This is a summarized version of a long debate I had on the epistemology of conscious experience, or the debate between Direct Perception v.s. Representationalism on the PSYCHE-D mailing list in April - June, 2005. I have selected choice threads that led to interesting exchanges, with links to the original messages.
This debate is interesting not only as a discussion of the nature of consciousness, but also as a prime example of a paradigmatic debate. In fact, this issue is perhaps the ultimate paradigm debate, because the alternative paradigms present such radically different inside-out-inverted views of the world relative to each other, no wonder the opposite camps could never reach agreement even on the meaning of terminology!
See also the Cartoon Epistemology
Initial message
Lehar: refutation of Gibsonian concept
Brook: Proof of direct perception: Kick a chair
Rickert: Gibson's epistemological error?
Lehar: Theories v.s. Paradigms
Lehar: What would it take to convince you?
Sizemore: Conceptual v.s. Empirical Issues
Sizemore: Meaning of representation
Lehar: Ontology of Experience
Lehar: Summary Direct Percepton v.s. Representationalism
Brook: What more do you want?
Rickert: No clear meaning of Representation
Rickert: Computers don't compute
Chalmers: Terminological dispute
Lehar & Brooks: offlist debate
Lehar: Analogical or analytical?
Dalton: Representation is analogical?
Reason: Is the word "red" red?
Seager: How big is your experience?
Trehub: We do not perceive our perceptions
Lehar: More paradigmatic stuff
Lehar: Still more paradigmatic stuff
Brook: Hostile to sense data theory
Lehar: Epistemology of exeperience
Lehar: Concluding exchange
Hi everyone
Regarding the most interesting discussion on projection geometry related to visual conscious experience between (primarily) Alex Green and Brian Flanagan, I highly recommend the work of Steven Lehar. See http://cns-alumni.bu.edu/~slehar/Lehar.html
He has written several books and a target article in Behavioral and Brain Sciences on this and related topics, and his work is not only insightful but (as you can see on his website) extremely well illustrated with artistic color cartoons.
Steven will be speaking at the next Tucson conference a year from now. See http://consciousness.arizona.edu/Tucson2006.htm
cheers
Stuart
From Lehar's web site:
Sizemore:
So, the ploy is simply to say that the representation is information
and information is "used" by some other part of the nervous system. By
substituting some other word for "see" (here "use" is substituted) the
glaring absurdity is obscured. But the notion of "using information"
needs to be critically examined, and when it is, it can be seen to
raise the same issues as "garden-variety" representationalism. In what
sense do nervous systems "use information"? Does it mean any more than
events cause receptors to "fire" which cause other neurons to "fire,"
and so on? If not, then we are simply saying that somehow physiology
mediates the behavioral functions that we wind up calling "seeing,"
and "hearing," etc. and this only raises issues of infinite regress
and homunculi if representationalism is assumed. If it means more
than that, then we are probably back where we started. That is, in the
conventional view, we say that we see the world, and that we do so by
creating an inner copy that is seen. But we may easily say that
animals (human and non-human)"use information" in the environment
(colors, sounds, etc.) in order to behave in the ways that we observe
and that they do this by creating an inner copy of the information
which is what is "used." If "using" information in the environment
requires further "usage" than why doesn't "using" the internal
information?
The key issue is not the word "see" or "use", to describe how the
brain makes use of internally represented information, it is the
question of whether the internal representation is processed by other
internal processes in the brain, or whether it requires a copy of the
*whole brain* in order to "see" that represented data. Only the latter
formulation leads to the infinite regress.
If Sizemore claims that a picture in your brain requires a miniature copy
of your whole brain to "see" that picture, then surely this objection would
apply to *any* information represented in the brain, including verbal,
linguistic, and cognitive knowledge, all of which would require a miniature
copy of the whole brain to interpret or process that cognitive
information. Why is it that the homunculus objection is only raised
against pictorial data? What is so special about image data that requires a
homunculus to "see", when other data do not?
