In practice, the inner representation explains many things. From the way someone draws a house, a neurologist can localize lesions in the brain: hemispheric neglects and dyspraxias, for instance, will show up in the drawing, and a knowledge of the way information passes through the brain will localize the lesion.

I've been playing this game with your arguments since my uncle's phantom limb pain was presented, early in this thread. It's a poor philosophy that can't withstand contact with neurology, but that's what keeps happening.

Grant Hutchison

Well, as Ken has pointed out, it left a large excluded middle in your argument, that of representation without dualism. Which allowed you to argue a straw-man dichotomy, in which you contrasted a deprecated subset of representationalism with non-representationalism.
So I'm glad we've got that sorted out.

Grant Hutchison

Fair enough. I just wanted to note some of the argument forms common to the representational viewpoint to further illustrate the differences with alternate views. Representations are often posited to explain what is believed to be a gap. A colored house is seen, but no "color substance" is found in the house with a microscope, so a "representation in the mind" is posited to explain where the color was introduced. Representational arguments tend to be of the indirect form, "It's not out there, so by process of elimination, it must be in me." The argument is deductive and to be valid, requires all possibilities to be identified first and eliminated one-by-one until only one remains. That's difficult to do in practice.

This site has a diagram of a man looking at a house and the colored representation in his head:

The Representationalism Web Site

The Representational model holds that the stimulus available in the environment is impoverished in some way and that the brain must supply the missing pieces. For example, you commonly hear that the retinas being two-dimensional are unable to pass the three-dimensional information of the scene to the brain, yet the world looks three dimensional to us. Hence, the brain must have created a three dimensional representation. JJ Gibson's insight into this matter is that the three dimensional information is available in the ambient optic array (his phrase for the optical stimulus available to be detected) and that it is possible for organisms to pick up that information using any of a number of means including binocular vision, focus distance, parallax, loss of detail with depth due to atmospheric effects, and so on. Gibson thought that there would be no gap left that required a representation to plug in the explanatory framework.

But notice that none of that was in the example. The example consisted of a person, a tank, and the world outside the tank. The business about brains interpreting stimulations in terms of representations is being smuggled in without direct demonstration when it is the very issue in question.

The "it" (an inner representation) needs to be demonstrated. It wasn't anywhere in the example. This shows that to talk of representations is not to speak of what is inside the person, but of what the person does, says, and sketches in the particular circumstances. Yes, something goes on in the person, of course. Those details just don't make a difference in this example. We would call the sketch a representation regardless of the which private or neural strategies were in play.

In Ken's example, the neurologist never showed up to the party. We granted Ken's feat the award, “Having Sketched the House from Memory” without having to deploy the cranial saw and neural probes. In other words, we treated Ken as a “black box” meaning the exact nature of the physiological details weren't going to be a consideration in saying that he sketched from memory. As long as he sketched without being in the presence of the house and without aids such as photos, we would say that he sketched from memory.

Your argument here switches from the personal to the subpersonal or “wires and pulleys” view. From this perspective, the neurologist doesn't have to limit himself to representations; the notion of “mechanism” can apply as well. People sometimes sketch paintings in art museums, and so on, so it is natural to want to explain the mechanics of sketching from memory as sketching from some sort of private painting or representation.

But there can be any of a number of reasons human beings can produce sketches. As an analogy, consider a child's Spirograph. With the wheels and templates, the child can produce drawings of a variety of patterns without the implements embodying copies of the patterns produced. The tools extend the child's ability to draw. Sketching from memory doesn't have to be a matter of the artist making a knockoff copy of an inner representation. Nature can have all sorts of strategies, skills, and techniques in play as Ken plans, starts, reviews, revises, and completes his sketch. Perhaps memorizing a house is not always a matter of storing a picture, but of adding mechanism to produce house sketches.

All right, buster. Just for that I'm going to hit you with a Skinner quote. Skinner said of mentalistic explanations that “they allay curiosity and bring inquiry to a stop.” I am fond and excited by this subject. I find that representational explanations “bring inquiry to a stop” and dissuade us from further exploring and appreciating all the strategies and techniques nature employs.

Well, that's fine if we're restricting our research to Ken's ability to sketch. I don't see much grant money coming out of it, though.
But if we're interested in the Bigger Picture, we need to account for the fact that people with brain lesions in particular places produce sketches that are defective in characteristic ways. It's a strong argument for internal representations that get broken. The fact that I can stimulate your intact brain directly and produce complex hallucinations is the other side of that same representational coin.
Gibson's alleged insight that the data are out there in the world and therefore don't need to be represented internally is just a non-argument: there's no test for internal representation in there at all.

Well, that's surely your problem (and, boy, was it Skinner's problem) but quite evidently it's not my problem. Most of what I've been posting here has been an effort to show what a fascinating, exciting and rich field consciousness research is when it looks inside the head. That doesn't make what goes on outside the head any less fascinating, exciting and rich. They're at least additive, and probably synergistic, in terms of the interest they generate.
I've got both, you've got one. So the loss seems to be yours.

