Why Do We Believe What Science Says? 19 



and because they give reasonably simple and often quite 

 beautiful formulations of the relationships between objects 

 or events. One of the reasons they work so well is, para- 

 doxically enough, that we are now always prepared to find 

 out that they don't work. We are sure of our result only 

 when we use a scientific generalization to predict events 

 which past experience tells us are likely to come out right. 

 If we push beyond this famiUar area we are more than 

 ready to note any departures from our expectations and 

 revise our generahzation so as to take these new events 

 into account. 



Scientists as a group, therefore, differ from most other 

 adults in being much more interested in what they don't 

 know than in what they do know. Indeed, it is only by 

 constantly testing what they think they already know that 

 they can be at all sure that they really know it. It is this 

 constant testing of both old and new ideas about how 

 the world is put together that characterizes science and 

 sets it apart from almost all other human endeavors. 



The modern point of view toward the validity of scien- 

 tific statements has several important corollaries. In the 

 first place, it means that science cannot become interested 

 in statements that cannot be tested by seeing whether 

 they fit events which can be observed in what we call the 

 "outside world." Questions which cannot be tested in this 

 way are frequently referred to as pseudoquestions. Many 

 important theological questions are of this character. No 

 satisfactory test of the idea of personal immortality, for 

 example, has ever been devised. Questions in the realm 

 of esthetics encounter the same sort of problem. No one 

 has ever devised a scale for measuring the beauty of a 

 picture. About the best that could be done is to poll a 

 large number of people and find that a certain percentage 

 prefer the "Mona Lisa" to Raphael's "Dresden Madonna," 



