114 HUME IV 



first place there is the idea of an object ; and 

 secondly, there is the idea of the relation of ante- 

 cedents between that object and some present 

 objects. 



To say that one has a recollection of a given 

 event and to express the belief that it happened, 

 are two ways of giving an account of one and the 

 same mental fact. But the former mode of stat- 

 ing the fact of memory is preferable, at present, 

 because it certainly does not presuppose the exist- 

 ence of language in the mind of the rememberer ; 

 while it may be said that the latter does. It is 

 perfectly possible to have the idea of an event A, 

 and of the events B, C, D, which came between it 

 and the present state E, as mere mental pictures. 

 It is hardly to be doubted that children have very 

 distinct memories long before they can speak ; and 

 we believe that such is the case because they act 

 upon their memories. But, if they act upon their 

 memories, they to all intents and purposes believe 

 their memories. In other words, though, being 

 devoid of language, the child cannot frame a pro- 

 position expressive of belief; cannot say " sugar- 

 plum was sweet " ; yet the physical operation of 

 which that proposition is merely the verbal ex- 

 pression, is perfectly effected. The experience of 

 the co-existence of sweetness with sugar has pro- 

 duced a state of mind which bears the same relation 

 to a verbal proposition, as the natural disposition 

 to produce a given idea, assumed to exist by 



