required to describe what had passed through 

 his mind in the process of solution. It is es- 

 sential to the method that the same general 

 sort of problem be set many times in succes- 

 sion, and that the preliminary consciousness 

 intervening between the signal " Ready !" and 

 the propounding of the particular problem 

 should be described, as well as the conscious- 

 ness transpiring between the propounding of 

 the problem and the attainment of the solution. 

 Whatever else may be said of the method, it 

 has at least produced a large mass of data re- 

 garding matters which had previously been the 

 subject of only casual observation. The 

 method has been sharply criticized by no less 

 an authority than Wundt, on the ground that 

 it does not fuMl the essential requirement of 

 experimental observation. In a proper experi- 

 ment, as Wundt says, the observer knows be- 

 forehand exactly where his attention must be 

 directed; the field of observation is narrowed, 

 and the observation is consequently more 

 minute and accurate than in ordinary cir- 

 cumstances. In this new work, however, the 

 observer, who is also the person experimented 

 on, does not know beforehand exactly what he 

 has to observe, and, besides, must devote his 

 attention first of all to the solution of his 

 problem, and only secondarily to the observa- 

 tions which are desired. With this line of 

 criticism, which is evidently the old, familiar 

 objection to introspection in general, our au- 

 thor seems not to agree. He regards the work 

 so far done as a promising beginning, except 

 that too much has been attempted at once, and 

 that some of the experimenters have been con- 



tented with observations on what the thought 

 was about, instead of insisting on a descrip- 

 tion of the thought as a mere conscious fact. 



As to the results of this work, one at least 

 has been gained, and is freely admitted by 

 Professor Titchener. It will be remembered 

 that the problems set in any one series were 

 of one general nature, which was understood 

 beforehand. The thinker becomes adjusted to 

 this general task, as is shown by the fact that 

 the propounding of the particular problem is 

 usually followed promptly by a course of 

 thought leading to or towards the solution, to 

 the exclusion of numerous other associations 

 which might otherwise be recalled by the 

 words, etc., used in putting the problem. The 

 preliminary adjustment limits or directs the 

 play of association. Yet, usually, no con- 

 sciousness of the nature of the task can be de- 

 tected in the interval between the setting of 

 the particular problem and the reaching of the 

 solution. What consciousness there is of the 

 nature of the task comes in the preliminary 

 period, after the ready-signal; and, even here, 

 as the series of similar problems progresses 

 and the task becomes familiar, the conscious- 

 ness of it tends to be reduced, and finally to 

 disappear, though the adjustment to the task 

 is all the time improving. This result is val- 

 uable both as illustrating the relation of con- 

 sciousness to mental function, and as indi- 

 cating a dynamic factor in thought. In both 

 respects, the result is not entirely new, having 

 been foreshadowed, in another field, by con- 

 clusions of some of the early students of re- 

 action times (Exner, Cattell, Lange) ; but it 

 has now received a much wider extension. 



Another curious result is the frequent oc- 

 currence, in these experiments, of states of 

 mind in which one is clearly aware of the 

 task in hand, or of the solution, or of some 

 other fact, but is unable to detect any image 

 or sensation, or anything which can be de- 

 scribed except as the " thought of " so and so, 

 or the " knowledge that " so and so. Some of 

 the experimenters, particularly Biihler and 

 the present reviewer, have been content to re- 

 gard this description adequate, and to con- 

 clude that such " thoughts " were elements of 



