Perception, action,and nonconceptual content

by Alva Noë e-mail

Department of Philosophy
University of California, Santa Cruz
Santa Cruz CA 95064

1. Perspectival self-consciousness and the mastery of sensorimotor contingencies.

To be a perceiver, one must be capable of keeping track of the ways in which one’s perceptual experience depends on what one does, and also, more generally, on one’s relation to the world around one. For this reason, perception depends on what Hurley [1998] calls perspectival self-consciousness.

Examples are ready to hand. An object looms larger in the visual field as we approach it, and its profile deforms as we move about it. As perceivers we are masters of the patterns of sensorimotor contingency that shape our perceptual interaction with the world. We expect changes in such things as apparent size, shape and color to occur as we actively explore the environment. In encountering perspective-dependent changes of this sort, we learn how things are quite apart form our particular perspective. Our possession of these skills is constitutive of our ability to see (and generally to perceive). This is confirmed by the fact that we can disrupt a person’s ability to see by causing changes in the patterns of sensorimotor contingency, even as we leave the rest of the perceptual apparatus intact. This is what occurs, for example, when one puts on inverting lenses of the sort used by Stratton [1897] and Kohler [1951] in their well-known experiments. Eventually one masters the new patterns — one accommodates — and vision is restored. Or consider the effects of cataract surgery to restore sight in the congenitally blind. Patients acquire relatively functioning eyes, but they are not yet able fully to see. In case studies patients are described whose eyes move around aimlessly in their sockets, unintegrated with the exercise of attention, or with the guidance of thought and movement ( e.g. by Gregory and Wallace [1963/1969]). Sight is only fully regained when the perceptual apparatus, the eye/head/body system, is successfully integrated into a sensorimotor framework. It is also noteworthy that one post-operative patient expressed astonishment at the way, for example, a round coin appears to change its shape when rotated [cited in Helmholtz 1909/1925].

These examples illustrate the ways in which the ability to see (or to perceive) depends on the ability to keep track of the interdependence of perception and action. Is this ability a nonconceptual one, as Hurley further argues [1998, ch. 4]? I would like, in what follows, to raise some doubts about this proposal.

2. Is the content of perceptual experience conceptual?

Nonhuman animals and human infants enjoy perceptual experience. As Hurley notes, however, they lack the sort of "richly normative conceptual and inferential capacities" possessed by adult humans [Hurley 1998, p. 136]. It would seem to follow, then, that perceptual experience is nonconceptual. But this does not follow. For it is far from obvious that animals and infants lack conceptual and inferential skills altogether. Indeed, as Hurley herself emphasizes [e.g. 1998, pp. 141-142], it is a condition of our treating an animal as a conscious subject of experience at all, as opposed, merely, to a locus of neural or psychological processes, that we view the animal as an integrated whole, situated in an environment, suitably responsive to features of that environment, with reasons it is capable of being moved by even if it is not capable of fully understanding. In short, it is only when viewed as simple agents, in possession of what are, in effect, rudimentary conceptual and inferential skills, that we can even make sense to ourselves of the idea that animals are full-blooded perceivers.

Hurley indicates that a hallmark of adult human conceptual skills is the ability to deploy concepts in a manner that is context-free and general. As she writes, "Someone with conceptual abilities who can judge that a banana is green and that a sofa is soft can also in principle judge that a banana is soft, that a sofa is green, that it is not the case all bananas are green, that if a banana is green then it is not soft, and so on" [1998, p. 138]. But it would seem that animal would-be conceptual capacities satisfy this "generality constraint," at least to some extent. For example, we ought, surely, to refuse to characterize a chimpanzee as recognizing another chimpanzee as of high status (something ethologists do regularly), if the chimpanzee’s capacity thus to recognize status were not to generalize to other conspecifics encountered on other occasions. And similarly, we would take a lion’s general unresponsiveness to the presence of gazelle, or its stalking behavior in relation to tree stumps, as reasons (although perhaps inconclusive) to refrain from describing the lion as stalking a particular gazelle on a particular occasion. In short, to the extent that we view an animal as subject to the constraints of normativity and holism — as flexibly responsive to its environment in ways constrained by intentions and primitive practical rationality — then to that extent we must admit that it possesses, to at least some degree, conceptual and inferential capacities that differ from our own only in degree.

