PERCEPTION 81 



modified to fit data. Theories, in other words, are the effect 

 of data in a way in which data are not the effect of theories. 

 In view of this conclusion it does seem safe to say that 

 science starts with facts, in the sense that it founds its 

 structure ultimately on facts. "Any scientific idea arising 

 in the mind of a scholar is based on a concrete experience, 

 a discovery, an observation, or a fact of any kind, whether 

 it is a physical or an astronomical measurement, a chemical 

 or a biological observation, a discovery among the archives 

 or the excavation of some valuable relic of an earlier civiliza- 

 tion." x Whether the events which thus constitute the basis 

 for science should be called "data" is, perhaps, questionable. 

 Properly speaking, an event becomes a datum only when 

 it is considered in the context of a problem, i e., only when 

 it is responded to by an organism. An event which is isolated 

 from the background of associated events by the selective 

 response of an organism becomes a datum. Hence a fact 

 cannot be a datum until it has been selected. But this may 

 be an attempt to introduce an over-precise terminology. 

 Certainly there can be no confusion in asserting that science 

 rests upon data. The problem which then arises has to do 

 with the methods for getting data. 



THEORIES OF PERCEPTION 



The act of getting data, at least in the natural sciences, is 

 ordinarily called observation, or perception. Clearly, if 

 knowledge is to be justified there must be some theory as 

 to the nature of the perceptual aGt. Prominent among such 

 theories is that of the ordinary man on the street. This 

 may be called the common sense theory, though it repre- 

 sents a somewhat sophisticated rather than a purely naive 

 theory. In other words, it is the position taken by any 

 observer who has recognized the fact of error, and who 

 therefore realizes that objects cannot be perceived directly 

 but must be known through the intermediary of something 

 which makes error possible. The theory is not at all clear 



1 M. Planck, Philosophy of Physics (New York: Norton, 1936), p. 89. 



