It seems best to start a new thread to continue the discussion that Bob French and I have been having lately on "Neurobabble in High Places" regarding how one relates our obvious ability to see things as curved (or straight) in the visual world, yet the postulated curvature of visual space is apparently not experienced as a curved visual world. My "diagnosis" is that in order to explain this apparent paradox we need to retrace the epistemological development of the idea of curved space, which does *not* appear to have been derived from visual experience (I believe Gauss is now credited with first advancing the notion of non-Euclidean geometries that would violate Euclid's postulate about parallel lines never converging).
In works I have read on the philosophy of space and time (Mach, Poincare, Nicod, Reichenbach, Grünbaum, and Nerlich) it has never been quite clear to me how a curved space would be perceived, particularly, how curved physical space would be perceived, except for the claim that it is locally Euclidean and therefore is perceived as being flat (ergo, "locally" Euclidean). But in the case of visual space, we perceive its entirety, not just a local region of it, so logically, we should be able to perceive its curvature. Yet most of the arguments for visual space being curved, as I have previously noted, seem not to be based on the perception of curvature, but upon discrepancies between perceived straightness (or parallelness) and third person observations of the test display stimuli employed to make those judgments (the so-called "alley experiments" originated by Blumenthal).
As geometrodynamics points out, physical space(-time) is curved because of a physical force: gravitiy. Why is visual space curved, moreover, of variable curvature as has been claimed by Bob, Indow, Koenderink et al? Is it attributable to the sphericity of the eyes themselves which, to some extent might introduce variable curvature as the eyes are deformed as they are turned?
In this regard, I was particularly interested in the NDE reports which Thomas Droulez shared from the research of Jean-Pierre Jourdan. I was particularly struck that patients reported something like an expanding bubble--this pointing to some sort of sphericality of visual space--and that the visual world as (apparently) seen from outside (from the vantage point of a "5th dimension") itself seemed to involve a kind of spherical construction as well. Perhaps Thomas or Jean-Pierre could supply more details of this apparent structure?
To get things started here, it might be good to briefly review where we left off. I was claiming that physical straight lines, apart from the case of illusions, are constituted as geodesics in a spherical space of 180 degrees by 120 degrees and that a variable curvature (given by a depth function) is needed when physical objects being viewed are at different physical distances from the eye. This results in a convergence of the lines in peripheral regions, such as if a long low building is viewed from a position midway along it. This effect is a bit difficult to determine due to lack of spatial acuity in the peripheral region, but an analogous effect occurs with the Helmholtz checkerboard which physically possesses a hyperbolic structure, but which appears as a normal checkerboard when viewed closeup so as to fill the whole of visual space. By the definition of a geodesic the correspondents in visual space of physically straight lines are the straightest possible even if they possess this convergence.
ReplyDeleteI am wondering if I am understanding the first proposition correctly. You write that "physical straight lines" are "constituted in a spherical space." Assuming that by the "spherical space" you are referring to is *visual space,* are you really meaning to say that there are *physical* lines in it?
ReplyDeleteBy "a spherical space" I am referring to visual space here, and no I do not hold that there are physical lines in it. I think that part of the problem here is vocabulary, and I don't know as there really is an ideal one, although it does help to try to define terms. I am using the vocabulary of the phenomenologists, notably Husserl, who at least claimed to be using the epoche, or bracketing, about ontological commitments when describing experiences, although admittedly Husserl also ended up in idealism. Perhaps you would find the phrase "correlates in visual space" less misleading, although that probably could be misconstrued as well. In any event I am trying to refer to our visual experience of the physical straight line and I hold that that experience is numerically distinct from the physical object, even though it is caused by it.
ReplyDeleteYes, "vocabulary" may be part of the problem, but behind it are conceptual problems as well. As much as possible I think ordinary language should be used first, and see how far we can get with it. In an effort to avoid ambiguity or contradictions we just need to be very specific, that's all.
ReplyDeleteFrankly I don't think Husserl or phenomenology in general will help us much in this regard, as I find their concepts artificial and difficult to relate to actual experience.
That said, we already have a method provided by Hans Reichenbach in his "Philosophy of Space and Time" that can serve as a model for visual space, and that is what he calls "coordinative definitions," something which I have already mentioned here previously. Basically that consists of defining geometrical abstractions by *coordinating* them with specific things, such as in examples I have given here before, e.g., a straight line = a ray of light in a vacuum (yes, I realize that in today's physics that may be a bit of an over simplification, but will suffice for purposes of illustration). The metric of space is defined in terms of "rigid bodies," etc.
Arguably, unless defined in purely physical terms (which would be tricky!) the "physical lines" to which you are referring are lines which we perceive but are defined by other forms of inspection/measurement. As long as those "conditions" are specified, we should be on fairly firm ground.
I really think other philosophical commitments might well be "held in abeyance" so that they do not interfere with analyzing the phenomena themselves--in this case, what is present in visual space.
So this brings us back to the basic question of how a visually straight line or a visually curved one is (ordinary) defined in the most basic sense(s) possible without immediately jumping into abstractions (as it were)?
I don't think that we have to worry all that much about criteria for physical straightness. Besides the coordinative definitions which you give Bill it is also possible to give operational ones, such as the ones from carpentry of sighting along a 2 x 4 viewed headon, or snapping a taut chalk line (the latter operation exploits the ambiguity in 'geodesic' between shortest distance and straightest). I am afraid that the situation for the visual appearances of these physically straight lines is not as straight forward though, since I cannot think of any defining operations which clearly can be exploited, in particular when viewing an object perpendicular to the line of sight (the line from the eye out to the object). Of course one could try to judge whether the visual appearance looks straight or curved, but this sounds a bit vague and subjective. Can you think of anything better Bill?
ReplyDeleteI would not consider the "operational" definitions you mention to necessarily be examples of how to define physical geometry via coordinative definitions, Bob, because one of them involves looking straight (sighting along the shaft), because it does not entail a measurement of some sort to determine its straightness.
ReplyDeleteThe chalk/string trick, though, is an expedient, the result of which presumably could be verified also by "sighting along." But again it raises the question of how it is also known that the chalk/string trick line is indeed straight? I suppose it is known to produce lines that are straight by being related to other measures of straightness. That would be a good question for a master carpenter!
On the other hand, the straightness (or geodicity) of a light ray traveling through a vacuum does not depend upon "looking straight"--which, of course, is the whole point of a *physical* geometry: it is intended to be "objective" and observer-independent, and is defined more in terms of physical parameters affecting the path of a light ray (e.g., the sun's gravitational field).
So perhaps one distinction that needs to be made is between measured straightness (and therefore measures of straightness) and perceived straightness.
The X axis of the visual field (as the axis perpendicular to the Y and Z axes is usually called in optometrics) is often identified by example with the horizon. This obviously does not permit the same visual behavior/observation as used in "sighting along." But it is arguable that a horizontally straight line also "looks straight" in some sense. In what sense, though?
I think the linguistic/conceptual problem is illustrated, Bob, by your phrase "visual appearances of ... physically straight lines" because there is one sense in which one is merely relating one's visual perception to another's ("third person" viewpoint, as it were). This use of "physical" refers only to the visual world, and only by inference to something outside it.
ReplyDeleteIf one rewords your phrase to read instead "appearances of ... straight lines," we may be getting closer to the heart of the matter, because "appearances" have definite properties that, in a sense, are "objective" in visual space, and prior to judgment. Measuring them is another matter.
It seems to me that the most obvious co-ordinative definition of visual straightness is that in the visual world, something is straight if it *looks* straight, simply because that perception alone represents a sort of "primitive" in the visual world: For all intents and purposes the line *is* straight because it *looks* straight. The straightness inheres in the visual world and is not with reference to a standard. (We suppose that that same "straightness" exists in the physical world, though independent of perception--an interesting sleight of hand, conceptually, but one that should be duly noted).
Bill Adams may have some thoughts about this because he is interested in ecologically valid ways of looking at things (pun intended). A while back I cited here his excellent blog synopsis on perspective, in which he mentioned the claim by anthropologists that the ability to "see" perspective depth as depth is something learned (or "cultural," anthropologists would say). Obviously this bears on the question of "sighting along," because it concerns deviations from the line of sight (the "depth" axis).
One other point: In sighting along, ordinarily one closes one eye. So it is a monocular perception, perhaps to avoid double images? The other thing is that in "sighting along" one is reducing a visual angle in the process, i.e., from something visually longer to something quite short.
I agree with most of what you say here Bill; for example that there is a distinction to be made between perceived straightness and measured straightness. There is a sense in which we never escape the visual world (which is identical with the the physical world according to naive realism), but I think that it is still useful to distinguish between what is objective in the sense of intersubjective, where comparisons are made between reports of experiences of different observers and what is objective in the sense of what is true of physical objects, in the sense of what is made of atoms and which are only indirectly learned about. Thus reports can be compared about what looks straight, and also about reports concerning tests for physical straightness, but these should not be confused. I take it that reports of tests for physical straightness actually give us evidence for the arrangements of atoms (which transcend our experiences), while reports (even intersubjective ones) about whether something appears straight have as their subject matter occurrences in phenomenal visual space. The attitude of naive realism just confuses things since it identifies the two matters.
ReplyDeleteIt still is not clear to me, Bob, whether in your formulation visual space *looks* curved. Is that your claim?
ReplyDeleteI am assuming that you are referring to the way that physically straight lines look here in the case of absence of illusions; otherwise you will have to clarify your question. I do construe physically straight lines as corresponding to geodesics (i. e., great circles) in visual space, and these do have a tendency to converge in peripheral regions. This effect definitely occurs when viewing the Helmholtz checkerboard closeup, and as far as I can see also occurs when viewing physically parallel lines perpendicular to the line of sight, although this latter effect is a bit difficult to discern due both to the lack of spatial acuity in peripheral regions of vision, and also the tendencies to make both head and eye movements here. If I have misunderstood your question please let me know.
ReplyDeleteAgain, Bob, you are talking about "physically straight" lines, without explaining what those refer to. When you look at a new moon, are you able to see that it literally *looks curved*? Do you believe therefore that the curvature of visual space itself can literally be seen in the same way?
