Realism

Realism

In the early history of philosophy, particularly in medieval thought, the term realism was used, in opposition to nominalism, for the doctrine that universals have a real, objective existence. In modern philosophy, however, it is used for the view that material objects exist externally to us and independently of our sense experience.

Realism is thus opposed to idealism, which holds that no such material objects or external realities exist apart from our knowledge or consciousness of them, the whole universe thus being dependent on the mind or in some sense mental. It also clashes with phenomenalism, which, while avoiding much idealist metaphysics, would deny that material objects exist except as groups or sequences of sensa, actual and possible.

The Polemic Against Idealism

At the close of the nineteenth century, idealism was the dominant Western philosophy, but with the opening of the twentieth century, there was an upsurge of realism in Britain and North America, associated in the former with G. E. Moore, Bertrand Russell, and Samuel Alexander and in the latter with William James (despite his pragmatism), the new realists, and later the critical realists. Before a discussion of realist doctrine, a brief survey may be given of its attack on idealism.

The Polemic Against Idealism

The claim that material objects cannot exist independently of mind had been made on various grounds. First, the analysis of perception, especially of illusions, was held to show that our knowledge was limited to groups of sensations "in the mind" or to products of the synthesis or interpretation of sensory data.

Later idealists, under the slogan "all cognition is judgment," stressed the role of judgment and interpretation in perception, concluding that objects as we know them must be largely or even wholly the work of the mind.

Second, physical objects cannot exist independently of the mind, for whatever is known is relative to the mind that knows it. This is the "egocentric predicament"—that one can never eliminate the "human mind" from knowledge and discover what things are like apart from one's consciousness or, indeed, whether they exist when they are not known, for the discovery itself involves consciousness and thus would be knowing.

physical objects

This may also be stated in terms of the doctrine of internal relations—that the nature of anything is grounded in and constituted by the relations it has with other things; no two related things could be what they are if the relation between them did not exist, and so, as a special case of this, physical objects could not be as they are apart from their relation to the mind that knows them.

STATUS OF THE OBJECTS OF PERCEPTION. Concerning the analysis of perception, realist philosophers have devoted considerable attention to showing that in perception we obtain knowledge of external physical objects either directly or by means of sensa.

Their accounts of perceiving and their solutions to the problems raised by illusions and other facts of perception differ greatly, but they agree in rejecting the view that things cannot exist unperceived. G. E. Moore's influential "Refutation of Idealism" consisted in an attack on this thesis, which, following George Berkeley, he stated as "esse is percipi" ("to be is to be perceived").

Refutation of Idealism

He claimed that in maintaining this the idealists had failed to distinguish between the act and the object in sensation. They had confused the sensation of blue with its object blue or, when claiming to distinguish them, inconsistently treated them as identical.

Sensations are alike in being acts of awareness but differ in what they are awareness of. Once the object is distinguished from the awareness of it, there is no reason to deny its existence unperceived.

Further, in no other situation have we a better claim to be aware of something distinct, so that if sensations are not cases of awareness of objects, no awareness is ever awareness of anything, and we cannot be aware of other persons or even of ourselves and our own sensations. Fundamentally, Moore's thesis concerning sensations rested on introspection; it has been denied on a similar introspective appeal by upholders of the adverbial analysis of sensing, and Moore himself later had grave doubts about it.

concerning sensations

Commonsense realists would say that he conceded too much in talking of sensations and interpreting "being perceived" (percipi) as "being sensed" (sentiri); the proper starting point is our awareness of material objects. But Moore was no doubt accepting the usual conclusions from the argument from illusion.

From his analysis arises the question: "What is the object of sensation?" The answer, "A sense datum," posed the problem, which he never solved, of the relation between sense data and material objects. It was met by others with some form of representative realism or, more usually, phenomenalism.

Phenomenalism, however, particularly if coupled with the adverbial analysis of sensing, means the abandonment of realism. The idealist stress on judgment in perception was at first little discussed, but critical realism and the sense-datum theory later offered more plausible alternatives.

plausible alternatives

THE EGOCENTRIC PREDICAMENT. The realist attack on the egocentric predicament involved considerable discussion, particularly in the United States, and led to some close argument—for example, in attempts to show that the idealist principle led to self-contradiction or circularity when developed.

