Friday, September 24, 2010

Space in the Causal Theory of Perception

John Smythies has kindly provided some of Lord Russell’s actual utterances about how he conceived the relationship between physical space and perceptual space in his book Human Knowledge: Its Scope and Limits, a basic and even now classic source for philosophical analysis of this topic. The first passage is from the chapter “Mind and Matter”:
The objects of perception which I take to be ‘external’ to me, such as coloured surfaces that I see, are only ‘external’ in my private space . . . When on a common-sense basis, people talk of the gulf between mind and matter, what they really have in mind is the gulf between a tactual percept, and a ‘thought’—e.g. a memory, a pleasure, or a volition. But this, as we have seen, is a division within the mental world; the percept is as mental as the ‘thought’. (Simon and Schuster edition, p. 228)
The second excerpt is from the previous chapter, “Space in Psychology”:
All this [the physiological account of perception], I say, has long been a commonplace, but it has a consequence that has not been adequately recognised, namely that the space in which the physical table is located must be different from the space we know by experience. (p. 222)

In these two passages Russell does not refer to the theory of perception as such, that is, the “physical theory of perception” as he calls it in the chapter “Physics and Experience,” or “causal theory of perception” in his book The Analysis of Matter, devoting a chapter to it under that name. Russell makes explicit reference once again to the proverbial table which so often serves as an example in these armchair exegeses (perhaps being a pedagogical tradition dating from medieval scholasticism?):

The colored surface that I see when I look at a table has a spatial position in the space of my visual field; it exists only where eyes and nerves and brain exist to cause the energy of photons to undergo certain transformations. (The ‘where’ in this sentence is a ‘where’ in physical space.) The table as a physical object, consisting of electrons, positrons, and neutrons, lies outside my experience, and if there is a space which contains both it and my perceptual space, then in that space the physical table must be wholly external to my perceptual space. This conclusion is inevitable if we accept the view as to the physical causation of sensations which is forced on us by physiology . . . . (p. 220f)
In saying that the conclusion “is inevitable,” I would refer to my recent remarks about data “demanding” one theory over another, because the inevitability of the conclusion (= theory) here not only depends upon but is in a sense "demanded" by the causal chain of empirical events, all save the last, arguably, as at most a percipient can only report that he indeed “sees the light” (pun intended), but no observer can also see his visual space and independently verify that it exists (yes, just the sort of argument that behaviorists used to use against what they miscalled "introspection"). Of course Russell was an avowed monist so it is reasonable to assume that dualism plays no part in the “physical causation of sensations” to which he refers.

Russell is at pains to say that confusion has resulted from conflating different senses of the word “space.” (p. 201), but assures us that percepts are nonetheless in the brain: “Percepts, considered causally, are between events in afferent nerves (stimulus) and events in efferent nerves (reaction); their location in the causal chains is the same as that of certain events in the brain.” (p. 209)

So without evidence to the contrary, it seems reasonable to assume that when Russell refers to two different spaces (physical and perceptual) he means space not in the cosmological sense, but in sense 11 of the word space in OED, “Specific or limited extent,” because he is claiming that perceptual space is “private,” or conversely, that it is separate from external physical space, i.e., external to the body.

In spite of that, I suspect that Russell may have conflated two senses of “space,” because merely in separating perceptual space from physical space he seems to be using sense 11 of “space,” yet elsewhere in the context of his argument seems to be employing OED sense 7, “Physical extent or area; extent in two or three dimensions.” They are not synonymous senses of the word.

Lord Brain basically just restates Russell’s main point, and while Broad (1923) says it is “possible” to have “a space-like whole of more than three dimensions, in which sensa of all kinds, and scientific objects literally have places,” and that “scientific space would be one kind of section of such a quasi-space, and e.g. a visual field would be another kind of section of the same quasi-space,” he does not seem to explain (1) the necessity of such a “space-like whole” nor (2) how it is “possible.” By referring to "places," again Broad seems to be using sense 11, not sense 7 of the word "space."

Where then does this leave us, as other than Broad’s imaginative proposal, what compelling evidence is there to support it since he put forth the idea 1923?


  1. I have a lot of admiration for Russell, but perhaps it should be pointed out that he shifted his positions on some of these matters several times during his long careers. As far as I can see in the 1920s, for example in his "The Analysis of Mind" he advocated a position of "neutral monism" whereby both matter and minds were thought of as being composed of sense data albeit under different arrangements. Later, he definitely ascribed to a version of the causal theory of perception, and definitely rejects naive realism in his "Inquiry into Meaning and Truth." It is not clear to me whether his latter account was still supposed to be monist though.

  2. 1. Yes, Russell was remarkable in that he was one of the few philosophers who changed his mind from time to time!
    In the case of Broad he adumbrated that phenomenal space and physical space are different cross-sections of a higher-dimensional space—that is he appeared to use 'space' here in the way I use it. But in 1923 he hedged his bets by using the ambiguous term "quasi-space". However, if 'quasi-space' A and 'quasi-space' B are cross-sections of a 'space-like whole' of more than three dimensions, the word 'cross-section' here only. I suggest, makes sense in a real space context. I knew Broad well and had many conversations with him in the 1950's. Curiously enough he never took the next logical step in 1923 of consulting with a cosmologist about his idea about higher-dimensional "space-like wholes". Yet, near the end of our discussions, having read "Analysis of Perception" he advised me that I should do so myself.
    I also knew H.H. Price even better (both of them were subjects in my experiments with mescaline) and he unequivocally held that phenomenal space and physical space were independent spaces that had no spatial relations with each other, only causal ones. Therefore he did not invoke any higher dimensions. Moreover his "phenomenal space' included only images and the Self.

  3. 2. It also seems to me that the senses that Bill gets for "space" in the OED need looking at further. Take sense 11 of the word space in OED, “Specific or limited extent,” What does that mean? Usage like "There was only enough space for one table in the room"? I do not see what that has got to do with phenomenal space. Certainly a person's phenomenal space is of limited extent, but that is an observable property phenomenal space has. When Broad says "have places" he presumably means "are located".
    Then take sense 7 "Physical extent or area; extent in two or three dimensions.” This does not seem to me to refer to physical or cosmological space which does not have 2 dimensions. Sense 7 seems to me to refer to things like football fields.
    I suggest we need to define "space" (phenomenal)] as "that in which visual (and somatic) sensations are located and extended."
    and "space" (physical) as that in which the planet earth is located and extended"

  4. Bill asks for any compelling evidence to support the theory that Broad was one of the originators. This is rather a large subject that I cover in detail in my books "Analysis of Perception" and "The Walls of Plato's Cave". I would be happy to answer questions about any particular points from these books that anyone cares to put.

  5. This comment has been removed by the author.

  6. Hello, everybody.As Pr. Smythies invites us to ask some questions about the description and model he develops in his book "The Walls of Plato's cave", I may ask this one, to begin with: what is exactly the ontological status of the (apparent) observer (that is refered to as "O")and the(apparent) void/vacant space/distance in (Ganzfeld or other) experiments in which subjects seem to report, while they have their eyelids shut, a "distance" of a few inches between an inner vantage point (the "observer")and the visual images that appear in their visual field? Isn't that purported "distance" or "void" between two entities a mere interpretation formed by falsely applying our usual exteroceptive categories of "subject-object spatial separation/distinction" used in cases of environmental inspection to a situation in which there actually isn't any real spatial distance between the inspecting "subject" and the inspected "object"? I mean: when we close our eyes and inspect our visual field, speaking of a vacant space or void between a visual screen and an inner observer can be quite misleading (because we can be accused of re-introducing a subtle version of the infamous homunculus fallacy). Wouldn't it be better to avoid any reference to an inner "observer" and say that this apparent "distance" is only an inner duplicate of our multimodal (proprioceptive-kinetic and visual in that case) apprehension of "distance" in usual daily perceptual situations, combining/blending in a single continuous holistic "phenomenal space" (with no segmentation and no real "void" between separated entities)the body image of our intra-cranial "volume" and eyeglobes with our curved 2D visual field, thus producing the impression that there is a "foreground", an in-between "void", and a distinct "background".

    Another correlated question: isn't it possible that the "screen" of our visual field is entirely flat (2D)and that our sense of depth is a learnt mental construction (perceiving the 2D visual sense data as 3D objects)acquired through proprioceptive-kinetic effort, during the ontogenetic cognitive development?

    After all, if we resort to a televisual analogy, when you look at your flat television screen, the images are contiguous on a geometrical plane and they have no depth at all, and yet, your mind is able to "reconstruct" a sense of depth in order to appreciate the depicted visual scenes.

  7. OED sense 11a of the noun "space" is "an area or extent delimited or determined in some way." The main limitation that Russell adduces is that perceptual space is *private.* One of two citations OED related to vision that OED gives for it is from 1879 G. C. HARLAN Eyesight iii. 37 "The most sensitive portion is a small space directly in the line of vision, called the yellow spot." That refers to a delimited area within VS, but by extension could logically include all of VS.