ReplyDeleteI thought that we were clear about the concept of physical straightness Bill, but evidently there is still more clarifying to do. At the risk of reiterating some points previously made, by physical straightness I am referring to a property that physical objects (e. g., arrangements of atoms) have completely apart from our perceptions of them. A separate epistemic issue is how we know about this straightness, and here I would invoke operations such as the carpentry and optics ones previously discussed. The physical property that is curved for your new moon example involves regions of illumination on the lunar surface and the relative positions of the sun the moon. This results in a curved appearance in the visual space of an observer. Since at least typically this just involves a small region of the space, and does not stretch across the whole space, it does not particularly involve any issues about geodesics, where the effects which I am talking about occur in peripheral regions.
ReplyDeleteI don't think carpentry and optics alone can suffice to define physical straightness (at least in Reichenbach's sense of a "coordinative definition"), because both include rather than exclude perception. What sort of "arrangement of atoms" would be a "straight" one? How would its straightness be ascertained by a physicist? We must distinguish what a carpenter does (and optician, for that matter) from what a physicist does by way of measurement.
ReplyDeleteI fear that I still have not quite conveyed the sense of "looking" straight (or curved). This characteristic is something literally embodied in visual space, a priori any judgment or comparison with a standard, much as Gestalten are, which exist because of perceptual straightness/curvature. Whatever physical circumstances looking straight (or curved) may depend upon (such as those you describe) are secondary to that perceptual fact.
Are you saying that the radius of curvature in visual space is so great that it is not actually seen as curvature in the sense of the curved appearance of the new moon?
What I have been saying about physical straightness is admittedly a little bit vague, and the concept also admits of degree; e. g., some physical lines are straighter than others. Nevertheless something like the carpenter's chalk line I strongly suspect is a good enough exemplar of physical straightness for our purposes. What you say concerning the radius of visual space strikes me as a confusion between the surface being projected on and properties of what is being projected. For example, either a galaxy in space or a person's face can be made to fill a 10 inch television screen. Here the size of the screen is the analog of the radius of visual space, i. e., one can display something that is indefinitely physically large or small on the same space.
ReplyDeleteThe use of the word "physical" with reference to the carpenter and his chalk line seems to be an ordinary use of the word "physical" and thus reflects a form of naive realism, not physical realism (i.e., physicists don't define straightness with reference to straight edges created by carpenters).
ReplyDeleteBy radius of curvature I am referring to the (putative) radius of curvature of visual space, just as a cosmologist speaks of the radius of curvature of space-time: Within the latter are objects (like planets and their orbits and atoms) that have a comparatively smaller radius of curvature than space-time itself. Analogously in visual space, the radius of curvature of the moon would (presumably) be smaller than that of visual space itself. My point is to compare the curvature of the new moon with the curvature of visual space as a whole.
It is not clear to me, and as I expressed in criticism I have voiced here previously of various empirical measures of visual space, as to whether the putative curvature is (1) a function of projective geometry and/or (2) reflects intrinsic curvature of visual space. My present line of thinking is attempting to tease out which is which. (You may recall, Bob, my previous proposal that the curvature of visual space may be an artifact of mapping different projections to each other, much as a curve can be described by tangents.)
It looks to me like we have two separate issues going now Bill, the issue of how to define physical straightness, and issues about the global curvature of visual space, and I thought I might just address the issues of physical straightness now and we can return to the issues of curvature. I agree that most carpenters are presumably naive realists, although I strongly suspect that they have not thought that much about the issues being talked about here. They presumably are also physical realists and think that what they are constructing continues to exist when not being perceived. My point about the chalk line is that it is actually a physical test, going from tautness (i. e., shortness) to a claim about straightness, and thus does not just involve a perceptual sense of what is straight. As I was saying, I think that this is good enough for our purposes, since in practice I think that this works very well. If you want to get into the physics though we can do that as well.
ReplyDeleteThe issue is how to define visual straightness, using by analogy Reichenbach's method of coordinative definition for physical straightness. More the problem seems to be what constitutes physicalness in this context and visual straightness on the other.
ReplyDeleteVisual curvature as measured by the alley experiments is just that: a measurement, more precisely, a mapping of visual straightness. Whereas a ray of light is a quantity in of itself that can be directly measured, the alley experiments are an *indirect* measure of something that happens in perception--perceived straightness/parallelness--because we have no direct means of measuring the latter.
The "naive realism" to which I referred is not that of a carpenter, but the notion that someone is seeing something as physical: There are no physical objects in the visual world, therefore no physically straight lines in it. The definition of a light ray traveling in a vacuum defines physical conditions in terms of physical quantities as defined by physics, and neither chalk nor string are defined as such except in the ordinary use of the word physical (i.e., as different from something mental "in the head" etc.)
In other words I think we must be very careful to avoid equivocation, in this case, that arising from using different senses of the word "physical" together. Conflating them only leads to confusion and muddle.
ReplyDeleteChalk and string are both made of atoms, and as far as I can see are just as physical as light. If you want to define physical straightness in terms of properties of light rays, such as by aligning with a laser beam, that is fine with me though.
ReplyDeleteThese are not my distinctions but those made by physics. Whether or not something is made of atoms does not concern the fact that "chalk" and "string" are not physical properties. A "light ray" as defined by physics would be more like dispositional property, because whereas we can see chalk and string, we cannot see light rays as defined by physics. Is that also your understanding of the situation, Bob?
ReplyDeletePieces of chalk and of string are physical objects and are made of atoms. A light ray is more like a physical process, and I at least think that it is a process involving changes in electromagnetic fields, which are also physical. The verb 'see' is theory laden, I think assuming the truth of naive realism, but there is at least some reconstructed sense in which we can 'see' chalk and string but not light, although the passage of light enables vision; i e., we cannot 'see' without it.
ReplyDeleteAre you saying that photons do not exist as elementary particles?
ReplyDeleteQuite a while ago now we discussed whether or not naive realism actually constitutes a bonafide "theory," and my conclusion is that it does not. How then is "seeing" laden by it in a theoretical sense?
Seeing "with" light in a physical sense is, again, quite a different thing than seeing light itself. Why are you comparing the two?
Regarding photons, I do not hold the Copenhagen interpretation of wave particle duality, but instead a position of wave particle unity where I do identify them with electromagnetic wave packets, as outlined in my paper "Wave Particle Unity and a Physically Realist Interpretation of Light" on David and my website quantumrealism.net.
ReplyDeleteBy making a distinction between a mental seer and what is physically seen, as far as I can see our usage of the verb "see" does assume naive realism; i. e., it assumes that we are immediately aware of at least the front surfaces of objects being seen,that these objects continue to exist in the same format when not perceived, and are physical in a sense which is not compatible with their just being mental sensations in our minds. Whether you want to classify the grouping of these statements as being a theory is I guess up to you; to me it is at least a rudimentary theory although certainly one that I do not hold and one which I believe we have overwhelming evidence against. Maybe we need a defender of naive realism to join the discussion here though.
Regarding your last question, by saying that light enables vision all that I was referring to is that we cannot see if no light is present; try looking at something in complete darkness.
What is a "mental seer," Bob, in contrast to things "physically seen?"
ReplyDeleteIt is not just the verb "see" that embodies naive realism, but arguably the whole (non-technical) language of perceiving itself.
I believe that a convincing argument can be made, following Wittgenstein and Moore to some extent, that naive realism is literally built into (the experience of) perception, except that the perceiver doesn't realize he is naive--at least that is the case for most people. Why is this so? Because, as I have stated before, the *act* of perceiving itself is perceived, it is contained in perception: We literally *look* around at things in the visual world. The "looking" is part of the act of visual perception, not something outside it. We experience our eyes turning and focusing (or not), for example. They are just as much perceptions and thus a part of visual perception as a whole as are the surfaces we perceive are visual surfaces when they are seen rather than touched.
Was none of this clear from my previous remarks on this point or is there perhaps something to which you take exception here, Bob?
I am not sure how much this is going to help Bill, but maybe it will help clarify where we agree or disagree. I hold that visual space is part of our conscious mind, and thus is mental in character, even though it is taken as being physical according to the naive realism of common sense. I certainly agree that naive realism is built into ordinary perception language, and thus if one just analyzes ordinary language as J. L. Austin (and his student John Searle do) one ends up endorsing naive realism and may also claim that other positions are incoherent since they do violence to ordinary language. Moore, as far as I can see, was a naive realist, but I really don't know what to say about Wittgenstein. His work is full of aphorisms which are subject to different interpretations. John Findlay, who knew him personally, once told me that he was a solipsist all his life, but I at least read his later work as being hard materialist, and I do not know what to say about the Tractatus, he evidently disagreed with Russell's reading.
ReplyDeleteAs far as I can tell we are not consciously aware of all eye movements, in particular saccades,although we may be aware of other ones and head movements.There is a sense of looking, or perhaps better 'attending' whereby we can concentrate on just part of visual space, but this remains just part of our conscious mind. I believe that a physical world also exists, comprised of atoms and other particles, but we only learn about it by the indirect methods of science. I hope this helps some.
In an attempt to keep talk of visual space as empirical as possible, probably it would be best to set aside ideas about it being "part of our conscious mind" or "mental in character" for the time being, because neither of those points really advance our knowledge of its nature and structure very far, at least beyond generalities. In fact, relating it to physical structure may lead to misconceptions (as I have argued).
ReplyDeleteThe "visual" is somehow ambiguous in this regard because on the one hand we recognize the visual world as something we see (or don't see if we close our eyes or become blind) yet we also just call it "the world," as somehow existing independently of us. Most people don't think of what they see as being something mental. It is only through a fairly elaborate process of thinking about what we see that we come to believe otherwise.
In other words, one could say that naive realism is true in the perceptual world, if not of its relationship to some physical state beyond it. That's the rub, as I see it. To just say that naive realism is false is to belie the facts of perceptual experience (i.e., within the perceptual world). We do indeed seem to directly see the visual world simply by looking at it. Do you experience that differently in some respect, Bob?
Wittgenstein was preoccupied with Gestalt phenomena in the 1920s and remained so off and on for the rest of his life. He writes about them in his "Philosophical Investigations" and elsewhere in his later philosophy. If one sees the latter as extending/correcting Moore's "defense of common sense," then he is basically talking about the verities of the perceptual world, because he does not talk about science or its theories, except as they relate to certain philosophical problems. As nearly as I can determine, he never talks of materialism nor of solipsism. In fact, in his "Brown Book" he explicitly rejects solipsism, as I recently quoted here.
