Back to CQ Homepage

Back to Archives' Table of Contents

The Commentators Respond

Jim Garson            John Bickle

Mark Churchland           Sue Pockett

Peter Lloyd           Andy Clark

U.T. Place           Gary Schouborg

John Bickle (post 2)           Sue Pockett (post 2)

U.T. Place (post 2)           Thomas Alexander           Bo Dahlin

Jim Garson writes:

      I have a problem with your view about computers presupossing sense datum theory. True the classical doctrine has information coming in. But isn't that a triviality? I mean the brain just is affected by the input vector (i.e. activity values of all sensory neurons throught time). It is a big leap from that uncontroversial fact to the idea that there are sense data coming in. First of all, not all of the classicist's input vector qualifies as sensed, for a lot of the information provided fails to be part of our awareness. Even parts that COULD BE in awareness are often not as attention changes. So the classicist's inputs are not identifiable with sensations. Second it is not clear to me that even the inputs of which we become aware count AS sense data, for it is far from clear that the inputs could even possibly be encountered AS sensations. For that, one would need to encounter inputs related to others in the right kinds of ways and this would involve finding the data within a cognitive level theoretical structure of some kind. Note that the standard classicist hopes to provide a functionalist account of mental states. So one expects the classicist to be either mum on sensation or to provide a functionalist account of sensations. But if the latter choice is taken, then raw inputs by this very theory cannot count as the sensations, since their individuation AS sensations requires a framework of interrelations provided by the functionalist theory that undergirds the identification of mental states. I am reminded of a parallel moral to be drawn about computers. Imagine we have a simple caclulator set up to do addition and subtraction. The raw input might be arrays of 0s and 1s (or more carefully, switch settings). These have no status as NUMERICAL inputs until we encounter their computational roles defined by the machine. Computer states do not take on their intentional properties in a vacuum. This is the sense that inputs mathematically construed (i.e. intentionally construed) are never "raw". This is an important Kantian theme in classical computationalism that seems to me to challenge features of "whole hog" sense data theory. Further evidence that the horse is long dead.

      John Bickle Writes:

      I wanted to correct a factual error in the comments you quote from Sergio Chaigneau. While his conceptual point about inputs to the LGN is correct--80-90% of synapses on LGN relay cells come from axons other than retinal ganglion cells, the majority come back from visual areas, primaerily V1 (primary visual cortex), but also some from sites higher up in the visual streams) and from brain stem nuclei that compose the (somewhat misnamed, especially concerning functional significance) "reticular formation" (used to be called "reticular activating system). These nuclei are important for general arousal states. LGN relay cells also receive extensive projections from cells in the reticular nucleus of the ventral thalamus, a thin network of exclusively GABAergic (hence inhibitory) neurons that project extensively back into dorsal thalamic nuclei. (The region of the reticular nucleus that receives collaterals from LGN relay axons bound for V1 and projects densely back into LGN (onto both relay neurons and intranucleus inhibitory interneurons) is called the perigeniculate nucleus.) No doubt that there are some synapses from "motor areas" (but there are a lot of motor areas, and without a more precise description it is impossible to evaluate even the approximate truth of Sergio's claim), but the bulk of motor projections back into the thalamus don't go to LGN. It is a mistake to cite Varela as an authority on the anatomy of the thalamus. The definitive work is Ed Jones 1985 book, The Thalamus. It has a chjapter on LGN and on reticular nucleus. You might also take a look at the chapter from Mason and Kandel in Kandel, Jessell, and Schwartz's definitive Principle of Neural Science (last volume published in 1991, a new one either now out or shortly comning out.) If you wish more detailed accounts of the neuroanatomy of thalamic nuclei, I can send you those as well (they are by real neuroanatomists, published in real neuroscience journals, not the kinds of things that cogntiive scientists and philosophers read and discuss--to the latter's detriment). My computational neuroscience group has published a paper in abstract form describing results from a computer model of LGN-V1-reticular nucleus circuitry. The results are strongly suggestive that this circuitry provides a mechanism for stimulus-driven selective visual attention. (The abstract is in Society for Neuroscience Abstracts 23 (1997), pg. 1589 (entry 620.7). We're currently preparing a paper for submission to the Journal of Computational Neuroscience describing the model, the results, and their implications for a mechanism of selective visual attention. If you wish, we can send you a copy when it is ready to submit.