  8. The first five senses of "space" in OED refer to time. Those five senses are grouped under Roman numeral I. Under Roman numeral II are grouped senses "denoting area or extension," particularly "general or unlimited extent." I erroneously gave as sense 7 what I inteded to give as sense 8 (because I neglected to notice where the examples changed to another sense--mea culpa), i.e., " Continuous, unbounded, or unlimited extent in every direction, without reference to any matter that may be present; this regarded as an attribute of the universe, describable mathematically (in modern science usually conflated with time: cf. SPACE-TIME n. 1); (as a count noun) a mathematical construct or model of this." Clearly this is what Russell had in mind as "physical space."

  9. As for specific questions about higher dimensional models in which one part is physical and the other perceptual, I would ask what empirical evidence there is for these two sections, and that they are indeed separate by substance (physical vs. perceptual)? Is this more than just a geometrical analogy suggested by cosmology to convey the idea of substance dualism just as Zollner in his day invoked the mathematically-possible fourth dimension as an explanation for psychic phenomena?

  10. Welcome Thomas Droulez! I shall defer to John Smythies to reply to your question, except to say that I believe VS is the source of all our ideas about space, and that in some sense it is three dimensional, and that we perceive that dimensionality in binocular VS, if in a circumscribed way (as we have already been discussing). Therefore I reject the view that the sense of 3-D is achieved by a relationship with another sense modality (somatosensory in this case) but share with J. J. Gibson the notion that VS is oriented and defines the observer as a unique position in VS (as we can literally *see* our own bodies) as well as that which is observed surrounding that position, indeed as a sort of geometrical origin to VS.

  11. One further point: As the newly sighted blind or resighted blind so vividly express in their comments, even though different spatial relations are given in VS, the ability to judge distances and things like the quality of hardness are a matter of correlating vision and touch. One could call that "perceptual learning" (in keeping with Elearnor Gibson's formulation of it).

  12. I would also like to welcome Thomas to the group. One issue to raise here involves equating two dimensionality with flatness. I agree that visual space is two dimensional but do not think that it is flat, at least in the sense of possessing a Euclidean metric. For example in viewing a random dot stereogram with two eyes, the resulting visual experience is not flat, but we do not perceive the interiors of objects being perceived either, and thus I deny that the experience is three dimensional either.

  13. This comment has been removed by the author.

  14. Thank you for your greetings. Yes, I agree that two dimensionality should not be equated with flatness. I mentioned that our visual field was curved in its metric, but I misused the term "flat" : what I meant was "absence of depth", of course (but I still had in mind the analogy with the TV screen). To give an answer to William Rosar, I would say that what I put into question is not the fact that various bodily parts appear in our visual field (including our torso, our legs, feet, arms, hands and even the sides of our nose and parts of our eye sockets --as in Ernst Mach's famous drawing, or, more recently in Steven Lehar's description of our visual field)but rather the idea that we are supposedly able to draw a distinction between an observing "subject" and an observed "object" merely and exclusively by inspecting the visual field. If we were purely visual beings (with no other sensory modalities),all we could experience would be occasional discrepancies between the behaviour of easily and frequently moving images (corresponding to the "bodily parts")and usually more stable and less frequently moving images (corresponding to the "environmental objects"). The distinction between observing "subject" and observed "object" relies on the link between the above-mentioned discrepancy and the proprioceptive components of agentivity: that is to say, when, in the course of one's individual development, one begins to put into relation the conscious effort needed to voluntarily move one's limbs with that discrepancy between these two types of images (one discovers that some images are controlable -they are synchronized with the somatosensory experiences : the movements of the partially perceived bodily parts- while some others are not -the environmental objects).In other words, it seems to me difficult to discriminate oneself as "observer" without discriminating oneself as "agent", and this seems impossible to accomplish on a purely visual basis.

    As far as the notion of depth and the passage from 2D visual presentation to 3D visual presentation is concerned,there is indeed a difference between monocular and binocular vision, but I do not know if this is the real crucial step for our sense of depth. I've read a long time ago about something remarkable: what do you think of some people with early congenital cataract who, after having undergone a corrective chirurgical operation, reported (in the early post-chirurgical phase of visual reeducation) having (in their binocular vision) the impression that the images they inspect in their visual field are stuck on their eyes, with no perceived distance? Would you make a distinction between the evaluation of distance and the acquisition of a sense of spatial depth? Would you say that these people do have a sense of visual depth but simply have difficulties to appropriately evaluate distances?

  15. Indeed welcome Thomas! I will work on answering your very 'deep' questions shortly.
    I agree that OED sense 8 is what Russell meant by physical space. But I am not sure about the phrase "without reference to any matter that may be present": I though that that the emptiest of physical space was still the site of the spontaneous generation of particles and anti-particals of matter from the energy field by quantum processes? But that may be a mere quibble.
    Re Bill 's question about evidence for the two space/two substance theory. I will take this up in more detail shortly but here I would like to note that a scientific theory is judged on two criteria. The first is that it explains the data better than its rivals. The second is that predictions made by the hypothesis are confirmed by experiment, These need to be treated separately.

  16. More on Broad and Price.
    The recent discussion about C.D. Broad and H.H. Price started me thinking why they came up with the versions of the multidimensional theory that they did.
    Broad, when doing philosophy, always reminded me of a Pharoah slowly and ponderously building a pyramid out of elements (words/stones) shaped with enormous care to avoid any possibility of misunderstanding, never ever taking any unnecessary risk. Yet, in 1923 he suddenly, in an unusual and uncharacteristic burst of speculation (brilliant though it may have been but speculation nevertheless) comes up with his idea. He must have been scared to death. So he disguises ‘space’ as ‘space-like’ and ‘quasi-space’, so to allow himself a line of retreat.
    Broad was a small round person who admitted that he had no interest whatever in the arts and who lived for pure logic alone. His favourite country was Sweden where he spent a lot of his time. He was interested for intellectual reasons in parapsychology and, like Bernard and myself, was a member of the Society for Psychical Research.
    2. Henry Habberly Price was a totally different type of person. A dark slight gentle Welshman he lived with his devoted sister in a tall narrow Oxford house. He possessed a delightful old-fashioned courtesy and a keen appreciation of the arts. He once told me that he deplored the modern tendency to neglect the inner life. He was also a leading member of the SPR. He felt that the multidimensional theory was no wild speculation, but that it gave an opportunity for people better to appreciate the inner life that he felt was so important. Broad in contrast had no interest in the inner life at all.

  17. First replying to Bob, I, too, tended for a time to favor the conclusion that VS is a surface, but now, I am not so sure. Epistemololgically we need to be very careful about sawing off the proverbial limb upon which we are sitting, because as Thomas has now posed, this begs the question of why we ordinarily say what we see is three dimensional? Note that I expressly say *see* rather than perceive, because I believe that there is a visual basis to tridimensionality, not dependent upon but only enhanced by the other sense modalities. I shall return to this later, though.

  18. There are conceivably other alternatives to VS besides it being a flat or curved surface, one of them being a sort of perspective relief, as Mach suggested, which is neither flat nor curved. I am not convinced that the experimental or anecdotal evidence is sufficient to conclude that it is curved, and that that claim may be a attributable to the role of shifting eye positions (and changes in optical projections as as result), as I have stated already. The depth of stereopsis is qualitatively different from that of monocular depth cues, and what the newly (or resighted) blind may simply be experiencing is a lack of stereopsis (perhaps due to inexperience with ocular convergence), because I dare say it is virtually impossible to think of the visual world around us as being a single surface in binocular vision. Nothing more dramatically illustates this than looking at a complex Random Dot Stereogram, at first seeing nothing but random noise, and then rather suddenly for a complex shape with different depth planes to "stand out" from the background. Trying to explain this in terms of surface doesn't seem to do justice to the visual facts.

  19. John's fascinating little memoir of Broad and Price suggests that the idea that sensations are "in" one section of a multi-dimensional manifold suggests that for them this idea was something like an intuition, because it seems underdetermined by facts, whether perceptual (and/or psychic) or physical. As I have emphasized already, the physical picture of the world is hopelessly contaminated (as it were) by sensations, as Mach and Stebbing explained. Schroedinger was of the opinion that sensations had been intentionally omitted from the physical picture of the world, but then acknowledges that this is not literally true in quoting Democritus as he did (see my previous posting). If the physical picture is so hopelessly entangled with ideas involving or derived from sensations, on what logical (or topological) basis is there to spatially separate them? To do so would seem akin to separating Siamese twins that share a number of limbs and vital organs. In other words, there is something arbitrary about the proposal in light of this epistemology, and I wonder if Broad and Price really took into account the state of affairs acknowledged by Mach, Stebbing, and Schroedinger (to cite their contributions in chronological order)?

