PERCEPTION 95 



character of the end-object. But his knowledge of can 

 not be solely inferential, for he could then never check the 

 validity of his inference. In order to do this he must, at 

 least in certain assumed situations, know directly. Sim- 

 ilarly, in order to know the ways in which the M-, B-, and 

 P-operators function, he must have a knowledge of them 

 which is other than inferential. Hence, he must, at least in 

 certain assumed situations, know M, B, and P directly. 

 The character of these assumed situations has to do with the 

 notion of normal operators, and will be discussed presently. 

 Here all that need be insisted upon is the double function 

 of the operators. They may be considered, indifferently, 

 either as parts of the object, justifying inference to other 

 parts, or as parts of the awareness, conveying direct infor- 

 mation about end-objects. 



The variability in the position of the "screen" separating 

 the awareness and the object can be made clear by an exam- 

 ple. Suppose one is attempting to determine the temperature 

 of a gas by immersing a thermometer in it. He inserts the 

 thermometer, waits a few moments, and reads off the num- 

 ber coincident with the height of the mercury column. He 

 then says that he is aware of the heat of the gas. But this is 

 not quite accurate. What he is directly aware of is only the 

 reading of the instrument; properly speaking, he should say 

 that from this reading he infers the heat of the gas. But he 

 soon realizes that there may be distortions in what he judges 

 to be the reading of the instrument, since he is obliged to 

 observe this through the intervening air and relative to his 

 own spatial position. Hence, he says that he is directly 

 aware only of the coincidence as it appears very close to his 

 eye and with all distorting influences removed. By these 

 two corrections the perceiver has moved the "screen" 

 separating the awareness and the object from the surface 

 of the end-object to the surface of the physical medium. He 

 is directly aware of the appearance of the recording instru- 

 ment, and from this he infers the character of the end-object. 

 However, he soon realizes that even this is not quite accu- 