      Mark Churchland writes:

      As a side point, '# of synapses' is not necessarily a good measure of the ability of one area to drive another. It is often the case that feedback projections outnumber feedforward projections. However, such feedback connections may be much weaker. This may be visible anatomically (smaller synapses/ farther out on the dendrites) or it may not. I don't wish to imply that feedback connections are never as poweful as feedforward ones. I only wish to point out that '# of synapses' can be a misleading metric of connective strength.

      Sue Pockett Writes:

      With regard to the word qualia, I'd suggest you examine the traditional psychologists' distinction between sensation and perception. My understanding of the word qualia is that it could usefully be replaced by the word sensation, used in this restricted trade sense

      Peter Lloyd Writes:

      1. I can't see how a non-sensationalistic epistemology can ever get off the ground. The mind is (amongst other things) an information-processing system that uses information successfully in interacting with its environment. Therefore the mind possess information. But it is a fundamental characteristic of information that it is *not* endlessly reducible. The analysis of information comes to an end with raw data. Therefore the mind possesses basic units of information, which we might call 'raw data' or 'sensations'.

      2. Talking to pragmatists in the past, I've got the impression that they think of sensations only as 'naive' sensations. For instance, if you think of your visual field as being consisting of pixels of visual sensation like a TV screen, then that's what I'm calling a naive view of sensations. In fact, the sensations that make up the visual field must be more complex and subtle than that. Nevertheless, the visual field still consists of sensations. (I am thinking here of Hubel & Wiesel's experiments, and my own experiences of migraine 'castellations', which suggest larger visual elements as the raw units.)

      3. As far as I have gleaned, the main attraction of pragmatist epistemologies is that it explicitly accommodates the crucial role of mental *activity* in the acquisition of raw data about the world. I do not, however, see any conflict between perception's being active and its yielding sensations. Just think about a robot exploring its environment: it carries out probing actions, such as switching on its lights and cameras, and gets raw data back. Likewise, the mind performs actions, such as listening to sounds, and gets raw sound sensations.

      4. Of course, sensations do not come *into* the mind from outside it (as John Locke seemed to think). They are constructed internally in response to incoming stimuli. I acknowledge that some philosophers would dispute this, but to me it is incontrovertible. For instance, the qualia of red does exist in the rose I look at: what the rose provides is elecromagnetic radiation in a particular range of wavelengths, which yields a sensation of red in my conscious mind. That red sensation is the raw datum.

      5. Another thing with which anti-sensationalists seem to have difficulty is the presence of pre-conscious processing of sensory data. This pre-processing does *not* count against the sensationalist hypothesis. Take, for example, the filling-in of shapes, such as the lines of a broken triangle. Clearly the brain is carrying out pre-conscious processing of the visual data, to fill in the gaps. Nevertheless, what the brain then delivers to the conscious mind are indivisible, unanalysable units of raw sensation. Another example would be hearing phonemes belonging to a language with which we are familiar: those phonemes *are* the raw conscious sensations. (Aside: Some people I've conversed with just seem to unable to comprehend this, no matter how many different ways it's expressed. I guess it's because they're locked into a mind-brain identity metaphysic, in which - e.g. - the phoneme is identical with the signals in the audotory cortex and therefore necessarily a compound entity. You do not, however, have to buy into mind-matter distinction to accept my point here. You can take an analogy from social constructs: legally, I am indivisible person, yet that legal entity of the person corresponds to a compound entity of the physical body.)

      Andy Clark Writes:

      At times, I felt you just might be failing to distinguish the idea of sense data from the idea of MERE INPUT. The point about sense data was that they were meant to be both RAW and yet somehow REALLY CONTENTFUL: a combination that looks ultimately incoherent. All that information theory really needs is the idea of an input to the system, and a set of actions to select from on the basis of that input.