  20. A few points in reply to Thomas and Bill; first I think that binocular depth perception just involves an enhancement of phenomenal depth cues (from a variable dynamic metric structure) which are already present in monocular vision. In reply to Bill's question about why in ordinary language we talk of seeing three dimensional objects, I think a partial answer is that ordinary language assumes the truth of naive realism, and thus takes a three dimensional physical object to be the immediate object of perception. Descartes gives an example in his Second Meditation of looking at people on a street below a second story window. Here we would ordinarily say that we see the people even if the view of their faces is hidden by hats and the rest of their bodies by overcoats. Thus, we say that we see an object even when most of the object is hidden from view, and I think that similarly when we say that we see something three dimensional it doesn't follow that we are actually visually aware of anything that is not just a two dimensional surface.

  21. The key is the nature of the "enhancement" to which Bob alludes, because clearly there is a qualitative and perhaps quantifiable difference between depth perceived monocularly vs. binocularly. The rubrics of "knowing more than we see" and "the presumptive world" of perception come to mind from cognitive psychology, but in the example from Descartes, the difference is that from another angle we *can* see those same people's faces under their hats. Similarly in VS we never see the whole surface of any solid at once, only one "face" of it from any given vantage point (in the case of transparent solids it would be interesting to analyze in what sense we see two sides at once--obviously the Cubists were quite preoccupied with that).

    If the notion of tridimensionality is given or built up on the basis of visual perception as I am maintaining, then we need to understand why that is so, rather than simply claiming that visual tridimensionality is somehow a misnomer. Clearly Nicod was thinking along these lines, but perhaps he did not fully appreciate how fundamentally our notions of dimensionality depend upon visual experience, and are indeed derived from it. We need to retrace our steps to see how three-dimensional geometry would arise from the visual world of experience.

    I believe the different and more ordinary senses of "space" in OED may serve as a guide to that end.

  22. The question of how vision defines both the observer as well as the observed was treated extensively by J. J. Gibson in his theory of ecological optics. For one, he noted that in the optic flow field (of perspective transformations), all changes with respect to the observer as a "station point" in the perspective, i.e. a point-of-view that is inextricably unique to the observer. That station point never changes, except that it becomes gradually higher in the world around it as we become taller.

    Many years ago I wrote a paper that reviewed some of this called "Sensory Localization as the Basis of the Self-World Dichotomy" in which I claimed that the observing self is defined through perception and does not need to be inferred, as some philosophers and early child psychologists argued.

  23. I just had occasion to visit a website devoted to documenting a dispute on the nature of picture perception between Gibson and Ernst Gombrich that lasted for years, and it might be said that the twane shall never meet, though there was some rapprochement between their views with time.

    But I believe much of what Gibson said is relevant to Bob's analysis of the significance of photographic distortions vs. visual ones, and I think it possible that John's invoking ICT and virtual reality analogies with respect to the neurophysiological findings in need of an explanation might instead actually be the first neurophysiological evidence in support of Gibson's theory of perception--namely, that perceptual information consists not of pictures or images, but of *invariants of structure* in light and that the visual system is attuned to those invariants, and is not primarily an image-making device, as it was thought to be in the age of Newton. We have already discussed in the case of associative agnosia a person is left with an intact VS that paradoxically tells them virtually nothing about what they are seeing. Clearly there is much more to visual perception than generating VS alone, and that is what Gibson devoted his life to elucidating, his starting point apparently being Gestalt theory.

    A visual system poised to respond to discrepancies or novelty might well be one that primarily responds to deviations from invariants in structure, what?

    Here is a link to the Gibson-Gombrich dispute site:

  24. 1. Tom asks some ‘deep’ questions! What is the ontological status of the Self (“O”) and the “space” that some say separates it from the visual field? Is the idea of inner space a false “interpretation”? Can we get rid of the Self altogether? If we do not, do we fall prey to the dreaded homunculus fallacy? These questions need long answers. I am just skimming the surface here.
    But before starting on that I have a question to ask Tom. What do you mean by “interpretation”? Do you mean the conscious response that a subject makes when the experimenter asks in a test using interoception “Tell me what do you see.” Or do you mean the unconscious processes that the brain makes in organizing the higher stages of vision?
    Then I feel we need to consider a more basic question before tackling Tom’s particular questions: and that is “How (and why) does the brain construct phenomenal space in the first place?” Evidence that the brain constructs phenomenal space comes from studies of blind patients. People with retinal damage keep their visual fields even if they are always black. People with damage to the visual cortex lose their visual space altogether and see nothing. So how does the brain construct this entity that is missing in cortical blind people?

  25. 2. I think there is now enough scientific data to enable us to state that the “common sense” theory of Naïve Realism is false, that the alleged process of the ‘projection’ of sensations is a myth, and that ‘qualia” (sensations) are a part of a human organism.
    By introspective methods many psychologists believe that they have identified an entity known as ‘the visual field’ that is a flattish, limited, 2D surface on which images are cast that portray external objects or events (or what the brain computes these external objects and events most probably should be). This structure is said to function as the last step in the long and complex mechanical process of visual perception. The problem is that no such structure can be found in the brain, which is commonly believed to house the final step of visual perception. Moreover, there is no functional need for any such a structure in the final processes of perception in the brain of any organism. The massive computational processes in the brain and do all that a brain does in analyzing the sensory input and using this to direct behavior, without making any such structure as a visual field constructed out of neurons or their electromagnetic fields. However, if we deny that such a visual field exists, are we not reduced to Naïve Realism—for, pretending that qualia do not actually exist (as Eliminative Materialism holds) does not seem to me to present a satisfactory way out of this dilemma? Furthermore, if I introspectively examine my visual experience, say the black field when in the dark, can I really get away with the judgement that this black curtain before “me” is really composed of neurons or their electromagnetic fields?

  26. 3. Some shorter comments:
    Any theory of the “Self” needs to account for the fact that our “Self” is also the dreamer in our dreams. In a dream a Self is located in a collection of visual images extended in a species of phenomenal space. This does not seem to me to involve any “interpretations” of “usual exteroceptive categories”. But it might involve unconscious brain mechanisms whose spatio-temporal categories are programmed by ordinary perception.

    We can ask what follows if we abolish the concept of a Self? Can we do a thought experiment and envision an organism that has every aspect of phenomenal consciousness, like we have, except there is no Self? This would not be an ordinary Zombie because these lack all qualia too. But could there be an extreme zombie? If qualia are phenomenological, but not epistemological, intermediates in perception, as Edmond Wright proposes, then might such an organism might be conceivable? If so this might be an argument that Selves exist.

    A comment on the “infamous homunculus fallacy". This famous argument was first stated, I believe, by Descartes and followed by Gilbert Ryle, Francis Crick and many others since. However, Fodor (1981) says “This is, however, a bad argument. It assumes, quite without justification, that, if receiving information from the external environment requires having an image, recovering information from an image must require having an image too. But why should we assume that?” (p. 77: 1981 “Imagistic representation.” In N. Block, Ed. Imagery. New York, Bradford.)

  27. I pretty much agree with your diagnosis of the situation today John, but I guess the question arises as to what to do about it if anything. I think that the mind body problem may be solvable, with all of the hints available from different disciplines about what is going on, but we need a strategy for what to do next.

  28. John has reviewed much ground that we have already covered here, notwithstanding certain inferences and conclusions that I have called into question and objections I have raised. For example I countered that naive realism is not a theory and that "refuting" it may accomplished little, and may actually entail throwing out the proverbial baby with the wash. One reason I have given for that opinion is that most of the evidence to refute it comes from what I am calling “third person naïve realism,” which is synonymous with that form of scientific objectivity that eschews what has been miscalled “introspection” in the context of psychophysics. Rather I believe it to be the case that we—collectively—may have painted ourselves into a corner, and first and foremost, have not gone back to the very beginnings of the so-called “mind-body problem” to see if, by chance, it is really a pseudo-problem, perhaps an artifact of analysis.

    I have suggested the Schroedinger has provided an important clue, namely, that perhaps scientific epistemology has partly created the problem by explicitly claiming to omit sensations from the physical picture of the world, only to find that sensations have formed it after all as Mach made clear in his writings.

    Understanding that state of affairs might get us on a productive track, rather than remaining entangled in what strikes me as being a typical philosophical problem that has no obvious solution, because the problem has been misstated. Peter Hacker has made a convincing case that some problems relating to consciousness are entirely the making of philosophers. Should we perhaps take that diagnosis seriously?

  29. You are certainly right Bill that there are other possible positions out there, many of which have been developed in considerable detail in the past. Perhaps it is something like Russell's earlier neutral monism, whereby the physical world is construed as being comprised of sense data which you are sympathetic with?

  30. We should be guided by logical consistency: Either sensations are part of the physical (picture of the) world or they are not. If not, then a dualistic proposal such as John's is valid, necessitating as it does a completely separate space-time system from that of the physical universe.