May I suggest, Bob, that you close your eyes and concentrate on the experiences of moving your eyes about. I doubt you will not notice them moving, even after you open your eyes and look around at the visual world. Attention is something above and beyond that, because we can attend to specific things in the visual periphery while fixating.
I actually like what you say about naive realism here Bill, within the visual world we are immediately aware of our visual experiences, and if we take the visual world as being physical, then we are immediately aware of physical objects. But, my position at least, is that everything here is just part of our conscious mind. The physical world comprised of atoms transcends all of this and is only indirectly learned about by the methods of science. Naive realism is false if it claims immediate awareness of this world; but maybe you claim that this world does not exist, I do not know.
ReplyDeleteRegarding Wittgenstein, he does explicitly talk about it a few times in the Tractatus. In 5.62 he says "This remark provides the key to the problem, how much truth there is in solipsis. For what the solipsist means is quite correct; only it cannot be said, but makes itself manifest" and in 5.64 "Here it can be seen that solipsism, when its implications are followed out strictly, coincides with pure realism. The self of solipsism shrinks to a point without extension, and there remains the reality coordinated with it." I also found several referenceds to solipsism in the Blue Book, which seemed to be subject to different interpretations, but maybe you can give me the citation from the Brown Book. In the Philosophical Investigations, I at least read him as holding a behavioristic position (and to the extent that this identifies a person with their body, as being a hard materialist), and I read the beetle in the box discussion in paragraph 293 as saying that even if we had a conscious mind(the analog of the beetle) we could not talk about it. Admittedly though all of this is rather aphoristic and hence subject to different interpretations. Different people have told me very different things about Wittgenstein.
Regarding eye movements, I do find that I am aware of some of them, but not the saccades.
Unfortunately I don't think the claim that visual space is "just part of our conscious mind" helps us very much, because, for one, it is not really a claim susceptible of empirical investigation and because there is a very clear sense in which we do not regard the visual world as mental, but other things within our experience instead.
ReplyDeleteThe few exceptions where Wittgenstein touches on solipsism only show that he gave it very short schrift, and though interested in it as a philosophical problem (mainly because of how it relates to idiocy of so-called "private" languages), apparently did not consider himself a solipsist (see H.O. Mounce's monograph on the "Tractatus"). In the "Blue and Brown Books" he writes (p. 58), "Now when the solipsist says that only his own experiences are real, it is no use answering him: 'Why do you tell us this if you don't believe that we really hear it?' There is no common sense answer to a philosophical problem. One can defend common sense against the attacks of philosophers only by solving their puzzles, i.e., by curing them of the temptation to attack common sense; not by restating the views of common sense." It has even been said that Wittgenstein's whole later philosophy was implicitly directed against solipsism.
Inasmuch as Wittgenstein was so preoccupied with Gestalt phenomena from the 1920s onward, I can hardly see how that is consistent with him adhering to behaviorism.
There is a tendency to pigeon hole Wittgenstein, and I think that is a mistake--as it is a mistake to pigeon hole views expressed here that may only superficially resemble traditional philosophical positions.
When the average person looks around the world and says something like "I wonder what could have been in my mind to have gone out doors without my jacket?" clearly his question to himself is not referring to the visual world, or any part of the perceptual world, but to something "mental."
So returning after this lengthy digression in our discussion to the question of how to arrive at a coordinative definition of visual straightness and curvature, how would you go about doing that, Bob?
BTW, as Smythies and I now concur, the jury is still out of whether we perceive saccades or not (some of the evidence of so-called "saccadic suppression" may not hold up as it turns out).
ReplyDeleteI agree that it is better to get out of our digression, so with respect to your question about visual straightness, the best example that I have is looking perpendicular to the line of sight at physically parallel lines (say that have been measured to be equidistant apart, and hopefully we can avoid a digression on this)and see whether the visual appearances of the lines tend to converge in peripheral regions of visual space. Geodesics (great circles) of a sphere, should tend to converge at both ends.
ReplyDeleteHow do you link converging lines in the frontal plane (x-y plane) with perceived (visual) curvature and distinguish them from the converging lines in a flat plane projection?
ReplyDeleteAre you saying that parallel lines in the frontal plane should *look* curved as well or not? It seems to me that convergence and curvature are two separable visual variables.
ReplyDeleteThe converging lines on a flat plane are not geodesics (i. e., there are straighter possible lines on the plane), while I hold that they are the straightest possible in visual space. You can have non-converging lines on a sphere (e. g. the lines of latitude on a globe), but with the exception of the equator they are not geodesics.
ReplyDeleteCurvature can both be described externally, in terms of properties in a higher dimensional space, or internally in terms of a coordinate system embedded in a space itself. If a space has an internal spherical metric (positive curvature) then its geodesics will converge; all the way to a point at the poles. It is not possible to project a planar figure onto a sphere without distortions.
Thanks for supplying these distinctions, but that was not what I was asking, familiar as I am already with them. What I was asking is on what basis one is able to distinguish definitively between converging lines in visual space due to projection from converging lines being geodesics (like great circles)?
ReplyDeleteFor all intents and purposes VS is like a surface or relief, a bounded one at that, and one that ostensibly has only one side. How would one know that even if, by means of measurement suggesting a non-Euclidean metric, that it was a function of the intrinsic curvature of the space and not the changing angles and slopes that we perceive in VS? I posed this question a few months ago, but got no response.
In other words, there is no possibility of geodesics intersecting in VS because it is bounded (by comparison, the surface of a sphere is a 2-dimensional unbounded surface).
ReplyDeleteIt is true that if visual space is just part of a sphere that geodesics will not intersect at two points (at opposite poles) since the space is too limited. I am not sure that I understand your first question, but is it how do we know that for physically straight lines (say, discovered to be straight by some physical test; pick your favorite one),the analogues in visual space are also geodesics (i. e., the straightest possible)?
ReplyDeleteI'm not sure we can even say that VS is "too limited," because one would have to say, "compared to what?" Perhaps we should not presuppose without very good reasons that VS is a section of a larger surface (in spite of my interest in the visual "expansion" in NDEs that Thomas has kindly related here).
ReplyDeleteIn the frontal plane about the most we can do in VS is define a geodesic as a line that looks straight and extends from one side of VS to the other (whether top to bottom or side to side).
The problem I have with claims of metric curvature (going back to Luneberg) is, as I have expressed here previously, that the measurements of perceived straightness/parallelness, related as they are to lights point arrays then measured by experimenters according to some measurement standard of straightness, do not control for eye movements, and changes in perceived angles and slopes as a function of different eye positions. This could readily confound the data, because the perception of straightness with eyes moving vs. stationary may not be the same, reflecting changes in projection.
Thus my reference to the probability (prediction) of there being both "apparent" straightness and "apparent" contours, reflecting a kind of visual "averaging," much as one keeps their eye on the trajectory of an object flying through the air but that is also falling to the ground due to gravity: One is "tracking" a curve, but in the case of straight lines, one is tracking a changing slope as the eyes rotate/elevate, etc.
I have had to ramble some in expressing these points, but perhaps they will get across anyway?
With respect to your question about what visual space is limited with respect to, I would say a complete sphere, which I holds exists in a manner which has the potentiality to become actualized into visual sensations, but clearly this begs some questions about visual space having a spherical geometry to begin with.
ReplyDeleteYou are definitely right to raise some methodological issues concerning judgments of "apparent straightness" and certainly controls for eye movements is among them. Perhaps some old experiments ought to be repeated, and maybe some new ones (such as concerning judgments about apparent convergence of physiscally parallel lines in peripheral regions of visual space) be performed with better controls over factors such as eye movements.
As a general precedent/model, we might be wise to follow the lead of Einstein and those around him concerning the analysis of physical geometry (i.e., the geometry of the physical world) in that they argued (especially Reichenbach) that not just physical geometry but geometry itself might ultimately be an empirical problem, and not just a matter of an internally coherent set of axioms (Euclid). More than one in that group of scientists and thinkers felt that the geometry of space was something amorphous and had to be determined empirically, not just conceived rationally.
ReplyDeleteWith that in mind I have tried here to recover the basis of visual straightness and curvature rather than depending upon measurement to decide what is straight or curved in VS. That is because otherwise there is a potential contradiction: On the one hand, the alley experiments depend upon judgments of straightness/parallelness (the light points *look* straight and parallel) but then the theoretical outcome is that they are somehow "really curved" given the "actual" positions of the light points in array as noted by the experimenter.
The geometrical solution really amounts to a glorified "regression line" as one sees in statistics, a "best fit," in this case, yielding a curved geometry. There is a tacit (?) assumption in Luneberg's theory that geodesics in VS are both judged and *seen* as being straight/parallel, but that assumption seems made without any real justification, other than to reduce visual straightness/parallelness to non-Euclidean geodesicity--which, IMO, is a bit of stretch, and is unfortunately typical of many of the ad hoc assumptions made in mathematical psychology as a field. Thus my heading for this post "Looking vs. Being Curved."
Clearly additional experiments need to be conducted using fixation-contigent displays (i.e., displays that turn on when there is fixation and turn off if the eyes move, as we have at UCSD in Keith Raynor's eye tracking lab), specifically experiments to measure just how well one can make out straightness in the periphery by altering the linearity systematically (e.g., by curving the line segments outside the foveal region). The same experiments can then be performed using fixation-free displays to ascertain if judgments of straightness (and/or parallelness) are same/different as a function of eye position.
I thought that I should let members of the group know that I just got a note from Imogene Angell that her husband Richard "Brad" Angell died on Dec. 24 at the age of 92. Brad was one of the big advocates that visual space has a spherical metric and had an influential paper on visual space in NOUS in 1974.
ReplyDeleteMy condolences to Angell's widow and family.
ReplyDeleteIt would be interesting to know, though, how the sign of curvature in VS changed to its opposite, i.e., idea of VS being hyperbolic (Luneberg et al) being rejected in favor of it being spherical, and since then its geometry being thought to be of variable curvature (Indow et al).
