CONDITIONAL JUDGMENTS AND PROPOSITIONS 275 



So long as P can follow from other grounds besides VJ/, we are not at 

 liberty to convert &quot; If S is M it is /&amp;gt;,&quot; simply -, just as we are not at liberty to 

 convert &quot;All S s are P t &quot; simply (i 18). And so far as the form of expression 

 goes so far, therefore, as formal logic is concerned the statement that &quot; If 

 S is M it is P&quot; does not in the least suggest that the fact of S being M 

 is the only possible ground of 5 being P. Thus, while S M conditions or 

 involves SP, S P does not condition or involve SM. 



It is pointed out that the only reason for the onesidedness of this relation 

 must lie in the fact that 5 M contains something over and above the ground 

 of 5 /&amp;gt;, something irrelevant to the latter ; that if it contained only the ground 

 of S P we could pass back from S P to SM as securely as we passed from S M 

 to SP, inasmuch as the necessity of conditioning does not differ from the neces 

 sity of being conditioned. It is pointed out, furthermore, that there are such 

 reciprocal hypotheticals known to us, e.g. &quot; If a triangle is equilateral it is 

 equiangular&quot; and indeed all those propositions in geometry which admit of 

 a simple, &quot; geometrical &quot; converse (i 18). And it is suggested that our general 

 aim should be, in every department of knowledge, to discover truths, laws, 

 principles, which we could formulate in reciprocal hypothetical propositions 

 whose antecedents would give a most exact and perfect knowledge of their 

 consequents by revealing to us the adequate and only possible ground of the 

 latter. At these we are invited to arrive by eliminating from the antecedent 

 everything that is irrelevant to the consequent ; the assumption being that if 

 we could do this we should have a reciprocal hypothetical, inasmuch as any 

 single, individual consequent will be seen, if we know it fully in all its relations, 

 to be grounded in one single, individual, corresponding antecedent. In the 

 inductive sciences this same attitude is expressed in the statement that the 

 relation of cause and effect is reciprocal ; that if we knew any phenomenon 

 fully we should see that it could not be indifferently the outcome of any one 

 of a number of alternative causes, 1 but must have sprung from one only : 

 and the existence of coroner s inquests is given as an apt illustration of 

 the soundness of this contention. 



To the theoretical soundness of this view, in so far as it regards the ob 

 jective facts of causality in the world, no exception can be taken. But in so 

 far as it regards the reasons or grounds on which we base our knowledge and 

 our inferences about these objective facts, it seems to confound the latter 

 with our necessarily limited and imperfect knowledge of the former. It is not 

 a true view of human knowledge whether looked at as a process, or as 

 a product which regards our knowledge of any given truth as imperfect so 

 long as we can prove that truth in more ways than one, so long as we can 

 connect it as a consequent with more than one antecedent. No doubt, the 

 different channels of proof for one and the same truth its different logical 

 antecedents must be themselves inten elated as parts of one and the same 

 rational system so that if our knowledge of all things were direct, intuitive, 

 and adequate, we should see how every single element of the whole of reality 

 involved everything else but then our knowledge would not be human : it 

 would be divine. 



At such a stage, or in such a state, of knowledge as that, besides the 



1 C/. Part iv. 221. 

 18* 



