J.— PSYCHOLOGY 201 



we see a flattened ellipse ; looking with both eyes and no tube we see a 

 circular shape, and the flattened ellipse which is the retinal shape is not 

 an element in our experience at all. If it is not an element in our experi- 

 ence, we cannot make a judgment about it of any kind, and the hypothesis 

 of perceptual judgment is in no better position than that of reflective 

 judgment. 



I know that it has sometimes been stated (e.g. by Gelb) that we can see 

 an object in its stimulus character by adopting a special ' critical ' attitude. 

 I think this, however, is a mistaken observation based on the undoubted 

 fact that some subjects can, by adopting a special attitude, see the object 

 more nearly to its stimulus shape ; they can reduce the effect of its real 

 shape on the apparent shape, but I do not find that they can reduce it to 

 zero, nor do I know of any evidence that anyone else has found this. 



We can test this question by asking our subject in the above experiment 

 to try to place his head, not at the height at which the ellipse looks circular, 

 but at that height at which he thinks it makes a circular retinal image 

 (prohibiting him, of course, from partially closing his eyes or otherwise 

 altering the conditions of perception). Now, if the retinal shape (the 

 ' sensation ' of the transmission theory) were itself an element of experi- 

 ence, this should be an easy task. In fact, the subject does not find it 

 easy. He has no immediate knowledge of when the retinal image is 

 circular. He has immediate experience of the phenomenal or apparent 

 shape, and on this he must base a judgment. If he knows nothing about 

 effects of this kind, he will generally judge that the ellipse is making a 

 circular retinal image when it looks circular to him. If he is better 

 informed, he will judge that the retinal image is circular at some angle of 

 inclination at which the ellipse looks elongated along its vertical axis, 

 but any adjustment he makes is merely a guess and generally wildly 

 inaccurate. Even if he happens to guess more or less correctly, he will 

 say : 'I think this must be about the position, but the ellipse doesn't look 

 circular here.' How could this be, if the retinal image were itself 

 transmitted to the brain ? 



We are led from consideration of this experiment to the same con- 

 clusion as was arrived at by Wertheimer as a result of his experiment on 

 phi-movement, that the ' sensation ' corresponding to the conditions of 

 local retinal stimulation, as an element in a complex perception, is a mere 

 fiction. Although it is clear that the conditions of local retinal stimulation 

 affect the resultant perception, we can find no trace of evidence that they 

 do so by being transmitted to the brain as ' sensations.' 



HI. An Alternative way of treating Visual Perception. 



I have examined this experiment in detail because it seemed as good 

 a text as any on which to hang a homily against the transmission theory of 

 vision. Other, more familiar, psychological facts might have been used. 

 Indeed it may be argued that even the familiar fact of the perception of 

 depth cannot be explained on the transmission theory without doing 

 violence to obvious facts of experience. Yet, for many reasons, our 

 minds tend to cling to the transmission theory. Most of all, I think, 



H 2 