      Re the knowing-how/knowing-that business, I agree with Garson that we need to beware fruitless distinctions. I think the point about data and process can be put somewhat differently (and this bears on what I was doing in BEING THERE) viz, as a difference between inner encodings that are distant from action control (in that the route from the encoding to the action is computationally expensive) and ones where the encoding is formatted in a way that is directly tailored to the control of appropriate actions. An example of the latter would be an encoding of proximal food location that JUST IS a motor command to move an arm to a spatial location,and to grasp and ingest what it finds there. Ruth Millikan's PushmiPullyu reps seem to me to be in this ballpark. Notice that this distinction applies as easily to connectionist systems as to classical ones.

      Ullin T. Place writes:

      A few points in response to your `Beating of an undead horse'. The first relates to the difference between James' "big blooming buzzing confusion" and my Wundt and "physical" v. "mental" pleasure and pain cases. What is true is that James identifies what he takes to be an actual case where experience (as a whole - yes) remains uninterpreted, because the child hasn't yet developed the required concepts. Wundt's two forms of experience do not involve uninterpreted experience in this sense. It's simply that according to him there are two different ways in which the SAME experience can be interpreted which implies that the experience and its interpretation are two different things. There is no reason to hold on to this view that any uninterpreted experience exists, except perhaps momentarily before an interpretation is arrived at or when switching from one interpretation to another.

      The pleasure/pain case is slightly different. Here the suggestion is that in the case of "physical" pleasure and pain the emotional response DOES NOT DEPEND ON the way the experience is interpreted. Again there is nothing that requires the actual existence of uninterpreted experiences.

      My second point relates to Broadbent's use of the term "evidence". When I use this term in my own work, I always put it in quotation marks. This is because, according to me, it is not evidence in the ordinary sense of that word. It is precisely because the use of the term `data', together with the phenomenalist theoretical framework in which it is embedded, treats sense-data as evidence in the ordinary sense that leads me to say that sense-data do not exist. The same incidentally goes for qualia, if it is taken to be part of the definition of a quale that it is a functionless epiphenomenon. But if you say that sense-data are data only in metaphorical sense or if you allow that qualia have a vital function in the process that leads to sense perception, I am happy to use both expressions and say that a sense datum is a private sensory experience and that a quale is a property of such an experience by which we recognise the stimulus situation confronting us as one of this or that kind.

      What is wrong with treating sensory experience as evidence for the belief that one is confronted by a situation of this or that kind in one's external environment is that we ordinarily use the term `evidence'

      (a) when talking about the relation between two statements or sets of statements, the evidence on the one hand and the hypothesis it is evidence for on the other,

      (b) where the evidence consists in one or more observation statements and where the hypothesis for which the observation statements provide evidence is something that cannot itself be directly observed.

      In the case of the relation between sensory experience and the categorization of it as an encounter with a situation of this or that kind, neither of these conditions apply.

      (a) In the categorization of sensory input there are no statements involved. Sensory experience and the categorization for which it provides the evidence are neural processes which occur in the brains of animals just as much as in the brains of humans. Even in the human brain identifying the kind of object or situation with which one is confronted is a distinct process both from that of naming the object or situation and putting what is observed into words in the form of a statement.

      (b) Contrary to the opinion of the phenomenalists, in the ordinary sense that word we DO directly perceive the objects and situations in our stimulus environment for whose presence sensory experience and its qualia provides "evidence".

      Contrary to the view expressed by Ryle in THE CONCEPT OF MIND, there are cases where we can quite properly be said to observe our sensations and other private experiences. After filling a particularly deep cavity in one of my teeth recently, my dentist asked me to check any pain I might subsequently have to see whether it was caused equally by hot and cold stimuli (good) or only by hot (bad, particulary if throbbing). This, however, is a rather sophisticated form of observation which we learn only AFTER we have already learned to observe what is going on in the world around us. When I say I rejected the doctrine of sense-data more than fifty years ago, what I rejected was the idea that in observing what is going on around us, we begin by observing our sensory experience, formulate those observations in the form of a sentence in a private sense datum language and then use those private observation sentences as evidence for the existence and nature of what we NEVER observe, namely the objects and situations in the world around us.

