Perception

views updated

PERCEPTION

The term perception may be used generally for mental apprehension, but in philosophy it is now normally restricted to sense perception—to the discovery, by means of the senses, of the existence and properties of the external world. Philosophers have been concerned with the analysis of perception—that is, the study of its nature and of the processes involved in it—and with its epistemological value—that is, how far, if at all, it can be regarded as a source of knowledge about the world. Their answers to these closely interrelated questions have been formulated in various theories: the commonsense theory and other kinds of direct realism, the representative or causal theory, critical realism, the sense-datum theory, and phenomenalism. This entry will be devoted to the main features of perception that underlie the various theories and that have raised philosophical problems and controversy. It will discuss both the initial evidence that may be analyzed without recourse to scientific findings and the causal and psychological process revealed by scientific investigation.

Initial Evidence and Analysis

reflective examination

As percipients we are all familiar with perception, and so the first evidence should come from reflection on our own experience. The following points may thus be made about perception.

First, it is awareness of the external world—of material objects, to use a technical term for physical objects in general, animals, plants, and human beings insofar as they are perceptible (their bodies, in fact). The main characteristics of such objects are that they are external, independent of the percipient, and public, meaning that many people can perceive them at once. Perception, in being the awareness of such objects, may be contrasted with imagery, bodily sensations, or having dreams.

Second, perception is, or seems to be, intuitive—immediate and normally undoubting, a direct face-to-face confrontation with the object in sight or a direct contact in touch. Nor are we normally conscious of any processes of reasoning or interpretation in it. On the rare occasions when we reason or we have doubts about what an object is, the reasoning or doubts are about the identity or character of something already perceived—for instance, a rectangular red object or something white on the hillside.

Third, perception is variable in quality and accuracy; we may fail to notice something, to see clearly, to hear distinctly, and so forth. Three types of variation may be involved: variations in attention, in what we notice or discriminate; variations in quality or distinctness (for instance, where there is nearsightedness or fog); and variations in liability to err—we may misidentify what we perceive or mistake its qualities.

Fourth, perception nevertheless normally gives us knowledge of material objects and properties. With a few fairly obvious tests, like touching and looking closely, or using the evidence of other percipients, we can establish certainty or else correct the first sight or hearing.

Fifth, perception often issues in some judgment or assertion (to others or perhaps only to oneself)—for example, "There is a green fly on the roses" or "Here's the milkman"—but it may not.

illusions

Illusions, comprising illusions proper, hallucinations, and cases of the relativity of perception, have traditionally been the most important origin of the major problems of perception. The two main claims of the argument from illusion are (1) illusions show that perception is never absolutely certain, that tests are never final, and (2) the appearances we are aware of in illusions, especially hallucinations, cannot be identified with the real properties of objects and must therefore be private objects of awareness, or sensa (indeed, all perception involves awareness of sensa that in correct, or veridical, perception belong to the object or correspond to its properties).

The first claim was long thought to rule out perceiving as a source of knowledge; instead, one had to turn to pure reason, or rational intuition, which was held to provide mathematical knowledge. But, since the absolute certainty of mathematics came to be generally ascribed to its ultimately analytic, or even tautological, character, the tendency now would be to stress the negligibility of the possibility of error in tested perception and to use a different standard of certainty and knowledge concerning matters of fact, one that allows perceptual statements to qualify.

The second claim, concerning the existence of sensa, is vital in that almost all theories of perception either found their analyses on it (as does the sense-datum theory) or seek to controvert it or explain it away (as does commonsense realism). The seeds of this conflict already lie in the results of the reflective examination. Insofar as perceiving seems to vary in quality and accuracy, it is easy to say that in illusions we merely see the object looking different from what it is. But if perception is a direct intuitive confrontation, the illusory appearance must be a genuine existent, perceived as it really is, a sensum in fact; "looking different from what it is" must be interpreted as "presenting sensa different from the standard ones." In any case, some phenomena—for example, the integration of hallucinatory images with a perceived background—are difficult to explain without supposing awareness of private sensa in all perception, and almost all the phenomena require scientific and psychological findings for their full explanation, thus pointing beyond this initial evidence.

perceptual consciousness and perceiving

The occurrence of illusions may lead to ambiguity in the use of perceive and allied terms. Thus, in double vision a man may be conscious of two bottles where there is only one. Do we say "He perceived two bottles" or "He perceived one bottle"? Each alternative has been adopted philosophically, and to avoid ambiguity, it is safer to distinguish (1) "X is perceptually conscious of Y " from (2) "Y is present to X 's senses (or light from it is acting on his sense organs)" and use "X perceives Y " only when both are meant.

This recommendation is claimed to have the further advantage of enabling us to discuss as perceptual consciousness the state of mind (or mental act) occurring in both veridical perceiving and illusions. Perceptual consciousness of, for example, a dagger might occur when only a stick was present or even, as in Macbeth's case, when nothing was there. The notion of such consciousness as a common factor in perceiving a real dagger, in having hallucinations of one, and in mistaking something else for one fits in best with dualist theories, such as the sense-datum theory (especially H. H. Price's version) or critical realism, since it suggests that the contents of such consciousness differ from the external object perceived. Direct realists are suspicious of it; for them having hallucinations is something (imagery, perhaps) quite different from normal perception, even if confused with it, whereas in illusions they want to stress that one is perceiving the real object present—seeing a stick as a dagger or the round table as elliptical.