The egocentric predicament was claimed to have no idealist implications. To infer from it that nothing exists outside consciousness is simply fallacious—that one cannot discover X does not mean that X does not exist or even that it is unreasonable to suppose that X exists.

Indeed, if it were true that things could not exist apart from a person's consciousness of them, neither, presumably, could other persons; the predicament would imply an incredible solipsism.

incredible solipsism

Nor is there any evidence of the lesser conclusion that objects outside consciousness would be quite different. No conclusion about the degree of distortion introduced by our consciousness follows from its ubiquity, and it may be negligible; one can only try to discover the degree by comparing various methods of knowing. (Distortion by the method of observation may be serious in atomic physics, but the same argument that establishes distortion there shows it to be negligible for objects larger than atoms.)

The predicament is sometimes stated in terms of the privacy of experience—a person can never know anything that is not a content of his private experience. This, however, is question-begging in that it simply denies the ordinary assumptions that we are aware of other persons and external public objects.

There may be grounds for denying these assumptions in certain cases, but such grounds rest on evidence of causal processes and of illusions, evidence that is largely obtained from other persons, or with the aid of public objects, or from comparisons with perceptions of public objects.

public objects

Further, though more dubiously, Wittgenstein has argued that if we had only private experiences, not only would they be incommunicable, but also we could not describe or speak about them even to ourselves, for the use of language implies rules that are communal and have to be established and checked with respect to public objects.

Against the doctrine of internal relations it was claimed that relatedness is compatible with independence, that the same thing can enter into a variety of relations without losing its identity.

This seemed so obvious that James confessed to finding it "weird" to have to argue for it. (Anticipating a contemporary approach, he accused the idealists of confusing linguistic or conceptual differences with factual ones; in referring to two relations of an object, our phrases and thoughts differ, but there is no corresponding difference in the object itself.)

confusing linguistic

As the realists were defending what in their eyes was obvious, they were forced into detailed criticism rather than into the kind of positive thesis that can be readily summarized.

This battle was certainly won by the realists in that few English-speaking philosophers in the twentieth century espoused idealism. Indeed, to anyone coming from contemporary discussions, the controversy has an air of unreality.

Partly this is because in a climate of thought that respects common sense and science, realism seems so obvious a starting point that it is difficult to explain how the idealist view ever seemed plausible; partly it is because current idioms, issues, and logical presuppositions are so different from earlier ones.

seemed plausible

Granted, however, that material objects exist independently of our perception, the difficulties facing a realist account of this perception still remain and cause serious divisions among realists.

Direct Realism

Direct realism is the general view that perception is a direct awareness, a straightforward confrontation (or in touch, contact) with the external object. It may be further subdivided according to the various attitudes then taken toward illusions and hallucinations.

In contrast, there are the various types of indirect or dualist realism, which claim that perception is primarily of mental representations of the external object, as in traditional representative realism, or that our perception of the external object is by means of private, mental sensa.

Direct Realism

NAIVE REALISM. Naive realism is the simplest form of direct realism and is usually alleged by philosophers to be an innocent prejudice of the average person that has to be overcome if philosophical progress is to be made. It is normally stated in terms of sensible qualities or sensa.

When we look around us, we can distinguish various colored, shaped expanses that we suppose to be the surfaces of material objects, we may hear various sounds that we suppose to come from such objects, we may feel something smooth and hard that we suppose to be a table top, and so on.

Naive realism claims that these suppositions are all correct—that the shapes, colors, sounds, and smooth, hard expanses (the sensible qualities) are always the intrinsic properties of material objects and in sight and touch are their surfaces.

intrinsic properties

Such a claim can easily be shown to be erroneous by the argument from illusion. When A looks at the table from above, he sees a round expanse; when B looks at it from a distance, he sees an elliptical one. Without selfcontradiction, however, the round and elliptical shapes cannot both be the surface of the table—that is, an intrinsic property.

Similarly, when C, who is color-blind, looks at a red book, he sees a black shape that, again, cannot be the surface of that red book; when D, a drunkard, sees snakelike shapes on the bed, they are not real snakes.

