Sight and touch : an attempt to disprove the received (or Berkeleian) theory of vision

Preface

The question discussed in the following pages is one of small compass, but no one versed in philosophywill estimate its importance by its extent. The theory assailed is in fact the stronghold of scepticism; for if consciousness is once proved to be delusive, there is an end to all appeals to its authority : doubt must reign supreme. It is to no purpose to say that it is not consciousness that is proved to be delusive, but an inveterate belief which is mistaken for a deliverance of consciousness; for it is practically the same thing whether consciousness itself deceives or something which is undistinguishable from it. It is of little use to prove that a certain witness is trustworthy, if in doing so we also prove that his evidence is falsified before it reaches us.

There is indeed only one dogmatic system conjistent with the Berkeleian theory of Vision, and that is the Berkeleian Idealism. Yet dogmatists of all schools have accepted the theory. The only detailed attack upon it on philosophical grounds with which I am

IV PREFACE.

acquainted is that of Mr. S. Bailey : " A Review of Berkeley's Theory of Vision, designed to show the unsoundness of that celebrated speculation" (London, 1842); and to this I owe some important suggestions. Sir David Brewster is known to be a decided opponent of the theory, but he has not attacked it on its philosophical side. I make no apology therefore for offering the following refutation, in which, as far as I was able, I have examined every aspect of the question.

Since these sheets were in press, the New Sydenham Society has issued the valuable work of Professor Donders, " On the Anomalies of Accommodation and Refraction of the Eye," which includes, in addition to his own important researches, all the essential information on the normal phenomena of adjustment and refraction, for which references had been given to foreign periodicals.

Note. — References given at second-hand are enclosed in square brackets, and follow the name of the work from which they are taken.

Contents

Page Introduction, 1

Chapter I. The Paradox and its Proof— Statement of the Question, 3

II. Of the Laws of Association, . . . .13

III. Application of these Criteria, . . . .19

IV. Illusions of Sight, and Conditions of its Perceptions, 36

„ Phenomena Inexplicable on the Berkeleian Theory, 48

V. Is Touch Perceptive of Extension and Distance ? . 60

VI. Sight the Sense properly Perceptive of Distance

and Trinal Extension, 75

VII. Perception of Distance in Vision with one Eye, . 85

VIII. Of Binocular Vision, 106

IX. What we Learn from Experience or Association, . 134

X. Cases of Persons Born Blind, . . . .140

XL Of the Evidence derived from Infants and Animals, 163

XII. Conclusion 174

SIGHT AND TOUCH.

Introduction

If we were chaUenged to point out a single discovery in mental science which is universally admitted, we should at once name the "Theory of Vision" of Bishop Berkeley. Its success has been, indeed, extraordinary. From the time of its first publication it has been accepted almost without question. Adam Smith indeed whispered a doubt, and Sir W. Hamilton, with singular tenderness and timidity, insinuates that although the theory seems to be satisfactorily demonstrated, it is yet in certain circumstances, "provokingly found totally at fault." But the only decided assailant of the theory in recent times was Mr. S. Bailey, and he was at once set aside as an impracticable sceptic ; Mr. J. S. Mill, who entered the lists against him, deeming apparently that in such a contest it needed but the very slightest effort of his skill to overthrow his adversary. At the present moment every one who has tasted philosophy, even summ is labris, is firmly convinced that he sees, not persons and things of various bulk, and at divers distances, but merely a variety of colour, or at the best, a flat picture of no perceptible magnitude, and at no perceptible distance. Yet the profoundest metaphysician, when he opens his bodily eyes, is mastered by the same belief as the unlearned : he cannot see what he knows he does see, and he cannot help