One source of the desire to withhold "richly normative conceptual and inferential skills" from nonhuman animals and infants is our adherence to a much too exalted conception of our own conceptual skills. We think of concept-possession on the model of the possession of concepts such as that of square, to possess which a thinker must know the criteria that govern (and justify) its application. But not all concepts are like this. As Wittgenstein’s [1953] considerations on rule-following demonstrate, at the base of our conceptual practices are conceptual skills that do not fit this Socratic or Fregean model.< 1 > When I judge that something is red, for example, I do not do so on the basis of criteria. I can give no reason for my judgment that it is red other than the fact that, for example, I can see it. I judge an argument to be valid because I recognize it to be an instance of modus ponens. I do not then owe an explanation of what it is that makes modus ponens valid. That is, my grasp on validity does not in general require this. (This is true even if it is the case, as it is, that there are standpoints from which one can reasonably ask, What makes a thing red (e.g. what physical properties)?, What makes modus ponens valid (e.g. what metamathematical properties)? For crucially, one is not required to take up this sort of physical stance when deciding on the color of a thing, or a metamathematical stance in evaluating an argument’s validity.)

The significance of this point is that our possession of such basic conceptual skills is strikingly situation-dependent and context-bound. We can tell by looking that a thing is red, or an argument valid, even if we cannot articulate the reasons why. But this situation-dependence and context-boundedness are an important respect in which animal and human conceptual skills seem to be on a par.< 2 >

A second source of our unwillingness to admit animals and infants possess primitive conceptual and inferential skills is that we hold to an unreasonable caricature of what it is to make use of a concept in thought and experience. We think of concepts as brought into play only in the context of what we might call explicit deliberative judgment. But conceptual skills can also enter thought as background conditions on the possession of further skills of one sort or another. For example, the chimpanzee’s possession of the concept of a kin-group member is exhibited, let us say, in its differential treatment of its relatives even if the chimpanzee never engages in anything like explicit deliberative judgment that so and so is kin. This interpretation of the chimpanzee’s cognitive accomplishment is compatible with its being the case that the chimpanzee may lack knowledge usually taken to be necessary for possession of kinship concepts (e.g. knowledge of the biological basis of kinship, etc).

This way of thinking about concept possession suggests that concepts can enter into an experience not so much because they are judged to apply, but because their possession is a condition on the having of that experience. We would not credit a person with the visual experience as of an ant-eater, if we did not believe the person has the concept ant-eater. This would be so even if no deliberative judgment is made in the context of perception.

3. The importance of taking one’s perceptual experience at face value.

To be a perceiver is to possess the ability to keep track of how things are by having experiences which one appreciates as relevant to how things are. This point is a phenomenological one. It is a basic fact about perceptual experience that we take ourselves, in looking around say, to have access to the world. Our experience presents itself to us as a form of contact with the world.

Of course, it is not the case that things always look the way they are. Nor is it the case that when one has a visual experience, one is always inclined to judge things to be the way the experience represents them as being. Experience, it has often been noticed, is belief-independent [Evans 1982] It can look to me as if the two lines of the Müller-Lyer illusion differ in length even when I have drawn them myself and know them to be of identical length.

Nevertheless, it would be a mistake to infer from this fact that perceptual experience is, as one might put it, belief-indifferent [Noë 1999]. For perceptual experiences, of their very nature, raise questions about how things are. In particular, they raise the question of what one ought to believe, on the basis of the experience, were one to take the experience at face value.

The idea that perception requires us to be able in this way to take our perceptual experiences at face value is of fundamental importance. First, it points to a way in which perceptual experience must be thought of as occurring in a space of reasons, to use Sellar’s [1956] and McDowell’s [1994] phrase. That is, only one capable of appreciating how a perceptual experience presents the world as being could actually have the experience.

Second, and of particular relevance to the present discussion, it enables us to understand that our ability, as perceivers, to keep track of the ways our perceptual experience depends on what we do — perspectival self-consciousness — is, in effect, a special instance of our ability, as perceivers, to keep track of how things are. Perceptual experience raises questions not only about how things are, but about how we stand in relation to how things are. In keeping track of how what we do affects what we experience, we are keeping track of what our experience tells us about the world.

The price we pay for losing track of the ways in which what we do affects what we see is that we lose the ability to take our experiences at face value. The loss of this ability is tantamount to a kind of functional blindness.

It is true, then, that perception depends on perspectival self-consciousness. But that is because perspectival self-consciousness is, so to speak, an aspect of perception.