But even if perceiving a round table as round or in perspective as elliptical is taken as immediate confrontation needing no further analysis, seeing a stick as a dagger (or a piece of wax as a tomato or a bush in a fog as a man) can hardly be equally simple and immediate. In such cases and in hallucinations one has to admit that one seems to see an object quite different from that present to the senses. This can fairly be described as perceptual consciousness of the (ostensible) object (dagger, wax, or man) and distinguished in analysis from actually perceiving an object (dagger, wax, or man). And in view of the subjective similarity it is but a short step to suppose that perceptual consciousness of X also occurs in perceiving X as X, the difference between illusory and veridical perception of an X lying not in this common consciousness but in whether X is present and acting on the sense organs. Any philosophy of perception should analyze this perceptual consciousness and explain how it may occur without the presence of the corresponding object.

analyses of perceptual consciousness

Three major analyses are integral parts of the theories of perception mentioned above. First is the traditional notion that perceiving—that is, perceptual consciousness—is the interpretation of sensations as properties of external objects. Second, the sense-datum theory claims that perceptual consciousness is taking for granted that the sense datum one is sensing belongs to a material object. Third, the analysis of the critical realists, though stated as an analysis of perceiving, amounts to saying that perceptual consciousness is taking an intuited datum or character complex to characterize an external object.

The essential difficulty in these three analyses is that they contradict introspective evidence by splitting up perceptual consciousness into the awareness of some private data, recognizable by analysis as such, and the act of interpreting them as, or taking them to be, objects or object properties. In experience there is no such core of sensing or intuiting data distinguishable from the consciousness of a material object, even if only subsequently; still less is there any passage of the mind from awareness of sensation as such to object perception. And if some critical realists are less liable to this difficulty because they do not treat their data or character complexes as existents readily distinguishable from material objects, they do this only at the expense of obscurity or disagreement as to what the data are. Attempts at a remedy must be postponed until the psychological processes in perception are considered.

A fourth analysis is the idealist claim that all perceiving is judging, which is really an analysis of perceptual consciousness but is easier to follow if stated in terms of perceiving. It is that perceiving (perceptual consciousness) consists in making a judgment, which has an implicit sensory basis, about the real existence of an object or property. Thus, perceiving a tomato on a plate or perceiving that the dog has hurt its leg are the sensorily grounded judgments "There is a tomato on a plate" or "The dog has hurt its leg." The "perceiving that" description of a perception (for example, "He saw that the dog was hurt") certainly seems to suggest judgment, though the form may be misleading and may only be for emphasis of the feature noticed. But the main reasons for this analysis are (a ) that perceiving is true (veridical) or false (erroneous) and only judgments or assertions can be true or false and (b ) that perception is more than just sense experience, for we identify and interpret what is given (that is, it involves inference from implicit data, and the conclusion of such an inference must be a judgment).

One may object that truth characterizes what is asserted, not the asserting—the judgments, in the sense of propositions, to which perceiving may lead, but not the act of perceiving itself. Perceiving may be proper, correct, clear, or accurate, but not true or false. Many other things we do may be done correctly and be liable to errors without being forms of judging, such as playing the piano, playing games, tying knots. False judgment is not the only form of error. Also, the idealist doctrine that all perceiving is judging is open to the general objection to the first three analyses above, particularly because the nature of the implicit data or sensory grounds is very obscure. Attempts to elucidate it—for example, Brand Blanshard's—turn them into sensa. Furthermore, the term judgment suggests something intellectual, explicit, and considered, with consciousness of the evidence for the assertion—conditions inappropriate to much perception. Also, we may correct a faulty judgment on learning the truth, but such knowledge does not enable us to correct illusory perceptions; we still see the mirage, and the railroad tracks still appear to meet in the distance.

Fifth, there is a causal analysis of perceptual consciousness—namely, that it is inferring that one's sensa are caused by an external object. This may be associated with representative realism but is not essential to it; representative realism's main thesis is that the sensa and the consciousness are externally caused by objects that the sensa "represent." One may accept this thesis along with any of the analyses of perceptual consciousness—the causal inference it involves is subsequent to the perceiving, and so is a claim about perception. The difficulties of supposing that perceptual consciousness consists in such an inference from effect to cause are that (a ) we are not conscious of such an inference; (b ) if we started only with private sensa, any inference to external causes would be too difficult and complex to be automatic and unconscious—it would have to be conscious; and (c ) it leads to paradoxes, such as that children, being ignorant of the supposed causation of perception, cannot therefore perceive or be perceptually conscious of anything.

conceptual analysis of perceiving

A rather different approach to perceiving is adopted by those who advocate conceptual analysis—that is, a close study of the ordinary meaning (or use) of expressions. This analysis is naturally associated with commonsense realism, for ordinary language tends to reflect ordinary views on perception or at least what once were such views. Such analysis, however, may well indicate features of perception that are not normally realized and so supplement or even correct reflective examination of an introspective kind.