Now I acknowledge a profound philosophical issue here, that applies to
*any* explanation of mental function, that is, after we are done
explaining the mechanism of perception or cognition, there is the
question of how come we get to *experience* that information
processing. The mechanistic explanation of neural signals and sensory
processing says nothing of the experience of perception and why we
have it. Even if we have actual pictures in our brain, how is it that
we *experience* those pictures? Why doesn't a computer experience the
image data that it processes? That is indeed a profound issue, but
again, it is one that applies to cognitive and verbal information
processing just as much as to pictorial processing.
But whatever the explanation might be for the
"Ultimate question" of consciousness, the indisputable fact remains
that we *DO* in fact experience the operation of our brain during
mental processing.
Steve
Lehar >
GS:
Lehar >
Glen Sizemore >
Lehar >>
Sizemore >
I acknowledge that placing a representation in the head does not solve the
whole problem of how we see. It *is* however a *prerequisite* for being
able to see something that that something must be represented in your
brain. Is it not?
Sizemore >
Yow! That's pretty radical man! I'm not sure I can debate you if we don't
even share this much in common!
Sizemore >
Wow! Thats pretty bizarre!
Do you agree with the robot metaphor, that people are like a robot
that receives sensory input, stores it in internal representations,
and processes that information in order to compute an appropriate
behavioral response? Do we at least agree on that metaphorical image?
If not, then I don't think we can have a meaningful exchange beyond just
agreeing to disagree.
Steve
Ok, NOW I understand where you guys are coming from! Its the Gibsonian /
O'Regan, "organism interacting with the environment" idea.
Well, in the first place, Representationalism has never been shown to be
false. Au contraire, mon ami, the *principle* of representationalism has
been actually demonstrated in robots that use video cameras for sensory
input, computations in a computer brain, and behavior by way of servo
actuators. Now admittedly current robots are pretty primitive
and "robotic", and very different from animal perception. But if anyone has
any ambiguity about the meaning of terms
like "information", "representation", and "processing", just look at a
robot and the concepts become perfecty clear.
Robots offer an *existence proof* that the *concept* of representationalism
is *feasable* at least in principle.
So lets have no more talk about not knowing the meaning of information or
representation or processing. Those terms are perfectly clear, and
obviously workable in principle.
Furthermore, in the absence of *compelling* evidence to the contrary, a
representationalist assumption is the most *reasonable* understanding of
perception, given the eye that appears to work like a video camera, the
optic nerve that sends data to the brain, given the complex wiring
suggestive of computation in the brain, and motor neurons from the brain
back out to the muscles as if to produce behavior. The representationalist
thesis comes directly from inspection of the wiring of animal bodies, that
looks for all the world *AS IF* it were a representationalist system! It
may not be, but until the strong contradictory evidence comes in, that *IS*
the most *REASONABLE* initial assumption!
Now, the Gibsonian / O'Regan / direct percepton concept, on the other hand,
has *NEVER* been demonstrated in ANY kind of artificial system, and there
is very good reason to believe that the whole concept is totally incoherent
and impossible *IN PRINCIPLE*!
How would you even build a robot that works by direct perception??? Would
you equip it with video cameras for eyes? If so, what do you do with the
data generated by those cameras? You can't send it to the brain, because
that would be computation and representation again, so we can save
ourselves the expense of video cable and computer brain. But what would
drive the servo-actuators? Where does that signal come from? And how does
the direct-perception robot project its experience out of its body into the
world so as to produce behavior *as if* it were actually *seeing* that
environment *without* representations and computations???
HOW WOULD IT BE DONE????
The concept is so vague as to be totally meaningless! Gibson himself
refused to discuss what gets sent from the eye to the brain, or what kind
of computations might occur in the brain. In fact Gibson even denied that
the retina records anything like an image! But he NEVER OFFERED ANY
ALTERNATIVE EXPLANATION for how behavior works besides a few vague
mumblings about being tuned to invariants in the environment. Gibson spoke
as if the computation of perception occurs OUT IN THE WORLD, rather than in
the brain. But there is no computational or representational *machinery*
out in the world, the only machinery is found in the brain, right where the
sensory nerves terminate, exactly AS IF the sensory organs were sending
data to the brain for processing!
