596 



SCIENCE. 



[N. S. Vol. XX. No. 514. 



cessity of exhibiting this assurance which 

 makes it impossible to subordinate logical 

 problems, and forces us at last to questions 

 of epistemology and metaphysics. 



As I am interested here primarily in de- 

 termining the field of logic, it is somewhat 

 outside my province to consider the details 

 of logical theory. Yet the point just raised 

 is of so much importance in connection with 

 the main question, that I venture the fol- 

 lowing general considerations. This is, per- 

 haps, the more necessary because the prag- 

 matic doctrine finds in the concession made 

 regarding the test of validity one of its 

 strongest defences. 



Of course a judgment is not true simply 

 because it is a judgment. It may be false. 

 The only way to settle its validity is to dis- 

 cover whether experience actually provides 

 what the judgment promises, that is, 

 whether the conclusions drawn from it 

 really enable us to control experience. No 

 mere speculation will yield the desired re- 

 sult, no matter with how much formal va- 

 lidity the conclusions may be drawn. That 

 merely formal validity is not the essential 

 thing, I have pointed out in discussing the 

 relation of logic to mathematics. The test 

 of truth* is pragmatic. It is apparent, 

 therefore, that the formal validity does not 

 determine the actual validity. What is 

 this but the statement that the process of 

 judgment is not itself the determining fac- 

 tor in its real validity 1 It is, in short, only 

 valid judgments that can really give us con- 

 trol of experience. The implications taken 

 up in the judgment must, therefore, be real 

 implications which, as such, have nothing 

 to do with the judging process, and which, 

 most certainly, are not brought about by it. 

 And what is this but the claim that judg- 

 ment as such is never instrumental? In 

 other words, a judgment which affected its 

 own content would only by the merest acci- 

 dent function as valid knowledge. We 

 have valid knowledge, then, only when the 



implications of the judgment are found to 

 be independent of the judging process. We 

 have knowledge only at the risk of error. 

 The pragmatic test of validity, instead of 

 proving the instrumental character of judg- 

 ment, would thus appear to prove just the 

 reverse. 



Valid knowledge has, therefore, for its 

 content a system of real, not judged or hy- 

 pothetical implications. The central prob- 

 lem of logic which results from this fact is 

 not how a knowledge of real implications is 

 then possible, but what are the ascertain- 

 able types of real implications. But, it 

 may be urged, we need some criterion to 

 determine what a real implication is. I 

 venture to reply that we need none, if by 

 such is meant anything else than the facts 

 with which we are dealing. I need no 

 other criterion than the circle to determine 

 whether its diameters are really equal. 

 And, in general, I need no other criterion 

 than the facts dealt with to determine 

 whether they really imply what I judge 

 them to imply. Logic appears to me to be 

 really as simple as this. Yet there can be 

 profound problems involved in the working 

 out of this simple procedure. There is the 

 problem already stated of the most general 

 types of real implication, or, in other- 

 words, the time honored doctrine of cate- 

 gories. Whether there are categories or 

 basal types of existence, seems to me to be 

 ascertainable. When ascertained, it is also 

 possible to discover the types of inference 

 or implication which they afford. This is 

 by no means the whole of logic, but it 

 appears to me to be its central problem. 



These considerations will, I hope, throw 

 light on the statement, that while knowledge 

 works, it is not therefore knowledge. It 

 works because its content existed before its 

 discovery by the knowledge process, and be- 

 cause its content was not effected or brought 

 about by that process. Judgment was the 

 instrument of its discovery, not the instru- 



