The New Science of the Mind

by Mark Rowlands

  5 The Vehicles of Perceptual Disclosure

  Consider the disclosing activities of a visually unimpaired subject. It is tempting, indeed it is typical, to think of these activities as restricted to processes occurring in the eyes themselves and subsequent neural processing operations. Such processes are, of course, vehicles of disclosure, not objects of disclosure. I am not aware of these processes: they are processes with or in virtue of which I am aware of other things. Relative to those things of which I am aware-the empirical objects of my awareness-the status of these processes is transcendental. These processes form part of my causal disclosure of the world.

  However, my causal disclosure of the world is not restricted to these inner processes. In addition to the various neural processes occurring in my eyes and brain we can also identify various activities that I perform in the world; and these activities also form part of my causal disclosure of the world. It is useful to break down these activities into three different, but partially overlapping, kinds: (i) saccadic eye movements, (ii) probing and exploratory activities involved in the identification of sensorimotor contingencies, and (iii) the manipulation and exploitation of the optic array. These categories do not exhaust the kinds of extraneural disclosing or revealing activities appropriate to visual perception. But they are, I think, absolutely central.

  Saccadic eye movements When I perform visual tasks, my eyes engage in various movements-saccades. Yarbus (1967) has demonstrated that (i) different tasks result in quite different scan paths, and (ii) the pattern of saccadic eye movement is systematically related to the nature of the visual task. In a famous series of experiments, Yarbus (1967) asked subjects, prior to their viewing of a painting, to perform certain tasks. The painting showed six women and the arrival of a male visitor. Subjects were asked to either:

  1. View the picture at will.

  2. Judge the age of the people in the painting.

  3. Guess what the people had been doing prior to the arrival of the visitor.

  4. Remember the clothing worn.

  5. Remember the position of objects in the room.

  6. Estimate how long it had been since the visitor was seen by the people in the painting.

  Yarbus demonstrated that the required task had an impact on the visual scan path that the subject took: different tasks resulted in quite different visual scan paths. Subjects asked questions concerning the appearance of people in the painting-for example, questions about their ages-focused on the area around the face. Subjects asked questions concerning the theme of the painting focused on various points throughout the picture. And different themes also resulted in different scan paths. For example, subjects asked what the people doing before the visitor arrived employed a different scan path from those asked to estimate how long it had been since the visitor was last seen by the family. In general, Yarbus showed, the scan varies systematically with the nature of the task.

  Saccadic eye movements, and the more general patterns of search in which such movements are situated, are part of the vehicles of perceptual disclosure. Being a person I saw only last week is an empirical mode of presentation of an object. So too is being a person I haven't seen for years. My saccadic scan paths, in the sort of context described here, are part of the means with or in virtue of which the world is disclosed as falling under one or another empirical mode of presentation-for example, as containing a collection of people who have not seen the visitor for many years rather than a collection of people who saw him last week.

  Saccadic scan paths are, of course, not objects of awareness-typically, we have little or no idea what our eyes are doing when we extract information from a visual scene. Phenomenologically, we are typically not aware of the eye movements, but of what these movements help us reveal: an object falling under an empirical mode of presentation. And this phenomenological point concerning what we are and what we are not aware of is grounded in the deeper point concerning the nature of intentional directedness: as revealing or disclosing activity, intentional directedness passes all the way out to the objects revealed. In this case, it travels through the saccadic eye movements out to the world itself.

  Sensorimotor activity Recall the enactive or sensorimotor account of visual experience outlined in chapters 2 and 3. Such an account accords a central role to a certain sort of activity: the probing or exploratory activity required to identify the sensorimotor contingencies pertinent to a given visual scene. Earlier, I cast doubt on whether enactive accounts could properly be regarded as versions of the extended mind. Here I simply want to focus on the role of exploratory visual activity. Whether or not enactive accounts are of a piece with extended accounts, it is still true that the sorts of exploratory activities invoked by enactive accounts are among the vehicles of perceptual disclosure. And this latter claim is all we need for present purposes.

  Suppose, to use an example of Dennett (1991), you are looking at a wall of photographs of Marilyn Monroe, a la Andy Warhol. Your foveal vision subtends no more than three or four of these photographs, and your parafoveal vision is insufficiently precise to discriminate Marilyns from squiggly shapes. Nevertheless, it seems to you as if you are confronted with a wall of Marilyns, and not three or four Marilyns surrounded by a sea of squiggly shapes. The wall of Marilyns, in its entirety, is phenomenologically present to you.