As for O'Regans concept of using the external environment as a
representation of itself, that idea is DEMONSTRABLY FALSE, because probing
the world with visual saccades, especially in the monocular case, is
**nothing like** accessing a memory, internal or external, because every
saccade presents only a two-dimensional pattern of light. The three-
dimensional spatial information of the external world is **by no means**
immediately available from glimpses of the world, but requires **the most
sophisticated** and **as-yet undiscovered** algorithm to decipher that
spatial information from the retinal input.
Secondly, O'Regan's concept of direct perception is totally inconsistent
with the *phenomenal experience* of vision, where we do in fact experience
the world as a spatial structure, and we perceive individual saccades to be
located at the location in the global framework of space that we perceive
that saccade to be located.
Thirdly, the absurdity of O'Regan's concept is highlighted by the condition
of *visual agnosia* (specifically, *apperceptive* agnosia) which is a
visual integration failure. An agnosic patient can see individual features,
but cannot distinguish a picture of a bicycle from a picture of
disassembled *parts* of a bicycle. They can see a wheel here, and a wheel
again, but they cannot tell whether the second wheel is not just a second
glance at the first wheel, or if it is a separate wheel, they cannot see
the spatial relation between the two wheels. In fact, the experience of
visual agnosia is *exactly as if* vision worked as O'Regan proposes, with
individual glances picking out features in the world in the absence of a
global framework or spatial representation to store or record those
features.
The condition of apperceptive agnosia is the absence of a visual function
whose existence O'Regan effectively denies!
So don't be complaining about the vagueness of terms like information,
representation, and processing. Those terms are **perfectly clear**, and
are demonstrable in **actual physical robots** that receive sensory input
and compute behavioral responses.
Instead, the real concern is the vagueness of terms like "active
interaction with the environment" and "responsive to invariants in the
environment" IF NOT by way of sensory input and internal representations.
What does that even ***MEAN***?
Until the *principle* of direct perception is demonstrated in a simple
robotic model, the concept is totally incoherent and ill-defined!
Steve Lehar
I have been lurking in the recent discussion but I want to say that,
modulo a few excess asterisks, Steve Lehar's message on
anti-representationalism seems to me to be exactly right. Theorists
who present themselves as denying the existence of representations
almost always, on closer inspection, turn out to be denying that
certain kinds of representations exist (Brooks) or to be denying that
representations are as plentiful or play as big a role as the
tradition would have it (O'Regan, Noe, Clark), or whatever. Some just
refuse to talk about central issues at all (Gibson). Mounting a
flat-out denial that there is anything in the brain that stands for,
indexes, refers to, even pictures items other than itself is a pretty
tough task. (Actually, Lehar might be a bit hard on O'Regan on this
score, though some of his rhetoric invites Lehar's kind of response.)
Andrew
I think the argument against representation should be made based on
physics. Namely, what is the nature of the environment-organism
interaction allowing the brain to make a copy of it, how exactly can a
copy of this environment be made? Is it a tape-like recording? A CD?
And despite its enormous capacity, can the brain really afford to hold
a copy of the entire universe together with whatever extra brain power
is needed to analyze it? Judging by nature's preference for parsimony,
the answer is no.
Andrew Brook >
Yes, I am more accustomed to debating people who contest the notion of
*spatial* representations in the brain, given that neurophysiology has
not (yet) found "pictures" in the brain. It threw me for a loop to
find people like Glen Sizemore who deny representationalism
altogether! Even the most ardent supporters of Gibson's theories
generally take care to disclaim his most radical views (Bruce & Green
1987 p. 190, 203-204, Pessoa et al. 1998, O'Regan 1992 p. 473)
although they present no viable alternative explanation to account for
our experience of the world beyond the sensory surface.