Such examples may be multiplied indefinitely and dispose of naive realism as thus stated, but commonsense realists would say that the doctrine misrepresents the views of the average person and that philosophical discussions of it beg the question in favor of dualism by speaking of sensible qualities or sensa as distinct from physical objects.

physical objects

NEW REALISM AND THE SELECTIVE THEORY. The new realists—E. B. Holt, W. T. Marvin, W. P. Montague, R. B. Perry, W. B. Pitkin, and E. G. Spaulding—are notable chiefly for a common realist platform published in 1910 and expanded in 1912 and for their polemic against idealism. Their realism was carried to the Platonic extreme of claiming real existence for logical and mathematical entities, and they had difficult and conflicting views about consciousness.

Without, however, pursuing these, we may note their main attempt (by Holt) to deal with illusions, which is a version of what is often called the selective theory. The essential points of this theory are, first, all the various appearances of an object are its intrinsic, objective properties and are directly apprehended by the percipient.

For example, the table that looks round to A and elliptical to B is intrinsically both round and elliptical; the mountain that looks green close up and blue in the distance is both green and blue. There is nothing private or mental about such appearances, for they can be photographed, as can mirror images and various optical illusions.

optical illusions

Second, the function of the nervous system and of the causal processes in perception is to select and reveal to the percipient one property from each set of properties, for example either the elliptical or the round shape of the table.

One difficulty in this is that it does not really account for error. If we are always directly aware of actual characteristics of objects, what sense does it make to talk, as we do, of illusions, mistakes, or misperceptions?

Another lies in the weakness of the selective theory compared with the generative theory, adopted by dualist realism, which states that the sensible qualities, or sensa, are "generated," by the action of the object on the sense organs and nervous system and thus are not intrinsic properties of external objects.

nervous system

The usual reasons for preferring the generative theory are, on the one hand, that it is self-contradictory to say the table is intrinsically both round and elliptical or the mountain is intrinsically both green and blue.

Furthermore, objects must be incredibly complex if they are to possess all these shapes and colors, plus, presumably, qualities corresponding to the queer appearance of objects when one has taken mescaline or suffers from giddiness or double vision. On the other hand, it is not clear how the nervous system specifically responds to or selects one of the various shapes, colors, and so on.

This is particularly so in such cases as color blindness, drugs, and double vision, where the different appearances are the result of differences in the percipient and where the pattern of light waves can be detected as already differentiated for the shape and color normally perceived.

various shapes

The generative theory, however, fits the facts of the causal processes quite well; it is natural to suppose that the generation of the sensory experience and its sensum occurs at the end of the causal chain that extends from object to brain by way of sense organ and nerves.

This is confirmed by the reproduction of such experiences in mental imagery (presumably because the appropriate brain activity recurs), by the sensations resulting from electrical stimulation of the brain, and by the time lag that may occur between an event and our perception of it—all things that the selective theory cannot explain. Also, the generative theory can explain how voluntary selection occurs.

When we turn our head to look at X rather than Y, we are allowing light from X rather than Y to strike our eyes and thus bring into being the sensa appropriate to X. As to photographing appearances, the photograph corresponds to the retinal image, not the sensum—that is, it reproduces not the perceived appearance but an intermediate cause of it; to enter into human experience, it must, in turn, be perceived by generating sensa.

generating sensa

PERSPECTIVE REALISM AND THEORIES OF APPEARING. The first objection to the selective theory—that it makes objects possess contradictory qualities—might be met by stressing that shapes, colors, and other qualities are not intrinsic but relative properties. The table is round from here, elliptical from there; the mountains are green in this light, blue in that light, and so on.

This idea has been coupled with direct realism in a number of similar theories: perspective realism (E. B. McGilvary), objective relativism (A. E. Murphy), or the theory of appearing. (This last name was given by H. H. Price to a view put forward by H. A. Prichard. Roderick M. Chisholm, however, uses it more widely, and it is convenient to class all these views as theories of appearing.)

Their central point is that direct realism can deal with illusions, or at least perceptual relativity, by saying that sensible qualities are not possessed by the object simpliciter but are always relative to some point of view or standing conditions.

standing conditions

We always perceive sensible qualities in some perspective—spatial, even temporal (we see the distant star as it is from here and now), or illuminative (the object as it is in this light). (In such theories the shape, color, and so on are possessed by the object at its own location but are perceived subject to perspective, meaning from a viewpoint. In contrast, Bertrand Russell had a phenomenalistic theory of "perspectives" that were spread through space as possible sensa and actualized by or in the percipient.)

Such perspective-realist statements as "The table is round from here" sound forced, for the natural word to use is looks, not is, and it is possible to express this kind of direct realism in terms of looking or appearing.