B

9 SIGHT AND TOUCH. I

seeing what he knows it is impossible to see. Seeing is, for the time, believing; but on deliberate reflection, sense is fairly overpowered by reason. This is truly a brilliant victory of science. An universal persuasion that nothing really exists, would be scarcely more surprising. It is an ungrateful task to attempt to deprive psychology of its only acknowledged triumph, and to wrest from philosophers the cherished analogy which never failed to support them in bidding defiance to probability and common sense. The adventurous critic who does so must encounter, not only the prejudice in favour of an established doctrine, but the not less powerful prepossession of logical minds in favour of what appears to be close reasoning. Yet this theory, whether true or false, is, in truth, the shame, not the glory, of psychology ; for it is a discovery in its domain made, not only without its help, but in spite of it, by physical reasoning. It has hitherto baffled psychology, either to verify it or to shake the physical basis on which it rests. But if the closest attention fails to discriminate ideas, not only successive, but of diverse senses — fails to distinguish perception from recollection — fails to discover ideas known to be constantly present — if psychology has been forced to make this ignominious confession at its first encounter with physics, it is time that it should abandon all pretence to be a science of observation, or, indeed, a science at all. The present state of the question is, moreover, a disgrace to philosophy. The theory, we are told, is at variance with common facts, and yet the proof appears to be "as satisfactory as anything in the whole compass of inductive reasoning."* By exposing its fallacy, we shall not only remove a reproach from philosophy, but shall sweep away with it not a few erroneous doctrines, of which it has been the chief or sole bulwark.

* Hamilton on Reid, p. 183.

THE COMMON SENSE VIEW.

CHAPTER I.

THE PARADOX AND ITS PROOF — STATEMENT OF THE QUESTION.

Let us first consider the prima facie evidence of common sense. When I feel an object with my finger, there can be no sensation produced unless some force be exercised; for instance, by pressure, which is followed by an unknown change in the nerve of touch. In order to exert this force, the object must be within a very minute distance ; and if that distance does not exceed a certain small limit, each sensibly distinct point of the object affects but one distinct sensitive spot. In this respect, sight differs from touch only in the much greater variety of distance at which an object can so affect the organ. In another respect, however, the two senses are psychologically contrasted, there being in the sense of sight less consciousness of ourselves as affected, than in any other sense. In some of the senses we are conscious of little but the affection of our organism ; as in taste and smell, which are not accompanied with the perception of an external object. In touch we perceive something external, but are at the same time conscious of an affection known as the feeling, of hardness, &c. ; in sight, on the contrary, we are not conscious of any feeling whatever. We see external things, their figure and colour; for though colour is in one sense a sensation, we are not conscious of it as such — red is perceived as red, in consequence of the constitution of our organs, but it is perceived as belonging to and in the place of the object seen. A striking testimony to this peculiarity of sight is found in the metaphorical use of terms borrowed from the different senses. Do we speak of knowledge, pure and simple, we use

b2

V SIGHT AND TOUCH.

the verb to see; while, on the contrary, feeling is in common speech synonymous with sensation. Now, what is the extent of this perception of sight? Ask the common man how he knows that objects exist without him, that they are of such and such a figure, that they are at a distance ; he will at once answer — "Do I not see them?" "But how do you know that the tables and chairs you see are not an illusion?" To this he will doubtless reply by an appeal to the sense of touch and resistance. But it is easy to distinguish the sphere of the testimony of the two senses. For, again, let us ask the man who has felt a pressure or resistance, how he knows that the cause is external, and it is invariably to sight that he appeals. Thus, if touch assures us of the substantial reality (objective existence) of the objects of sense, sight gives the most decided testimony to their externality and distinctness from ourselves. In touch, moreover, we perceive something resisting — something, therefore, outside us, or rather, which is different from the feeling organ, yet at no distance from it, for outness and distance must not be confounded* Indeed, most of the words which express apposition without interval are borrowed from touch. The object of sight, on the contrary, is perceived as at some distance : we can move towards it or away from it, or interpose our hand between it and the eye. This is the doctrine of common sense. To believe it is philosophical heresy ; it is to prove one's self incapable of appreciating a fine distinction or a subtle argument. The orthodox doctrine, the discovery of which is Berkeley's chief claim to immortality, is the following.