      That, of course, means that I rejected - here following Wittgenstein - the notion that the observation sentences which provide the foundation of empirical knowledge are sentences in a sense-datum language describing the private sensations of a single individual. What it did not mean is that I denied either the possibility of describing private experience or the idea that empirical knowledge has to be anchored to observation statements. With regard to the former, I have been insisting for more than forty years that our ability to describe our private experience is parasitical on a prior ability to describe what is going on in the public world. With regard to the latter, I have long assumed, but rather more recently begun to insist, that the observation statements which anchor our language to the reality it enables us to depict are statements describing a publicly observable state of affairs (events disappear too quickly) on whose correct description any competent speaker of the natural language or technical code in current use will agree. It is because I take this principle as axiomatic that I describe myself as a behaviorist. See `A my radical behaviorist methodology for the empirical investigation of private events' BEHAVIOR AND PHILOSOPHY, 1993, 20, 25-35.

      One final point in this connection. The relation between a sensory experience and the categorization of the current state of the stimulus environment for which it provides the "evidence" is a straightforward causal relation; whereas the relation between evidence in the ordinary sense and the hypothesis for which it provides evidence is a logical relation. Logical relations such as this can, of course, act as causes in persuading an individual to accept (or sometimes reject) the hypothesis for which it is evidence. But that does not alter the fact that logical relations, as such, are not causal relations. The analogy between the two cases is that in both, it is important for the individual to GET IT RIGHT. The difference is that in the experience-categorization case what the individual has to get right is what it is he or she is currently observing; whereas in the evidence-hypothesis case what the individual has to get right is a verbal description of something that is NOT currently available for direct inspection.

      Another difference is that all the might of natural selection is mobilised to ensure the conformity of our perceptual categorization to the way things are; whereas, except in a handful of cases where getting it right is a matter of life or death, there are only a few relatively weak social sanctions to ensure that our hypotheses are and remain consistent with the available evidence.

      Sergio Chaigneau's mention of J.J.Gibson reminds me of my own excitement when, as a very inexperienced psychology teacher at the University of Adelaide, I read Gibson's first book THE PERCEPTION OF THE VISUAL WORLD when it appeared in (?) 1951. Here for the first time was a psychologist doing experimental work within a conceptual framework entirely consistent with what I had learned from Austin's `Sense and Sensibilia' lectures - so different from the ghastly conceptual confusion of the Gestalt Psychologists, whose work had been endlessly thrust down my throat during my psychology course at Oxford in 1947-9 and which was the principle target of my critique of the phenomenological fallacy in `Is consciousness a brain process?'.

      During the winter of 1955 after I had returned to Oxford from my four years at Adelaide and while I was waiting for `Is consciousness a brain process?' to appear in print, I had the privilege of getting to know Gibson personally. He had a visiting appointment at the Institute of Experimental Psychology where I was registered as a candidate for the D.Phil., a degree which I never managed to obtain. I tried to persuade him, unsuccessfully as it turned out, that his position would be more consistent if he dropped the phenomenological veneer and stated it in a straightforward behaviorist way. Interestingly, I was supported in this by his wife, Eleanor Gibson, who not only worked on perception in animals, but had been a student of Clark Hull at Yale. I have a copy of my correspondence with J.J.G. during this period on file on my computer and could e-mail it to you, if you're interested.{I was and he did. see the Gibson-Place Correspondence on this Website--WTR}

      You might also be interested, in connection with Ruth Milliken's deployment of Ryle's `knowing how' and `knowing that' distinction, in a section of my chapter on `Ryle's behaviourism [sic]' in W.O'Donohue and R.Kitchener (eds.) HANDBOOK OF BEHAVIORISM which is forthcoming from Academic Press. In it I discuss the distinction and suggest that it marks a failure on Ryle's part to study the grammatical objects of psychological verbs with the same thoroughness with which he explored their aspectual characteristics. This left room for Roderick Chisholm to introduce his linguistified version of Brentano's intentionality, thereby generating a new piece of conceptual confusion for philosophers to pick over.

      This, of course, needn't undermine Ruth's thesis which I would express in my behavioristic way by saying that getting one's propositions right depends on a great deal of contingency-shaped learning of semantic conventions which in turn depends on the, part contingency-shaped, part innate, pre-linguistic categorization ability found in animals.