    It is not clear to me that sensations are necessarily "mental," as they are now loosely said to be under the woolly notion of qualia, and apparently heedless of all the arguments against that view mounted by Mach and others a century ago. How "mental" is VS in the case of associative agnosia in which visual recognition has been disabled? What exactly is the nature of the "consciousness" manifest in VS if not a certain state of being or existence?

    In ordinary language we do not talk about the (perceived) world around us as being a product of mental activity, let alone our own, nor do we think of that world as being something "private" as Russell would have it. So how did it get to be mental and private?

    Perhaps then one fundamental error is to construe sensations as being mental, and Mach's view of them, which influenced Russell's, is indeed a kind of neutral monism in which sensations are only mental when they are associated with thought, or as Mach would say, "things of thought."

    Mach was an advocate of what Richard Avenarius called the "elimination of introjection" (not to be confused with Paul Churchland's "eliminative materialism"). On this view by a mistake of reasoning sensations have been introjected (internalized) by philosophers into the person, more specifically, the person's brain. The elimination of introjection simply entails recognizing that introjection is a fallacious non-existent process, the converse of the so-called "projection of sensations" that John critiqued in a paper specifically devoted to that some years ago.

  31. I once had an argument with a fellow graduate student in philosophy at Boston University on this topic. We both agreed that there was nothing "neutral" about Russell's "neutral monism" but then it turned out that we disagreed about whether the basic ontology was mental or physical; he held that it was physical and I held that it was mental. It sounds Bill like you would agree with my friend. I on the other hand am a physical realist about such primitive entities talked about by physicists as atoms, electrons, and photons, and I don't see any evidence that they are in any way either comprised of or comprise sensations.

  32. The strong claim that I now propose is that it is impossible to separate reality from what we call sensations, that reality is actually a property of sensations, just like color, space, and time. Through various mistakes in reasoning made in the course of developing the physical picture of the world--or more precisely, the physically realistic picture--that picture seems devoid of them, when they are really there at every turn. To reprise what Schroedinger quoted from Democritus in the imaginery dialog between the intellect and the senses about what is "real": "'Ostensibly there is colour, ostensiby sweetness, ostensibly bitterness, actually only atoms and the void,' to which the senses retort: 'Poor intellect, do you hope to defeat us while from us you borrow your evidence? Your victory is your defeat.'"

    As Schroedinger notes, "The observer's senses have to step in eventually. The most careful record, when not inspected, tells us nothing. So we come back to this strange state of affairs. While the direct sensual perception of the phenomenon tells us nothing as to its objective physical nature (or what we usually call so) and has to be discarded from the outset as a source of information, yet the theoretical picture we obtain eventually rests entirely on a complicated array of various informations, all obtained by direct sensual perception." How is that conclusion to be refuted?

  33. I keep thinking of a remark made by Sherlock Holmes in the "Adventure of the Priory School" when he is stumped at one point: "It *is* impossible as I have stated it, and therefore I must in some respect have stated it wrong." He then asks Watson, "Can you suggest any fallacy?"

  34. Yes, common sense "naive realism" is much too amorphous to be called a theory! But naive realism is also used as a term by many worthy people to describe a theory of perception. We need to be sure that we do not mix these up.
    May I suggest that we have taken this about as far as it will go and we should agree to differ on certain points? I agree with Bob that "we need a strategy for what to do next". I suggest a focus on some specific points that need our critical survey e.g.
    1. Many neuroscientists, especially Francis Crick, have looked for a specific anatomical locus in the brain for "consciousness" e.g. the reticular nucleus of the thalamus, the cholinergic nuclei of the midbrain and, lately the claustrum. Recent evidence suggests that this is a mistaken strategy and "consciousness' is a function of the entire cortex......
    2. Most neuroscientists focus entirely on the electrical or electromagnetic functions of the brain as the 'source' of consciousness, The work of Hyden, Sickles (thank you Bill for steering his ms my way) and others suggests that other systems may be involved too, such as protein synthesis. This may explain two hitherto mysterious features of brain function — the monumentally complicated signaling pathways neurons use and the vast array of neuromodulators, particularly peptides. To this we can add the new discoveries about the active role of glia in the brain. Why all this extravagance? Is this to do, in part, with the emotions? We know very little about the neural bases of emotions. What goes on in the brain when a sudden surge of emotion affects every system? What is going on in the brain during an attack of mania, when every cognitive function is powerfully affected in a particular way? I have suggested that the brain's adrenergic system based on the C1-C3 nuclei on the medulla may play a key role in this. Other systems may be involved as well, and so on. There is much to dig up here. When we have more knowledge of this basic material we may be in a better position to tackle higher level problems such as how the brain 'creates' space.

  35. I could not agree with John more for the need to focus on certain key issues that are relevant to our topic, but I don't think elucidating the neural correlates of emotions will be of much use in understanding the structure of VS.

    When I said in my penultimate comment above that it might be worth entertaining the possibility that what we ordinarily call "reality" is a property of sensations, I meant that quite seriously, and for a very specific reason: The existence of the psychiatric condition known as derealization, in which the sufferer feels that the world is no longer real. If the "sense" of reality can actually disappear from our experience in this way, so that the contents of VS no longer seems "real," what remains? If VS can seem unreal, what about physical reality? Can it too be derealized?

    We tend to think of reality in black and white terms and it may be that the assumption that things can be more or less real needs to be examined. Why should we ascribe greater reality to the brain than to our own experience that is subject to losing its sense of reality?

    Need I say that so-called "lucid dreaming" can seem very real to a lucid dreamer, thus its name. This is why I am proposing that perhaps reality is actually a certain condition that is achieved in some way, or created. Again, I offer this wary that I may be conflating different senses of "reality" in what I have just outlined!

  36. Given the nature of this blog format it is possible to have several reasonably detailed discussions on separate topics occur at the same time. I would be happy to contribute to any of them if I think that I can make an intelligent input. I think that it would be good to move on to some other topics though.

  37. I agree with Bob here, but that different threads need to be explored in parallel, because they may be interrelated. In saying that, I hope all of you share my commitment to be prepared to question every assumption, and not just take things on faith, regardless of how fundamental certain assumptions may seem.

    In this regard, I am reminded of something I sent John some months ago from Whitehead's "Science and the Modern World" in which he writes: "Science has never shaken off the impress of its origin in the historical revolt of the later Renaissance. It has remained predominantly an anti-rationalistic movement, based upon a naive faith.... Science repudiates philosophy. In other words, it has never cared to justify its faith or explain its meanings; and has been blindly indifferent to its refutation by Hume."

    Something to think about!

  38. Correction: That's "blandly indifferent," not "blindly" (Freudian slip!)

  39. Bill asks " If the "sense" of reality can actually disappear from our experience in this way, so that the contents of VS no longer seems "real," what remains?"
    I can speak as a expert witness here as I experienced a short, sharp (and terrifying) attack of derealization during my first mescalin experiment in 1951. "What remains" is the VS looking exactly as it did before—but none of it is real. The unreality is not produced by some subtle change in the ways things look. It is the complete disappearance of a property of qualia that we do not normally notice as a property at all—their reality. I felt not just ordinary terror but abysmal angst. So I hurridly swallowed the antidote—sodium succinate. Some 20 minutes later reality returned in an instant and I leant forward and grabbed Edward Osborn's now real foot with enormous relief. Curiously enough some months later during my second mescaline experiment none of this recurred.
    So what are we to make of the kindred phenomenon of depersonalization? Here it is not the visual field that becomes unreal but the Self. Some people like Dennett and Hume deny the existence of the Self. The phenomenon of derealization argues powerfully for the reality of the Self—for how can one lose that which one does not have??

  40. The derealization that John experienced in his first mescaline experience lends support to my suggestion that reality itself is qualitative in nature, as is its antithesis unreality. The Leipzig Ganzheit psychologists (Felix Krueger et al) would call it a "quality of the whole" ("Ganzqualität"). They argued that Gestalten (configurations) are a special case of such "whole qualities." These qualities are indescriminantly being lumped together today under the vague catchall of "qualia," probably because no one is even aware of the tremendous amount of conceptual and experimental work that was done in Leipzig psychology before the rise of Nationalisocialism. Sometimes in reading about Q.M. I think that they are describing conditions that are not fully real in any ordinary sense, but they don't come out and say that.

    With the converse, depersonalization, though the self is perceived as unreal, there is still something like an ego that is aware of that fact. Schilder talked about the "central ego" (Zentral Ich) as being at the core of the personality, and within the bodily self (Körper Ich), and as with ketamine, there is a sense in which one experiences one's bodily self as no longer belonging to oneself (paradoxical as that may sound on the face of it).