  The explanation supplied by the enactive account of this sense of phenomenological presence is simple and elegant. First, the impression we have of seeing everything-the wall of Marilyns in its entirety-derives from the fact that the slightest flick of the eye allows any part of the wall to be processed at will. This gives us the impression that the whole wall is immediately available (O'Regan and Noe 2001, 946). Is this impression erroneous? It would be erroneous only if seeing consisted in the production of an internal representation isomorphic with the part of the world seen. If, on the other hand, we accept that seeing consists in combining the results of environmental probing with knowledge of laws of sensorimotor contingency, we are indeed seeing the whole scene, for probing the world, and knowledge of these laws, is precisely what we do and have as we cast our attention from one aspect to the next.

  Second, in addition to our ability to direct our attention, at will, to the visual world, the visual system is particularly sensitive to visual transients. When a visual transient occurs, a low-level "attention-grabbing" mechanism appears to automatically direct processing to the location of the transient. This means that should anything happen in the environment, we will generally consciously see it, since processing will be directed toward it. This gives us the impression of having tabs on everything that might change, and so of consciously seeing everything. And if seeing consists in exploratory activity combined with knowledge of sensorimotor contingencies accompanying such exploration, then this impression is not erroneous. We do, indeed, see everything. The suspicion that we do not derives from a residual attachment to the idea that seeing consists in the production of an internal representation that maps onto the outside world.

  Casting one's attention at will to any part of the visual scene, or having one's attention drawn by a visual transient to a part of the scene: these are both examples of probing or exploratory activity.' Like saccadic scan paths, these sorts of activities are vehicles of perceptual disclosure: they are that with which or in virtue of which, in part, the visual world is disclosed in the way that it is. That is, the activity is that in virtue of which a given part of the world is revealed as falling under one or another empirical mode of presentation-for example, a falling under the empirical mode of presentation "wall of Marilyns" rather than the mode of presentation "wall of indeterminate shapes." These probing, exploratory activities are not, typically, things of which we are aware when we visually experience the world: they are things with which we visually experience the world. That is, they are activities with which the world is disclosed to us as falling under one or another empirical mode of presentation. That is, these activities are among the vehicles of caus
al disclosure of the world; part of the means by which, in the case of vision, our intentional directedness toward the world is achieved or effected. As such, our visual consciousness both lives through and travels through these activities no less than it does through processes occurring in our eyes and brain.

  Manipulation of the optic array Consider, now, Gibson's (1966) account of visual perception outlined earlier. A key component of Gibson's account is the idea that by acting on the optic array, and thus transforming it, the perceiving organism makes available to itself information that was, prior to the action, present but not available. When an observer moves, the optic array is transformed, and such transformations contain information about the layout, shapes, and orientations of objects in the world. More specifically, by effecting transformations in the ambient optic array-by transforming one array into another systematically related array-perceiving organisms can identify and appropriate what Gibson calls the invariant information contained in the optic array. This is information contained not in any one static optic array as such, but in the transformation of one optic array into another. The proper function of the transformation of one optic array into another is to transform the status of this invariant information from the merely present to the available.

  The manipulation of the optic array, manipulation that is carried out by movement on the part of the perceiving organism, is another vehicle of perceptual disclosure. The perceiving organism need not be, and typically is not, aware of its manipulative activities: these activities are vehicles of disclosure, not objects of disclosure. That is, the activities are ones with or in virtue of which the perceiving organism becomes aware of certain features of its environment; they are not, typically, activities of which the organism is perceptually aware. Phenomenologically, the perceptual awareness of the organism passes through the activities to the world that those activities, in part, disclose to it. Crucially, where does the organism's perceptual disclosure of the environment take place? It takes place, in part, wherever the activities take place. And these activities do not stop short of the external optic array. One cannot manipulate an external structure unless one's manipulation reaches out to that structure.

  Like saccadic scan paths and sensorimotor probing, manipulation of the optic array is one of the vehicles of causal disclosure of the world. That is, in the case of visual perception, manipulation of the array is one of the means by which our intentional directedness toward the world is brought about; it is one of the means by way of which a part of the world is revealed as falling under some or other empirical mode of presentation. Our visual consciousness both lives and travels through these manipulative activities no less than it occurs in processes occurring in the eye and brain.

  6 The Return of Otto

  The story so far looks like this. Intentionality-intentional directedness toward the world-should be understood as revealing or disclosing activity. This is the activity in virtue of which the world is presented to the subject as falling under one or more aspects or empirical modes of presentation. If intentional directedness consists in disclosing activity, it takes place wherever this disclosing activity takes place. Such activity exists in many places. Processes occurring in the eyes and the brain can be part of this disclosing activity: they are certainly part of the activity in virtue of which the world can be disclosed to a subject as being a certain way-that is, as falling under a given empirical mode of presentation. However, there is little reason for thinking that the disclosing activity constitutive of intentionality is restricted to processes occurring inside the brain. Certain things we do in and to the world can, no less than neural processes, be part of the activity that discloses the world to a subject as falling under a given empirical mode of presentation. When Merleau-Ponty's blind man uses a cane to disclose the world as containing objects in a particular orientation to him, this is part of the activity through which he discloses the world. When a visually unimpaired subject employs a certain saccadic scan path in order to identify salient information in a visually presented scene, this is part of the disclosing activity of that subject. When he probes and explores the world in a manner that reveals its sensorimotor contingencies, this probing and exploration forms parts of his disclosing activities. When he manipulates the optic array in order to make available information that was, prior to this, present but unavailable, this too is part of his disclosing or revealing activities. If intentionality consists in disclosing activity, then this intentionality is not restricted to processes occurring inside the brain.