What is Angell's most compelling argument, given (what I believe to be) the basic flaw in the interpretation of most of the experimental results based on judgments of straightness/parallelness (i.e., that they do not control for eye movements and thus different retinal projections being "averaged" as it were)?
This is just a partial answer to your questions Bill, and if you have not already done so, I encourage you to read Angell's paper in NOUS 8 (1974) p. 87. One point is that the theory that visual space is spherical is the older theory going back to the work of Thomas Reid in 1764, and thus a better question might be why the Luneburg theory ignores the evidence that visual space is spherical.
ReplyDeleteAngell does address the Luneburg theory when he makes a 4 fold distinction among Ap actual geometric properties of physical objects; Jp judgments of geometrical properties of physical objects; Av actual geometric properties of visibles; and Jv judgments of geometric properties of visibles. He then claims that the Luneburg theory is about Jp and claims that visual space is spherical are about Av and thus that the two have different subjects and therefore may be compatible. I don't know how convincing you find this and I once pointed out to Angell that I thought that his theory had a problem with binocular depth percedption; e. g., with accounting with the sense of depth with random dot stereograms, which certainly seems to be a geometric property of visual space itself.
With respect to Luneburg I know that he was a mathematical consultant to Adelbert Ames Jr. at the Dartmouth Eye Institute in the late 1940s. As you I am sure know Ames is famous for his demonstrations of ways to fool the eye with respect to depth cues ; e. g., the Ames rooms, and I remember at an introduction to psychology class at Dartmouth they trotted out Ames's trapezoidal window which appears to be rotating in the opposite direction to which it is physically rotating. I once asked Ames's son Adelbert Ames III, who at the time was a physician at the Mass General Hospital in Boston, about Luneburg but he had never heard of him. I have subsequently found some references to his work in physical optics (he had a book "Mathemtical Theory of Optics," but my impression is that he was primarily a mathematician, and thus perhaps was not aware of the earlier work on the geometry of visual space, but I am not sure.
Of course ideas about the space of sight being spherical (or oblate) are ancient, but a geometrization based on actual psychophysical measurements is much more recent, thus Luneberg's model based on Blumenthal's alley experiment data, which initially seemed to be counterintuitive in light of that tradition.
ReplyDeleteLooking at Angell's paper again (after the passage of over 25 years), it is apparent that none of his thinking controls for the role played by changing projections as a function of eye position. Also he assumes with Reid that great circles will look straight within a sphere, which is IMO quite debatable, given that the vault of the heavens had traditionally been described as curved (recall my quotation from Mach in this regard).
The main difficulty, as I see it, is that in Euclid's day there was no projective geometry, so the problem is how to reconcile visual appearances that are ostensibly a function of projection (if one buys that they produced by optical projection on the retinas) with the shape of the space itself and, moreover, the problem of how to disentangle the one from the other (more recently this has been a problem in cosmology, as I have stated before).
If I had to evoke a single geometrical picture to describe what I imagine the situation to be, it would be that in perceiving, say, the walls of an ordinary size room interior, it is rather like a plane being rotated on the surface of a sphere, and that VS in fixation is ostensibly a flat space, not curved (and that includes taking projection into account). In other words, the frontal plane is rotating on its own axis (as it were--a simplification for purposes of illustration to be sure).
The most fundamental problem, IMO, is the assumption that VS exists as geometrical structure independent of the role played by eye movements in determining its geometry. That is the same as assuming that a cube exists independently of looking at it from different angles.
ReplyDeleteThere is a work, Euclid Optics (distinct from his Elements), that does contain a projective geometry, although I gather that there has been some disputes about its origins.
ReplyDeleteI have some of the same problems that you have,
Bill, with respect to some of Angell's examples. For example it requires bodily rotational movement to see the horizon come back on itself, and head motion is required to see each of four ceiling angles as being clearly obtuse. Some other examples, such as Helmholtz's checkerboard can clearly be perceived at one glance though. I agree though that further experiments should be done for examples such as some of these that do control for eye movements.
The Helmholtz checkerboard is something of an anomaly IMO, because it looks barrel-distorted at one distance, yet rectilinear when viewed at a closer one. I'm not sure that just invoking a curved geometry is the only (or best) explanation for that fact. The basic question is: How does something that is visually curved at one distance get "straightened out" at a closer one?
ReplyDeleteMy hypothesis is that the alley experiment "curvature" (or "discrepancy" as I would call it more neutrally) is to be explained in terms of what might best be explained as "apparent straightness," and that just as there are apparent curves in VS, there may be apparent straight lines. But both may require a local factor rather than a global one, as geometrizations have presupposed--in other words, something more in line with Gestalt processes that "rectify" (as it were) curves under certain conditions. I am unaware of anyone having investigated this, though.
I am now in contact with Keith Rayner at UCSD and he has assigned a graduate student to see if doing some new experiments controlling for eye position is something they would like to undertake.
Since the effect is progressively great towards peripheral regions of visual space, the real variable here is how much of visual space is subtended by the checkerboard. One can control for any issues about absolute distance away (maybe due to squinting or something) by enlarging the size of the checkerboard. I am glad that you are looking into some new experiments controlling eye positions. I would be quite interested in any results, and also if there is anything that I can do to help (such as providing background information) please let me know.
ReplyDeleteBy contrast the visual angle in the alley experiments is rather small and nearly foveal. So whereas the Helmholtz effect depends upon comparatively large scale features of VS, the alleys result from smaller scale ones, presumably implying two different radii of curvature because of that.
ReplyDeleteObviously to systematically explore this and the Helmholtz checkerboard would be done with computer displays today which can produce both flat and 3-D images, and also permit continuous interactive capabilities for the experimental subjects.
My guess is that ultimately some sort of tensor analysis will be required to make sense of all this, because if contours are being "unbent," one needs some force responsible for that, unless it can merely be shown to be a function of optics.
Bill, what you are describing here at least sounds to me to be very close to what I have been trying to describe all along; that visual space holistically possesses a spherical metric structure but locally possesses a variable curvature to account for phenomenal depth effects. I try to handle this with differential geometry in my NOUS paper.
ReplyDeleteNot exactly, Bob, because I don't see that your geometrical model involves what amounts to an inverse transform that would rectify contours such as in the Helmholtz checkerboard, and analogous to a Mercator projection, make right angles out of obtuse ones. Is that something that can be achieved optically do you know?
ReplyDeleteThat is, the hypothetical transform (or projection) is, in this case, not "area preserving," i.e., by being rectified, the checks are losing relative area though gaining in size.
ReplyDeleteClearly more investigation needs to be done here Bill. One place to start; I did a Google search of "Helmholtz checkerboard" which showed some recent papers published in Perception, and you might want to check them out. My thinking, at least roughly, was that if a projection from the right angles of the mercator projection onto a sphere results in converging lines, then a projection onto a sphere with non-converging lines should be from a hyperbolic figure such as Helmholtz's checkerboard, but clearly this needs more spelling out in detail. I don't understand your distinction between size and area when you speak about the checks losing relative area but gaining in size, so maybe you can clarify that. Also, I once saw a discussion of Helmholtz's checkerboard in a book, I believe on painting and photography, and if I can find the reference I will give it to you.
ReplyDeleteI think that the book I was referring to is "Optics, Painting, and Photography" by M. H. Pirenne, but I don't have a copy here to make sure.
ReplyDeleteThis is perhaps the most interesting of the Perception abstracts:
ReplyDeletePerception. 2007;36(9):1275-89.
Straight lines, 'uncurved lines', and Helmholtz's 'great circles on the celestial sphere'.
Rogers B, Brecher K.
Department of Experimental Psychology, University of Oxford, South Parks Road, Oxford OX1 3UD, UK. bjr@psy.ox.ac.uk
Abstract
Helmholtz's famous pincushioned chessboard figure has been used to make the point that straight lines in the world are not always perceived as straight and, conversely, that curved lines in the world can sometimes be seen as straight. However, there is little agreement as to the cause of these perceptual errors. Some authors have attributed the errors to the shape of the retina, or the amount of cortex devoted to the processing of images falling on different parts of the retina, while others have taken the effects to indicate that visual space itself is curved. Helmholtz himself claimed that the 'uncurved lines on the visual globe' corresponded to 'direction circles' defined as those arcs described by the line of fixation when the eye moves according to Listing's law. Careful re-reading of Helmholtz together with some additional observations lead us to the conclusion that two other factors are also involved in the effect: (i) a lack of information about the distance of peripherally viewed objects and (ii) the preference of the visual system for seeing the pincushion squares as similar in size.
This is relevant IMO because it invokes the role of eye position (though the wording of the abstract is rather confusing, to say the least, with its apparent contradictions). I see that Koenderink and his associates have also been investigating this as well. I've read Pirenne's book but do not have a copy.
ReplyDeleteBut the problem is that the checkerboard is seen in *both* conditions in VS, so the transformation (rectification) is entirely a function of size relative to the viewer. Since it is a flat image that changes shape ostensibly only in two dimensions within VS (height and breadth), one must account for both perceptions of it. Invoking projection on a spherical surface therefore IMO only begs the question of how it is rectified (or "straightened" to use the word in the literature on it), but it might be interpreted that the shape of VS is not isotropic/homogeneous at different scales.
By "area preserving," in this instance, I am referring to the fact that as the checkerboard increases in visual angle, the pin-cushion polygons are stretched into squares, so the area increases (and is therefore not "preserved").
I agree with you Bill, that there are ways that a space with an internal spherical metric may not actually be a sphere from an external three dimensional point of view, in particular if it is not homogeneous, and I suspect that a move like this would be necessary if its immediate neural correlates are in the visual cortex.
ReplyDeleteYou are right that when the checkerboard is viewed so as to fill at least most of a viewer's visual space the squares appear to be at right angles and to have equal areas. I think that this results from the combined effects of the opposite process of marginal distortion in wide angle photography (or with the distortions of the Mercator projection)whereby large regions in a peripheral region of a plane surface are reduced in size when projected onto a sphere; together with a "straightening out" of hyperbolic lines which is the analog of straight lines on a plane being projected as converging great circles on a sphere.
It does not seem to me that the Helmholtz checkerboard effect necessarily requires curvature to explain it, if one thinks of some sort of normalizing process that tends to make irregular shapes appear regular (i.e., rectilinear). Such a process would be the converse of gravitation, i.e., it would unbend contours rather than bending them.