  41. Answer to John Smythies. Actually, I was not considering the “inner space” to be an illusion (and I think that you, John, gave really good reasons –cf: The Walls of Plato’s cave, pp.91-103- to admit the existence of an “inner space” and to reject the various strategies that have been striving to explain it away or to depict it as abstract phase space, etc), but I was merely expressing some doubts about its structure being three-dimensional and bi-polarized (with on one “side” an observer and on the other an observed “screen” with a kind of distance between the two). In short, I concur with you in admitting the existence of an “inner space” (and I agree that this “phenomenal space” cannot be strictly identical with the intra-cranial neurophysiological space), but I have some doubts about its structure and about the status of the so-called “O” and about the existence of a real distance/separation between this “O”’s vantage point and the visual “screen”, so to say.
    By “interpretation” (in the above-mentioned context of conscious inspection of our visual field) what I meant was : an a posteriori and potentially misleading verbal depiction of the experiential content that was a priori (ie: before any perceptual categorization, and before any use of verbal re-categorization) inspected in the visual field. I mean: don’t you think that it is possible that the subjects described an Observer-Observed distinction because they were misled into thinking that the binocular ‘depth effect’ in the (either non-imaginal or imaginal) visual scenery of our visual field (cf the analogy with the TV screen) necessarily implies the presence of two real “poles” or points of reference to account for the formation of a perspective (on the “flat” visual screen itself) because the subjects (as everyone of us) were taught (in infancy, while they were still learning the acceptable way to refer to the categorized “objects” in terms of subject-object distance, especially when dealing with “exteroceptive” sense data) that when something appears in their visual field it becomes some thing (a “thing”, an “object”) by contradistinction with a subject being able to either designate it (for others) or to reach for it. What I am suggesting is that maybe this learnt descriptive distinction (I think that it is “externally” learnt in the case of the visual field thanks to the intervention and teaching of others, but that it is “internally” discovered in the case of the body image, through the proprioceptive experience of motor effort and resistance) was imported, as an ingrained habit of thought, into a domain in which it is not appropriate (ie: when considering the visual field itself, without any judgments about any referential “object” as ob-jectum). What I sometimes observe when I carefully inspect the micro-texture of my own visual field is an array of faint or bright tiny pixel-like dots (tiny squares I would say) that first are grouped together in non-imaginal geometrical patterns and then form varying moving shapes with contrasting colors, and then identifiable figurative outlines, etc… but all that seems to occur on the same “level” and in a kind of continuum, without any sudden “gap” or “void” forming between an “inner” observer and its private “screen”.

  42. This comment has been removed by the author.

  43. This comment has been removed by the author.

  44. That being said, since you mentioned David Hume, I must now say that, while I really cannot discriminate the presence of a “self” in visual experience, I do not yet share Hume’s criticism about the illusory nature of the “self” and about its being an abstraction used to unify and tie together the various synchronically heterogenous and diachronically discontinuous conscious events occurring in introspection. I even strongly disagree with him on that point. I have intellectual affinities with Maine de Biran’s position, rather : the conscious “self-other duality” first (ontogenetically speaking) appears with the motor experience of effort and resistance, and then is extended to mental varieties of effort (in the “flow of consciousness” some thoughts or mental images are controllable while some other occur without being consciously willed). So, as you can see, I fully accept the fundamental and developmentally essential “self-other duality”, but, if we admit that psychology should not be equated with ontology, then the question is : does this “controller/observer-controlled/observed duality” emerge from a state of fusion in an underlying continuous unitary space, or does it reveal/unveil the existence of a latent (mute, tacit but objective) individuated substrate different than the “phenomenal space” of its “sense data”?

  45. William Rosar aptly reminded us of Schilder’s distinction between the “Zentral Ich” and the “Körper Ich” (we can find similar distinctions in Thomas Metzinger’s “phenomenal self-model”, by the way), and I think that indeed, the “self” is a complex multi-layered construction(cf: Metzinger, Jeannerod, etc). Thus the question is : what is the most fundamental layer, the elementary sine qua non component for perceptual self-awareness and then for reflective self-consciousness? Maybe development psychology and comparative cognitive ethology can help us here : even before mirror visual self-recognition (that only some evolved mammals and birds can manage) there are many more primitive developmental steps including situated embodiment, controlled agency,etc… My question would be : can we have an inspection of the visual field without having the impression that “something is doing the inspection”? If yes, then the Observer-Observed duality is not intrinsically a part of primitive visual experience and it certainly appears much later.

  46. The depersonalization experiences that John Smythies mentions clearly point towards the idea that, indeed, the “self” isn’t nothing in our conscious experience (as some radical eliminativists would want it to be). “For how can one lose that which one does not have?”: I agree on that point. But saying that it is not nothing does not necessarily entail that it must be some thing, that is to say that it must be a thing distinct from other things…As John explained: resorting to the existence of an homunculus is as useless as it would be useless, in a description of the images appearing on a TV screen, to say that one image is a separated “subject” and another is a separated “object”. Maybe the “self” should be viewed as an image among other images, as a quale among other qualia (I agree that the term “qualia” can sometimes be too vague to be deemed useful). It would be interesting to investigate the way the various fields of perception look like during the depersonalization experiences and the way the emotional feelings manifest themselves : even slight differences in sensory background or in emotional undertones could give us indications about what is lacking. I suspect that it cannot be an entirely noetic/intellectual process with no phenomenal content that gives rise to the experience of what we call the “self”. There is something more primitive. There certainly is a “way it is like”, a “phenomenality”, a specific “quale”, when the “self-other distinction” is experienced, and thus shouldn’t it be possible to isolate it as the ‘missing quale’ and to analyze what it adds to the phenomenal properties of our sensory and emotional experiences, and especially for our visual experience?

  47. I think that it might help the debate that is going on here to define what senses of 'real' and 'unreal' are being used. Sometimes they are used to refer to whether something exists or not, but sometimes, as with a hallucination, to refer to the existence or non-existence of a physical object of perception. There also appears to be an issue of whether a self exists as a separate entity apart from a phenomenal space of which we are aware. In order to avoid regresses, I think that we are better off not claiming that they are separate entities. I also think that it would be good to set up a separate section if we are going to continue to debate these subjects.

  48. I agree that most of the '3 D space' in ordinary perception is a virtual space like the '3D space' in the case of the new 3D TV (which is all really on a 2D surface). One can view this surface in ourselves very clearly while observing the stroboscopic patterns, which are patterns spatially distributed on a very clear 2D surface in front of 'us'. The reason "I" feels set back from this surface may be that I can attend to different parts of the surface—the center, periphery and points anywhere on it as though I was set above it. Whereas, if I was really located in the plane of the patterns (no 'space' between), I could not do this any more than a Flatlander can see his plane world from 'above'. Perhaps this sense of 'separateness' of the Self derives as an 'illusion' from this basic attentional mechanism. That is "I" may 'feel' above the plane, as that is the only way such differential attention could be obtained. On the other hand, ancient Hindu (highly sophisticated) psychology equates the Self (Atman) with the Brahman (which would need, as Bob says another section to discuss.
    To answer Bob, I do not think postulating a separate Self involves the danger of a regress because of Fodor's powerful argument against the "little green man in the head' argument that I have already quoted.

  49. There is more than one regress argument to be wary of here. Besides the one about questioning the need for additional eyes in the head raised by Descartes in his Dioptic, Gilbert Ryle gives the following one in Ch. 7 of The Concept of the Mind in the context of a representational account of seeing a horse race."This means that having a glimpse of a horse race is explained in terms of something else, the patchwork of colors. But if having a glimpse of a horserace entails having at lease one sensation , then having a glimpse of color patches must again involve having at least one appropriate sensation , which in its turn must again be analyzed into the sensing of yet an earlier sensum, and so on forever. At each move having a sensation is construed as a sortof espying of a particular something, sometimes gravely called a 'sensible object,' and at each move this espying must involve the having of a sensation. Do you think that Fodor can adequately handle this type of a phenomenal regress John?

  50. First in response to Bob, I gave the caveat that different senses of "real" and "reality" might be conflated with respect to derealization. However whether these senses are mutually exclusive is another matter. One sense, for example, might be contingent upon another. I doubt the origin of the word "reality" was in determining whether things existed or not, and that is probably a later sense of the word than whether or not an existing state of affairs is the "reality" of the matter, etc. Physical science seems only to use the first mentioned sense in constructing a physical ontology, not the latter (except perhaps informally)

  51. With regard to the use of "virtual" in the context of VS, I think it does not apply for the reason I have given before: "Virtual" only makes sense in the context in which one can compare the "virtual" simulation of something with the "real" thing. We cannot do that in the case of VS, so as to compare it with something else.

    Exactly how to determine the dimensionality of VS is, of course, one of the main topics of this blog, but thus far, no criteria or method has been agreed upon for how that is to be accomplished. I have maintained that our conception of dimensionality is based on our visual experience, so logically it should be possible to reduce those conceptions back to their visual origins, much as Einstein sought a physical interpretation of geometry in what was called "physical geometry," e.g., a straight line = a ray of light in a vacuum (this is an example from what later became the basis of so-called "geometrodynamics").