4. Attention, perception and conceptual content.

To think of perceptual experience as nonconceptual is to think of it as unconstrained by what we are capable of grasping in thought. This idea goes very naturally with the notion that the representational content of a perceptual experience is much richer and more detailed than that which we can grasp in thought. Commitment to something like this idea comes out very clearly in Peacocke’s conception of scenario content [Peacocke 1992]. The scenario content of a perceptual experience is the way of filling out the space around the individual such that the experience is veridical. Scenario content is nonconceptual, according to this view, because the perceiver need not, in order to have the experience, possess the concepts needed to capture in thought the ways of filling out the space that would make the experience veridical. For example, a creature without the concept sphere, or the concept degrees left might have a visual experience as of a sphere so and so many degrees to the left of center, an experience that would be made veridical by the presence of a sphere in that location.

This conception of the representational content of experience is phenomenologically wrongheaded and empirically ungrounded. It will be helpful to discuss these points in turn.

We misdescribe the character of our perceptual experience if we suppose that we do in fact have in consciousness all at once the environmental detail that would, as it were, make the experience veridical. For experience, as we know it, is intrinsically indeterminate. Objects in the center of attention and focus are indeed experienced in detail, but we do not experience the whole visual field this way as sharply focused and in uniform detail. At a given instant, the rest of the visual field remains in the background as indeterminately present.< 3 > To fill out the space around a the perceiver in such a way as to make the current experience veridical would, then, be to attempt something impossible — to make an indeterminate physical space.< 4 >

How can we square these phenomenological observations with the fact that we take ourselves, in experience, to come into contact with a densely detailed, high-resolution environment? This is a delicate point that requires careful elaboration. First, it is true that we take ourselves in experience to come into contact with the detailed environment. It does not follow from this that we take our conscious experience to represent all that detail. Rather, we take the detail to be there, in the world, and we take ourselves, in experience, to have access to that detail. Second, experience as we actually encounter it (in contrast with the fantasy of experience described in some philosophical theories), is not a momentary occurrence, but a temporally extended encounter with the environment. (If you are asked to reflect on your visual experience, you will probably look around you and think about what is there.) Third, experience in this more full-blooded sense — that is, experience thought of as a temporally extended episode of encounter with a densely detailed environment — is bound up with further capacities for thought and is, thus, conceptual. Perception is itself an activity of exploration of the world and this activity draws not only on our sensorimotor skills, but, more generally, our understanding of how things are.

This conception of experience as a mode of activity, and these criticisms of nonconceptual scenario-content gain support from recent work in psychology on the relation between perception and attention. In a broad range of experiments, psychologists have shown perception to be highly attention-dependent. Perceivers do not experience features of the scene — even features that are in the center of the visual field and that are of great interest — if attention is directed elsewhere. Two examples will suffice to clarify the point. In one study, perceivers are asked to watch a video tape of a basketball game and they are asked to count the number of times one team takes possession of the ball [Simons & Chabris, in press]. During the film clip, which lasts a few minutes, a person in a gorilla suit strolls onto the center of the court, turns and faces the audience and does a little jig. The gorilla then slowly walks off the court. The remarkable fact is that perceivers (including this author) do not notice the gorilla. This is an example of what has been called inattentional blindness.< 5 > In a second study, due to O’Regan, a perceiver is asked to describe changes occurring in video clip of a drawing [O’Regan & Noë 2000].< 6 > Because the changes occur very slowly, and thus do not attract attention, they tend not to be perceived. The net effect of the changes, however, can be very substantial, such as a change in color of the central object in the image. This is an example of what is known as change blindness. < 7 >

The liability of normal perceivers in this way to fail to perceive what is going on around them owing to misdirected attention does not get us into trouble too often.< 8 > For our perceptual systems, and our cognitive system, are designed to work together to enable us to direct our attentional resources, and our perceptual resources, to where they are needed. In general, by the flick of an eye, or the turn of the head, we are able to acquire the perceptual information we need, when we need it [O’Regan 1992]. This attunement on the part of perceivers to the environment enables us to preserve a deserved feeling of contact with that environment, even though, at any given moment, we are conscious only of fairly sparse amounts of detail.

These facts have two relevant upshots for our current investigation. First, the attention-dependence of perception further demonstrates that we misdescribe our experience as having the sort of scenario content Peacocke imagines. The actual content of a brief episode of seeing is much more sparse than that.