ReplyDeleteAs I see it, the problem with all geometrizations is analogous to curves in statistics: They may be either real or artifacts of analysis. The trick is how to be certain, and is my impression that there is virtually no overlap between studies of perceived curvature vs. straightness and attempts to geometrize VS, because there seems to be a tacit supposition that VS is non-Euclidean (favoring a positively curved geometry).
If we regard the observer being *in* perceptual space, it might be better to think that we are *inside* a sphere, if indeed that is the case (though I am as yet convinced of that), much as we are also inside the physical universe. Einstein wrestled with Descartes' idea of an "Archimedean point" outside the universe from which to behold it, and realized that there is no such observation point (or "God's eye view," as we discussed here previously).
With that in mind we should probably think in terms of internal geometry. If one shifts the point of view to the interior of a sphere (which is more consistent with the "vault of the heavens" appearance of curvature in VS) the question comes back to whether we really perceive VS as curved or not, and I don't think you have really answered that question, Bob. Do you believe that VS looks curved to observers relative to it?
We seem to be going in circles with respect to this question Bill, so it might help to say more to try to clarify it. Are you asking whether, from an internal geometry perspective, geodesics in visual space appear to be curved, or something else?
ReplyDeleteAgain, returning to the title of my posting ("Looking vs. Being Curved") it is really a very simple question that I am asking, Bob, but I'll try to rephrase it so that it is, hopefully, perfectly clear: Does an observer *see* the curvature of VS in the same way that they see the curved side, say, of a coin held up to the eye? If your answer is "no," perhaps you can explain why the curvature of VS is not seen as curvature.
ReplyDeleteFirst off, the curved edge of a coin is not a geodesic, and thus there are straighter lines in visual space. As far as I can see in general, with the exception of optical illusions, such as the Hering Illusion, the analogues of physical straight lines are rendered as geodesics in visual space. Also, as far as I can tell these have a tendency to converge in peripheral regions when viewed perpendicular to the line of sight, but there are no straighter lines. I really don't know whether this entails in some sense that they "look" curved or not.
ReplyDeleteAll of the previous remarks (and your example of the edge of a coin)concern one dimensional lines. If you are asking whether visual space (that I at least hold is topologically two dimensional) "looks curved" this is another question, and at least raises the issue of "curved with respect to what"?
Leaving the question of geodesics aside for the moment, Bob, how do you know that there are straighter lines in visual space than the curved edge of a coin held close to the eye? Presumably you are (tacitly) adopting some criterion of straightness and curvature. Is it perceived straightness/curvature--in other words edges/lines that *look* straight/curved?
ReplyDeleteI assume that you are not talking about the case of really squinting (which I am not able to do) so that the coin takes up essentially all of visual space here; if not I have misunderstood your question. In that case, a physically straight line, as reconstructed (or whatever verb you prefer here) in visual space appears (and looks) straighter than the edge of the coin. Admittedly this criterion is subjective (in more than one sense) and also a bit vague, but I am afraid that it is the best we can do.
ReplyDeleteI have come to conclude that there is a strong rationalist bias in the existing geometrizations of visual space, which exhibit Quinian theoretical under determination for various reasons I have already outlined.
ReplyDeleteIf anything it is geometry that needs to be "reconstructed" from visual perception, as John Smythies was advocating by starting the blog by synopsizing Nicod, rather than the converse, as if rationality somehow knows more about what it is to be visually straight than vision itself: The same line in visual space cannot be both straight and curved at the same time, otherwise there is classic contradiction.
IMO this needs to be sorted out before progress in conceptual analysis can be made, let alone new empirical work. That is what I have attempted to do with this posting.
With respect to Quine, I have met him before, and while I disagree with him on his behaviorism and his hard materialism, I partially agree with his attack on the analytic synthetic distinction in that I don't think that it is always sharp. I also like the way he cites Otto Neurath's metaphor of rebuilding a sinking ship while it is at sea; in other words there is an intertwining of theory and observations in many matters and you cannot pursue one to the complete exclusion of the other.
ReplyDeleteAs far as a conceptual analysis in geometry goes, you might take a look at a paper I had published in Philosophical Studies, 49 (1986) p. 213, arguing that Euclidean Geometry is analytic. It at least argues that there is more to Euclid's definitions than is sometimes made out. But as far as both physical geometry and phenomenal geometry go, I think it definitely takes a combination of working with concepts and observations to make any progress. So, if more relevant empirical work can be done, such as with the experiments you were talking about controlling for eye movements, I would strongly encourage you to continue the pursuit of them.
Of course my main point is two-fold: (1) the geometrization of visual space is under determined by data and (2) over determined by theory (geometry). One of the main goals of the blog has been to review the *observations* upon which the geometrizations have been derived. If we cannot agree on what a straight line is in visual space vs. a curved one, how is this supposed to have been accomplished without some arbitrary definitions being substituted for what is observed?
ReplyDeleteOne point concerning Quine is that while he explicitly talks a lot about the indeterminacy of radical translation, and there is a Quine Duhem thesis that theory falsification is difficult because of the role of auxiliary hypotheses, I don't believe that he was the first to say that theories are underdetermined by data. Any account saying that scientific reasoning is inductive in character says this.
ReplyDeleteNevertheless it is clearly true, but I don't see that introspective methods are any worse than the rest of science in this regard. Also introspective methods can be intersubjective in the sense that they can be repeated among different subjects and consistency of answers can be looked for.
With respect to judgments of curvature one point is that these are routinely made under a naive realist outlook; e. g. with verbal descriptions of what I see - that the coin is round or the ruler is straight. It is true, that under a naive realist outlook independent tests can also be made for roundness and straightness (comparing with a straightedge etc.) that are not available for introspection, which by definition is just concerned with appearances. Still, I think that the visual appearance of a coin is verbally judged to be round, or that the visual appearance of a ruler is judged to be straight, is some sort of evidence (perhaps not apodictic) that the relevant reconstructions in visual space have these properties.
Quine (with whom I corresponded years ago) was doubtless not the first to point out the pitfalls of empirically under determined theories in science, since it is perhaps the main thesis of Duhem's "The Aim and Structure of Physical Theory," if formulated somewhat differently.
ReplyDeleteWhat more concerns me here is the converse, namely, theoretical *over* determination which tends to override data that don't "fit" a theory and lead to theory laden observations. Even in your account above, Bob, it is difficult to know what you consider to be an observation with respect to VS, because virtually all your statements are couched in theoretical terms, seemingly with the intent to equalize all epistemic frameworks rather than acknowledging the primacy of sensory data when it comes to knowing the structure of VS.
We have already agreed that in the visual world, the observer sees directly what he looks at, so naive realism is irrelevant to things going on inside the perceptual world. Why do you then bring it up again?
(Meanwhile, I do hope that our two resident philosophers will comment on Lothar's new posting that promises to approach our topic from a different and novel perspective.)
ReplyDeleteUsing language is unavoidablein a post, and you may be right that essentially all language is to some extent theory laden, but maybe you can tell me which phrases in particular of mine you fine to be "couched in theoretical terms."
ReplyDeleteRegarding the "visual world," as far as I know J. J. Gibson coined this phrase in his 1950 book (if I am mistaken please let me know) and I thus find the term to be laden with Gibson's philosophy of perception, which I admit I have never been able to make much sense of, but maybe you can help me out there. Again, if your usage is different from Gibson's please let me know.
Above you refer to the "naive realist outlook," Bob, but it is not clear to what that refers to in this instance, since in the alley experiments (for example) the experimental subject is making judgments based on what he sees in his visual space (/visual world). What is naively realistic in that situation, which consists of making an experimental observation? The experimental Ss are not making theoretical statements but observations.
ReplyDeleteThe term "visual world" was not coined by Gibson but was already in use in the 19th century. It is really the notion from which visual space is an abstraction. I have used both terms along with perceptual world more or less interchangeably, and with minimal theoretical baggage implied or intended.
But clearly in the perceptual world, when we open our eyes, our only awareness of the process of perception is turning or moving our head (or body), moving our eyes, squinting, fixating, closing them, etc. There are no other perceptible processes mediating visual perception in the perceptual world, and we see visual objects directly for all intents and purposes. In the visual world seeing something "indirectly" would mean perhaps seeing a reflection in a mirror.
Thanks for the clarification on the phrase "visual world" Bill. Phrases, and often theories, often have a longer history than we realize. I do have a question though regarding the alley experiments. You say that the experimental subject is making judgments based on what he sees, but what type of judgments? I know that Angell in his paper claims that the instructions call for judgments of the relations of physical objects, rather than of their visual appearances per se. I also know that Luneburg cites the experiments, and the divergence in results between distance alleys and parallel alleys as confirming evidence for this theory that visual space possesses a hyperbolic metric structure.I also know from the size constancy literature that the nature of instructions can make a big difference in the amount of constancy reported.
ReplyDeleteSince you worked directly with Tarow Indow on a reproduction of the experiments do you know what instructions were used; i. e., to make judgments of actual physical distances or of just the visual appearances?
The Ss are asked to adjust the light points in the "parallel" alley condition until they are (1) straight and (2) parallel to the S's line of sight. In another condition Ss are asked to adjust the light points in parallel arrays until they are equidistant. Presumably whether called "physical objects" or "appearances," the results are the same because the experimental task requires the Ss to see these things which are thus in their visual space.
ReplyDeleteThis makes it sound Bill like Angell was right with his assertion that it was judgments about physical length which were used in the alley experiments. I don't know whether or not you have looked at these; but two works which talk about the different effects of instructions on reports of size constancy which I used when I was originally researching this are Gilinsky's "The Effect of Attitude upon the Perception of Size," in American Journal of Psychology 68 (1955), p. 173, and Epstein's book "Stability and Constancy in Visual Perception."
ReplyDeleteNo judgments of length are entailed in the alley experiments, but only as I said (1) whether the arrays are straight and (2) parallel, or, alternatively, (3) the points points of light in the alleys are equidistant (in the frontal plane).