    We thus need a comparable procedure for equating straightness with visual things, and I have already suggested that something that "looks" straight is a logical candidate, since it is indeed often the case that things that look straight also *are* straight by other criteria.

    So like physics needed a physical interpretation of geometry, we need a perceptual interpretation of geometry. Nicod made a start towards that, as I have said already more than once.

  52. Ryle's rather silly analysis just points up what I have stressed before, namely, that the *act* of perceiving is itself perceived, most often as described as "perceptual behavior." Gibsons also said that one can, a la a reversible figure, study visual space as perspective, or alternatively but not at the same time, as the perception of objects.

  53. When I talked about the 'virtual reality' component of the VF in my recent paper in "Perception", I was referring to the (fluctuating) part of the VF that has been demonstrated to be provided by the memory and prediction mechanisms of the brain, not by the retina via the LGN and V1. The latter supplies the 'reality' component.

  54. If the "fluctating" part of the visual field has been shown to be provided by memory, does the percipient knowingly experience those portions as conscious visual memory? Otherwise you must be referring to "unconscious" memory, John, which is a much less appealing proposition.

    What has been identified as a "prediction mechanism in the brain"? The rather woolly notion of virtual reality is just a way of "hyping" what is ordinarily known as simulation (as in the USAF's original flight simulator training programs). In theory it is only a surrogate for something that could potentially be really experienced. If you are saying, that the memory is supplying conscious visual data that could ortherwise be experienced like the "filling in" of a scotomas based on surrounding visual field data, that seems reasonable--except that we would not call it memory in any ordinary sense.

    The principal difficulty is that most displays are either wholly virtual, or televisual, not partly both. Introducing "virtual" data might potentially also be a source of error in the visual system, so that might be an undesirable a trade off to take into consideration.

  55. This comment has been removed by the author.

  56. 2. The next series of experiments was reported by Kovács, Papathomas, Yand and Fehér (1996). These workers took two photographs, one of a monkey's face and the second of a leafy tropical jungle. They converted these into two pastiches each composed of portions of each photo, so that in the location where one photo showed part of the monkey's face the other showed leafy jungle. Then each pastiche was shown separately to each retina, so that retinal rivalry occurred. Under these circumstances, the subject did not see what was actually there—that is the two pastiches alternating—but rather a complete monkey face alternating with a complete leafy jungle. Clearly the brain had suppressed the improbable mixed pastiche in favor of what it was familiar with (and thus computed what was more probable). Many other experiments, based on stimuli such as moving plaid patterns, have shown this phenomenon where the perception of an improbable input is suppressed by the brain, and replaced with the perception of what it computes to be more probable ones.
    Even more crucial experiments (Kleiser, Seitz and Krekelberg 2004) have shown that, during a saccade (rapid movement of the eye), information coming from the eye is suppressed, and what we see is largely virtual reality created by the brain from processed memory. [Note: when the memory appears as a memory image in consciousness it is of course a conscious memory. But the processing of ALL memories is unconscious].

  57. This is part 1 of this post! Should be put above 2 !

    1. The best way to answer Bill's comments is to present the evidence that both the brain and modern digital TV use mechanisms involving virtual reality:
    (based on my recent paper in "Perception").

    "The first body of evidence that supports a TV theory of perception came from a series of experiments that suggest that we do not perceive what is actually ‘out there’ but what the brain computes is most probably ‘out there’, or, as Crick (1994) put it, “What you see is not what is really there; it is what your brain believes [substitute ‘computes’] is there.” Ramachandran and his colleagues carried out the first of these experiments. They showed that scotomata—the blind patches caused by local injury—are not perceived as patches of nothing, but that the brain fills them in with a continuation of whatever pattern surrounds them.

  58. 3. These authors expressed it thus:
    “When you look into a mirror and move your eyes left to right, you will see that you cannot observe your own eye movements. This demonstrates the phenomenon of saccadic suppression: during saccadic eye movements, visual sensitivity is much reduced.”
    In other words, 'filling in' has a temporal as well as a spatial dimension. This function is based on a widespread neural network, that includes the superior colliculus and parts of the thalamus. The purpose of this would appear to be to suppress the violent swings in the visual input (the world swirling round one) and resulting severe vertigo that would otherwise result. Since saccades occur very frequently, this means that virtual reality plays a major role in normal everyday seeing.
    We now also know something of the mechanism by which the brain effects this function of mixing reality and virtual reality in visual perception. In this the brain uses technology already worked out by television engineers for their own purposes. These engineers found out that the efficiency of digital television transmission can be increased in the following manner. The scenes televised consist of a focus of attention (foreground), where the action is, plus a lot of background, where nothing much happens. The computations on which digital television is based are expensive in terms of computational time and monetary cost. To reduce this, the computations are concentrated on the foreground (’reality’), as transmitted by the TV camera, and much of the background is provided from memory combined with computations on what is likely to happen next, given the previous history of the program (‘virtual reality’) provided by the computers in the system. The former is accurate but expensive: the latter is relatively cheap but ‘fuzzy’. The art of TV engineering is to provide the optimum mixture of these two processes.

  59. 4. Interestingly enough the brain does much the same. The input from the retina that carries information about ‘reality’ is carried by the optic nerve and optic radiation to the visual cortex where the axons terminate in layer IV. The input from the brain’s neurocomputationally active memory and prediction systems, that mediate the virtual reality component of visual perception, is carried by cortico-cortical connections to the visual cortex that terminate in layers I and II. [In other words the higher visual cortex is continually monitoring the visual input and looking for the unexpected. If the brain regards what comes in as too improbable it may suppress the retinal input and supply consciousness with its estimate of what should have been seen—like the complete monkey picture and complete leaf picture described above. During a saccade the entire retinal input may be momentarily suppressed and replaced from the cortex].

  60. 5. The balance of these two systems is mediated by the neuromodulator acetylcholine. The cell bodies of these cholinergic neurons are located in the Nucleus Basalis of Meynart that is located in the basal forebrain. The axons of these neurons project to the cortex, where they activate stimulatory nicotinic receptors in layer IV, and also inhibitory muscarinic receptors in layers I and II. Now, when nothing much interesting is happening in the environment, there is a low level of activity in the Nucleus Basalis, in which case the cortico-cortical input to the visual cortex is active mediating virtual reality. Then, when a new stimulus is received that indicates that something important may be happening outside (for example “predator!”), this nucleus is activated, releasing acetylcholine at its axon terminals. The projection to layer IV then activates stimulatory nicotinic receptors on the bodies of those pyramidal neurons that receive the retinal input, which is thereby promoted. The simultaneous cholinergic projection to layers I and II stimulates inhibitory muscarinic receptors on the bodies of pyramidal neurons that receive input from cortico-cortical projections that mediate ‘virtual reality’, which is thereby inhibited. One function of all thi, during saccades,is to prevent the severe vertigo that would otherwise result. Thus the brain can now expend its energies in analyzing the new stimulus for salience. It seems to me that Naive Realism (NR) has difficulties in accounting for this data. It is implausible to claim that we see external objects directly when our eyes are still, but indirectly when we execute a saccade."

    Members of this blog are invited to see this for themselves by doing the saccade experiment themselves—first looking at their own eyes in the manner described and then looking at the eyes of someone beside you who is doing the same saccade observation.

  61. If the brain is engaged in virtual reality, it is also engaged in virtual naive realism, because as I have already noted, we are unable to compare the external physical world with our perception of it, whether generated on the basis of retinal stimulation, or generated internally by the brain, which is what would be required to make virtual reality a valid analogy.

    Again, there is a lapse into what I have been calling "third person naive realism" in John's account above, because there is a presumption that the experimenter is seeing those individual pictures presented to the retinas directly, whereas the experimental subject is not. Invoking the notions of "more probable" and "computation" presupposes that the brain both computes and judges probabilities, and it is just these claims that I have been at pains to deny, for the reasons already argued by Raymond Tallis in his little book, "Why the Brain is Not a Computer," i.e., that they entail anthropomorphism of the brain, turning it into a kind of homunculus.

    I wonder if alternative explanations could be found for the visual phenomena and neurophysiology above, but have not been considered? If the brain's main task is regulating the body and bodily behavior, then these functions could be interfering with direct perception in order to maintain balance and regulate levels of physiological arousal, respectively. In other words, there is something to be said for behaviorial neurobiology!

  62. The facts are that the visual field experienced is being continually filled with material that comes from two sources: (1) the retina via the LGN to visual cortex layer IV and (2) the memory banks in higher cortex to visual cortex layers II-III. These are seamlessly blended into the one coherent picture we experience. Most of the scientists working on this phenomenon call it ''virtual reality". But I would be receptive to any suitable alternative name you care to suggest that does not carry the metaphysical overtones that you do not like.
    The well-known skeptical argument (that we cannot know what the external events are because we cannot compare this final picture with external reality because we cannot put them side-by-side to compare them) is not, I suggest, relevant to the question of the origins of the picture itself.
    Furthermore the Skeptical argument presumes that the only way we know that A is a picture of B is to put them side by side. I do not think that we have to do this. If I put an object A before you and 1,000 different pictures of an object that I did not disclose what it was of, would you, adhering to your principle that two objects must be compared side by side, be unable to find out that the I,000 pictures were all of A? Avrum Stroll has written extensively on the topic that we should not take skeptical arguments to unreasonable extremes.

