18 COMMON SENSE (aND SCIENCE) 



predictive capacity preserve themselves through ages of human 

 thinking. 



The concept "man" relates yesterday's experience of one man to 

 today's experience of a second and to tomorrow's predicted experi- 

 ence of a third. But how? The concept shows me an element of iden- 

 tity in all the men, but by itself offers nothing intrinsically useful for 

 prediction. "Relate" then must have some further meaning. Effective 

 organization and prediction of experience implies not merely con- 

 cepts but relations among concepts. In my mind the concept "man" 

 is associated with a group of other concepts referring to many as- 

 pects of "human behavior." Through such associations the concept 

 *'man" takes on the full measure of its meaningfulness. "Man" may 

 be old or young, strong or weak, tall or short, white, black, red, or 

 yellow— but by and large "man" shows "human behavior." Organized 

 by the concepts of human behavior, my past experience of "man" 

 permits me to guide my dealings with this man noio by means of 

 forecasts of the various results likely to be produced by alternative 

 courses of action. Only by way of the confirmation of many of those 

 forecasts can the appropriateness of my concepts be demonstrated 

 and their denotations become firmly established. Our selection and 

 understanding of concepts is required for, and inversely determined 

 by, the statement of the relations we seek. 



FROM CONCEPTS TO COLLIGATIVE RELATIONS 



What I shall call colligative relations so link together certain ele- 

 ments of our past experience that the appearance of such elements 

 in future experience is made subject to prediction. Typical colligative 

 relations are: "The burned child dreads the fire" or "Water seeks its 

 own level" or Boyle's law— "At constant temperature and with con- 

 stant quantity of gas the product of pressure and volume is a con- 

 stant." A scientific law is often a colligative relation, as Jeffreys 

 indicates : 



The test of a scientific law is its capacity to account for sensations 

 already recorded, and its interest lies largely in its capacity to pre- 

 dict new ones. 



The adjective colligative is essentially neutral and descriptive. 

 Literally, to colligate is to bind together. Every relation must repre- 

 sent some kind of binding-together, and to this extent the adjective 