Second, these considerations bring forcibly to mind the fact that what is seen depends on activity on the part of the perceiver that is at least quasi-conceptual. Attention is a way, in experience and thought, of identifying, discriminating, carving out features of the environment from the background. Moreover, attention is, as it were, gist-dependent. Where you look, how you inquire, depends on how you take the scene, on how you understand it.< 9 >

5. The holism of thought, action and perception.

I have urged that we recognize that perspectival self-consciousness is not so much a precondition of our ability to perceive, as it is an aspect of that ability. I have also urged that our ability to perceive, more generally, should be thought of as drawing on our conceptual and inferential capacities. In making this claim I am advocating a more thoroughgoing holism than that which has been advocated by Hurley and others. To view perceptual experience as a phenomenon of the personal or animal level is, I suggest, to view it as integrated with broader capacities for intentional action and thought. In advancing this more far-reaching holism, I hold fast to the idea that perception is, in essence, a capacity to learn how things are on the basis of active exploration and inquiry.

More important than the claim that perceptual experience is conceptual, is the thought that there is no sharp division between the conceptual and the nonconceptual. Our mental lives are conceptual, I propose, at least to the extent that they are understood to be phenomena of the personal level.

It may be, however, that there is also no sharp line between the personal and the subpersonal. This thought can be illustrated in connection with Hurley’s important conception of the person or animal as a dynamic singularity (1998, p. 2). An animal is an active locus of mutual interaction with an environment, a system of dynamic feedback relations embedding a nervous system in a living body which is in turn situated in an environment. The interplay between perception and action that we have discussed is an aspect of this mutual determination of organism and environment.

Armed with the dynamic singularity conception, Hurley brilliantly demonstrates the inadequacies of a more linear, input-output conception of the relation between perception and action. For reasons that should by now be clear, however, I question whether she is right in taking this conception to operate, as she does, at the subpersonal level alone (1998, p. 2).

Perceptual experience, I have argued, is the activity of exploring the environment making use of one’s mastery of sensorimotor contingencies [O’Regan & Noë, under review]. Is this activity subpersonal activity? It must be in part. After all, there is no doubt that subpersonal mechanisms causally enable vision and other cognitive capacities and that, therefore, the conception of the person as a dynamic singularity depends on an account of those subpersonal processes. But it also seems clear that such an account must advert to personal-level activity as well, and not only because it is personal-level phenomena we seek to explain. We have already considered that our ability to keep track of the dependence of what we experience on what we do is a feature of our more general capacity, as perceivers, to keep track of how things are on the basis of our active exploration of the environment. Perceivers are familiar, at least implicitly, with the way appearances change as a result of movement. This familiarity reveals itself in the automaticity with which we squint, pat our pockets in search of our glasses, and move closer when we are trying to get a better look at something. To be a perceiver is, among other things, to know how to do all this. In this way our capacity to keep track of the dependence of perception and action is firmly integrated with personal-level conceptual and inferential skills.

I do not wish to deny the fact that we need to distinguish the level at which considerations of normativity and holism applies from the level at which, as it were, we are interested in blind causal processes which enable. This is an important distinction, and keeping it clearly in mind is necessary if we are to avoid the fallacies involved in reading subpersonal structures and processes onto the personal level, or vice versa. One example of this sort of fallacy, noticed by Dennett [1991, but see also Pessoa, Thompson and Noë 1998] is that of believing that because there a place on the retina where there are no photoreceptors (the "blind spot"), then there must be a process of neural perceptual completion whereby the brain fills in the gap.

But matters are more complicated than this makes them seem. For one of the central upshots of the way of thinking of perception that I have been advocating here is the importance, for perception, of embodiment. Vision is precisely the capacity to explore the environment making use of our distinctive visual apparatus. Our ability to see is constituted by our mastery of the ways in which, for example, the movements of eyes like ours give rise to systematic patterns of change in stimulation. But the character of these sensorimotor contingencies depends, in the most nuanced of ways, on the construction of the whole eye/retina/brain/head/body system. If this is right, then it is not possible sharply, once and for all, to distinguish a level of implementation, of blind causal process, from a level at which intentionality comes into play. For the abilities that constitute our personal-level capacities are precisely abilities to deploy our bodies (and our minds) in a plethora of different ways.

This is the basis of an argument against multiple-realizability and so against a certain kind of orthodox functionalism. If this general line is right, then only creatures with bodies like ours could have perceptual experience like ours, for having perceptual experience of this kind is just a manner of exercising our bodies in various ways.



<1> I take it that Wittgenstein’s claim that there is a way of following a rule that does not depend on a further act of interpretation (PI § 201) expresses the thought that sometimes the application of a concept is more immediate than this.