Much attention along these lines has been directed to the categorization or classification of perceiving. Previous philosophers have referred to perception in various ways: as an act, even an operation, as a process, and as a mental state. None of these is satisfactory. "Act," at least as activity or operation, suggests listening or watching rather than just hearing or seeing; "state" and "process" suggest something long-term, and "process," like "activity," suggests something open to public observation—yet whereas one may observe X looking at Y one cannot observe X seeing Y. (One can perhaps claim that the best description of perception is "mental act," which would put perceiving in a special category with realizing, noticing, deciding and so on, but mental acts as such are suspect to these philosophers.) One suggestion is that perceiving is simply having an experience, but this neglects the active side of recognizing and identifying involved in it. A more popular suggestion is that perceiving is a skill or art, or, rather, since seeing X or hearing Y occur at a definite time, perceiving is the exercise of a skill.

Oddly enough, the evidence for this is not linguistic. We may speak of a skilled observer, one who can direct and coordinate a series of perceptions, but not a skilled perceiver; we do not say that X is an expert at the art of seeing or hearing things. Rather, this suggestion is based on the fact that perceiving can be improved by learning and experience, so that one recognizes things easily, avoids mistakes, or can make allowances for such factors as distance. Although this may occur to one on reflection, however, its full and precise extent has been established only by psychological investigation. As soon as one seeks out this and other psychological evidence about perceiving or even asks how one learns by and exploits experience in perceiving, one is carried far beyond language and conceptual analysis to a scientific study of the subject. Also, to maintain that perceiving is the exercise of a skill brings one back to the suggestion that it is an operation or activity.

More striking perhaps was the earlier claim of Gilbert Ryle that "perceiving" is an achievement verb, like "finding" or "winning," and indicates the scoring of an investigational success. This means that perceiving is not an activity or process, though it may be the successful termination of the activity of looking for something; it is instantaneous, not something that takes time or can be observed. Ryle's aim was to attack representative realism and its associated dualisms of mind and body, sensa and object, by claiming that (a ) perceiving, usually thought to be a private mental activity because it is not an overt one, is not an activity at all and thus provides no evidence of a mental world and that (b ) since it is not a process, perceiving is not the final stage or effect of a process, particularly not of the causal process from object to person. Hence, there is no need to suppose that science proves that perceiving is awareness of private sensa.

These are not very convincing arguments. As to the first, winning or scoring involves some activity such as kicking a ball. Likewise, perceiving involves experiences of colors or sounds and the psychological processes discussed below; these are normally claimed to be mental. The second is a non sequitur—instantaneous success may be the end and result of a causal process. Thus, scoring and finding may be observed and may be the result of a process or series of activities; other conditions may also be required but do not rule out their being effects. More generally, if perceiving is an achievement, what are misperception, illusion, failure to see properly, a casual glance? An analysis of perceiving must take these into account and not apply only to veridical perception. Ryle also failed to show how perceiving is related to the causal processes representative realism emphasizes. Thus, if instantaneous, perceiving can no longer be the relation across time and space that direct realism would need to claim in view of the factually verified time lag, the time taken by the causal transmission from a distant object. Indeed, contrary to Ryle's intention, a dualist interpretation of his claim is possible. Perceptual consciousness is instantaneous; when it is also successful, that is, when its content corresponds to the properties of the object causing it, it is perceiving; when unsuccessful, it is misperception or illusion.

The Causal Processes in Perception

the causal chains

The causal processes involved in perception form causal chains from the external object to the percipient's brain. In sight a complex system of light waves, sometimes emitted by the object but normally a differential reflection of light from the object's surface, travels from the object to the percipient's eyes. This system is diversified in intensity and wavelength according to the shape, brightness, and color of the object surface and, on striking the eyes, is focused so that an image of the object is cast (upside down) on each retina. Each retina has a mosaic of more than 120 million receptors, which are activated by the light cast on them in this image. The light causes chemical changes in the receptors; these changes, in turn, cause electrical impulses to pass along the nerve fibers that lead from the receptors to one of the two visual receiving areas of the brain. The impulses set up activity there and in certain other association areas; this done, the person then sees the object. More than one million such fibers form the optic nerve from each eye, and each fiber consists of a succession of cells that are made to conduct by a chain reaction; the resultant impulses can be picked up and reproduced on a cathode-ray tube.

In hearing, a pattern of sound waves is emitted or reflected from the object and strikes the eardrum; this causes vibrations to be transmitted through a series of bones to the liquid filling in the inner ear, thereby setting up vibrations in the basilar membrane of the cochlea according to the frequency (waves per second, corresponding to pitch) and intensity of the sound. The receptors in the cochlea then transmit electrical impulses along the nerve fibers to another receiving area in the brain. (These impulses are not at the frequency of the incoming sound.)

