No. 558] CAUSES AND DETERMINERS 



351 



that is the essential point for clearing one's thought. 

 Two diverse cases that require this clearing are of fre- 

 quent occurrence. (1) A question expressed in general 

 terms is so conceived by one person as to require for its 

 answer a certain experiment, while another person 

 understands it in such a way that it requires another ex- 

 periment; thence arises argumentation at cross pur- 

 poses. Clear statement of the problem as an experi- 

 mental situation reveals at once that two diverse ques- 

 tions are under discussion, and gives either immediate 

 agreement, or a method of solving the difficulty by ex- 

 perimentation. 



(2) Questions or propositions as to causality or ex- 

 planation are frequently so framed that they exclude an 

 answer by any conceivable experiment in changing con- 

 ditions. The attempt to state them as an experimental 

 situation at once reveals that they do not belong to (ex- 

 periential) science. Questions expressed in general terms 

 are frequently so understood that no experiment or 

 series of experiments could answer them, though the 

 same questions may be so interpreted that they are 

 answerable by experiment. When one side of a discus- 

 sion understands the question in one of these ways, the 

 other side in the other way, the resulting confusion is 

 dispelled by the attempt to formulate the question as an 

 experimental situation. Our second practical rule aids 

 powerfully in clearing up such matters; it will therefore 

 be taken up before passing to illustrations. 



2. Causation as the Production of the Difference 

 between Two Specified Cases 



Xothing in science appears so productive of confusion 

 and disagreement as attempts to state causes or deter- 

 miners of things. Clearing of thought results if one 

 adopts, at least as a preliminary measure, the rule to 

 search for the causes or determiners of the difference 

 between two specified cases. 



The production of an event or a result requires, as a 