  63. Good point, Bill, about the role of these brain mechanisms in modulating arousal. That is exactly part of what they do. Yu, A.J. and Dayan, P. (2002) (Acetylcholine in cortical inference. Neural Networks. XV, 719-30), who discovered this cholinergic mechanism, linked it, as you can see from the title of their paper, to inference. In my subsequent paper I linked it to information compression and virtual reality mechanisms instead. But these mechanisms are also involved in arousal, and the direction of attention with repercussions in many physiological systems.

  64. I was not arguing from skepticism, John, but criticising the argument from analogy, the analogy being between the technology imaging media known as "virtual reality" and what the visual system is doing. The difference between "virtual reality" imaging technology and vision is that in the case of virtual reality one can, in principle at least, compare what it shows and what we might normally perceive "in reality." That is not possible in ordinary vision.

    The best example of that is the original purpose of virtual reality, i.e., simulation software used for training pilots. The purpose of that software was to train pilots without the dangers of actually flying. That is what was meant by "virtual," because it only simulated the visual part of the experience, as a surrogate for an experience that they could otherwise actually have.

    Where the analogy between that and ordinary vision breaks down is that unlike the pilot training on flight simulation software who could in principle compare it with really being in the cockpit of an airplane he was flying, or learning to fly, as visual percipients we have no means of making that kind of comparison with external reality, nor of knowing which part would be "virtual" versus non-virtual--something that is evidently only apparent to the vision scientist--again, "third person naive realism."

    This is not skepticism but just showing the pitfalls of taking analogies too far and too literally. There is a lot of this in science because scientists do not observe careful reasoning when it comes to the use of analogies. That is the only reason for Raymond Tallis's pithy little lexicon, "Why the Brain is Not a Computer: A Pocket Lexicon of Neuromythology"--the "neuromythology" in question being fallacious reasoning that arises from taking technological analogies too literally.

  65. Back to a previous subject for a moment. Can the brain be said to compute or do only people compute? If we take the case of a reflex knee jerk one can hardly say that the person is responsible for the movement of the knee. His spinal cord does that. The movement is entirely subconscious. Likewise the computations ascribed to the brain by neuroscientists in their description of how e.g. perception works is also entirely subconscious (as when the brain estimates the probability that an incoming sensory message is veridical). Therefore the person cannot be held responsible for them. A person is only morally responsible for his voluntary acts. A schizophrenic who kills someone because of his paranoid delusions is held to be "not guilty by reason of insanity". The paranoid delusions are brought about by faulty brain computations are they not?
    Returning to your points above. You say
    "as visual percipients we have no means of making that kind of comparison with external reality, nor of knowing which part would be "virtual" versus non-virtual--something that is evidently only apparent to the vision scientist--again, "third person naive realism."

    I suggest that you do have a way to find out "what part is virtual" and "which part is non-virtual" that does not depend on having a vision scientist beside you telling you what is going on ("third person naive realism"). That is when you perform the simple experiment I described earlier—watching your eyes in a mirror when you perform a saccade. You will notice that you cannot observe the movement at all. That is the virtual reality. The rest of your face that you can see is normal—that is reality. You can also see in the mirror perfectly well the eye movements of a person standing beside you executing saccades—that is also reality.

  66. Yes, ultimately it is people who perform computations, not brains. To the extent that we have been able to automate the process of computing with machines, computers can carry out our computations for us. But computers are created by people, not the other way around. Attributing "computations" to the brain is an example of the fallacy of division, in which one mistakenly attributes to a part a property or activity of a whole, the part being the brain in this instance, the whole being a person.

    "Reflex" is a technical term that is really only used to describe involuntary movement, so that is not at issue. People obviously are not voluntarily schizophrenic. But it is people who are schizophrenic, not brains.

    The failure to see eye movements in a mirror may have as much to do with the limits of foveal vision being able to "catch" them moving, because one is clearly conscious of moving one's eyes. If this reflected some sort of "virtual reality," one would be able to compare being able to actually see one's eyes moving in a mirror. But one cannot do that. Surely that is not sufficient to support the strong claim (hypothesis by way of analogy) that the brain is involved in virtual reality simulation activity?

    I would also pose a more general question: What is the appeal of likening human behavior and brain function to the activity of man-made machines and technology?

  67. Bill, You will have an uphill job convincing the vast majority of neuroscientists that brains do not compute, since the word "neurocomputation" is strongly engrained in the entire discipline, and is likely to stay that way. Its use does not seem to me to impede the continual flow of new discoveries about how the brain works made by the neuroscientists using this term. Even if humans are ultimately responsible for the computations computers make (by writing the soft ware), humans are not responsible for organizing whatever the brain does—the brain does that itself.
    Take for example the newly discovered cliques in the (mouse) brain. One of these is composed of distributed neurons in the hippocampus that fire whenever a nest is noticed by the mouse. The nest may be of any shape, size, colour, made of any material, located anywhere—so long that it can function as a nest. In other words, one could say, a clique recognises concepts. What word would you use to describe what the clique is doing? Does it recognise a nest? Or does it enable the mouse to recognise a nest? Or both? What the clique is doing, of course, is firing whenever a nest is seen. The clique then sends a message "nest present" to the premotor cortex that plans movements, as well as to motivational, sleep and other centers. The whole distributed network than makes a decision (what else could you call it?)— "climb it nest" (or perhaps "ignore nest") and send that to the motor cortex. Behaviour results. Perhaps we should develop a new terminology "recognise (b)" for what the brain does and "recognise (m)" for what the organism (in this case a mouse) does?

    I do not think that your explanation for failing to see one's own saccadic eye movements in a mirror explains why you can see these same eye movements perfectly well in a person standing beside you executing similar saccades.
    (Note that the explanation I gave is not mine—it is that of the experimenters).

  68. Fortunately in such cases one can choose their battles, and convincing neuroscience that ascribing computation to the brain is not one I wish to wage. Besides it has already been refuted on logical grounds in Tallis's little book, so he has already done the job, if anyone has paid attention to him (probably they have because the book has gone through several printings and republications). So it may take mainstream neuroscience a while to notice, though.

    Both the beauty of neurobiology and its pobtential as a source of fallacious reasoning is that one can correlate brain events with behavior, and then claim that the behavior is being caused by a certain area or nucleus, the so-called "localization of function."

    A century ago philosopher T.K. Österreich in his book "Die phänomenologie des ich in ihren grundproblemen" (The Phenomenology of the Self and its Fundamental Problems) already employed the term "pseudo-localization" in a critical way with reference to our general tendency to localize functions and feelings in certain parts of the body, even if they are not really (experienced) there. The chakras are perhaps the best example of that tendency, but neuroscience now has a brain swarming with the little homunculi like those that William James lampooned in his critique of "faculties" in his "Principles of Psychology"--over a century ago.

    As for "cliques" in mouse brains (a perfectly silly term), why not just be content to say that activity in the neuronal complex in question is correlated with nesting behavior? Why must one then go on to say that recognition and concepts are also there? At best that would only meet the criteria of an operational definition, but certainly does constitute proof of there being concepts or a sense of recognition "localized" there, much as the wishful thinking of researchers would like that.

    Obviously one cannot stop people from spouting rubbish, and as long as it makes no difference, they have no incentive to stop. It was only at the subatomic level that it was found that Newtonian mechanics seemed to break down and require a new theory. Perhaps neuroscience has not yet reached such an impasse--that and possibly there being something of the "Emperor's New Clothes" going on as well? That is to say, neuroscience exaggerates and overstates its claims routinely, but no one seems astute enough to notice!

  69. Neuroscience should be evaluated by the same standards used by the "skeptics" in evaluating psi, such as the Amazing Randi, who is a professional magician. A wealthy patron should be approached to offer a million dollar award to anyone who can demonstrate conclusively that any brain, human or other, is performing even one computation according to the correct definition of the term given in Raymond Tallis's book.

  70. A brief memoir on the term "computation" as used in cognitive neuroscience: When I started in cognitive psychology as an undergraduate in the mid-1970s, "computation" was not used as it is now, even though much computer modeling of cognitive (mental) activity was being done already.

    There was much talk of the "system" and "processing" (as in "stages" thereof) but mainly in the context of "information processing," especially due to the influence of a text by David Rumelhart (then UCSD), "Introduction to Human Information Processing," an approach strongly promoted also by his research cohort Donald Norman (then UCSD). Notice the phrase is "human information processing," not "brain information processing."