We say, for instance, that we see an orange, which we can feel by putting out our hands, and if we choose, can divide and suck; we see, that is to say, its figure, size, colour, externality, and distance. In truth, says Berkeley, we see no such thing. By sight, we are aware of a yellow figure, which

* See Hamilton on Reid, p. 177.

THE ORTHODOX THEORY. 5

however has no resemblance to the figure perceived by touch ;* of a certain degree of dimness or clearness, distinctness or confusion, and of a certain sensation in the muscles of the eye. But these ''visible ideas" being constantly associated with certain ideas of touch — as softness, roundness, distance, magnitude, and figure (which have no resemblance whatever to the former, nor any natural connexion with them) — these latter come to be immediately suggested by the former, and that so certainly and quickly, that we mistake them for objects of the same sense, and say that we see the figure, distance, magnitude, &c. ; whereas these attributes and that of outness being implied in the idea of an orange, we do not really see the orange, except in the same sense in which we may be said to see "anger or shame in the looks of a man,"-f- or in which we "hear a rock,";f when the word is pronounced. "It is a mistake to suppose that we see and feel the same object."§ He " would fain be informed how we come to know that the same object causeth ideas by different senses." || Berkeley, logically enough, went further than his successors in accounting in the same way for the idea of number given by sight. That we see one orange only, even with one eye, is, according to him, the result of association. In order, however, that I may not be charged with putting an interpretation upon the " inaccurate "^f language of Berkeley, which his disciples would disown, I shall quote and use Mr. Mill's own statement, viz., "That the information obtained through the eye consists of two things — sensations, and inferences from those sensations : that the sensations are merely colours variously arranged, and changes of colour; that all else is inference, the work of the intellect, not of the

* "Theory of Vision," sees. 117, 127. f Ibid, sec. 65.

t Ibid, sec. 136. § Vindication, sec 21. || Theory, sec. 108.

^ Berkeley's language is really accurate enough, but is, of course, tinctured with his own idealist views.

eye."* "When we judge by the eye of the remoteness of any object we judge by signs; the signs being no other than those which painters use when they wish to represent the difference between a near and a remote object. We judge an object to be more distant from us by the diminution of its apparent magnitude, that is, by linear perspective ; or by that dimness or faintness of colour and outline which generally increases with the distance, in other words, by aerial perspective."-)- Mr. Mansel's statement agrees with this, except that he allows that the convergence of the optic axes may perhaps aid us in estimating the distance of near objects. We shall see hereafter that the signs enumerated by Mill may be eliminated or reversed, one and all, without the judgment of distance being rendered one whit less decided and irresistible. In this respect, he has not improved upon Berkeley, who enumerates as the signs of distance — 1st, the convergence of the axes; 2nd, confusedness or faintness; 3rd, straining of the eye, besides other secondary signs. The principal difference between him and his successors is, that he held (sec. 158) that planes are no more the immediate objects of sight than solids; whereas the modern doctrine is, that the immediate object of sight is a plane variously coloured. In a word, the theory has been exactly stated by Swift, "Vision is the art of seeing things invisible."

I shall endeavour to show that this theory proceeds on assumptions which can be proved to be false ; that it supposes processes which are opposed to all analogy and must be pronounced impossible, and that its results are in contradiction to experience. It is necessary however to make one or two preliminary observations. First, that the only account we can give at best of any perception is to show the mode in which the appropriate affection of the organ is produced. This, in the case of taste, for example, is by the solution of * Dissertations, vol. ii., p. 89. f Ibid, p. 87.