Physical objects simply are such that they appear different from different positions, and we see them as they appear from a viewpoint or in certain conditions. Thus, we may see the round table looking elliptical from here, but even so it is still the table that we see.

different positions

Thus far the theory is trite and does little more than state the situation in a way that dualists could accept and then claim to analyze. To be distinctive, it must, as its essential characteristic, separate directness and incorrigibility.

Sense-datum theory links the two, assuming that if we see an object directly, we must see it as it actually is. Thus, when the round table looks elliptical, we do not see it directly; what we see directly is an elliptical datum belonging to it.

In contrast, theories of appearing must simply claim that seeing an object directly is compatible with variation or even error in perception, so that we still see it directly when according to viewpoint, lighting, and similar factors, it appears really different from what it is. (Some might object that the theory cannot admit that perceiving is ever erroneous.

similar factors

Perspective realism treats all properties as relative and all perspectives as equal—the table is round from here, elliptical from there, but not round in itself; similarly all appearances should be treated as equally valid. Nevertheless, it seems more plausible to treat some appearances as privileged; in some conditions we see the real shape, the round object appearing as it is—that is, round.

It may be considered a weakness of the perspective theory that it does not take into account the fact that objects do seem to have real [measured] shapes and volumes absolutely, not relative to a viewpoint.)

The approach of theories of appearing may deal plausibly with perspectival and similar variations, but it has two main defects. First, not all variations are of this nature. In double vision or mescaline illusions there seems to be existential appearing—there may appear to be two or even many tables when we look at one table.

double vision

Price has argued that this cannot really be a case of directly seeing one table, for it differs significantly from seeing something merely with different properties, such as seeing a brown table instead of a black one. Also, many illusions are the result of subjective factors, so that it is difficult to say that one has a genuine perspective.

Talk of physiological perspectives is little help. "The bottle from here" is not on a par with "the bottle as it is to someone who has taken mescaline," for mescaline may cause a range of different experiences.

Similarly, when a sentry at night is convinced he sees the enemy approaching but only a shadow is there, is he directly seeing the shadow in some special perspective, such as "the way it is to an anxious sentry" or "looking like a man"?

special perspective

Another anxious sentry might see it as a shadow and say it does not look like a man. And in a full hallucination there is no object at all. Second, theories of appearing cannot deal plausibly with the causal processes in perception since they have to adopt the selective theory.

Further, we do know with varying degrees of completeness why things suffer perspectival distortion or how they cause illusion. The explanations concerned are often in terms of the causal processes and so seem to call for the generative theory and the abandonment of direct realism.

COMMONSENSE REALISM. In the tradition of Thomas Reid, revived by G. E. Moore, many twentieth-century British philosophers defended what they took to be a commonsense view of perception.

Thomas Reid

Moore's defense was primarily of the certainty of such simple perceptual statements as "This is a hand"; he argued that denial of these statements leads to inconsistency in beliefs and behavior and that the grounds for their denial involve propositions less certain than they are. However, his analysis of such statements in terms of sense data led away from direct realism and the commonsense view of the nature (as opposed to the reliability) of perception.

Defense of common sense became particularly associated with the Oxford linguistic analysts. Strong critics of the sense-datum theory (unlike Moore), they also reject the traditional naive realism as unfair to common sense—after all, we do not think that everything we see is the surface of a physical object (certainly not lightning flashes or rainbows) and are quite ready to admit that we often see things looking different from what they are.

Although quarreling with the common philosophical uses of appear, direct, and real, they maintain a direct realism not unlike the theories of appearing and attempt to show in detail that in so-called illusions, including reflection and refraction, we do actually see the physical object concerned.

physical object

Criticism has been made of the view that hallucinations are indistinguishable from normal perception, and more positively it may be claimed that hallucinations are mental images confused with perceptions owing to such special circumstances as drugs or fever.

It is doubtful whether this can explain all the cases, and the role of the psychological processes—for example, in attention or in the influence of expectation and past experience—throws doubt on the directness of perceiving.

Some attempt has also been made to deal with the causal processes, but not very convincingly. Attacks have been made on the dualist interpretation for making it seem that we perceive something in our heads and not external objects and for the view that perceiving involves awareness of sensations.

awareness of sensations

But linguistic analysts have said little of a positive nature; their main attitude is that the causal processes are at most only the conditions of perception and are the concern of the scientist but that the philosopher is concerned with perception itself, which is a skill or instantaneous achievement, not a physical process or the final stage of one.