      Gary Schouborg writes:

      As I recall your discussion of the relationship between sense data and qualia, let me suggest that each is a theoretical concept employed to answer a different question from the other. Sense data are employed to explain how we can adjudicate conflicting beliefs. Qualia are employed to explain or characterize 1st-person experience. The Given is relative to beliefs / hypotheses / interpretations about it. The Given is not irreducible, but that which disputants will accept as settling their differences. Similarly, one does not need Absolute Leverage to stand up on her own two feet, only something sufficiently stable to do the job. Even in this relative sense, I don't believe The Given is an immediate given, but a theoretical concept. Seems to me, we don't begin with sense data, but will real chairs, streams, sticks, etc. Since we found our judgments changing, such as with the classic bent stick half immersed in water, we have developed a theoretical concept, sense datum, to explain how such judgments can differ and how they might be adjudicated. Your discussion about sense datum should therefore be placed in that context, to wit: if sense data cannot explain how judgments can differ and how they might be adjudicated, what can? Seems to me, this leads ineluctably to some interactionist view. Note that to say we begin with real chairs, etc. is not to espouse naive realism, which overlooks the mind's contribution to perception. It is only to identify phenomenologically, where we in fact begin. Related to this is a principle which I am currently touting as crucial: Innocent Until Proven Guilty. I believe Plato got us in a mess by saying we begin with opinion and try to move to knowledge by justifying our opinion. This has led us in a fruitless search toward the chimerical grail of foundationalism -- thinking we know nothing until we've provided incorrigible evidence for it -- as if each of was guilty before the law until we proved we were innocent. Rather, I think the way we actually do things is according to the principle of Innocent Until Proven Guilty -- we say we know something unless we have some good reason for doubting it. I think this perspective sheds an entirely different light on knowledge and The Given. To the question, then, of -- Do we know anything? -- the answer is, Of course we do. Asked for an example, we can come up with most anything -- e.g., that this Mac before me is there even when I'm out of this room. It is not for me to prove this, but for you to provide a reason why I should doubt it. As to qualia, I am working on an article, Being and Function, that argues that consciousness explains nothing, only function does. For reasons we do not, and probably never will, understand, cs accompanies some functions. For those who find this an unacceptable counsel of despair, let me point out that we accept essentially this position with regard to the question, Why does anything exist rather than nothing at all? This most medieval of questions is now all but universally considered to have no answer. Not that it is unanswerable in principle -- who can say? -- but it is beyond the scope of explanations that we seem to be able to provide. Unlike the medievalists, we now limit ourselves to explaining only the shifting forms and relationships of things, not why anything is there in the first place. I increasingly see cs as parallel. Functionalism is the (?) science of cs, but the existence of cs itself is beyond explanation, just as is the existence of contingent being.

      John Bickle writes:

      With regard to your last CQ, I have some things to say about Hubel and Wiesel's work, although I'm not sure how much it really pertains to the sensations versus perception distinction in psychology. (With regard to that, I'd suggest looking in any popular text for Sensation and Perception classes in Psychology departments.) Hubel and Wiesel won the 1980 (or 1981) Nobel Prize (they split it with Soerry) for their work from the late-1950s and early 1960s on information processing in visual cortex. Their initial work followed up on the work of Stephen Kuffler, who used electrodes to measure activity in retinal projections to thalamus (lateral geniculate nucleus) upon presentation of a visual stimulus. Basically, they used his procedure to measure the receptive field properties of neurons in primary visual cortex (dubbed V1 by some authors, and Brodman's area 17 by others). They worked on anesthesized cats. At first, they were using slides with black-white contrast as visual stimuli. After almost a year of just trying to get their experimental set-up working, they initially got really disappointing results--activity in V1 neurons seemed almost random. They did get some activity to one slide, however, and discovered that a shadow was falling on it when it was presented, creating a bar of light black-white contrast. They then started showing bars of lights and measuring V1 activity, and discovered that V1 cells do indeed have bars of light at particular orientation and location as their receptive fields. For example, if a neuron was most active on presentation of a vertical bar of light in the upper left quadrant, it would discharge more and more vigorously as the visual stimuli approached that location and that orientation. In later studies (early 1960s), they worked up a theory of how V1 neurons get their receptive fields, based upon the receptive fields of LGN relay neurons, their projection to stellate cells in V1, and the latter's projection to simple cortical cells (in V1). With their work was born the idea of a hierarchy of visual processing areas, where lower cortical regions extract simple information from the visual stimulus (e.g., bars of light at particular orientations and locations) and project this information to higher areas, which extract increasingly abstract information (curves and edges in V2, etc.) Visual processing splits into two streams past extrastriate cortex, a ventral stream projecting into inferotemporal cortex extracting information about a stimulus's identity, and a dorsal stream projecting into posterior parietal cortex extracting information about motion and location. I have a very brief description of these streams in a review essay I published in Philosophical Psychology in December 1997, and the paper I'm going to send you talks in some detail about LGN-V1 projections