In smell and taste there is a chemical stimulation of receptors in the nose and tongue by particles of the substance perceived, and the receptors, in turn, send neural impulses to another area of the brain. For touch, the brain is linked to receptors all over the skin, some of which respond to the pressure of direct contact with the object, some to heat, and some to cold (or, rather, to rate of change in skin temperature). Other receptors in the skin and the body respond to a wide range of stimuli by transmitting to the brain impulses that ultimately cause a sensation of pain. There are also other senses—for example, a kinesthetic sense by which receptors in the muscles send impulses to the brain so that the position of the limbs is sensed or unconscious adjustments are made to guide and make efficient voluntary movement. There are also receptors in the vestibule and semicircular canals behind the ear that assist balance and give us information about head position.

The chain process (object-[waves]-receptor-nerve impulses-brain activity) is a necessary condition of perception of an external object, for if it is interrupted by damage to the sense organ, no perception occurs. It is not a fully sufficient condition in that other areas of the brain must be suitably active so that the person is conscious and minimally attentive—that is, not wholly absorbed in thought. The interesting question is whether or how far the chain process is necessary and sufficient for perceptual consciousness of an object, granted conditions of consciousness and attention. At least the brain activity is clearly necessary, but theoretically one might insert stimulation at some point on the chain and thus cause experiences the same as those that would normally be attributed to the external object. This apparently happens naturally in illusions and hallucinations, including phantom limbs, and electrical stimulation of the appropriate areas of the brain may cause sensations of color, smell, or touch. (The sensations are not like the contents of perceptual consciousness of objects, but this difference may be due to the comparative crudity of the artificial stimulation by an electrode; also, activity in the association areas is necessary for normal perception.) Thus, it seems probable that suitable activity in the nervous system is a necessary and sufficient condition of perceptual consciousness, though it may be that some kind of external stimulation, even one quite unlike the object perceived, is required to trigger it.

time lag

Causal processes take time. In the case of distant objects this is marked. Thus, because sound waves travel much more slowly than light waves, the flash of some distant gunfire or explosion may be seen appreciably before the sound is heard. Even at its great speed light takes eight minutes to reach us from the sun and four years and four months from the nearest star. Consequently, we may well be "seeing" a star long after it has disintegrated, for the perceptual consciousness occurs at approximately the time of the arrival of the star's light on Earth. But as time is required for the sense organ to be activated and for the nerve impulses to travel to and spread in the brain, there is a slight but variable time lag in all perception; an accurate estimate is not possible but the delay is probably of the order of one-tenth of a second for nearby objects.

uniformity of nerve impulses

One surprising fact is that the nerve impulses are of a similar type for all the senses. All that travels along any nerve from any receptor to the brain is a sequence of such impulses varying normally between 10 and 100 per second. The frequency variation is, in fact, a mark of intensity; the stronger the stimulus, the more impulses per second. Consequently, what distinguishes causation of an experience of sound from that of smell or an experience of a high pitched sound from that of a low one is not the impulse itself but the connections of the nerve fibers excited and conducting—where they start in the sense organ and where they end in the brain. (Though if one imagines a cross section across a bundle of nerve fibers, the pattern of some conducting and some not conducting can be regarded as a changing code.) Thus, excitation by nerve impulses of one tiny portion of the brain results in awareness of a loud shrill sound, excitation of another in awareness of a blue line. Various areas of the body are mapped in the brain, a group of receptors (or sometimes an individual receptor) in the skin and tissues corresponding to each point in the cerebral receiving area. Similarly, the retinal image is reproduced point by point in the brain, though with each half reproduced in a separate area, duplicated there, and distorted. Again, a strip of brain tissue is activated at different points according to the frequency of the sounds heard, as if it were a keyboard.

complexities

There are nevertheless many complexities in the system, only a few of which we can mention here. The nerve connections are intricate, with feedback fibers from the brain to the incoming sensory fibers and cross connections between the sensory fibers; in the grouping of receptors and in the brain there is summation—several nerves join one that conducts only when all or most of them do. (In fact, neurologists constantly use such terms as selecting, integrating, summating, and coding.)

Binocular vision involves retinal disparity (a slight difference in the images cast on the two retinas) and the operation of two visual receiving areas reached by crossed-over nerve fibers so connected that the left-hand receiving area receives the signals coming from the right half of each retina and the right-hand area receives those of the left half. As a result we somehow normally see one object with depth and solidity rather than two two-dimensional ones.

Constant small eye movements are necessary for vision, with a shifting of the retinal image and of the resultant pattern of impulses in the fibers of the optic nerve, yet the object is seen as steady.

Most of the impulses reaching the brain from the eye come from a small portion of the retina (the fovea) that has relatively many receptors giving great distinctness; for exact vision the image is focused on the fovea by eye movement.

Color vision is particularly complex, and its mechanism is disputed. All the colors we know can be produced by suitable mixtures of red (long-wave), green (medium-wave), and blue (short-wave) light. White light can be formed by an appropriate combination of three colors or even of two widely separated ones. (Light, or "spectral" colors, mix differently from paint colors.) The simplest theory is that there are receptors in the eye reacting to each primary light color (red, green, and blue) and the brain, by summating the three color inputs, is enabled to cause the final color sensation. Thus, grass looks green because it absorbs red and blue light but reflects green, and a buttercup is yellow because it absorbs blue but reflects green and red, which combine to produce yellow. There are many difficulties in this theory. For instance, no receptors for blue can be positively located in the eye, only for red and green ones; the light from a green surface actually contains a mixture of wave lengths, with green predominating; the light wave lengths cover the spectrum of all the colors of the rainbow; the brightness and purity of the color also affect its hue. A final theory must therefore be very complicated.

