2o6 SECTIONAL ADDRESSES 



light to the eyes than a brightly lighted piece of black paper, it does not 

 necessarily look less white than the black paper, although it may do so if 

 the shadow is very deep. The seen whiteness is in between the ' real ' 

 whiteness and the stimulus intensity of the retinal image. Again, in this 

 tendency to see objects in their real whiteness irrespective of illumination, 

 we find wide individual difi'erences. 



These effects have been generally described under the name ' the 

 constancy tendency.' I do not much like this name since the distinc- 

 tive feature of these effects is not that phenomena (or appearances) tend 

 to remain constant while stimuli change. It is easy to arrange an experi- 

 ment (such as that of the inclined lantern screen already mentioned) in 

 which the stimulus remains constant and the phenomenon changes. A 

 more fundamental feature of the effects seems to be that phenomena are 

 determined not only by local stimuli but also by the perceived ' real ' 

 characters of the objects causing the stimulus. I have, therefore, 

 suggested that we should call these effects ' the tendency to phenomenal 

 regression to the " real " characters of objects.' I have no wish to 

 quarrel with those who prefer the term ' phenomenal constancy,' but it 

 is convenient to stick to one name, so for the purpose of the present 

 address, I shall speak of ' phenomenal regression,' and I shall call the three 

 effects above described : phenomenal regression for shape, for size, and 

 for colour, respectively. 



When we include under one name (whether ' constancy ' or ' pheno- 

 menal regression ') these three tendencies to see objects more or less in 

 their real shapes, sizes, and colour, irrespective of their inclination, 

 distance and illumination respectively, we are implying that these three 

 effects are all of the same nature. For this assumption, we need better 

 evidence than the mere fact that all three can be described in similar 

 terms. The direction in which to look for this evidence is suggested by 

 the fact that in all of them there are wide individual differences. If now 

 the individual who shows a large tendency to see things in their real sizes 

 tends also to see inclined objects near to their real shapes and objects in 

 different illuminations near to their real albedos, we have positive evidence 

 to justify the natural suspicion that these are simply different aspects of 

 one general tendency. 



We can easily determine by experiment whether or not this is the case. 

 If we test a group of subjects in their tendency to phenomenal regression 

 for shape, for size, and for whiteness, we find that those who have a large 

 tendency to see the ' real ' size of an object tend also to have a large 

 tendency to see the ' real ' shape and the ' real ' whiteness. The cor- 

 relations between these tendencies are about o-6, which shows that they 

 have a considerable factor in common. We can thus speak of individuals 

 as having high phenomenal regression if their perceptions of apparent 

 shape, size and whiteness are largely determined by the ' real ' characters 

 of the objects looked at, while those whose perceptions are determined 

 relatively more by the conditions of retinal stimulation (i.e. who see 

 objects getting much smaller as they go farther away, and so on) we shall 

 describe as those of low phenomenal regression. 



It may seem fantastic to suggest that there are such large individual 

 differences in the way the world looks to different people since certainly 



