QUANTITATIVE ESTIMATES OF SENSORY EVENTS 327 



case against the prevalent interpretation of these experiments is complete. 

 It depends simply on the fact, demonstrated in the beginning, that they do 

 not provide any practical criteria for associating any sensory magnitude with 

 number in the particular manner which is essential for measurement, so 

 no relation between sensory magnitudes and stimulus magnitudes can 

 possibly be derived from them. If the explanation I have suggested should 

 prove wholly, or in any important aspect, unacceptable, we are simply left 

 for the time being with no explanation of the experiments. The old one 

 will not do ; and the sooner writers on psycho-physical problems stop 

 describing and interpreting their experiments in terms of sense distances 

 and sensation scales the sooner will it be possible to seek a true interpretation 

 of their results freed from the tendentious influence of a falsely suggestive 

 terminology. 



The foregoing general considerations cover all the other types of experi- 

 ment in which the observer associates numerals with perceived stimuli 

 composed in different ways, such for example as the well-known experiments 

 of Dr. L. F. Richardson and his colleagues. 



This Report is already too long to permit discussion of such methods in 

 detail, but it can be said of all of them that whatever may be the criterion 

 by which the observer assigns a number to an observed stimulus relation, 

 or makes some equivalent decision in connection with it, such as marking 

 a point on a line to correspond to the ' position ' of a percept in a series 

 ranging between two extremes, it cannot consist of the intuitive perception 

 of some quantitative relation between psychological magnitudes, for there 

 are none to perceive. In all these experiments, as in those we have con- 

 sidered more fully, the guess, estimate, or judgment, whichever it is, relates 

 to stimulus magnitudes and not to sensations. It may be a guess or estimate 

 based on direct association with known cases of the same type of stimulus 

 relations. This is a process we perform almost every day when we estimate 

 lengths, weights, temperatures, etc., without measuring them. It depends 

 merely on direct association, and in the case of those things with which 

 we are very familiar may often be effected with considerable accuracy. 

 Those experiments which are not explainable on the basis of direct associa- 

 tion must have for their criterion the perception of sameness of relation 

 shape between some elements of two or more relation structures. 



The foregoing discussion centres round the possibility of measuring 

 sensation intensity as an A magnitude — that is, as something expressible in 

 terms of units of its own kind. A few words must now be devoted to the 

 possibility of treating sensation intensity as a fi magnitude, defined by an 

 arbitrary relation to stimulus intensities. In the first place, assuming it 

 can be done, it would serve no purpose whatever. It would merely result 

 in our being able to say that the intensity of sensation corresponding to the 

 stimulus I is the intensity of sensation corresponding to the stimulus /, 

 which would not help us much in any psycho-physical problem. The 

 utility of a temperature scale, which, as we saw earlier, is defined by an 

 arbitrary relation to the properties of a standard thermometer, is that the 

 thermometer can be used to measure the temperatures of other bodies. 

 It would clearly be useless to define the temperature of a resistance 

 thermometer as a function of its resistance, as we do, if it could only be 

 used to measure its own temperature. All the definition would mean is 

 that the temperature of the thermometer when its resistance is R ohms is 

 the temperature corresponding to a resistance of R ohms. This would be 

 the position as regards sensation intensity. We could not use the sensation 

 scale established by the definition in one standard sensorium to measure 

 the sensations in other people's sensoriums, because our criterion of equality 