Unfortunately, scientists generally claim that the study of the causal processes requires representative realism, and even if the average person does not bother about them, an adequate philosophical theory cannot ignore the causes and conditions of perceiving, particularly since the explanation of illusions depends on them.

Indirect or Dualist Realism

Many realists are persuaded by the argument from illusion and by their study of the causal and psychological processes in perception to reject direct realism and to distinguish between external material objects as the causes and ultimate objects of perceiving and private sensa that are the mental effects of brain processes due to the action of those objects on the sense organs.

mental effects

The classic form of this general view was the representative realism (also called the representative or causal theory) of René Descartes and John Locke, which is still maintained in principle by many scientists. From Berkeley on it suffered much criticism, and its defects led to its being unpopular among philosophers. Modern attempts have been made, however, to remedy these defects and to propose an acceptable theory.

The resultant position we shall discuss as critical realism. Although they start from an analysis of perceptual experience and do not argue from the causal processes underlying it, supporters of the sense-datum analysis who are not phenomenalists are forced into one of these kinds of dualist realism.

REPRESENTATIVE REALISM. In what is loosely called "seeing a table," light rays reflected from the table strike the eye, cause chemical changes in the retina, and send a train of impulses along the optic nerve to the brain.

chemical changes

The resultant brain activity is then said to cause the mind of the percipient to be directly aware of private sensa (Locke called them "ideas") that represent the shape, color, and other visual properties of the table.

A similar account is given for the other senses. The essential point is that perceiving proper is the direct awareness of sensa; perceiving external objects is redefined as perceiving sensa caused by them, and so all our awareness is strictly limited to sensa.

"Represent" is usually interpreted in accordance with the doctrine of primary and secondary qualities—that is, the sensa resemble the object in spatiotemporal properties but not insofar as colors, sounds, smells, and other secondary qualities are concerned. Modern analogies of "representing" are the relation between a map or radar screen and the region they cover or between television or movies and the studio events reproduced.

Merits of representative realism. Representative realism has important merits. It is the easiest inference from the scientific account of the causal processes up to the brain in all perceiving and fits other scientific evidence.

Thus, color blindness and deafness are the result of defects in the sense organs that so affect all subsequent stages in the causal transmission that the resultant sensa are different from normal. That electrical stimulation of the brain causes sensations of color, smell, and so on, according to location, seems to confirm the theory, and it can easily accommodate the time lag in perception.

Further, by holding that representation does not amount to resemblance in the case of secondary qualities, it can be made to fit the distinction between the world as we see it (that is, the sensa grouped as ostensible objects) and the scientific account of material objects, which is in terms of colorless, tasteless, and smell-less elementary particles.

Representative realism also accounts for illusions, dreams, images, hallucinations, and the relativity of perception. Relativity and many illusions result from changes in the stimulation of the sense organs because of distance, medium, angle of sight, and other relevant factors; such changes affect all that follows and so vary the sensa caused.

Other illusions are the result of misinterpretation of sensa. In imagery and dreams the brain activity that occurred in corresponding perceptions is reactivated as the result of internal causes and so brings about the recurrence of similar sensa. (The reactivation may be only partial, and the resultant data may be consciously or unconsciously altered by the mind.)

Hallucinations are also imagery. Since the images are of a similar character to normally perceived data and are the result of a similar immediate cause in the brain, it is easy to see how they may merge in integrated or triggered hallucinations or how perception may be imaginatively supplemented.

The standard explanation of phantom limbs—that they are sensations caused by irritation at the stump of nerves normally coming from the amputated limb—is also accommodated. As perception is confined strictly to the effects of the causal chain, interference with it en route may readily deceive us.

Finally, representative realism has also traditionally been part of the widely accepted interactionist or dualist account of the relation of mind and body: The body affects mind in perception, mind affects the body in voluntary action. Not all who accept that theory realize that they are saddled with representative realism.

Defects of representative realism. Despite its merits, representative realism has some serious defects. If, as it claims, our perceiving is strictly awareness of the mental ideas or sensa, it is difficult to see how we can break out of the circle of sensa and observe external objects.