  The discussion, so far, has supplied two things. First, there is the general model of intentionality as revealing activity. Second, there is the application of this general model to perception-in particular, visual perception. However, the amalgamated mind-the conjunction of the embodied mind and extended mind-is a thesis about cognition in general, not merely perception. Therefore, the next stage is to switch focus and apply the model of intentionality as revealing activity to cognitive processes more generally. With this goal in mind, let us return to Clark and Chalmers's case of Otto.

  Wanting to see the exhibition he has just read about in the newspaper, Otto consults his notebook, and sees that the Museum of Modern Art is on 53rd Street. On Clark and Chalmers's version of the extended mind (at least as this is usually understood), the entry in Otto's notebook is identical with one of his beliefs. For reasons outlined earlier, I do not endorse this claim. The sentence-tokens in Otto's book are the wrong sorts of tokens to be identical with belief-tokens. Therefore, we should not identify the sentences in Otto's book with token cognitive states. According to my version of the extended mind, on the other hand, Otto's manipulation of his notebook counts as part of his process of remembering the location of the Museum. If the account of intentional directedness defended here is correct, then it is clear why this should be so. The manipulation of the book forms part of the vehicles of causal disclosure of the world. That is, the activity of manipulating the book is part of the means whereby, in the case of memory, Otto's intentional directedness toward the world is brought about. The manipulation of the notebook is, in part, that in virtue of which the world is disclosed to Otto, in memory, as falling under a specific empirical mode of presentation. That is, in virtue of Otto's activity, the museum is disclosed as falling under the empirical mode of presentation of being located on 53rd Street. The intentional directedness of Otto toward the world both lives and travels through his manipulation of the notebook. For this manipulation is nothing more than what, in part, reveals the world to him, in memory, as being as certain way.

  This is why it is legitimate to regard Otto's manipulation of his book as part of his process of remembering. The manipulation is part of the vehicle of his intentional directedness toward the world-intentional directedness that, in this case, takes the form of remembering. There is, as I argued earlier, no intentional directedness in the void. Intentional directedness always takes place through some or other vehicles-intentional ether, if you like. In the case of Otto, the vehicles are, in part, brain processes: the processes, for example, that allow him to detect the sentence that is written on the page and form beliefs about the content of this sentence. The processes are also bodily: for example, ones that allow his arms, hands, and fingers to move in such a way that they can manipulate the book. But they are also, finally, environmental processes-processes of manipulating the book in such a way that information that was hitherto unavailable to Otto now becomes available. In the case of Otto, all of these processes-neural, bodily, and environmental-form proper parts of the overall process of disclosing the world, in memory, as falling under a given empirical mode of presentation. But to talk of disclosing the world, in memory, as falling under a given empirical mode of presentation is simply to talk of remembering the world as falling under a given empirical mode of presentation. To remember is simply to disclose the world in memory. Therefore, in the case of Otto at least, all of these processes-neural, bodily, and environmental-can properly be regard
ed as forming part of the overall process of remembering.

  This way of understanding the case of Otto has another advantage: it allows us to defuse what many take to be an important objection to the extended mind. One objection to regarding Otto's consultation of his book as part of the process of remembering is that Otto's access to his book is fundamentally different from Inga's access to her memories. Otto's access to his book is perceptual; Inga's access to her memories or beliefs is not. In considering this objection, Clark and Chalmers accept the premise of the objection: Otto's access to his notebook entries is indeed perceptual and Inga's access to her beliefs is not. But they deny that this disqualifies the notebook entries from counting as among Otto's beliefs. The difference in mode of access is not sufficiently significant to result in this disqualification. Consider, they say, a cyborg of the Arnold Schwarzenegger Terminator genre. The terminator's access to its stored information-both about the world and its internal states-takes the form of a quasi-perceptual display on a virtual visual display unit (VDU). This does not seem to preclude ascribing to the terminator, when the appropriate virtual display is activated, the belief that, for example, the fleeing suspect is indeed John Connor. Whether or not an item counts as a belief is not something essentially determined by one's mode of access to it. Rather, as we have seen, for Clark and Chalmers, it is the item's functional role that is decisive.