HOW WE KNOW WHAT WE SEE OR FEEL. 7

the sapid matters in the secretions of the organ. With regard to most of the senses indeed we have not yet advanced even so far. We do not know, for instance, the exact effect produced by sound on the organ of hearing. But in no case can philosophy be called on to show how that affection can issue in such and such a perception. The opinion that a resemblance must exist between the perception and its antecedent will scarcely be explicitly maintained by any philosopher of the present day. Secondly, we must observe that the question is not about the origin of the idea of space, but about its empirical apprehension; nor is it the question, whether the eye can measure distance up to the limit of distinct vision, but whether it is capable of giving us the fact of distance at all, without the help of any other sense. As this distinction between the perception of distance absolutely and that of degrees of distance is very important, and as they have been explicitly confounded by Mr. Mill, I shall develop it a little more fully.

It is conceivable that sight might convey the idea of objects being at a distance, and even of one object being farther than another, yet without our having the least notion of the measure of that distance. For, consider the case of colour, which is admitted to be the proper object of vision, and which in fact, as colour, is simply a mode of sensation. Well then, let us present to a person not colour-blind several shades of green together: he will readily distinguish and arrange them in order of depth, lightness, &c. ; but to recall them to mind and reproduce them is another matter. Ask the same person to define the depth of a particular shade seen apart from others, or which comes to the same thing, to say how much lighter or darker it is than one he saw yesterday, and in nine cases out of ten he will judge wrongly. Few persons, except those who are obliged to attend to these sensations, are able to reproduce them with any degree of accuracy. In illustration of this, it may be mentioned, that in the discussion between Mr. Airy and Sir David Brewster about the solar spectrum some years ago, the former, himself a practised observer, asserted that "the eye has no memory for colour." Brewster, in his reply, stated that his own memory of colour was perfect* If the question be of the pitch of a sound, we shall find still fewer persons able to give an accurate judgment. Yet it would be absurd to say that every man who cannot accurately designate shades of colour is colour-blind ; or that unless a man can tell what note or chord he hears when a piano is struck, he is insensible to music, and all tones sound alike to him. If he can distinguish colours or tones when in juxtaposition the defect is not in the sense. It is the same with perceptions, as from the nature of the case it ought to be. We perceive resistance, but who can accurately measure the amount ? We perceive muscular effort, and hence weight ; yet how few can estimate either effort or weight, except in the roughest possible way. The explanation is plain. To know what colour we see, what tones we hear, what effort we exert, means this, and no more: — to compare the colour, tone, effort, or weight with other sensations or perceptions of the same kind, and assign them their proper place in the scale. To speak of perceiving what an effort or tone is absolutely, is as unmeaning as to speak of the absolute place of an object. To know, then, what we see, hear, or feel, is not a matter of direct perception, but requires comparison and judgment. But as the other terms of the comparison are not simultaneously perceived, there is required further an effort of memory. But neither the power of reproduction nor that of judgment is proportional to that of perception, or in any way dependent on it. Instead, therefore, of the one operation of perception, we have three or five distinct processes ; and the power * Philosophical Magazine, vol. xxx., pp. 74, 153.

CONDITIONS OF PROOF. 9

of perception may be conceived to exist in the absence of all the rest, as is very probably the case with some animals.

Having thus relieved the question of irrelevant matter, let us consider what is required in order to prove that the visual perception of distance is the result of association with perceptions of touch.

First, it must be proved that sight actually does not perceive distance. This may be shown a priori or a posteriori. The only a priori proof possible is a physical one — one, namely, which should show that the immediate object in perception must be identical whatever the variation in the circumstances supposed to be perceived. An objection founded on the fact that the variations are such as contain no necessary connexion with the idea of distance, is perfectly worthless, since the process of every sense is equally beyond our ken : nowhere is there any necessary connexion between the idea and its antecedent. A posteriori, it might be shown that persons deprived of sight are capable of the perceptions in question ; while those who possess sight and not the locomotive faculty, are not. Secondly, it must be proved that distance is perceived by touch, or the locomotive faculty. This may be done by the observation just mentioned; or we may have recourse to the Instant ice variantes in proximo, and show that the accuracy or power of perception of distance is proportional to the energy and exertion of the locomotive sense. Thirdly, the fact of the association between the perceptions of touch and the sensations of sight must be established; and fourthly, it must be shown, that the variations in the suggesting sensations correspond with those in the distance perceived. In stating these conditions, I have not transgressed the limits which philosophy and logic alike pronounce indispensable.