Grant Hutchison

I just enjoyed a TED conference video, neuroanatomist Jill Bolte Taylor describing her own stroke: How it feels to have a stroke, YouTube presentation, about 20 minutes.

Talks a bit about left and right hemispheres, which might be apropos for this topic (though probably covered long ago). Mostly it's just fun, for her presentaion skills, and educational, for her knowledge and unique point of view.

"...and then I realized: I'm having a stroke. I'm having a stroke! And
then the next thing my brain says to me is, 'Wow... This is so cool!'"

__________________
0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 ...

I hope you can by now anticipate how harshly I would deal with the appending of an "ism" to the word "representation" and calling the result science. That is the wrong way to do physics, and is probably even the wrong way to do philosophy (the idea that we must "choose a camp" with which we "align our concept of reality"-- it's just plain silly). Fortunately, we do not call our particle physicists "particlists", our tsunami experts "wavists", or our condensed matter theorists "matterists". We do have geologists, but they do not "believe in geologism", they just find rocks interesting!
I for one am not "holding" anything when I adopt a representational model, I am, well, adopting a model. And I am doing it for the same reason as anyone adopts a model-- for the explanatory advantages it offers, when coupled with a research tool like sketch analysis or neuroscience or all the above. Cognitive researchers ought to know better than to "hold" beliefs, if they don't then they have not achieved the revolution that physics has benefited from since the days of Galileo.
Well, as I understand the way radiation fields store information, how parallax works, and why we have two eyes, that all just sounds pretty silly to me. Nevertheless, I have no doubt the brain uses a host of visual cues to reconstruct a 3D representation from correlations stored in the radiation field.
This is not an argument that there isn't a representation, it is merely an argument that the brain has more help in generating that representation than extremely naive people who know no optical physics may have suggested. Also, that Gibson can create a list of gap-fillers in no way proves that no gaps remain. Optical illusions, for example, have already been discussed, and the ways they circumvent every single physical effect in Gibson's list.

But even if there were no gaps, the action of the brain to assemble and unify that information so that it could act on it or perceive an understanding of it could still be fruitfully described as "making a representation". Indeed, I see the main difficulty in that representation as being in deciding what to ignore, not what "gaps" have be filled in the sensory input.
On the contrary, the "direct demonstration" is in the value of using the model. I have already expressed that value-- almost every single word we are using is defined using representational models. You said that the "example consists of a person, a tank, and the world outside". Well, here's the news flash: those are all representations! A truly non-representational approach that was true to its own assumptions would have to simply say that there was a reality that was interconnected in ways that we might representationally refer to as a person, a tank, and a world. You see, it's hard to get through even a single sentence in this discussion without invoking a representational approach.
It was everywhere in the example, just as it has been everywhere in your own statements on this thread. The very use of language is the use of internal representations, unless you think of what we are doing right now as pushing on keys that stimulate others to push on other keys. That seems mighty clunky to me, I'll take the representational approach to communication any day.So now you are saying that there are representations, but they are not "inside" the person. So "non-representational" actually means "outer-representational"? You are merely choosing different representations, you have not said anything about the use of representations. Your argument has morphed from "don't use representations, look at the behavior", to "OK use representations, but make sure they only manifest in behaviors". Again, this is nothing but a projection-- the man looking at a shadow and saying "reality is two-dimensional".
Grabbing at straws now. Of course the details make a difference, as we would call a different sketch a different representation.

__________________
Physics doesn't predict the future, it predicts the past that hasn't happened yet.

There are two kinds of delusions, the obvious kind that clearly don't work, and the insidious kind that clearly do.

The information is certainly all there, in the incoming data, and I've never understood why Gibson's rediscovery of this has been seen as particularly insightful.
What is interesting is how little of this external information is actually getting into your head at any given moment: the resolution of your retina is very poor indeed outside the fovea, a limitation you compensate for by saccadic movements of your eyes, shifting from one fixation point to another, and entirely suppressing visual processing while you make the saccadic twitch. So we tile together a series of high-definition patches surrounded by low-definition impressions, punctuated by frequent brief episodes of blankness, to produce what seems like a coherent and stable view of our environment.
The patchiness of our actual data reception has been neatly demonstrated using computer technology. If you monitor someone's eye movements using sensors in a set of spectacle frames, you can make changes to a screenload of text during their saccadic eye movements, when their visual processing is suspended. An onlooker watching over the subject's shoulder sees a screen full of rippling, changing text; the experimental subject perceives that they are observing a stable screen of text which is somehow both static and yet never quite the same when a particular bit of text is re-examined.

If you want to prove to yourself that there are gaps in your "continuous" visual perception, just look in a mirror, and shift your gaze from one eye to the other of your reflected face. You can't see your eyes move; all you see is the end result of the shift in gaze. And yet there's no discernable perceptual gap while your gaze shifts.

