164 WHAT IS SCIENCE? 



be more perfect than the rest of mankind. But any 

 danger of paying too much or too little heed to pronounce- 

 ments put forward on behalf of " science " will be 

 avoided if the distinction on which so much stress has 

 been laid is borne in mind. On questions of means to a 

 given end (if they concern the nature of the external 

 world) science is the one and only true guide ; on ques- 

 tions of the ends to which means should be directed, 

 science has nothing to say. 



THE CERTAINTY OF SCIENTIFIC KNOWLEDGE 



I have thought it better thus to start with a considera- 

 tion of the limitations of science ; not because the greater 

 danger lies in the neglect of those limitations, but merely 

 to convince the reader that I am not blind to their exist- 

 ence. Actually, in this country at least, the greater 

 danger lies in the other direction, in refusing to accept 

 the clear and positive decisions of science on matters 

 which lie wholly within its bounds. Why is there any 

 such danger ? It arises, I believe, from two sources not 

 wholly independent. The first source is a disbelief that 

 science is really possessed of any definite knowledge. 

 Scientific experts seem to differ as much as experts in 

 other subjects, and may be heard in any patent litiga- 

 tion swearing cheerfully against each other. The second 

 source is a general distrust of the " theorist " as com- 

 pared with the " practical man." The chief points that 

 have to be raised will appear naturally in a discussion of 

 these two errors. 



It may be thought that the first " error " has been 

 implicitly confirmed by our previous discussions. For it 

 has been urged that there is a strong personal element in 

 science and that complete agreement is to be found only 

 in its subj ect matter and not in its conclusions. But while 

 it is perfectly true that a theory, and even to some extent 