ReplyDeleteEffects of "mental set" (as it were, aka "Einstellung" in the Gestalt literature) are relevant to interpreting distance in perspective, and Bill Adams might well have some relevant things to say about that, if he feels so inclined, because perceived length (and therefore estimates of distance) are ambiguous in perspective.
As Indow (et al) found the "distance" judgments (i.e., judgments of equidistance between points in the alleys) resulted in different "objective" configurations. Of course, he did not take into account that the eye scanning paths and fixation points involved in perceiving/judging equidistance may be different from those in judging straightness/parallelness--thus the different results. That is why eye movement is a variable that needs to be controlled for in the alley experiment paradigm (and which has not been).
ReplyDeleteI think that there is still an issue Bill with the concept of "equidistant;" does this mean judged to be physically equidistant or for the visual experiences to be equidistant? It sounds like the instructions were for judging physical equidistance, or am I mistaken?
ReplyDeleteThe Ss are asked to judge equidistance purely on the basis of appearing equidistant. So the Ss adjust the light points so that they appear equidistant.
ReplyDeleteI can still imagine a possible ambiguity in ways to take these instructions between appearing to be physically equidistant, and for the visual appearances themselves to be equidistant. This in turn raises at least two questions 1. was there any evidence from quite disparant results that some participants were interpreting the question in different ways and 2. how close were the results to those of physical size constancy; i. e., where the physical distances between pairs of lights remains the same as a function of distance away from the viewer?
ReplyDeleteThe "distance" judgments were just a corollary of the parallel judgments, and the experimenter made/makes it clear to Ss that they are to appear parallel or equidistant, not how they should appear such in perspective (e.g., like converging train tracks).
ReplyDeleteThis issue about being physically equidistant vs. visual appearance in perspective sounds pretty basic Bill, and I find increasingly that I am not clear on the actual results. For example, physically are the alleys constructed to be further apart as a function of their distance away (so that, in perspective, the distances visually appear the same) or do they remain physically approximately the same distance apart here (as would be expected if it is judgments of physical distances which are called for)? Maybe a diagram would help here, or maybe we should have a separate section devoted to discussing the alley experiments.
ReplyDeleteIn the alley experiments the lights are on a series of horizontal tracks, each one a little further away in the line of sight from the observer to correct for perspective. The Ss move the lights horizontally until they appear parallel to their line of sight. From the experimenter's point of view the resulting arrays after being positioned by the Ss appear to be in the shape of a funnel, flairing out from the Ss's observation point (which is nearly level to them). The judgments of equidistance are made using the same set up, only the Ss are told to make the lights the same distance from each other horizontally. It's really just two sets of instructions for the same experimental set up, because the lights are only movable horizontally, not in depth (at least this is the traditional alley experiment--there may well have been other variations devised).
ReplyDeleteAgain, this points to an interesting inconsistency in talk of the perceptual world. I have referred previously to "third person naive realism," the presupposition that, say, an experimenter is seeing things the way they "really" are whereas the Ss are seeing an illusion, distortion, e.g., a penny which is round appears round only from a certain orientation, otherwise it appears elliptical, etc.
This "version" of "third person naive realism" creates what amounts to a dichotomy within the visual world between what is "physical" vs. what is "visual." I suspect that this usage historically goes far back into the study of vision and optics, but in fact, it is an artificial division within the visual world and thus within the theory of perception, because what is being called "physical" is really just another visual viewpoint. Though we have been round and round about this distinction here before, I don't think my point has been taken.
This helps some Bill, but while you describe what happens with the parallel alley instructions, you don't say what happens with the equidistance instructions. Do these also result in a funnel-like arrangement from the experimenter's point of view or something else, and if so what?
ReplyDeleteYes, both result in flaired curves, the D-condition (equidistant) slightly less curved than the P-condition (parallel) from the experimenter's P.O.V. In the D-condition the Ss are asked to make pairs of lights equidistant laterally, but still in rows extending pair-wise away from the Ss line of sight--and thus ostensibly equivalent to parallel pairs of light points, even though it produces to two different curves. For a geometrical model to explain this would require a space that literally changes shape with slightly different configurations in the same place. I don't find that to be a particularly parsimonious explanation!
ReplyDeleteAlready David Noton and Lawrence Stark had shown in a 1972 paper (in "Science") entitled "Scanpaths in Eye Movements during Pattern Recognition" that the eye movements in learning and then recognizing the same picture are not the same, so it is not surprising that they would be different (as I predict) for these two different experimental tasks.
Of course it never seems to have occurred to any of these experimenters that if the alleys look one way to the experimenter and another way to Ss, their configuration is dependent to some extent upon P.O.V.--thus implicating the role of projective geometry. Truth to tell, I always found the whole experimental paradigm to be a bit daft!
ReplyDeleteThanks for the information Bill. I am not at all sure what to say about the parallel alley results, but I would be inclined to explain any deviations in the distance alley results away from those given by a projective geometry as being due to the tendency towards size constancy, which is greater in binocular vision than monocular vision.
ReplyDeleteThe problem with size constancy is that it is little more than a name, and to the best of my knowledge, has never been included as a variable in geometrizations of visual space. It is only a problem in mapping perceived configurations to "objective"/"physical" correlates, otherwise we would be unaware of size constancy within the visual world alone, where we take it for granted that most solids do not change size or shape.
ReplyDeleteHopefully Lothar will be shedding some light on Gestalt factors which also need to be taken into account in geometrizations, as I think there is a real ontological problem with Gestalten in general, because they only seem to exist in perception, not independently. If that is so, then we are back to the question of the reality status of certain basic spatial features of "objective" physical reality--but that is a discussion (and has been) for another posting.
I agree with you Bill, that the size constancy tendency is a name (and possibly also a description) of a phenomenon and not an explanation.I do invoke it though in my NOUS paper on the geometry of visual space, and there at least try to account for it in terms of my depth function acting on a sphere (i. e., for physically closer physical objects, their reconstructions in visual space are construed as possessing as possessing a shorter radius, and thus their area in the space is less for covering the same solid angle).
ReplyDeleteThe most basic question that any experimental psychologist would ask looking at a graph of a curve is: What variable is responsible for the source of the curve? Why is the function graphed not a straight line?
ReplyDeleteIt is an easy assumption to make that postulated positive curvature of visual space might be attributable to the fact that the eyes are round, though that would hardly explain things if the geometry of visual space possessed negative curvature as Luneberg claimed, let alone curvature of variable sign (ranging form positive to negative) as Indow claimed, and that is is a function of configuration, rather than the shape of eyeballs.
One might therefore argue that the curvature of visual space is in some sense dynamic and not solely a function of the projective surface of the retinas. But if that is so, where are those factors introduced? The model you give above, Bob, is less and explanation than a geometric description.
I think that the curvature is dynamic, and is a function of the physical distance away of an object in a given direction. I develop this in some detail in both my NOUS paper and in my posting on apparent distortions in photography.
ReplyDeleteIn physical space-time, the curvature is caused by gravitation, according to Einstein. What causes it in visual space? Is it due to the shape of eyeballs or not?
ReplyDeleteI have actually seen a book, W. H. S. Monck's Space and Vision published in 1872, which argues that the immediate neural correlates of visual experiences are in the retina, but such a theory has obvious problems in accounting for binocular depth perception. Thus, while suggestive, I do not hold that the shape of the eyeball is causally responsible for the shape of visual space. Further neural projections clearly do not have a spherical metric, and at most have a topological isomorphism with the geometry of visual experiences.
ReplyDeleteThus, while I am satisfied that something happening in the brain is causally responsible for the dynamic variable curvature for visual space, I must plead ignorance of what it is other than to say that depth cues must be tied into it. If anyone can make a constructive suggestion as to what it might be I would be very interested in hearing it.
All of this is very interesting, Bob, but it sounds like there is neither any specific cause for the curvature or even an obvious reason for it that might be useful to perception, since it is usually only discussed in terms of distortions or deviations from perceptual norms. Why should the brain introduce curvature into visual space then? Also, can you define "topological isomorphism" since isomorphism is ostensibly not a topological property?
ReplyDeleteBy a "topological isomorphism" I am referring to a mapping which maps close points to close points, but where a metric is not necessarily preserved.I don't know of a better word than "isomorphism" to refer to this, but if you can think of something better please let me know.
ReplyDeleteRegarding your first question, I am afraid that I don't have anything particularly intelligent to say, although I wish that I did. It might help if we had a better understanding of how the brain works with depth cues.
I'm not really up on point set topology, which seems to be where this might apply, and it is certainly relevant to talk about geometrical properties of the pattern of retinal stimulation being preserved or transformed as you say with respect to close points.
ReplyDeleteBut more and more the putative curvature of visual space seems like an unexplained anomaly, almost an artifact, without any obvious source as the curvature of the eyes). How does a real space change its shape without something like a physical force being involved?
For example, from the literature of neuropathology, I am unaware of topological changes such as stretching occurring in visual space, which would result in changes of shape and/or size. This just isn't reported. There is micropsia and macrosopsia, in which everything appears either too small or too larger, respectively. We hear only of global transformations such as that, or VS being inverted or reflected, though having just written that, I seem to recall reports of rectilinearity of VS being changed, so that there are no longer right angles, but just trapezoidal shapes instead. Again that is a global transformation, not a local one. What then is preserving this global geometry?
You ask a good question here Bill about what is preserving a global geometry, but I am afraid that I don't have a good answer. One could try to speculate about unknown forces and things, but that sounds pretty ad hoc. I think that the best that we can do for now is just to try to identify relevant considerations that might help to either home in on an explanation, or to eliminate possibilities. One thing that I think would help, that you previously mentioned you would look into, would be to repeat both the alley experiments, and possibly ones on the Helmholtz checkerboard controlling for eye movements.
ReplyDeleteIn the development of CGI (Computer-Generated Imagery) for animation purposes, to simulate life-like movement and behavior of virtual objects, certain physical parameters were introduced into the computer imaging models to achieve that (the results can be see at your local movie theater or in any video game).
ReplyDeleteJust as the Gestaltists believed, I think we need some sort of field effect in order to explain the robust geometry of visual space. The mistake of the Gestaltists was, as I have noted, to assume that the field was being generated by the brain (as was refuted by Lashley in his experiments of inserting plates into the brains of animals). William Sickles, who was aware of those experiments, investigated other forces/fields in the course of his own Gestalt-oriented research.
