zoo SECTIONAL ADDRESSES 



by making some supplementary hypothesis to explain how the pheno- 

 menal shape gets changed from its expected identity with the stimulus 

 shape ? A suggestion that has been made is that the flattened elliptical 

 shape as projected on the retina does produce its own ' sensation ' of 

 shape, but that this sensation becomes changed by a process of judgment. 

 Plainly it is not a judgment in the ordinary sense of reflective judgment 

 since the subject is carrying out no conscious process of inference, but it 

 is claimed that there is a more primitive unconscious process of perceptual 

 judgment which is always modifying our sensations in accordance with 

 the knowledge we have of the nature of the objects producing them. 



There are, I think, many objections to this explanation. First, it does 

 not explain the fact which has to be explained. This is not that the 

 subject judges the ellipse he looks at to be circular, but that, at that 

 inclination, he sees it as circular. He may quite correctly judge it to be 

 really an elongated ellipse, and if he is well informed about such experi- 

 ments (but not otherwise) he may judge it to be making a retinal image 

 which is a flattened ellipse. He may or may not make such judgments, 

 but his immediate experience is of a circular shape. 



Secondly, a process of judgment is affected by the knowledge that the 

 subject has about the facts relevant to the thing which is being judged. 

 This, however, is not the case here. A subject who has not done these 

 experiments before will think that his retinal image is circular when the 

 inclined ellipse looks circular, whereas one informed about the nature of 

 the experiment will know that it is not so. The angle of inclination of the 

 line of vision at which it looks circular will be found to be in no wise 

 affected by the presence or absence of such knowledge. We can, how- 

 ever, make the subject see the inclined ellipse in its retinal shape by 

 giving it a dark structureless background and making him look at it with 

 one eye through a blackened cylinder. Now it will look like a circle only 

 when the retinal image is circular, and under these conditions it will look 

 circular whether or not he is told that he is really looking at an inclined 

 elongated ellipse and not at a circle normal to the line of vision. Again 

 knowledge of what he is looking at does not make any difference ; what 

 is necessary for the apparent shape to be intermediate between the stimulus 

 shape and the real shape is the actual presence of perceptual cues which 

 indicate that the real object is an elongated ellipse with its long axis away 

 from the observer. 



A last and most fatal objection is that this hypothesis requires at least 

 that a sensation corresponding to the retinal image should have been 

 transmitted to the brain in order that a judgment about it may take place. 

 There is, however, no indication that this supposed sensation has any 

 existence as an element of experience. Thesubject doing this experiment 

 under ordinary conditions — with both eyes fully open — is quite unable 

 to see the flattened ellipse which is the shape of the retinal image. He 

 can do so, as we have seen, by altering the conditions of perception, as, 

 for example, by monocular observation through a tube. This, however, 

 is no reason for saying that in any sense he sees this retinal shape under 

 ordinary conditions of vision. Obviously we cannot discover what is 

 seen under one condition of perception by finding out what is seen under 

 another condition of perception. Looking with one eye through a tube, 