The auditory receiving area gets impulses from both ears. This enables us to locate the source of a sound. If a sound is to the left, then sound waves reaching the left ear differ in phase (that is, timing of the wave crests) and in intensity from those reaching the right ear. The brain apparently combines the different inputs so that the location is done unconsciously and we just hear the sound as if it came from a certain direction.

limitations of the senses

Radiant energy is known to range from short cosmic rays to long radio waves, but the eye responds only to visual light, which is a narrow band occupying about one-seventieth of the whole range. Even then we cannot distinguish light of different polarizations, as bees and birds apparently can, or see very small objects or fine structures. Similarly, in hearing we can distinguish only waves between 20 and 20,000 cycles per second; dogs, cats and rats can hear higher notes. Our sense of smell is obviously very inefficient compared with that of most other animals. Hence, though we can extend our range of observation by microscopes, infrared or X-ray photographs, radiotelescopes, and so on, it is clear that our senses themselves are very limited as a direct source of knowledge of the external world.

the causal argument

The causal argument maintains that the existence and character of these causal processes refute direct realism and force the adoption of a dualist position. Perception of an external object cannot be the direct contact or immediate confrontation it seems to be, since it requires this causal chain from object to the percipient's brain and is prevented if that is interrupted—for instance, if the optic nerve is cut or one of the small bones in the ear does not move properly. In this sense directness or immediacy must mean no intermediary and no possibility of interruption. The causal chain suggests that perceptual consciousness and its objects are generated, or brought into being, by the causal process, presumably by its last stage, the brain activity. In other words, insofar as perceptual consciousness is intuitive, it is awareness of some content or object quite distinct from the external object.

This suggestion is supported by various points. First, the time lag—perceptual consciousness may occur after the external object has disappeared or moved, so its content cannot be identified with the object. Second, the possibility of perceptual consciousness without any external object at all or without one at all similar seems confirmed by the production of sensations by stimulation of the cerebral cortex and seems actualized in hallucinations. Third, the enormous complexity involved shows that the subjective simplicity of perception is illusory, at least insofar as a relation to an external object is concerned. Fourth, illusions and the relativity of perception are often explicable in terms of the causal processes. Unless the contents of perceptual consciousness are generated and conditioned by the causal process, one has to attribute bizarre and contradictory properties to the external object. Fifth, the simplicity and uniformity of nervous impulses show that they cannot transmit all the various secondary qualities that make up objects as we know them. These qualities must thus characterize contents of consciousness generated by the causal process. (This point is supported by such other limitations as the purely mechanical transmission through the bones of the ear.) Hence, it follows from the fourth and fifth points that one must abandon the other assumption of direct realism—that even when we are not perceiving them, objects continue to exist with the exact qualities we normally observe in them.

There is a good deal of resistance to these conclusions. One obstacle is that they seem to require a self-refuting type of representative realism. This fear is unjustified. It must also be noted that granted the dualist conclusion that the causal process generates the sensory experience whose content is (numerically) different from the external object, the nature of that experience and its content is still open. It may be that the awareness is of sensa, or it may be a full-fledged perceptual consciousness of percepts or ostensible objects. One is not even forced to adopt a mind-body dualism, though it is normally thought that sensa or percepts are mental. One might claim that though apparently distinct objects, they are in fact only the contents of sense experience, not existing apart from the sensing of them (adverbial analysis), and that they and the brain activity are two aspects of the reaction of the organism or person as a whole. This would mean that sensa are only a correlated aspect of brain activity, not effects of it, though still conditioned by the rest of the chain. In this way one might bypass one of the notorious difficulties of ordinary dualism—the unique and obscure causal relation supposed to exist between material brain and immaterial mind.

Sometimes, however, the opposition takes the form of denying the relevance of the scientific evidence to philosophy; it tells us only what the causes of perceiving are, not what perceiving itself is. Philosophers must investigate the latter and leave the causal processes to the scientist. But scientists normally hold that these processes require the adoption of representative realism, thus giving them philosophical relevance; also, those philosophers who wish to concentrate on the nature of perception alone usually come up with some answer (the sense-datum analysis or a view that perceiving is the exercise of a skill or an investigational success) that is compatible with or even supports a dualist interpretation of the causal processes. But, above all, to achieve full understanding of anything so vital as perception, one must consider its causes and conditions, particularly as their study has traditionally been claimed to transform our concepts of perceiving itself and of our knowledge of the external world.