As I have said before, we really need a biophysicist on board here! Perhaps Simon may have some thoughts about this, because he has been thinking along these lines?
Bob, several exchanges ago you asked if you could be of help with the proposed testing of alley experiments controlling for eye movement. In parallel with the experiments we need to develop a geometrical model that would (perhaps) involve modeling VS as rotating around a hemisphere. My concept is to "reverse engineer" the geometry of visual space (as it were) in attempt to identify/isolate the factors most immediately responsible for it (I could send you some preliminary computations made for that purpose done years ago by a Caltech mathematician, based on Indow's data curves.) Such a model could be generated via computer modeling which, in turn, could be checked interactively with the experiments.
ReplyDeleteDo we also need an expert on geometrical optics for this purpose?
Bill: I was much interested in your remark—
ReplyDelete“I am unaware of topological changes such as stretching occurring in visual space, which would result in changes of shape and/or size. This just isn't reported…. I seem to recall reports of rectilinearity of VS being changed, so that there are no longer right angles, but just trapezoidal shapes instead.”
The old psychedelic literature contains a lot of possibly relevant material. I refer to Beringer’s “Das Meskalinrausch” (1927), Rouhier’s “Le Peyotl” (1927) and Kluver’s (1928) “Mescal”. I have not read these for 50 years so details are hazy, but I recall that they record a remarkable range of sensory distortions. Many of these relate to illusory movements, as when previously still objects start to move (e.g. a dead fish on a dinner plate starting to writhe about): or the opposite (e.g. when a lighted cigarette being whirled round in a circle is seen as ten stationary glowing balls spaced out along the path of the circle). But some refer to shape. From my own experience I recall that the walls of the room changed to a, not rhomboid, but a pyramid-lying-on-its-side shape, and one open door clearly fused with the wall beside it.
In somatic sensation also there may be remarkable changes. Parts may grow or shrivel. The sensed limbs may detach themselves from the body and lie on the floor. One subject reported that his body came apart at the waist and the lower half danced off by itself. Hard objects may be felt to be soft and malleable. The subject then feels an irresistible desire to indulge in the novel experience of moulding stones and kneading the walls of the room.
ReplyDeleteAlso a comment on CGI: I have been trying, in vain, to find out for some time if any digital TV systems, when processing the information that will go to building up the final TV picture, subdivide the computations into those concerned with movement, with color and with shape? Or is it all done in one process where these three operations are combined? My friend Max D’Oreye is a Belgian TV development engineer. He told me how TV systems use information compression technology but has failed to come up with an answer to my present question.
One point on Lashley's experiments. I believe that they were directed at a theory of Wolfgang Kohler which tried to identify visual experiences with direct currents in the visual cortex, which sounds like a version of a dual aspect theory to me. While Lashley may or may not have succeeded in refuting such a theory, as far as I can see it does not refute a theory of spatial substance dualism such as mine (and I believe John's) since these do not numerically identify the visual experiences with brain processes per se.
ReplyDeleteThe reports from "psychedelic" experience pose an interesting challenge for the study of visual form, to be sure. These metamophoses that are typically reported effectively are removing some of the constraints on the behavior of objects in the visual world, just as you describe, John. But these may not constitute actual local alterations in the geometry of visual space per se. What would expect of such alterations of visual space per se would presumably be more like scotomata, as they would be constant in one place in the visual field. For example, if the metric was not constant, every time an object moved into one region of visual space it would perhaps shrink or grow in size, or consistently change in shape in some way.
ReplyDeleteIn that very long paper by Schilder on the vestibularis that was published in the 1930s, I seem to recall that some of his patients reported some peculiarities of VS (I'll check the article). If we think of the vestibularis as being analogous to functioning like gravitation in relationship to the construction of the perceptual world, perhaps even controlling/regulating the global geometry of the perceptual world, this makes a certain amount of sense, when one realizes that they are responding to whole motions of the head or body relative to the environment.
One must distinguish between CGI and digital image processing, because computer-generated images may just model the physical behavior of objects. For example, the famous spaceship "air" battles in STAR WARS used actual WWII fighter jet footage as a model to generate the behavior of the spaceships.
ReplyDeleteIn a digital visual array pixels are "addressed" by a place code (where they are to appear in an array on a computer display). Such place codes are not a function of any encoded shape, motion, or color encoding per se. It is in some ways not much different from a half tone color photograph made up of dots. Simon could probably explain this more fully.
The Gestalt theorists thought that Gestalten might be a function of EM field effects in the brain, not currents. What Lashley and Sperry found by inserting both insulating and conductive material in the brains of cats and monkeys after they had previously learned certain visual discrimination tasks was that it did not affect their ability to perform them. They thought that dispelled the idea of any field effect being responsible for isomorphism, but Köhler rebutted, and I don't think they ever countered his arguments. Obviously it is a more complicated issue than any of them at first thought!
ReplyDeleteHere is a very rare report of the type of 'local' psychedelic effect you mention Bill.
ReplyDeleteAuthor: Shanon B.1
Source: Journal of Consciousness Studies, Volume 10, Number 2, 2003 , pp. 3-31(29)
This paper examines the standard conceptualizations of the notion of hallucination in light of various non-ordinary phenomenological patterns associated with altered states of consciousness induced by psychoactive agents. It is argued that in general, the conceptualizations encountered in the literature do not do justice to the richness and complexity that the psychological phenomenology actually exhibits. A close inspection of this phenomenology reveals some pertinent distinctions which are usually not made in the scientific literature. On the one hand, the discussion is based on first-hand experiences and, on the other hand, it is grounded in empirical and theoretical cognitive investigations of the phenomenology of human consciousness. Theoretically, the discussion is grounded in an approach highlighting the centrality of experience, meaning and action in cognition.
Document Type: Research article
Example 1: The golden city
ReplyDeleteThis is one of the very first, and most powerful, experiences I have had with Ayahuasca. It took place during day time when I was sitting on a small bench looking at a grove of the Amazonian forest. When the brew had its effect, I saw an enchanted city, all constructed of gold and precious stones. It was of indescribable beauty. The scene that I was seeing appeared to be in front of my eyes and, at the same time, separated from me — just as a scene in the theatre would
be. Every now and then I would turn my head aside and away from the scene of the vision. Returning my gaze, I would come back to the same visionary scene I had inspected before this turn.
Example 2: The enchanted forest
The occasion was the same as described in Example 1. Sitting there by the Amazonian grove, I saw the forest in front of me full of animals — both zoological and mythological. Amongst others, these included dragons, tigers and big birds. With my eyes open, I was sitting viewing the forest as if it were a stage. It was as if a screen were raised and another world made its appearance. Indeed, it was as if
the forest was revealing its mysteries to me. It was all blissful, and very real. The experience coupled together exquisite aesthetic gratification, a profound sense of personal well-being and a feeling that a new dimension of reality was revealed to me. Now, almost ten years after this has taken place, I still consider this experience
as one of the most beautiful, most enriching ones of my entire life
Affiliations: 1: Benny Shanon, Psychology Dept, The Hebrew University, Jerusalem, Israel ., Email: b.shanon@mscc.huji.ac.il
Thanks for sharing these, John. The sense of separation reported in the last account (as if 'on a stage' in a 'theatre') seems to signify that the experience is in toto somehow other than normal vision, which does not appear as if on stage--or a perhaps a fourth dimension? So that sense of the psychedelic vision being framed, as it were, and separate from the observer is telling us something--but what?
ReplyDeleteIt occurs to me that the mundane character of most everyday visual experience almost serves to partly define it as *not* being otherworldly; rather, it mostly turns back on itself recursively like a closed system, as if by its very structure specifically excluding the visual phenomena that typically characterize visionary experience (or "transport" to use Huxley's word), mostly due no doubt to what Gibson might call perceptual probabilities: Things tend to be the same in the perceptual world from one experience to the next, and thus lead to certain kinds of expectations (aka "expectancy effects").
The resulting mind set of everyday perceptual experience of the "ordinary" world seems to give credence to skepticism, may indeed be its very foundation, i.e., "things like X don't happen, and therefore are not real." But that reality is perceptual reality, because the quantum world seems to be pretty bizarre by comparison!
Well, John, I would not dismiss your experience as hallucinatory just because it diverges from the ordinary experience of everyday life. No, but I would dismiss it for one or other or both of two reasons. (1) It does not add up or work out even on its own terms. (The pictures Escher made of impossible stuctures are this way.) (2) It conflicts with other things that I have better reason to believe than what is claimed. (The claim that someone saw Elvis alive after 1977 would be this way, although working it through in detail would be complicated.) Now, the claim that what you saw looked as if it were on the stage, or on the screen, is what makes me suspicious as much as anything. For all I know in any hard line way, there may be wondrous beings present even on this Earth with all kinds of exotic attributes and powers, including invisibility to ordinary people. But then, if I took some drug that enhanced my faculites for a time so that I could see it, I would expect to see it in about the same way I used to see my brother's dog running around his house. I would not expect to see it as an overlay. Or have I misunderstood when I speak of it as an overlay?
ReplyDeleteDavid, sorry for the mix up. The body of my post is a report by Shannon not me. He, and not I, had the unusual experiences. The comments on them are his.
ReplyDeleteAs they are so unusual, in fact unique in my extensive reading on this subject, I feel we need to find other reports of this "shared" phenomenon before coming to any firm conclusions about them. There is quite a lot of work going on at present on ayahuasca so this topic may come up again.
Hear, hear, John. This really needs to be the topic of a new posting, more than a continuation of this one. Persons suffering from derealization sometimes report that things appear as if completely detached from them, as if on a stage, so this is not unique to hallucinatory experiences by any means. Rather, derealization appears to be common element in psychedelic experiences.
ReplyDeleteMethinks there is a note of skepticism in David's comments. Perhaps David would like to start a new posting using his somewhat cryptic remarks as a point of departure?
Well, I reckon I am kind of skeptical. But I am concerned to apply the skepticism at the right places and for the right reasons. In fact, in one way, I am the opposite of skeptical. I am prepared to accept even such experience under drugs as manifesting genuine reality, if it can stand up to sober criticism.