    Even the term "information" was not used in the technical way it is defined in information theory, but more in the common sense way. So there was talk of "people" computing in this context (and Rumelhart's book only uses the word "compute" on one page), but not brains.

    How the shift from people computing to brains computing occurred is something of a mystery to me because, as I have stated, it entails falling prey to the fallacy of division, and of course, taking an analogy literally. My presumption is that the computer industry, which funds much cog sci research, has heavily promoted the myth that computers are brains, and therefore, that brains are computers. This is the reason for the subtitle of Tallis's book "A Pocket Leixon of Neuromythology."

    Therefore the transition from people computing to the analogy that brains are computing being taken literally seems to involve either (1) a fallacy arising from analogical reasoning and a corresponding one involving the fallacy of division (2)the use of "transferred epithets" (as Tallis says) (3) "machinizing" the brain (Tallis's term) and/or (4) propaganda from the computer industry.

    Though he uses the term "computing" in this way loosely, Rama does not think the brain is literally doing computations and has as much as said so in lectures. One key difference is perhaps that whereas brains are alive, computers are not.

    My concern is what the use of the analogy *conceals*, more than what it explains--which is very little, other than to say that the brain is like a person's mind (N.B. another analogy), in that "it" recognize things, thinks, forms concepts, analyzes, synthesizes, etc., and that there are special compartments that do this (more or less like the old concept in Scholastic psychology of "faculties").

  71. Oddly enough, it does not appear that those invoking the inability to see one's eyes moving in a mirror seemed as evidence of saccaddic masking and the ability to see the eye movements of other's did not take into consideration the effect of foveation. If one gets very close to a mirror and move's ones eyes very slowly, you can see them moving. Moving back, one can really not even see one's own eyes at all, let alone their movement, because one's gaze has shifted out of a foveal region that would make it possible to see one's eyes moving. Naturally this would be different in observing someone else's eyes move, because one is able to foveate on the other person's eyes. There is no discussion of that rather crucial difference in the literature (at least of which I am aware).

  72. Re cliques. The interesting thing about a clique is not that, as you say, it can be regarded legitimately as a device for controlling behaviour, but that it is responding to stimuli that are united only by their collected ability to come from an arrangement that forms a nest. The word "nest" is ordinarily regarded as a word describing a concept, but, as you say, that may not be important. The real interest lies in the problem of how does the visual cortex have neurons in it that can respond only when a nest is 'outside'? One can see how increased convergance from an increasing number of stimulus-selective hierarchically-arranged neurons can produce neurons that react only when, say, a spotted yellow, round and moving jerkily from left to right object is present. Or even when only one particular face is present. But there are no constraints on the individualization of the stimuli from what will constitute a nest. So how is it done? The authors of this paper do not enlarge on this point.

    Re your remarks about the folly of ascribing 'computation' to brains (do you include Sejnowski and Churchland's "The Computational Brain" in this category?) I really need to read Raymond Tallis's book before developing my arguments further.
    I share your concern about the unfortunate zombification of humans that results from remarks from people like Marvin Minski's famous "The brain is a computer made of meat". But I feel the cure is not to deny that the brain computes in some sense but to deny that it does nothing but compute.
    Re saccades. The saccade effect does not operate if you move your eyes slowly. At the distance used in this experiment I can see my eyes very clearly and I can easily tell if they are moving or not (if slow). So I do notl to understand your argument. Nor do I understand why you cannot 'foveate' on your own eyes when you say you can 'foveate' on someone else's eyes further away. The authors of this paper describe further how momentary stimuli that pop up on the computer screen around the target of the saccade cannot be seen either. This also supports a momentary shut down of retinal input during the saccade. Finally the authors give a plausible reason why evolution should have produced such a mechanism. Without it one would be overcome with extreme vertigo, with the room swirling round every time, one executed a saccade.

  73. Of course, "cliques" ordinarily refer to people. To what do they refer to here? Homunculi, of course.

    My previous remarks about neurobiology were meant to refer to *behaviorial* neurobiology, which is really a form of behaviorism that eschews "introspection" and the "phenomenology" of experience, as Rama likes to classify such trivial things as the totality of consciousness experience.

    Trying to infer cognitive processes--even rudimentary ones--from behavior involves surpassing an impasse that seems insuperable without some form of communication to report experience--something of which our avian brethren seem incapable. So, instead, the behaviorist (neurobiologist) "asks" their brains, as they like to say--of course, metaphorically.

    IMO about all one can say is that the "clique" is linked to nesting behavior, without inferring much else, much as rodents seem to have a "cognitive map" in their hippocampus that responds only to loci in their environment that seem "scent coded" in some way (meaning they seem to be responding to scents in different locations in their environment).

    All notions of neural computation share the same logical defect, because they are founded on a fallacy. As Tallis writes, "[With] premature conceptual closure . . . under the influence of the language of neuromythology, we are misled into believing we know, and make coherent sense of, more than we in fact do." Thus the "Emperor's New Clothes" syndrome to which I referred, which can only be disspelled by recognizing that it rests of fallacious reasoning.

    One cannot turn one's eyes away from the reflection of them at a distance of, say, about 2 feet from a mirror, and still discern one's eye position in the reflection. That is because it is no longer in foveal view, yet the reflection in the mirror next to it of someone else's eyes can be foveated. These are simple limits of vision's capacity to resolve detail outside the foveal region, whether stationary or moving.

  74. I seems that I failed to describe the experiment clearly enough. You do not
    "turn one's eyes away from the reflection of them". You first fix your gaze on one eye, say the left, in the mirror. The left eye is in full foveal view. Then you perform a saccade to fix your gaze on your right eye. The right eye is now in full foveal view. During the saccade you will observe no movement of the eyes at all. Yet if someone does the same thing standing beside you using the mirror you can see the movement very clearly. Furthermore, if you hold your finger under the target eye and wiggle it ever so slightly during the saccade you can see its movement.
    BUT—there is one snag in this account given by the experimenters to explain their results. I have tried moving my eyes slowly from one to the other, when theory says you should see the movement clearly since it is not a saccade—yet I cannot convince myself that I can see the movement at all. So I suggest on Tuesday we take a further look at this phenomenon in the lab. We need to determine, inter alia, more about the conditions under which one can see the eyes moving of the person standing beside one.
    Re cliques, etc: I have ordered Tallis's book and will defer further discussion until I have read it.

  75. The relevant saccade paper is by Kleiser, Seitz and Kiekelberg (2004) Current Biology. 14 (5). 386-390 and R 195-197. They say "The stability of visual perception is partly maintained by saccadic suppression: the selective reduction of visual sensitivity that accompanies rapid eye movements". This needs further investigation.

  76. As I emailed John privately, I believe the claims of the saccadic suppression research have been overstated and perhaps misinterpreted. Even under the conditions described above, each eye moves away from foveating its reflection in the mirror. If one moves one's gaze ever so slightly away from straight ahead, one cannot even see what direction one's eyes are looking in the reflection. In general peripheral vision is much more sensitive to movement than foveal, which might account for noticing the wiggling finger and not being able to discern one's eye movements in a refleciton, respectively.

  77. There is perhaps a basic, fundamental defect in the idea of VS (or perceptual space as a whole) being *caused* by something else, namely, something physiological in the brain, more particularly in the visual cortex, because it tends to lead to a kind of creatio ex nihilo situation, which might be why there have been attempts to identify it with "activity" or "states" of the brain (although it is not clear what said "states" are; cf. Tallis), rather than its actual physical structure, which seems largely indistinguishable from other brain architecture, as John has duly noted.

    Again correlation does not mean causality, although the destruction of the visual cortex does seem to present *what seems to be* a clear cut cause and effect relationship. But if we entertain the possibility that VS is not *caused* by the brain, then its relationship to the brain is something other than causal and, perhaps to echo John's thinking, that relationship may be only spatial--in some bizarre way.

    Culbertson, as I have noted previously, localized VS at the other end of the visual process--at the surface of objects, brought about by an extended spatial-temporal system of connections that does indeed include the brain.

    Otherwise it would be comparable to physical space being created by something non-spatial, so unless VS is somehow part of physical space, we have this curious problem of bringing into existence space that wasn't there before which, physically, might involve major physical instabilities, much like the creation of matter from energy (to just pick a loose analogy), even perhaps the brain (and person) exploding in the process! No one has ever rigorously tried to "physicalize" VS as I proposed in my old 1986 paper, to think of it in terms of basic physical or quasi-physical (or biophysical) dynamics, in spite of all the loose talk by psychologists of figures "attracting" other figures or colors in VS, as if some form of EM was involved.

    So perhaps what needs to happen is a more careful analysis of the notion of causality (or at least the chain of causal inferences philosophy and vision science has made) and particularly that of action by contact. What is contacting what in bringing VS about? What delimits it from other spatial entities yet *spatially* associates (or literally juxtaposes) it with unlike (orthogonal?) sensory manifolds such as sound and touch?


Note: Only a member of this blog may post a comment.