The Psychological Processes

It is clear from experimental psychology that perceptual consciousness involves a whole range of adjustments and selective or quasi-interpretative processes. The main evidence for this lies in differences between what psychologists often call the phenomenal properties of an object (those we are perceptually conscious of) and its stimulus properties. In this context the stimulus is the pattern of light rays from the object striking the eye, of sound waves striking the ear, or of heat or pressure from touching the object. The stimulus properties are those that we should observe in the stimulus (such as shape, color, pitch) could we observe it directly and in itself. This is difficult to achieve, and in fact the evidence of cameras, tape recorders, and other instruments is used, plus knowledge of the nature of the object and reasoning from the laws of perspective or of physics generally. The difference between the two kinds of property is presumed to be the result of modifications by the percipient.

attention and selection

It is a simple fact of experience that the quality and accuracy of perception vary with our attention. We often look inattentively and fail to notice pronounced features of a scene, yet we may carefully observe and thus notice unexpected details—a mark on the wallpaper, a printer's error, a wrong note in a recorded symphony. From the evidence of other people, from photographs, and from other means there may be no doubt that these features appeared all the time in the stimulus properties even when we were unconscious of them. Besides confirming this, psychologists have shown how greatly what we do or do not notice depends on habits of attention or interests, on often unconscious "priming" or "set." A mother will hear her baby cry but not notice much louder noises; an architect may notice features of buildings, and a boy notices makes of cars, both being oblivious to much else. Thus, perceptual consciousness is very selective, and this selection is usually largely unconscious, though voluntary attention can greatly modify it. One special case of voluntary attention is of importance—"perceptual reduction" or "phenomenological observation," where we concentrate on the sensible qualities of what we perceive and not, as is usual, on the identification of the object concerned. An artist must do this when he has to paint a scene, and this kind of observation may reveal all sorts of previously unnoticed details of color, shape, and so on. It is open to question whether this kind of reduction reveals an element present in all perception—namely, sensing—or whether, and this is more plausible, it is simply a special kind of perception of external objects not found in normal perceiving.

errors and enrichment

Some errors in perception can be attributed to psychological factors—misidentifications because of careless observation, seeing what one expects to see rather than what is actually present, thinking that one hears the expected visitor coming when no one is there, and the like. These point to a common characteristic of perception and one apparent only when it goes wrong—the enrichment of perception by imagery and thought. Many psychological experiments have been concerned with this. For example, vague or ambiguous stimuli (pictures or sounds) are presented to different groups of people who see or hear them as definite objects or words, and the direction in which they are thus unconsciously supplemented or altered can be shown to be caused by suggestion or by the interests, emotional state, or physical state of the person. Another kind of case is the divergence between several eyewitnesses' accounts of an incident, which may all differ from a filmed record. Again, blind spots or other visual defects are often not apparent to the subject, who unconsciously fills in the gap (this happens to us all if we look with one eye, for there is a blind spot where the optic nerve leaves the retina). Extreme cases are hallucinations where the apparition is integrated with the background or casts shadows. Unnoticed supplementation by imagery, which is admittedly private and mental, seems strong evidence for the dualist claim that the contents of perceptual consciousness are similarly private and must be distinguished from object properties.

learning and cues

Our perception is clearly affected by learning and experience. Identification and discrimination afford obvious examples; one can learn to identify objects seen or photographed from unusual angles, to detect animals in natural camouflage, to distinguish different birds' songs. Driving a car involves perceptions of distance and relative speeds, perceptions that are acquired by experience. Psychological investigation has shown the role of learning to be far greater than this. Perception of spatial relations generally depends to a large extent on learning (normally unconscious and in childhood) to harmonize sight and touch and to use various cues. This is shown by various experiments, such as those with distorted rooms or inverted spectacles, and by the evidence of blind men who recover their sight. Among the various cues used for perception of distance and of solidity are shadows, aerial and linear perspective, parallax (or relative movement), and the interposition of objects. These assist binocular vision and enable us to see depth even with one eye.

figure-ground and gestalt

In perception our immediate consciousness is of an organized or structured whole. Some shape or feature stands out and is seen as the figure against a background, and if discrete units such as dots are presented, we see them as grouped or patterned in some way. This characteristic of experience has been particularly stressed by Gestalt psychologists, who produced much experimental evidence to show that we see wholes or structures (Gestalten, literally, "forms") and that perception develops by discriminating these in and from a background and not by synthesis of atomic elements or point sensations first perceived separately. Such organization of the visual field, though little affected by learning, is nevertheless largely the result of processes in the percipient himself. The clearest evidence of this comes from the reversals, or "alternating illusions," where the stimulus (picture or succession of sounds) is constant but is perceived differently at different times; thus, sometimes one pattern or shape stands out as the figure, sometimes another. Examples are the goblet that may appear as two faces in profile, Edwin Boring's wife–mother-in-law figure, Ludwig Wittgenstein's duck-rabbit, and the staircase that seems to be seen now from above, now from below.