ReplyDeleteWhether or not you are skeptical by temperament, David, what is it in this context that you are skeptical about? Surely you are not denying that people hallucinate on hallucinogenic drugs?
ReplyDeleteI talked with Benny Shanon for quite some time about the hallucination described above (at a conference on "consciousness" no less). He made comments similar to what David McGraw said here, that ayahuasca-induced experience is not so much hallucination as participation in a genuine alternate reality.
ReplyDeleteI pointed out to Benny that the elements of the vision, while distorted and fantastical compared to his ordinary experience, were nevertheless elements of ordinary reality: buildings, landscapes, animals, and shapes, edges, surfaces and colors with which we are familiar. Wouldn't that seem improbable for an arbitrary alternate reality? Isn't it more likely, I suggested, that the drug had induced distortions to ordinary perceptual experience and their common associations (in memory)?
Shanon's argument was that there was no sense of derealization at all; in fact the reverse was true. He felt completely connected in a visceral, emotional way to the vision/experience, which had a sense of beauty and awesomeness, but not strangeness or alienation. Yet, at the same time, experience was suffiently self-alien that he was convinced that it was not self-constructed, but independently existent. His argument for an alternate reality therefore was based on his intuition of self vs not-self, not on analysis of visual or other perceptual elements.
I think Shanon's point of view goes to the heart of the issue between appearances and reality (visual vs "physical" world).
With respect to Bill R.'s comments on CGI and virtual reality, I would like to draw a possibly useful distinction. When you put on a heavy duty VR helmet (more encompassing than a video game), and look to the side, you have the experience that you simply look to a part of the VR landscape that was already there. You do not have the sense that a new landscape has just been painted on the screen at the moment you turned your head and eyes, even though that is what actually happened. You feel that you have looked to the right, visually reorienting your mental exploration with respect to a pre-existing objectivity.
ReplyDeleteThe computer monitors your eye movements and neck muscles. As you move your head and eyes around, the unfolding of the landscape is precisely coordinated to your exploratory intentionality, just as in the real world.
The experienced VR landscape is real to the visual system, as real as any reality can be, by any criterion you could use (with the limitation, of course, that such VR systems are not yet sophisticated enough to be 100% compelling, but they are getting there!)
As far as the visual system is concerned, the VR landscape is not “virtual” at all. It is the set of objects you happen to be visually exploring right now. That reality is only called virtual because intellectually you know that later you will take the helmet off. If you somehow forgot that, or if you did not have the capacity to think of it, you would just accept the visually presented scene as the real visual world of physical objects that you were in at the moment. There would be no reason to doubt it. Seeing is believing.
For those who have not experienced the compelling visions of even mascalin as I have once, it is impossible to describe how real everything appears--so much so, in fact, that one may despair, as I did, whether things would go back to how they were before the drug was taken, as the experience lasts many hours.
ReplyDeleteHuxley attempted to capture something of the surprise and astonishment in his "Doors of Perception" because, as he explained, he was not ordinarily much of a visualizer, even with great effort, such that nothing could have prepared him for what he saw on mescalin (courtesy of the good Dr. Smythies and his colleague, Humphry Osmond).
In short, the hallucinations are as real as ordinary reality, there is no sense or clue other than their strangeness and beauty to suggest that they are anything but real, nor even that one is hallucinating. Indeed, seeing is believing, and it is as if one's everyday visual world had simply been replaced by an alternate world. There is no subjective sense that one is creating the visions. When one closes their eyes the visions are even more extraordinary, of vast cities of glass geometries shining in the most splendid colors that one could experience.
To liken such experiences to "distortions" IMO is to do violence to what are better described as the most profound sights imaginable--or even beyond imagination. That is to say, no one who has experienced these things would use the word "distortion" to describe them. "Transformation" seems even rather an inadequate term, because our quotidian world is simply replaced temporarily by another one altogether: It is just "another world."
Bill Adams's very useful remarks about VR support my contention that the term "virtual" is best confined to the *content* of visual space rather than to the visual world per se, as John has proposed. Visual space would only appear to be "virtual" if it were--somehow--regarded from a vantage point outside it by which to compare it to the physical world which, needless to say, is a perspective none of us is able to assume, and more than the Archimedian point Einstein sought in despair, and which then gave rise to theory of relativity.
ReplyDeleteAgain whether "distortion" or "transformation," one is still referring to relations within the visual world, not between it and the putative physical one.
ReplyDeletePerhaps I misunderstood David's comment above, but I took him to be saying that he was denying that people had such experiences altogether, not that they were just hallucinatory, as Bill Adams seems to be addressing. Perhaps David can clarify?
ReplyDeleteThe perceptual world may certainly be *analyzed* in terms of features in it such as buildings, landscapes, animals, and shapes, edges, surfaces and colors, as Bill Adams says, but that is to extrapolate from something that exists as a holistic entity (e.g., the visual world), that which we ordinarily call "reality" (at least most people, if not philosophers).
Again, if one conceptualizes what one sees under the influence of hallucinogenic drugs as being in some way related to that perceptual reality that we ordinarily call reality, one is once again relating percepts to percepts, rather, as I would maintain Lothar is doing, deeming one to be "veridical" and the other a "distortion." But I would again claim that these are relations *within* the perceptual world, not *between* it and the physical world.
The question remains what those percepts, whether ordinary or hallucinatory, are "of"? This is where the classical (or physical) realist steps in to save the day, assuring us that there is--somehow--an external reality beyond the senses that corresponds to what we perceive (with or without the aid of scientific instruments), even if it can only be known in the most abstract of ways through group theory.
One important finding Lothar reports in his book is that the formative processes of percepts (so-called "Aktualgenese" or "actual genesis") seem to be the same whether for ordinary seeing or clairvoyant "second sight." So in some sense the "machinery" of perception is operating in the same way in both conditions, which again is consistent with both Moncrieff's "clairvoyant theory of perception" and Bergson's notion of the brain as a "reducing valve" which Huxley invoked in his "Doors of Perception" relative to interpreting his mescalin experiences, and that reality is somehow much more complicated, something like a superposition of different states, of which we perceive only one (this is Saul-Paul Sirag's view in his theory of "multiple realities").
Hi Bill,
ReplyDeleteI understood David to question whether hallucinatory experiences are properly categorized as such, instead of being merely "alternate" experiences. I did not read that he was denying the existence of such experiences.
I thought you concurred that intense (so-called) hallucinatory experiences often entail an aspect of realism that is so phenomenologically convincing that it is not doubted, let alone denied. True for you, anyway?
Following on that idea, I am suggesting that a robust criterion for the "reality" (veridicality) of perception is belief, not correspondence. Such belief is engendered, when exploratory intentionality is well-satisfied by anticipated sensory input. That occurs in most ordinary perception, and in high quality VR environments, and possibly also, under the influence of certain psychoactive drugs.
Under that analysis, the question, "what are the percepts 'of,' does not arise, because the correspondence theory of perception is not presupposed.
Bill Adams
Bill,
ReplyDeleteI also wanted to ask you what degree of analysis of perception you deem to be appropriate? You reject analysis down to animals, landscapes artifacts, and visual elements, and prefer a "holistic" view of the visual world.
If we accept such a preconceptual visual world without analysis, how can we discuss it?
Bill
All good questions and points, Bill.
ReplyDeleteHopefully David himself will explain what he meant so we don't have to guess :}
In addition to the Gibson or quasi-Gibsonian criteria of normal perception which are met typically in mescalin experiences (such as John and I here have both had), Lothar discusses findings in his book that complement that view, that the actual genesis of perceptual experience is the same in both ordinary seeing and "second sight" (he does not use the latter term, though). That is pretty interesting IMO.
My view is that reality, whether physical or perceptual, is constructed, the result of a process of construction--of some sort. What Gibson describes by and large is perception *within* the perceptual world, not something going on between it and the physical world, because the "picture" he gives of the physical world is just a representation of the perceptual one (or simulation of it).
We have discussed the phenomenon of derealization some here already, and I believe it bears on your question, Bill, because the change that seems to occur in that experience--an experience by no means confined to hallucination, but one that can occur in many different psychological contexts more or less in normal experience--is that what we call real no longer seems real. What is changed, though, is not belief per se--that would be a secondary effect--but something about the *quality* of the perceptual experience itself is altered (as in altered state of consciousness), a qualitative change that may only perhaps be described as if things seem no longer real (!) That suggests that reality is quality of whole experiences, what the Second Leipzig School of Psychology called a "total quality" or "quality of the whole," not just a change in some feature of, say, visual experience, but of the whole visual world. These qualities of the whole (whole experiences) are akin to what von Ehrenfels called "Gestalt qualities," only rather than applying to Gestalten, apply to the whole of an experience itself.
Of course people hallucinate under the influence of drugs, and of course there are hallucinatory experiences. There is no question about that. The question is exactly about whether some few of these experiences under drugs (and if so, which) involve genuine awareness of some alternate or exotic reality, as oppose to being merely hallucinatory. I am not skeptical regarding the existence of such experiences, taken simply as episodes in consciousness. Instead, I am skeptical regarding the claim that some of these experiences latch onto to some other reality.
ReplyDeleteIt is not clear to me how one would go about answering this question, any more than one can determine "genuine awareness" of ordinary reality to be genuine other than through what Freud called "reality testing" which, to some extent, can even be performed during so-called lucid dreaming, such that the dream world demonstrates internal consistencies.
ReplyDeletePerhaps the most disturbing aspect of the hallucinatory experience is that most of the time people are unaware that they are hallucinating. The hallucinated world has simply replaced the ordinary one. That being so, one can readily surmise how such experiences may give rise to the conception of alternate realities.
But one doesn't need to hallucinate in order to posit other realities--the Surrealists maintained that this was what they were doing through poetry and art, achieving what they called *surreality,* a higher order reality, as it were, to the quotidian one.
Motivated by an attempt to account for paranormal phenomena, the theoretical physicist Saul-Paul Sirag proposed not a "many worlds" hypothesis, but a theory of "many realities." So this sort of speculative thinking is found in physics, not just metaphysical musing by those who have had hallucinatory experiences.
Further to the above, I am going to start a new posting using some comments from Wittgenstein's "Philosophical Investigations" as a point of departure, because we have now veered far away from the topic of my posting above.
ReplyDelete