QUANTITATIVE ESTIMATES OF SENSORY EVENTS 331 



E. ByDr.y.H.Shaxhy. 



It seems clear that the members of the Comimittee as a whole come out 

 by that same door as in they went, and that no general complete agree- 

 ment can be reached. The question we have been asked to decide, if not mal 

 posd, at any rate turns to some extent on the definition of the ternm 

 ' quantitative.' 



Mr. Guild's article confirms this view. With its conclusions I concur 

 on the whole, but I do not feel that his demonstration that sensory magni- 

 tudes are neither of class A nor of class B, disposes of the possibility that 

 they may none the less be magnitudes (of class X say). That they are not 

 magnitudes of practical importance to physicists or physiologists I agree, 

 but this need not in itself consign them to limbo. 



I suppose we should all agree that the conception of magnitude (like other 

 conceptions') has a sensory basis ; one thing is greater than another because 

 it looks larger or sounds louder or smells stronger. The measurement of 

 the stimuli producing these different sensory effects has been the task of 

 physics, and in the process it has been found that the introspective sensory 

 estimates, while in the main giving a correct grading, are rough and often 

 not even self-consistent ; they have therefore been superseded by the 

 ' objective ' physical modes of measurement by scales, balances and what 

 not. But this very fact indicates the nature of these intuitive estimates ; 

 no less than our balance, etc., they are methods of measuring stimuli. 



We look at the sensory experience directly, so to speak, instead of looking 

 at the vernier or spot of light or stopwatch. In so doing we perform a 

 measurement of the stimulus and our result is such a measurement, more 

 or less accurate, and nothing more psychologically fundamental than that ; 

 certainly not a measure of the sensation itself. It is true that we use changes 

 of sensation as our indicators of changes of stimulus, and if we have inde- 

 pendent measures of stimuli we are entitled to use these to deduce the laws 

 which our sensory indicators follow, e.g. (if Weber's Law holds) that our 

 series of just noticeable differences functions as a logarithmic scale of 

 stimulus-intensities. But this in no way justifies us in supposing that we 

 have measured the intensity of our sensations. To do so is merely to lay 

 down a postulate. Postulates may be useful or necessary ; Euclid's are 

 necessary if we are to cross the pons asinoriim. But this sensory postulate 

 is mere lumber, for we make no further use of it. It does not help us to 

 compare stimuli because the intuitive method of doing this does not require 

 it, and it leads to no information about the physiological mechanisms of 

 sense. 2.7.38 



(V) Statement by Prof. J. Drever. 



The Quantitative Relation between Physical Stimulus and 



Sensory Event. 



In my turn I have been invited to present the case for an affirmative 

 answer to the question whether sensation intensity is in any sense measur- 

 able, at the same time dealing with the main objections urged from the 

 opposite point of view. As the arguments on the other side are based on 

 general principles, so must also the answer to these arguments be similarly 

 based. The most fundamental principle of all is the principle that both 

 physicist and psychologist have a common starting-point in the world of 

 sense experience, and a common aim in the fuller and clearer understanding 

 of this world of sense experience. To that we shall return later ; in the 

 meantime let us consider Mr. Guild's arguments. 