      Sue Pocket writes:

      Hubel and Wiesel became probably the most famous neurophysiologists in the world (ah, how restricted fame is in these fields - if ya wants fame, become a rock star or a heavyweight boxer) by doing the following sorts of experiments. They took cats, anaesthetised them and opened their skulls. Then they recorded from single cells in the visual cortex while displaying various visual stimuli to the cat, through lenses calibrated so that they knew the stimuli were focussed on the cat's retina. What they found was that certain cells in the brain would fire action potentials in reponse only to certain very restricted kinds of visual stimulus - a light/dark edge in a certain orientation (say 45 degrees from vertical top to bottom) moving in a certain direction across the visual field, for example.

      They found that cells lower down in the brain (closer to the input end) registered simple features of stimuli (just edges, just movement etc) and cells progressively closer to the outside of the cortex registered progressively more complex features.

      This is of course a very simplistic account of a lifetime of experimentation - you can find more detail in any recent neurophysiology text, if you want it.

      Quite what this has to do with the present discussion is doubtful. I can see what your correspondent means, but he is perhaps overlooking the fact that the cats were anaesthetised and thus by definition not experiencing visual (or any other) sensations. So what Hubel and Wiesel were studying was preconscious processing, not consciousness.

      It would be technically possible to repeat these experiments using unanaesthetised but paralysed cats, but that would be universally considered by the majority of neurophysiologists to be unethical (and this judgement is institutionalised by ethics committees the world over and by the publication code of journals in the field, which would simply refuse to publish such expts)

      for definitions of the difference between sensation and perception try

      Lezak M.D. (1995) Neuropsychological Assessment (3rd Ed) Oxford University Press, New York, Oxford. pp 25-26

      (this is the bible in its field)


      Sims A. (1988) Symptoms in the mind. An introduction to descriptive psychopathology. Balliere Tindall, London Philadelphia Toronto Sydney Tokyo

      (shrinks spend quite a bit of time trying to figure out whether or not their patients are experiencing hallucinations).

      U.T. Place writes:

      I would like to comment on the sensation/perception issue. The traditional view of this matter to which I subscribe holds that sensation + concept = perception. This formula implies that there can be such a thing as a `raw', i.e., uninterpreted, sensory experience. As evidence that such a notion is needed, I would cite the distinction we draw between `physical' pleasure or pain, where the pleasure or pain reaction is a response simply to the quality of the sensory experience, and `mental' pleasure or pain, where it is a response, sometimes to the very same experience, once it has been conceptualised or interpreted, e.g., as a symptom of some fatal illness.