Grant Hutchison

That's quite interesting. I'm a monkey's uncle (rather than cousin) if that's not a key part of the real reason for the "full Moon on the horizon illusion". I'll have to try it. I imagine we can see someone else's eyes half-open during a blink more easily than we can see our own blink in the mirror.

__________________
Physics doesn't predict the future, it predicts the past that hasn't happened yet.

There are two kinds of delusions, the obvious kind that clearly don't work, and the insidious kind that clearly do.

The point there was that when people say “he drew a sketch of the house from memory” doesn't necessarily mean he sketched an inner representation. We make statements like that when we describe or sketch without being in the presence of the thing sketched or when other people do such things. That's just an observation about how the language of “he sketched from memory” functions. The language is agnostic to underlying details. It will not reveal those details. That requires a different sort of investigation.

“A broken representation,” however, suggests that part of the house Ken wanted to sketch was in one of the brain lesions, and that Ken was unable to “read off” that section of the sketch. All your example here shows is that the lesion mediates behavior. I think there are other alternatives to mere representation here. (Especially in any sense of representation related to the picture depicted at the top of The Representationalism Web Site with or without the little man in the head.)

As I understand it, a neural circuit can be trained to respond to patterns. Let's say that researchers construct and train a neural circuit to be selective for Halle Berry's face such that it could reliably detect Halle's face in a variety of photographs taken from different perspectives in a variety of contexts. Isn't it the case that the circuit can “recognize” Halle's face without having stored a multitude of representations of her face that it must compare to one by one until it gets a match? (Think of how fingerprints are identified on TV crime shows. The computer monitor shows fingerprints displaying one after another until a match is found.) Isn't our neural circuit's functional profile just tuned to be selective for Halle's facial patterns?

But there doesn't have to be a complex hallucination sitting there waiting for its cue to step onto a stage of an inner theater. Just because I errantly see something that isn't there doesn't mean I am correctly seeing something that is inside me. There might be any of a number of reasons I make perceptual misjudgments, just like the reasons Ken and I saw the house differently was partly do to the way we each tended to the features of the house.

His argument is that the brain does not necessarily have to make up a deficit in stimulation to, say, fill in gaps, explain three dimensional perception, and so on. His view of the brain was a very simplistic metaphor that it resonated to information in the environment.

That's why I see the "entire emobodied individual" as including the brain. (I just don't know much about the brain, so I don't write much about it.) I always took the "the proper subject of perception is not the brain" as a desire to widen the scope to include life in the environment and not to exclude the brain. Some people I've discussed this issue with in the past, however, seem bothered by the notion of perception, in a sense, extending out of the body and into the environment, which in my opinion, suggests a sort of nueral-supremicist bigotry.

No, it doesn't. You're cartooning the "representation" to make it fit into your Cartesian version. If we treat the brain as a black-box "information transducer", passing visual information to the final sketch, then the neural lesion is damaging the transduction process. The representation that is passed from eye to hand gets broken in a reproducible way, which is predictable from the site of the lesion, once we get to look inside the black box.

I haven't suggested anything different.

I haven't suggested that. You're cartooning again.

Formed visual hallucinations are rather more striking than mere "perceptual misjudgements". I knew a lady, for instance, who was followed around by the upper half of a yellow Confederate soldier for four days. The effect was quite robust to changes in environment and lighting conditions. It's a bit more interesting than Ken and you disagreeing about the roof of a house.

In which case, why the fuss? What you're describing there is a commonplace view, not a revolution.

To me it suggests a quite reasonable regard for the meaning of words.
If I were to propose that the meaning of the word "digestion" should be extended into the environment to include all sources of food and all means of sewage disposal, I would deserve equally short shrift. Using the phrase "intestinal-supremacist bigotry" would be comic, not cogent.

Grant Hutchison

Then "a representation" is not the best concept to describe our Halle Berry-detecting neural network. Dynamic mechanisms deserve a seat at the philosophical table as well..

On the contrary, the firing of a "Halle-Berry-detecting neural network" is as much a "representation" of external reality as is the projection of a picture of Halle Berry in some mental Cartesian theatre. Ken and I have pointed this out on several occasions, and you seemed to accept that there was more to the representative theory of perception than cartoon dualism when you wrote:Maybe you need to let us know what you mean when you say "representation" in this context, since it seems to be something different from anything I've encountered.

Grant Hutchison

Quite so!. The discovery of change blindness and inattentional blindness must have been equally as surprising. It was learning about the things you bring up here that finally loosened my grip on the traditional representational view. It started to dawn on me that Mother Nature was working like an engineer on a tight budget, not wasting time and energy painting inner pictures, but economically sifting for the ambient information most relevant to the task at hand.

I don't have a citation handy, but I remember Dennett mentioning an interesting twist to one of those experiments you mentioned: If they made a word track your eye's movements as it ma