The Subject of Consciousness
© C.O. Evans

Chapter 4

Unprojected Consciousness - 4.4.10
[table of contents]  [previous]  [next]

4. The Logical Dependence of Mental Images and Thoughts on Bodily Activity

[10] I align myself with psychologists such as Ribot and Bain, and philosophers such as Hobbes and Hume, 97 who stress the intimate connection between mental imagery and perception. The position I shall argue for consists of two claims. The first is that no one

-page 131-

lacking the requisite sense modality can enjoy the corresponding form of mental imagery. Thus, no one lacking sight can have visual imagery, no one lacking hearing can have auditory imagery, and so on. The claim must be understood as denying a logical possibility. The second is that it is not possible to pay attention to our mental images without inhibiting spontaneous sense-organ activity. Thus, no one can give his attention to a visual image unless he ceases to engage in activities such as 'looking about', 'looking for' and 'looking at'. His eyes may remain open while he has the visual image, but their being open is not itself a sense-organ activity. Again, this must be understood as a denial of a logical possibility.

Persons, such as Bain, Ribot, and I, who make these two claims (either explicitly or by implication), I shall describe as adopting the 'sentient approach' to mental imagery. With this approach I shall contrast what I shall call the 'phantom approach'. I give it that name because it expresses the view of those who, in Bain's words, hold that the image is 'a kind of phantom without definite seat'. The phantom approach consists of a denial that there is a logical dependency of mental imagery on sense perception. Those espousing the phantom approach might be prepared to concede a contingent connection between the sense modality and the related form of imagery. They could therefore agree that a man without sight from birth might as an empirical fact not have any visual images, but they would insist that there is nothing logically impossible in the idea of such a man having a visual image. The same would be said mutatis mutandis of the other sense modalities.

It is worth bearing in mind that on the question of disembodied existence the phantom approach would leave open the logical possibility that a disembodied being could have mental imagery. Such a possibility is ruled out by the sentient approach. Thus the truth of one or the other of these conflicting claims would have a bearing on the intelligibility of the notion of a disembodied existence. The loss of mental imagery would constitute perhaps a decisive impoverishment of the consciousness of a disembodied being granted that sense perception would already have been conceded to be lacking. 98

I begin by attacking the phantom approach, hoping in the process to clear the way for the sentient approach. The first point to notice is that the very notion of a mental image is dependent upon the notion of the corresponding experience. When a description is

-page 132-

given of an image, it is not ordinarily described as image. What is described is the thing which the image is an image of, and even when we wish to describe the image qua image we must resort to the language of perceived objects. Thus if we are dealing with a visual image, then the description will draw upon such concepts as 'look', 'perspective', 'colour', and 'shape'. But a person who had lacked the visual experience without which he could not be said to have these visual concepts in the full sense could not apply them to describe imagery, nor could he understand what was meant by imagery described in terms of them.

What I am saying amounts to this. The descriptions we give of our visual images of necessity make use of visual concepts. Now let us suppose that a man sightless from birth claims to have a visual image. 99 How can we ascertain whether his claim is true? Suppose we ask him to describe the image. There are two possibilities: (a) he may try but fail to give a description, or (b) he may in fact produce a description making use of visual concepts. If he cannot give a description, it seems to me we have every right to be sceptical about his claim that his image is visual. What, we may ask, makes him call it visual? What does he understand the word 'visual' to mean here? If (b) is the case, the position is different: sightless people can come to learn the correct use of visual words, such as 'red'; but correct use of the word 'red' does not entail that the user can see red or that he can see simpliciter. We have no guarantee, therefore, that when a man uses a visual term to describe his image, it must be a visual image he is describing.

