PRAGMATISM AND THE FORM OF THOUGHT 205 



of inference, the syllogism is dangerously deceptive, just because 

 it effectually conceals the evidences of its weakness. The syl 

 logistic axiom, the dictum de omni et nullo, pretending to represent 

 the essential form of thought in abstraction from all particularity 

 of content, is, in reality, without application to any content what 

 soever; for its terms require just that fixity and clearness which 

 the thoughts of men can never claim. 



The pragmatist theory, that all meanings refer ultimately to 

 correlations of stimulus and response, can be accepted only with 

 certain reservations, which may be summed up in the statement, 

 that such reference is never direct and never uni vocal. Let us 

 consider the latter qualification first. 



A concept is never univocal in its reference to a mode of con 

 duct ; that is to say, its meaning is never limited to the correlation 

 of a certain type of stimulus with a certain response. On the 

 contrary, its import invariably embraces a variety of actions 

 under different circumstances. To take a simple example, the 

 concept of the straight line means that when we wish to look at 

 one object we must take care that a second does not stand in 

 the way; a circumstance which, when it occurs, may be obviated 

 by moving either of the objects, by standing aside, or by changing 

 the attitude of the body. It also means that in order to hit an 

 object with a missile we must throw it in its direction; that in 

 order to reach a destination with the greatest prompitude, we 

 must travel directly toward it; that in order that a rope may 

 not sag it must be stretched taut ; and so on, practically ad infini- 

 tum. So also an apple means to us the eating of it, if it be sound 

 and sweet and our appetite be so inclined ; the paring and coring 

 of it, if need be; the removal of a worm or bruised spot perhaps. 

 And the case is not different with such concepts as joy and sorrow, 

 pity and scorn. We may add that even when the particular 

 situation is given, the concept never determines a specific appro 

 priate adjustment. The immediate one-to-one correlation does 

 not fall within the function of thought. That remains the func- 



