J.— PSYCHOLOGY. 187 



noted as characteristic of his daughter Armande, may thus be detected as 

 early as the third year. 



Perhaps the strongest impression left by the careful study of thought 

 in little children is the astonishingly early appearance of the elementary 

 thought processes. Time was when it was asserted that science could 

 not be taught to boys below the age of fifteen because inductive reasoning 

 was not possible before then. About a dozen years ago Dr. Cyril Burt 

 caused some surprise by showing that all the varied elementary processes 

 of reasoning can be traced in children's thought at the age of nine. I 

 venture to state that all or very nearly all the elementary thought pro- 

 cesses may occur by about the age of three, at least in children of 

 intelligence quotients of about 1.3 to 1.5, and taking as our clue to 

 elementary thought processes Spearman's classification of relations. 



Certainly it is important that one should be carefully on the look out 

 for these elementary thought processes. They occur sporadically. 

 Like all mental developments in young children, they may appear one 

 day and then apparently disappear and not function even when the 

 situation demands it for some days or weeks. And they no doubt need 

 favourable surroundings, such as a happy home and sympathetic parents 

 or brothers and sisters, if they are to show themselves. It is the lack of 

 observation under such specially favourable surroundings to which I 

 attribute the extraordinarily late date at which Piaget, in spite of his 

 most valuable investigations, places some appearances of thought pro- 

 cesses, which I shall explicitly refer to later. 



Suppose we take first, as fundamental elements of thought processes. 

 Spearman's classification of relations, and enquire when these are first 

 apprehended or ' educed.' Undoubtedly, as Spearman says, relations 

 may be apprehended before they can be expressed in language ; but we 

 will confine ourselves to the earliest appearance of language in which 

 these relations are implied. 



The possessive noun (e.g. ' Daddy's hat,' ' Mummy's hat ') occurs 

 at 1 ; 4, and probably implies a relation of attribution. 



The correct use of ' in ' and ' on ' at 2 ; 1 and 2 ; 2 reveals the education 

 of the spatial relation, and sentences in which it is omitted but implied 

 occur some months earlier. 



' With ' (I brush my hair with Daddy's brush) as indicating a medium 

 or tool is used correctly at 2 ; 3 and reveals a relation of agency — for which 

 I cannot find a true equivalent in Spearman's list. 



Perhaps the most surprising example of the grasp of a relation at a 

 very early age was a sentence by B at 2 ; 4. ' 'At too big (for) B, not too 

 big (for) Daddy.' Here is surely the first glimmering of an idea of 

 relativity. 



The beginnings of a grasp of a causal relation occurs already at 3 ; 8. 

 Thus ' What makes the water come again ? ' (asked when I forked the 

 garden path to let the puddles of water escape.) Cf. ' What keeps the 

 sky up ? ' asked by a boy of 3 ; 7, and again at 4 ; 1 ' Why doesn't the 

 ink run out when you hold up a fountain pen ? ' (See ' Scientific 

 Interests of a Boy,' Forum of Education, 928, p. 21.) 



That these examples of relations are not just fortuitous usings of words 

 the child has heard spoken by adults is indicated by the fact that the 



