SCIENCE AND NATURE 31 



of it. We judge men's thoughts by their actions. In 

 common life we generally use for this purpose one par- 

 ticular form of action, namely, speech : if a man says 

 "I see a table " I conclude that the thoughts in his mind 

 are the same as those in my mind when I say " I see a 

 table." And men are generally so truthful that we do 

 not often need to examine further. But sometimes we 

 may suspect that a man is wilfully lying and that the 

 relation between his words and thoughts is not normal 

 (although it is again a relation of which we have some 

 experience in our own minds), and we can often detect 

 the lie by examining other actions of his. Thus, if he 

 says that he cannot see a table, we may not be able to 

 make him change his assertion ; but we may be able to 

 induce him to walk across the room, after having 

 distracted his attention from the matter, and then note 

 that he, like ourselves, walks round the table and does 

 not try to walk through it. Such tricks are familiar 

 enough in attempts to detect malingerers in medical 

 examinations. But what I want to point out here is 

 that the method can only be applied to detect lies about 

 a certain class of matters. If a man says that he does 

 not believe that 2 and 2 make 4, or holds that an object 

 can be both round and square, I do not see that we have 

 any way whatsoever to prove that he does not believe 

 what he says he believes. And the distinction is clear 

 between the matters in which lying and imposture can 

 be detected and those in which it can not. As we detect 

 imposture by examining a man's actions, it is only in 

 thoughts and beliefs that affect his actions that we. can 

 find out certainly what he thinks or believes. There may 

 be actually universal agreement on the proposition that 

 2 and 2 make 4, but in this case the objection that we are 

 considering is valid. A single denier could upset that 

 universal agreement, and we should have no way of 

 discounting his assertion and proving that the agreement 