      This notion of `raw' unconceptualised experience is anathema to the Kantians and the phenomenologists; and there are at least three sets of considerations which lend support to their view. One is the relatively trivial point that you can't say anything about an experience until it has been conceptualised in SOME way. Another is the point that the qualia merchants are in danger of overlooking, namely, that an unconceptualised experience is like a unfertilised egg, an entity that has failed to fulfill its biological function. But it is the third consideration which, to my mind, is the most interesting. It is a point which is suggested by a lot of recent neurological and neuropsychological work, particularly the work that has been done on the functions of the extra-striate visual areas, V2-V5. Contrary to what is suggested by the adjective `raw', it is now becoming clear that a great deal of complex processing has to go on in assembling the experience, BEFORE it becomes what Broadbent (1971) calls "a state of evidence" capable of suggesting an interpretation/conceptualisation/ categorization. What seems to happen in visual areas V1-V5 is that there are specific neurons in these areas which are "tuned" to respond to features of the input which become more and more abstract and are triggered by retinal stimulation over wider and wider areas the further removed they are from V1. These features are things like an edge, a gradient of texture (interpreted as a surface at certain angle of slope relative to the horizontal - Gibson 1950) or a stationary object with a background moving to the right (interpreted as watching an object moving to the left - Gibson op.cit.) which are seldom, if ever, conceptualised as such, but which, when "bound" together with other such features result in a recognisable "image" of an object of some identifiable kind. When one way of "binding" a set of features together fails to yield an identifiable object, another way of "binding" the features may be tried and, failing that, the standard reaction is to look again, this time more closely.

      Morover, the phenomenon of simultanagnosia which results from lesions of this so-called "ventral stream" and which consists in an inability to perceive the relations between different objects in a visual array, even though the objects themselves are recognised normally, suggests that the interpretation of a complex visual array proceeds in two stages. In the first stage the individual objects are identified. In the second the experience/"evidence" is revisited in order to conceptualise the relations between them.

      The complexity of this process and that of the processes of response-selection and response execution which ensue, not to mention the linguistic processes of assigning a name to a concept or a concept to a name and of organizing and deciphering complex sentence structures, explains why it is that only PROBLEMATIC INPUTS (i.e., those that are either unexpected or significant relative to the organism's motivational concerns) are processed in this way. The task of separating the problematic from the unproblematic, alerting consciousness to the former, while either ignoring the latter or routing them automatically and unconsciously along well-worn channels to output, falls to the automatic-pilot or "zombie-within" as I call it

      Thomas Alexander posted this to the Dewey List, partly in response to my posting of the first CQ post on that list, and Frank Ryan's reply

      Just a note to Frank Ryan's fine posting. I agree with him, especially regarding the silly criticism that pragmatism "lacks foundations." You don't need "absolute foundations" to build a house--the degree of solidity of the foundation is related to the function, size and context of the building.

      But all this talk about "foundations" is recent (I never heard it used in graduate school). It is a metaphor, perhaps ultimately coming from Descartes' Discourse but popularized by Rorty in 1979 in Philosophy and the Mirror of Nature. It is a bad metaphor.

      Instead I propose that we talk about "roots" instead of "foundations. So many philosophies today are "rootless" instead of "rooted." Rorty, comes to mind. If philosophical systems are living and growing things (and they are), it makes more sense to speak of their "roots" than their foundations anyway.

      The person who used this metaphor irst, by the way, was Empedocles of Akragas who called the ultimate principles of nature (Earth, Water, Air, and Fire) "rhizomata"--the roots of phusis, of nature, of "that which is born." A far better example to follow than a mechanist like Descartes!

      Bo Dahlin posted this to the Dewey list:

      Dewey had a streak of radical experientialism: all philosophical questions should be answered from reflection on concrete, lived experience, not from abstract conceptual analysis. We do not experience "sense data", unless we take an artificial attitude to our experience, therefore sensationalism is false. Merleau-Ponty argues exactly the same.

      From this radical experientialist point of view, I perceive a fallacy in your argument, viz. to exclude thinking from experience in general. Dewey also had a tendency to look only at the outer "doing" and "suffering" aspects of experience, but in some passages he reasons in a way that implicates THINKING in itself as a form of EXPERIENCE.

      If thinking is admitted as a form of experience, we do not end up in absolute scepticism, because that which is lacking in sense-experience, viz. SELF-REVELATORY MEANING, is present in active, conceptual thinking. Naturally we may still be *mistaken* in our comprehension of the world, so there are still no grounds to be 100% certain of anything we know. This uncertainty is, however, contingent, not principal, as in scepticism. In active, conceptual thinking we always have a basis, a "foundation", on which we can proceed towards deeper and deeper *experiences* (sic!) of truth and reality.