238 HUMANISM 



XIII 



Mill, and Kant 1 It is difficult not to believe that its 

 success was largely due also to their intellectualist pre 

 judices and their unawareness of its real scope. For in 

 itself Hume s argument, though brilliant, is by no means 

 invulnerable. Indeed, with a little care, we may detect 

 in its proof several flaws and gaps. 



Hume s analysis of the way causes are imputed by 

 us does not go nearly deep enough. 



1 I ) He had no right whatever to start with events 

 and their sequences, and to assume that the problem 

 was how to connect them. Human activity penetrates 

 more deeply into the making of objects of knowledge 

 than either Hume or Kant suspected. It not only 

 turns sequences into consequences, but singles out 

 sequences and events by selection of the relevant, in 

 a way that is always risky, and must always seem 

 arbitrary to an intellectualism which is looking for a 

 fool-proof method of absolute cogency. Hume s em 

 piricism takes over uncriticized the pragmatic realities of 

 common sense, which has analysed experience into a 

 coming and going of things and persons in space and 

 time, and tries to distinguish them still further into a 

 series of impressions of which each is to be a distinct 

 existence. But to a more radical empiricism experience 

 presents itself as a continuous flow, out of which events/ 

 effects and sequences have to be singled out by 

 strenuous efforts, and the causal principle is an instru 

 ment of analysis. The determinate sequences, therefore, 

 for which causal connexions have to be discovered are 

 themselves creations of human attention and interest, and 

 do not exist as such, apart from our volitional activity. 

 Hence they cannot validly produce a basis for a denial 

 of that activity. 2 



(2) It seems to be profoundly vitiated by a confusion 

 between the historical origin and the logical validity of 



1 Cp. J. S. Mill, Logic, III. 5, n. Mill, like Hume, assumes that the 

 volitional theory cannot be true, if it is not certain previous to trial. 



2 Cp. Formal Logic, ch. xx. 3. It is clear that in correcting this funda 

 mental error of Hume s we dispose also of all the philosophies which have 

 assumed with him that the task of philosophy is to find principles of synthesis. 

 Kant s whole problem, e.g. disappears altogether. 