How can we tell what these objects are like; indeed, how do we know that there are such objects? If we try to verify the existence of the table by touching it, we simply obtain more sensa—tactile ones—and if we see our hands touching the table, we are just having visual sensa. Whenever we try to peer over the barrier of sensa, we just get more sensa. This difficulty undermines the analogies used in the theory.

Representation is conceived of as something like mapping or photographing, but we know a map represents or a photograph resembles an object because we can observe both and compare them; ex hypothesi, however, we can never strictly observe both objects and sensa to compare them. Observing objects is just observing sensa, so we do not know that objects and sensa resemble each other in primary but not in secondary qualities.

It is often said that representative realism not only leads to skepticism but is also self-refuting, cutting off the branch on which it sits. Its premises and evidence assume that we discover the action of the objects on the sense organs by observing them. Its conclusion—all our perception is of sensa—denies that we can do this.

However, there would be self-refutation only if the conclusion contradicted the premises, which it need not do if carefully stated. The theory may be regarded as really distinguishing two types of perceiving: perception in its everyday meaning, which is discovering about external objects by means of the senses, and perception proper—direct awareness of sensa.

It is saying that the first type really amounts to or, better, is really effected by the second type. Thus, granted that by perceiving sensa we do discover the nature of objects (at least insofar as their primary qualities are concerned) and their interaction, the first type of perception and the evidence it gives still hold good, and there is no self-refutation.

Nevertheless, the skepticism remains, for since our direct awareness is limited to sensa, we do not know that there are objects or what they are like; we only suppose or guess that and what they are.

Even though representative realism need not be selfrefuting, it is open to the charge of circularity if considered as an attempt to explain perceiving. It appears simply to transfer perceiving as ordinarily conceived (a face-to-face confrontation) from outside to inside the person; perceiving external objects is now put forward as perceiving private replicas of them, for we look at maps and television pictures in the same way that we look at the countryside. Even if we say perceiving objects is achieved by perceiving sensa, there is the same duplication of perceiving, which is thus explained in terms of itself.

Representative realism's view of the mind is rather crude, for it tends to speak almost as if the self or mind were a little person in the head looking at pictures of the outside world. It is not clear how sensa can exist in an unextended mind, since they apparently possess shape and size; nor is any serious attempt made to fit the psychological processes of perception into the general scheme.

There are special difficulties for those versions of the theory that claim that in perceiving objects we infer the existence or nature of external objects from our sensa. Apart from the inevitable dubiety of such inference, the main objection is that we are never conscious of these inferences nor are we aware of sensa as such—that is, as private mental data.

If we were, it is difficult to see how the notion of publicly observable causes would occur to us. But the representative theory may simply say that the sensa seem to be external (or externally caused) from the start and that any inference is justificatory to deal with skeptics.

CRITICAL REALISM. Critical realism is the name primarily given to the views expressed by the American authors of Essays in Critical Realism—namely, that the data in perception (that is, what is intuited, what we are directly aware of) are not actually part of external objects but are "character-complexes ... irresistibly taken, in the moment of perception, to be the characters of existing outer objects" (p. 20). In veridical perception these characters are the characters of external objects; in illusions they are not.

The authors were unfortunately divided over the nature of this datum or character complex, Durant Drake, A. K. Rogers, George Santayana, and C. A. Strong claiming that it was not a mental existent or any kind of existent, but only an essence, a mere logical entity or universal, whereas A. O. Lovejoy, J. B. Pratt, and R. W. Sellars held that it was a mental existent, a content of sensory experience.

It is difficult to grasp what the datum can be if it is not a mental content or existent, and so the second version is the more plausible and is adopted here. Although clearly dualist, it should not be confused with representative realism; in fact, it provides remedies for representative realism's main faults.

The critical realists held that the root of the troubles of representative realism lay in its failure to analyze perceiving or perceptual knowledge. Accepting the ordinary notion of perceiving as intuiting, which means a direct awareness or confrontation, and finding that because of the causal processes and of illusions such awareness was not of external objects, Locke concluded that it must be of intramental ideas and so imprisoned us in the circle of such ideas.

The more reasonable conclusion, however, would be that this ordinary notion of perceiving is wrong and that a more careful analysis is needed. This will show that an essential feature of perceiving, even as ordinarily understood, is that it is the way we discover the existence and nature of external objects—that it is, in fact, a claim, often justified, to knowledge.