How have these conditions been satisfied ? Berkeley has endeavoured to prove the first in the a priori method ;

the second he takes for granted ; the remaining two he does not seem to have thought of stating, much less of proving. Yet his work has been called "demonstrative." His deficiency, however, on the second head, others have endeavoured to supply, with what success we shall have to consider presently. Berkeley's argument is in substance the following : — It is physically impossible that the eye should be the organ of perceiving distance ; but it is a fact that by sight we do judge of distance. Consequently, this must be by the suggestion of some other idea. Now, it is admitted that distance is perceived by touch, and if not by sight, by touch only. This, Berkeley (with Mill) apparently considers too obvious to require to be stated. Hence follows his conclusion, that the supposed perception of distance is a suggestion through visual signs of a tactual idea.

It is my purpose to show — 1st, that the immediate perception of distance by sight is in no way impossible or anomalous; 2nd, that on the contrary, the assumed association of visual and tactual sensations does not exist, and is even impossible ; and 3rd, that neither touch nor the locomotive faculty or sense is perceptive of distance. I propose to show further, that the eye is perceptive of distance — 1st, by appeal to consciousness; justified 2nd, by the analogy of the senses and by physical considerations ; 3rd, by the analogy of the lower animals ; 4th, by special cases of congenital blindness cured in mature years.

First then, as to the physical difficulty, it is conceded, according to Berkeley,* that distance cannot be in itself perceived, being a line directed endwise to the eye. This has been generally regarded as unanswerable. It is in fact the sole positive argument advanced either by Berkeley or Mill. The latter puts it in these words, in a very clear and convincing manner: — "The distances of objects from us are re- * Essay, sec. 2. Three Dialogues, Dial. 1.

THE PHYSICAL ARGUMENT FALLACIOUS. 1 1

presented on our retina in all cases by single points ; and all points being equal, all such distances must appear equal, or rather we are unable to see them in the character of distance at all." "As the interval between an object and our eye has not any interval answering to it in the retina, we do not see it."*

Now, instead of regarding this argument as decisive, we ought to set it aside at once, as unsound in principle. It is absurd to impose a 'priori conditions on the perceptions of sense or other physical phenomena. Indeed, in all other departments of science it has been for centuries recognised that our business is to ascertain, not what the phenomena ought to be, but what they are ; not to deduce effects from their supposed causes, but to inquire, in the first instance, what are the effects.

If, however, we chose to indulge a priori speculation, we might demonstrate, in imitation of Berkeley, that the only qualities of sound which can be heard are pitch and loudness. To speak of tone, timbre, &c, of distinguishing strings from pipes by the ear is absurd ; since the vibrations of the tympanum can differ only in extent and velocity ; and, therefore, all sounds which are alike in these respects, i.e., in loudness and pitch, must produce identical sensations. Yet the fact is, that we can distinguish the tones produced by strings, pipes, reeds, by metal, wood, the human voice, &c, and even in each class can trace an almost endless variety of character. We can recognise a friend by his voice as well as by his face. A Berkeleian ought, by analogy, to maintain, that in these cases we do not hear directly the tone of our friend's voice or of his violin, but that the visual ideas of him and of his instrument are suggested to us, and falsely referred to the ear. And, no doubt, this would have been affirmed if it had been possible to discover any intermediate ideas to which * Dissertations, vol. ii., pp 95, 116.