perception of motion

Perception of motion was closely investigated by the Gestalt psychologists, who drew attention to the Phi phenomenon, which is the impression of movement between adjacent stationary stimuli that are activated in succession. This underlies the consciousness of movement on a motion picture or television screen and is used in illuminated advertisements in which if groups of lights are successively switched on for a brief time, one is perceptually conscious of a moving figure or even of words moving along. Intermittent illumination may also make moving objects appear stationary. Thus, when illuminated by the flashing light of a stroboscope, a moving crank in a machine may, if the flashing is properly adjusted, be seen as stationary and examined for defects; if there is a slight maladjustment of the flashing, it may seem to rotate slowly backward like the wheels of coaches in Western films. There is a clear distinction in these cases between the properties of the stimulus and the contents of consciousness. Figure-ground effects also occur in movement perception, such as when the moon seems to sail through the clouds or when one's stationary train seems to move if an adjacent one starts.

object constancy

The widespread phenomenon of object constancy in perception differs from the above in that the phenomenal properties of an object tend to remain constant or nearly so even though the stimulus properties vary considerably. Thus, when we look at a round object—for example, a dish—from an angle, it often still looks round and not elliptical, although by the laws of perspective the stimulus (light-ray pattern) or retinal image is elliptical, as would appear on a photograph taken from the percipient's viewpoint. (This causes complication in stating the argument from illusion and perspective realism.) Only if the angle is very marked does the dish look elliptical. ("Look" here refers to the sensible quality, not to what we judge to be the object's shape.) Similarly with size, brightness, and color—a man looks much the same size at ten yards' distance as at five even though the image cast on the retina is half as high in the former case; a white patch in the shade reflects less light than a dark one in bright sunlight, but it still looks white; a white patch in a yellowish light still looks white although it is reflecting yellowish light (one may be surprised by color photographs taken in the evening, for the camera cannot adapt itself to the yellower light).

In general, over a range of varied stimuli we tend to see something corresponding to the property of the object or at least some compromise between this and the stimulus property. Experiments show that this constancy depends not on knowledge of the object but on the visibility of its background, for if the background is cut off by a screen so that only the object is visible, constancy does not hold and the stimulus property is seen. It is as though we made unconscious allowance for distance, angle of sight, and illumination as revealed by the whole scene. But this is not a learned or intelligent adjustment; children and even chickens or fish apparently see things with constancy, though to some extent it can be counteracted by adopting a stimulus attitude (trying to see the stimulus property).

philosophical significance

The existence of these many complex processes that underlie perceptual consciousness and affect its content reinforces the causal argument by making even more incredible the direct-realist notion of perceiving as a straightforward direct confrontation with the actual properties of objects. If perception were a simple intuitive awareness of such properties, there would be no place in it for variations in quality; for the effects of interests, priming, and learning; and for the use of cues for enrichment by or integration with supplementary imagery, for constancy adjustments (especially where they produce a compromise between object and stimulus properties), for changing figure-ground effects, or for the Phi phenomenon.

The range of these processes is far greater than that which would be compatible with the usual analyses of perceptual consciousness—namely, that it is the interpretation of sensations (or inference from implicit grounds) or the taking for granted that a sensed datum belongs to an object. These views were mainly influenced by the possibility of error in perception, particularly in identification, although they took some account of the use of cues and of the role of learning. But they seem inadequate to cover the part played by attention and unconscious selection or by such organization adjustments as figure-ground, grouping, object constancy, or the Phi phenomenon, whereas some of the imaginative supplementation goes far beyond what can be called interpreting a datum. It is sometimes claimed that these adjust-ments are interpretations. But this is implausible, for they seem little affected by learning and are not intelligent since lower animals make them. Nor can many of the illusions or adjustments be overcome by knowledge of the facts or by conscious interpretation; where some counteraction is possible, as in object constancy, it is very difficult, and for most people the presence or absence of screens in experiments is compelling in its effect.

The final objection to such analyses concerns the alleged pure sensory data; interpreting or taking for granted, insofar as we are aware of it, is of something we are conscious of as distinct and external and which is thus already the effect of many of these processes. Normally, however, perceptual consciousness seems intuitive—that is, without interpretation and quite unanalyzable; except in perceptual reduction its content almost always consists of ostensible objects. All the same, psychological evidence shows that there is a range of subjective processes. The only answer seems to be a genetic hypothesis, not an analysis into elements. Perceptual consciousness is introspectively a whole but must be supposed to be a product of a range of selective, supplementary, integrative or organizational, and quasi-interpretative processes acting on a supposed basic sentience. But—and this is the point—both processes and sentience are unconscious and so may plausibly be regarded as cerebral activities or adjustments of the nervous system. However, since we cannot as yet give any precise neurological statement of these processes, we have to describe them as if they were conscious, basing the description on the difference between the input to the senses and the finished product, but this product (perceptual consciousness) does not reveal within itself the processes that may be supposed to form it.

The suggestion that perceptual consciousness is the product of many unconscious processes is controversial, and any general conclusions about perception are bound to be personal. Hence, the main attention in this entry has been on the facts that have to be taken into account in any fully adequate view of perception, and the reader is also referred to the statement of the various theories here and in other related entries. In this way one has the material for assessing the general view here adopted—namely, that the causal and psychological processes essential to perception, as well as its liability to illusion, require abandonment of direct realism for a dualist position. One must distinguish perceptual consciousness, whose content or objects are subjective and private to the percipient, from perception that occurs when this perceptual consciousness is caused by an external object with properties corresponding to its content. But one must not confuse this dualism with the traditional representative realism, which is only a variant of it, some form of critical realism being superior; the sense-datum theory's dualism of sense data and objects (perceptual consciousness is not thus analyzable, and its content consists of ostensible material objects); or the Cartesian mind-body dualism (it is possible also to adapt this view of perception to a double-aspect account of mind and body).