If we appreciate this from the start, we shall not be tempted by the apparently intuitive character of perceiving into an analysis that limits it to ideas, and if we remember that this knowledge claim is not always justified—that is, that there are illusions and errors—we shall avoid the other pitfall of direct realism, in which error becomes inexplicable.

The next step is to realize that though it involves an intuition or direct awareness, perceiving is much more than this. It also involves an active external reference, as is implied by the knowledge claim; we refer this intuited mental content or character complex to an external object—that is, we explicitly judge that it is, or is the character of, an external object or we unreflectingly take it to be this or we immediately react to it as if it were an external object.

These modes of reference are differently stressed by different writers, but the point seems to be that they occur in varying degrees according to circumstances. Our perception is sometimes an explicit identification or judgment, or at least it immediately issues in one—for example, we say, "Here's our bus" or "There's Tommy"; more often we just see that it is Tommy without formulating any judgment, or our perception that it is our bus and our starting to go and catch it seem indistinguishable, for the reference to the external object is manifest in an immediate physical response.

All the same, in contrast to the behaviorists, the critical realists stressed that there was an intuited mental content, the character complex of which we were directly aware. Attempts were made to fit the analysis in with current psychology by explaining how this external reference arose in childhood—the apparent externality of the content was with us from the beginning of perceptual discrimination, largely because the external reference was founded in physical response to the object.

There is some similarity between this "reference of an intuited datum to an external object" and the "taking for granted that a sense datum belongs to a material object" of Price's sense-datum theory, especially since both stress that no distinction between datum and object is drawn by the percipient at the time. But there is a difference in starting point and emphasis.

Price began with sense data, treating them as distinct existents and willing to allow that material objects consisted of them. This branch of critical realism began with knowledge of external objects, but, being mental, the content or datum distinguished within it was not regarded as capable of distinct existence and was very difficult—much more so than Price thought—to isolate even subsequently from the associated reference.

Also, reference covered a wider set of activities than taking for granted, for it also involved the bodily reactions. In order to stress the relative subordination of the datum, some critical realists spoke of perceiving external objects by means of, guided by, or mediated by, the datum.

Since critical realism can agree that the datum is generated, it is free from the difficulties of the selective theory and can share in the advantages of representative realism. In this version it seems able to avoid the latter's worst faults.

There is no self-refutation, for from the start perceiving is always perception of external objects by means of the intuited data, an analysis that does not deny that we perceive such objects.

There is no duplication or circularity, for the direct awareness of the datum is not a replica of perceiving; insofar as it can be distinguished at all, it is much less complex than perceiving, for it involves no identification with external objects and is not in itself directed on them—hence, the map and movie analogies are essentially faulty.

Common sense is not being offered an explanation of perceiving in terms of perceiving; it is being shown that perceiving is far more complex than common sense supposes, involving not only causal processes that bring about the datum or mental content but also the psychological processes of reference or response.

Moreover, there need be no skepticism. True, in perceiving we only take the datum to be an external object or its properties, and this may, of course, be erroneous. In a sense it is always erroneous in that the datum or content is never the object, but normally the taking or reference is correct to the extent that we are perceiving an external object and that the intuited characters also do characterize the external object insofar as primary qualities are concerned; to that extent we are perceiving actual properties or at least projections of them.

In general, the claim that perceiving is thus far veridical and amounts to knowledge is said to be the best hypothesis to explain the order and nature of our sense experiences. The realist claim is simply that once ordinary errors and illusions are ruled out by comparing the evidence of different senses or of different persons, the simplest explanation of the situation is that there are external objects causing the sense data or contents and corresponding to them in primary qualities.

And this is plausible because if we dismiss as incredible solipsism the view that only oneself and one's own sense experiences exist, then the only real alternative is phenomenalism, a view that has fatal weaknesses and really amounts to proposing a series of deceptive coincidences.

Critical realism is not fully satisfactory, however, particularly if regarded as a theory of perceptual consciousness—that is, as an account of the mental activity that goes on in perception. Thus, the alleged datum or character complex suggests a group of sense data and invites the objections discussed under the entry Sensa.

A closer examination is required not only of the concepts of datum and reference but also of the general relation of mind and body presupposed in perception and of the nature of mental contents; above all, the theory must take full account of the numerous quasi-interpretative activities that modern psychology has found to be involved in perception.