The retort might be made that these are no more than difficulties of verification. We know that sightless people who have their sight restored cannot at first distinguish one geometrical shape from another by sight, in spite of being able to distinguish them by feel. Bearing this in mind, we are not entitled to transfer our doubts about the sightless man's description of his image into doubts about the possibility of his having a visual image. It is theoretically possible, according to this reasoning, for the man without sight to have a visual image without he or anyone else knowing that his image is visual. But this point raises many difficulties of the same order as those raised by the question whether a newborn infant sees what we see. If it is allowed that the phenomenal character of our perceptions is altered by our ability to pick out and recognize

-page 133-

perceptual objects because of our possession of concepts, then it must be allowed that the phenomenal character of the alleged 'visual' image of the sightless man must be different from our own because no similar abilities and recognitional capacities are exercised in connection with it. Whatever the phenomenal character of his image we could still reasonably doubt that it could be visual. Other considerations, too, tell against this objection and these we shall come to shortly.

Perceptual concepts such as 'seeing' and 'hearing', entail the concepts of 'an organ of sight' and 'an organ of hearing'. 100 Let us continue to confine ourselves to sight. When we look at an object, the object of necessity appears from a certain perspective, determined by the location of our eyes. We see objects in a particular direction, and at a certain distance from our eyes. We see only an aspect of the object, visible from our particular perspective. We do not see, as it were, the object 'in the round' all at once. In other words, we cannot see an object from every perspective at once. That we are limited in this way is part of what we mean by 'seeing'. Now it is true also of a visual image that the description we give of it is logically similar to the description we give when we are seeing rather than imaging. As I have said, we describe the object of which the image is an image, and that means that we describe the object as though it were being seen. We would describe the object as though it were being seen from a certain perspective, and this entails the object's presenting a certain aspect, and being at a certain distance. 101 Not even in visual imagery can we entirely escape the limitations of perspectival appearance. The object as imaged presents an appearance from a position, in direction, at a distance, in the same way as does the object as seen.

The question arises: Can a person without sight satisfy these conditions for visual imagery? Or to put it another way: Would he be able to make sense of them? He certainly would have no visual experience of these perspectival characteristics, which visual perception and visual imagery have in common. I am not asserting that a sightless person lacks a spatial system, but only that he lacks an experience of perspective. He will have no experience of the real size of an object differing from its apparent size, nor of the characteristic

-page 134-

distortions of shape that result from foreshortening. Furthermore, I can have a visual image the phenomenal character of which is due to the fact that it is as if it were of an object I had seen out of the corner of my eye, or due to the fact that it is as if it were of an object the surface of which I was scanning with my eyes. The alleged image of a sightless person could not be an image with such a phenomenal character. In fact the images of a person without sight can have none of the characteristics associated with perspective that I have described. But in that case his alleged image will lack precisely the characteristics that make it visual.

The essence of my argument is that the concept of a visual image is logically dependent on the concept of seeing, and the latter, in turn, is logically dependent on the having of 'sight'. Its import is that a sightless person necessarily lacks the conditions necessary to the having of visual imagery. This is one more consideration telling against the phantom approach.

Yet another is that if the phantom approach were true, it would be difficult to understand why there are modes of imaging corresponding to each of the sense modalities. Why should there not be ten or eleven types of mental image, perhaps including X-ray images, and radio wave images? Furthermore, if the connection between a sense modality and an image is contingent, why should the image be named after the sense modality? Why should the images be described as visual, auditory, tactile, and olfactory, instead of being given some other description, if the description of the one did not logically depend on the description of the other?

[table of contents]  [previous]  [next]


97. Hobbes for maintaining that imagination is but decaying sense, and Hume for maintaining that there is no idea without an antecedent impression. There is a problem for these empiricists as to whether this is an a priori truth or a matter of fact.
back to text

98. See below, pp. 218-9, where this issue is taken up.
back to text

99. I shall from now on simply refer to a man as 'sightless', and take the qualification 'from birth' as read.
back to text

100. See Shoemaker, Self-Knowledge and Self-Identity, ch. 5, where this point is covincingly argued.]
back to text

101. The question of the perspectival quality of the visual image must not be confused with the question of the location of the visual image itself. This is separate question, which is dealt with in the next section.
back to text