Bibliography

general introductions

Perception is discussed, though usually less extensively than in the related entries in this encyclopedia, by most introductory books on philosophy. For example, see Bertrand Russell, Problems of Philosophy (London: Williams and Norgate, 1912) and Outline of Philosophy (London: Allen and Unwin, 1927); Charles Harold Whiteley, Introduction to Metaphysics (London, 1955); John Hospers, Introduction to Philosophical Analysis (New York: Prentice-Hall, 1953; London: Routledge and Paul, 1956); and Arthur Pap, Elements of Analytic Philosophy (New York: Macmillan, 1949). A clear and useful outline with emphasis on the associated epistemological problems is A. J. Ayer's The Problem of Knowledge (London: Macmillan, 1956). A more detailed introductory treatment is given by R. J. Hirst, The Problems of Perception (London: Allen and Unwin, 1959), Chs. I–VI; the later chapters of this book develop a more advanced treatment of perceptual consciousness and the scientific evidence on the lines adumbrated here.

perceptual consciousness

For analyses of perception in the sense of perceptual consciousness, see H. H. Price, Perception (London: Methuen, 1954) on the sense-datum theory; Durant Drake and others, Essays in Critical Realism (London: Macmillan, 1920), and Roy Wood Sellars, The Philosophy of Physical Realism (New York: Macmillan, 1932), on critical realism; Brand Blanshard, The Nature of Thought, Vol. I (London: Allen and Unwin, 1939), on idealism; and Roderick M. Chisholm, Perceiving: A Philosophical Study (Ithaca, NY: Cornell University Press, 1957), which is a causal analysis with a general consideration of perceiving and epistemological questions. Roderick Firth, "Sense-Data and the Percept Theory," in Mind 58 (232) (1949): 434–465, and 59 (233) (1950): 35–36, is critical of any such analyses.

conceptual analysis

For the approach by conceptual analysis see "On Seeing and Hearing" by Winston H. F. Barnes, "Identification and Existence" by Stuart Hampshire, and "Sensation" by Gilbert Ryle in Contemporary British Philosophy, Vol. III, edited by H. D. Lewis (London: Allen and Unwin, 1956); Gilbert Ryle, The Concept of Mind (London: Hutchinson, 1949), Ch. 7, and Dilemmas (Cambridge, U.K.: Cambridge University Press, 1954); Anthony M. Quinton, "The Problem of Perception," in Mind 64 (253) (1955): 28–51; and D. W. Hamlyn's discussion of psychological theories, The Psychology of Perception (London: Routledge and Paul, 1957). John Langshaw Austin's Sense and Sensibilia (Oxford: Clarendon Press, 1962) is an outstanding defense of common sense against the argument from illusion, but unfortunately he does not discuss the scientific evidence. Another direct-realist study is D. M. Armstrong's Perception and the Physical World (New York: Humanities Press, 1961).

scientific background

Useful elementary accounts of the causal and psychological processes are given in such textbooks as Ernest Ropiequet Hilgard, Introduction to Psychology (New York: Harcourt Brace, 1953); David Krech and Richard Stanley Crutchfield, Elements of Psychology, 3rd ed. (New York: Knopf, 1962); and Robert Sessions Woodworth and D. G. Marquis, Psychology (New York: Holt, 1947; London: Methuen, 1949). Rather more detail, with a philosophical discussion as well, is given in George McCreath Wyburn, Ralph William Pickford, and R. J. Hirst, Human Senses and Perception (Edinburgh: Oliver and Boyd, 1964). Two significant but unconventional psychological books that raise philosophical questions of interest to the advanced student are James Jerome Gibson, The Perception of the Visual World (Boston: Houghton Mifflin, 1950), and Friedrich August von Hayek, The Sensory Order (Chicago: University of Chicago Press, 1952; London: Routledge, 1953).

interpretations of scientific evidence

See the works by Price, Chisholm, Hirst and the critical realists mentioned above and Arthur O. Lovejoy, The Revolt against Dualism (La Salle, IL: Open Court, 1930), also a critical realist. C. D. Broad's Perception, Physics and Reality (Cambridge, U.K.: Cambridge University Press, 1914) and Scientific Thought (London: Kegan Paul, 1923), sense-datum analyses, are very technical at times. William Pepperell Montague, The Ways of Knowing (New York: Macmillan, 1925), gives a general comparison of the different realist (and other) views. See also the more sophisticated types of representative realism—namely, John Raymond Smythies, Analysis of Perception (London: Routledge and Paul, 1956), or Bertrand Russell, Analysis of Matter (London: Kegan Paul, 1927) and Human Knowledge (London: Allen and Unwin, 1948). Of Russell's two books, the first is often difficult, and the second is more for the general reader.