<2> Some might say that basic concepts such as red and valid are concepts that are applied on the basis of nonconceptual grounds. To understand such a concept is just to apply it in an appropriate condition, even though one has no concept of the condition itself. This has been argued by Peacocke. But this seems mistaken to me. The point is not that such concepts get applied on nonconceptual grounds, but that they are not applied on the basis of grounds at all. Explanation, as Wittgenstein said, comes to an end.

<3> I have benefited from the views of Sean Kelly on the main point of this paragraph.

<4> This impossibility is the source of more than one cinematic joke. For example, in Woody Allen’s Zelig, the character played by Robin Williams has gone out of focus, as it were, in real life.

<5> This term is due to Mack and Rock [1998]. A detailed study of the phenomenon is contained in this book.

<6> These demonstrations, along with other demonstrations of change blindness, may be viewed on O’Regan’s web site:

<7> For a recent review of the change blindness literature, see Simons [2000] and Simons and Levin [1997]. For a discussion of philosophical implications, see Noë, Pessoa and Thompson [2000] and O’Regan and Noë [under review].

<8> Although one might wonder how often this sort of phenomenon is the cause of road accidents and the like.

<9> This is an important theme in the change blindness literature. You are more likely to notice a change to a scene if the change alters the scene’s gist. See Simons and Levin 1997 for more on this.



Dennett, D. C. (1991) Consciousness Explained. Boston: Little Brown.

Evans, G. (1982) The Varieties of Reference. Oxford: Oxford University Press.

Gregory, R. L. & Wallace, J. G. Recovery from early blindness: a case study. In P. Tibbetts (ed.) Perception: Selected Readings in Science and Phenomenology. Chicago: Quadrangle Books, 359-388.

Helmholtz, H. v. (1909/1925). Physiological optics (Vol. 3). Translated by J.P. C. Southall. Rochester, New York: Optical Society of America.

Hurley, S. L. (1998) Consciousness in Action. Cambridge, MA: Harvard University Press.

Kohler, I. (1951). Über Aufbau und Wandlungen der Wahrnehmungswelt. Österreichische Akademie der Wissenschaften. Sitzungsberichte, philosophish-historische Klasse, 227, 1-118.

Mack, I. & Rock, A. (1998) Inattentional Blindness. Cambridge, MA.: The MIT Press.

McDowell, J. (1994) Mind and World. Cambridge, MA: Harvard University Press.

Noë, A. (1999) Thought and experience. American Philosophical Quarterly. 36, 3: 257-265.

Noë, A., Pessoa, L., Thompson, E. (2000) Beyond the grand illusion: what change blindness really teaches us about vision. Visual Cognition. 7, 1/2/3: 93-106.

O’Regan, J. K. (1992) Solving the "real" mysteries of visual perception: The world as an outside memory. Canadian Journal of Psychology. 46, 461-488.

O’Regan, J. K. & Noë, A. (under review) A sensorimotor theory of vision and visual consciousness. Behavioral and Brain Sciences.

O’Regan, J. K. & Noë, A. (2000) Experience is not something we feel but something we do: a principled way of explaining sensory phenomenology, with Change Blindness and other empirical consequences. Talk given at the ASSC Conference: THE UNITY OF CONSCIOUSNESS: BINDING, INTEGRATION, AND DISSOCIATION, Brussels, June 29-July 2, 2000. Available online at:

Peacocke, C. (1992) A Study of Concepts. Cambridge, MA.: The MIT Press.

Pessoa, L., Thompson, E. & Noë, A. (1998) Finding out about filling in: a guide to perceptual completion for visual science and the philosophy of perception. Behavioral and Brain Sciences. 21, 6: 723-802.

Sellars, W. (1956) Empiricism and the philosophy of mind. In Herbert Feigl and Michael Scrivens (eds.) Minnesota Studies in the Philosophy of Science. Vol. 1. Minneapolis: University of Minnesota Press.

Simons, D. J., & Chabris, C. F. (in press) Gorillas in our midst: sustained inattentional blindness for dynamic events. Perception.

Simons, D. J. & Levin, D. T. (1997) Change blindness. Trends in Cognitive Sciences. 1, 7: 261-267

Stratton, G. M. (1897). Vision without inversion of the retinal image. Psychological Review, 4, 341-360, 463-481.

Wittgenstein, L. (1953) Philosophical Investigations. Third Edition. Translated by G. E. M. Anscombe. Oxford: Blackwell

top | back to symposium index | Hurley's Reply