the suggestion could be attributed. As to the particular argument now before us, it may be applied to prove that distance cannot be directly perceived by any sense. If both ends of the line affect the same spot of the eye, touch labours under the more fatal defect of not perceiving the farther end at all. Locomotion does not even pretend to be a means of direct perception of an interval. By its means, indeed, we may be cognisant of certain points in the distance, but not as points in the same right line, and not simultaneously ; and as for distances within the reach of our arms, it may be said, as of the eye, that since different distances from the organ will produce the same effect, there can be no perception of distance at all. There is no affection of the organ answering to distance. If we grasp an object we do not perceive at what distance our hand is from our body. The fact is, that the argument implies that the antecedent to the perception of distance must be a sensible interval in the organ, in other words, that the organic affection must resemble the perception. But, as just remarked, we can assume no such a priori conditions ; and such a condition as this would exclude every other sense as well as sight.

These considerations would be enough to set aside the objection if the statement on which it is founded were true; but it is false. It might be questioned even as regards a single eye ; but the organ of vision consists of a pair of eyes, a fact of which Mr. Mill does not betray the least suspicion ; and distance accordingly has an interval corresponding to it on the organ — an interval, too, of a very peculiar kind, which cannot be confounded with that belonging to lateral distance. As, however, the further consideration of the physical question must be resumed hereafter, I shall now pass to the examination of the supposed association.

LAWS OF SUGGESTION. J 3

Chapter II. On The Laws Of Association

It is admitted that we cannot by any effort of attention recover or bring into consciousness the steps of the suggestion supposed. This at once separates it from all those phenomena which are alleged as analogous, as well as from all other proved associations. There is no stronger association than that between the ideas of common things and their names ; so that uneducated people find it hard to understand that it is only accidental. Yet, with very little pains, this association can be resolved, and even changed ; the very conception of a language different from our own implying such a resolution. Again, the association between written characters and the sounds they represent is, to persons who read or write much, one of the strongest. Yet here again there is no difficulty in distinguishing the association from natural connexion, and but little in establishing a different one. If it is true, again, that words, " do their work as signs . with the faintest possible suggestion of most of the sensible ideas which compose their meaning,"* it is also true, that we can, if we will, arrest this process, and fix our attention on any part of the sensitive perception, and can thus separate the part of sense from that of intelligence.

Since, then, it is not pretended that this analysis of the phenomena of visual perception can be "demonstrated" by observation, we are compelled to examine the principles of association in order to discover whether it is in accordance with them.

J 4 SIGHT AND TOUCH.

The fundamental laws of Association have been reduced to two — the law of Redintegration and that of Repetition. By the former, with which alone we are concerned, those thoughts which have at any time formed part of the same mental whole, have a tendency to suggest each other. Hence, first, they must have co-existed — they must have existed simultaneously — if a series of thoughts have intervened, it will effectually bar all direct association ; and, secondly, the tendency to suggest is mutual. If the thought A recalls B, then B also tends to recall A. This law is modified by certain secondary laws. If A have been associated with many other thoughts, and B with few besides A, then B will be more likely to suggest A than vice versa. For example, in reading, the sight of the words suggests their sound and signification, while the converse is not the case with any but constant readers. It is for the same reason that a man who does not readily recollect the names of persons he meets, yet may find no difficulty in remembering the person when the name is mentioned. So also with most persons the words of a song are not so readily suggested by the air as the air by the words.

Secondly, the association will be strong in proportion to the frequency of co-existence of A and B. Unless the attention has been peculiarly excited a single instance will not generally effect their connexion ; but after several the one may even invariably call up the other. It will of course be remarked, that whenever A occurs without B, the association is practically weakened, inasmuch as new thoughts will struggle with B into notice on the next occasion. To state the law generally, the tendency of A to suggest B is greater in proportion to the frequency of their connexion, and less in proportion to the frequency of A's connexion with other thoughts. Strictly speaking, B is suggested equally, whether A have other associations or not, but in the latter case it com-