114 PSYCHOLOGICAL METHODS OF TESTING INTELLIGENCE 



tirely out of consideration other compensatorily im- 

 portant portions. If, on the other hand, we keep the 

 idea of compensation in mind and therefore assemble 

 together tests that correlate only moderately with 

 one another, we have no way of knowing which com- 

 bination of tests does actually secure that mutual 

 balance on the basis of which as a whole a more cor- 

 rect quantitative expression of intelligence is to be 

 yielded. 



To the first of these points it can of course be re- 

 plied: we can select tests that show a high correla- 

 tion from such divergent realms of mental life that 

 we shall avoid the danger of testing only a very re- 

 stricted portion of the field of intelligence. But, in 

 opposition to this, attention must be directed to a 

 matter that has not been sufficiently heeded hereto- 

 fore. All test investigations, even if they pertain to 

 the most widely different realms of mental life, have 

 this in common that they are experimental modes of 

 procedure, so that all put the examinee into the same 

 mental condition that of being a subject. And this 

 involves not only very definite adjustments of atten- 

 tion, of mood, etc., but also in particular the habit of 

 mere reacting, of taking at a given moment a re- 

 ceptive attitude toward a problem set from without. 

 Consequently, spontaneous intelligence is excluded 

 by the conditions of the experiment ; there is not in- 

 volved that intelligence that sets its own problems, 

 that thinks out things independently beyond what is 

 immediately given, that anticipates explanations by 

 questions, and that in the real situations of life 

 quickly works out the best way of confronting a sit- 

 uation. And we have absolutely no way of knowing 



