308 



and if we stay away from such phe- 

 nomena as the toss of a single coin or 

 tlie decisions of a single mind, then 

 science can often offer "explanation by 

 analog^'," this being useful, interesting, 

 and curiously comforting. But this kind 

 of explanation is, fundamentally, a 

 complete illusion: and at the other ex- 

 treme the strict and formal abstract 

 type of a scientific theory contains 

 nothing whatsoever that constitutes, in 

 any sense, explanation. 



This is a rather shocking thing to 

 say— that science does not furnish any 

 really ultimate or satisfying explana- 

 tion. And this imperfection leads at 

 once to the question: Does science 

 have other important imperfections? 



Without claiming completeness, I 

 want to speak here of a total of five 

 imperfections. You will not be sur- 

 prised, I think, to have me say that 

 these are not, actually, so much im- 

 perfections in science as imperfection 

 in the views that are held by some con- 

 cerning science. Science is amazingly 

 successful at the surface, so to speak. 

 But at its logical and philosophical and 

 artistic core, it has, at least in my view, 

 a number of limitations which can be 

 viewed as imperfections. These are the 

 blemishes that make science a human 

 and endurable enterprise. 



For example, the fact that science 

 is superbly successful at dealing with 

 phenomena, but that it possesses the 

 inherent defect (which I assume it 

 shares with many other fields of 

 thought) that it cannot furnish ulti- 

 mate explanation, is, in my own view, 

 really not a defect at all, but rather an 

 example of the honesty and clarity that 

 comes with maturity. And again, this 

 defect has the virtue that it joins sci- 

 ence to the rest of life, rather than 

 separating it off in cold perfection. 



Second, it is an obvious imperfec- 

 tion that scientists themselves do not, 

 and apparently cannot, agree about 



SCIENCE 



certain of the deepest and most cen- 

 tral aspects of science. 



This imperfection of science I find 

 a most attractice one; for it reflects the 

 fact that science is not monstrous and 

 monolithic, but is a very human enter- 

 prise, exhibiting the same lively and 

 useful diversity which one finds in 

 philosophy, art, music, etc. 



Thirdly, you are all aware of the 

 nineteenth-century fear that science 

 was in the process of imposing purely 

 mechanistic and deterministic inter- 

 pretations upon all phenomena, in- 

 cluding ultimately the individual de- 

 cisions of an individual person. And 

 you are all aware— for this has been 

 widely publicized— that science has it- 

 self now abandoned the view. Science 

 recognizes that the individual events, 

 down at the level of electron, protons, 

 photons, mesons, etc., are all probabi- 

 listic in character, and individually 

 simply not predictable. Since all large- 

 scale events— the falling of a stone, say 

 —are ultimately composed of individ- 

 ual events, the large-scale events are 

 themselves, strictly speaking, probabi- 

 listic also. But the large-scale phenom- 

 ena are nevertheless dependable. And 

 this is simply because this large-scale 

 event is the net result of so incredibly 

 vast a number of small-scale events that 

 the eccentricities always average out. 



So it is an imperfection of science, 

 if you choose so to name it, that it is 

 essentially statistical in nature. This 

 means, for example, that perfect ac- 

 curacy is unattainable in any measure- 

 ment, that certainty is impossible in 

 any prediction. 



Mv fourth defect is related to the 

 fact that there are those who say, "I 

 will admit that science is no doubt 

 more strictly logical than any other 

 field of intellectual activity, but logic 

 is a cold and relentless master, and I 

 am not so sure that I want my life 

 dominated by it." 