Many of Gibson's observations on environmental affordances and invariants
were very valuable and insightful, even from a representationalist
viewpoint. In fact, it was Gibson's refusal to discuss physiology and
computation that released him from the burden of having to consider issues
of "neural plausibility", and that is why he dared to make such bold and
generally valid observations on the nature of perception.
But Gibson's profound epistemological error backed him into a corner which
is ultimately indefensible, which made him get defensive and dogmatic in
his later years, as often happens to those who commit themselves to
defending the indefensible.
Carreno >
I am viscerally sympathetic with this argument. Given what we know about
neurophysiology, it seems totally implausible that the brain could
construct a model of the world as rich and complex as our experience of it,
and maintain it in real time as we move about in the world. Carreno is
right: that does ideed stretch credulity to its elastic limit. But before
deploying Occam's razor we must first balance the scales, and take a full
accounting of the alternatives under consideration. For the alternative is
that we experience the world directly, as if bypassing the causal chain of
sensory processing. As incredible as representationalism might seem, the
alternative is even more incredible.
Lehar >
Fallacy of the excluded middle. There are other options. One is the
one I sketched in the message from which Steve quotes: that our
representations give us access to the world itself, not just to the
end point of the representing process. Our brain has the capacity to
work its way down the causal chain to experience the kickoff
point. How we can do this is a wonderful mystery but that we do it
seems, to me at least, pretty much beyond question. When I open my
eyes, I see the world around me. This is to be in epistemic contact --
to see, know, experience, be conscious of -- the world, the part of it
in my immediate vicinity anyway, not any representation or construct
of mine.
Brook >
Wonderful mystery indeed! Downright *miraculous*, wouldn't you say?
I mean, how would you demonstrate this in a simple robot model?
Experience or awareness of an object means having posession of
information about that object, color, shape, location, etc. Exactly
*as if* one had a miniature colored model of the object right there
inside your brain. Except we *DON'T* ?
So how would this work in a robot model? The video camera picks up a
2-D image which is sent to the computer brain, that extracts a few
features and performs a few computations, and suddenly, miraculously,
a three- dimensional colored data structure appears--not in the
computer brain in some kind of holographic 3-D imaging mechanism at
the end of the causal chain, but right back out there in the world!
With NO high-tech holographic imaging machinery involved!!! The image
just appears out there, disconnected computationally from any of the
hardware of the robot.
And when the robot closes its lens covers, the model out there
DISAPPEARS! As if it were causally connected to the computational
hardware downstream of the video signal, except it *ISN'T!*
I say again: Until the *principle* of direct percepton can be
demonstrated in an artificial sensory system, the whole idea is
completely implausible and ill-defined. I don't mean anything fancy,
just a simple demo like the representationalist robot. *HOW* does the
robot "work its way down the causal chain to experience the kickoff
point" ???
Steve,
I'm not sure what's going on here. We should be on the same side. Here
is why I think that some form of direct realism is virtually
unrejectable. When I open my eyes, I see a chair just as directly as
when I kick it, I kick a chair. In both cases, it is the chair that I
am in contact with. In neither case am I in (interesting, relevant)
contact with any intermediary. And -- here is why the view is
essentially unrejectable -- if you say, in either case, 'the contact
is not direct', then I will invite you to tell me what you mean by
'direct'? What could be more direct than seeing or kicking an object
in plain view in front of me? If this is not direct, what would be?
Simple robot model? I sketched how at the end of my message. In the
same way that vision systems since Marr's have been able to extract
three dimensional objects from two dimensional arrays, our vision
system not only extracts threre dimensional objects but allows us
'reverse infer' down the causal chain to be directly aware of
them. What more do you want?
Brook >
Yeah, thats what *I* thought when you chimed in to reject the extreme
Gibsonian (Sidemorian) version of direct perception! I guess that this
issue is not a clean binary choice, but there are a number of
intermediate positions between direct perception and
representationalism.
Brook >
Ok, stand in front of a chair, and before you kick it, touch your
finger to one eyeball (through the eye lid) and push it gently to one
side, until you see a double image. Now KICK! Now you see TWO chairs,
and TWO feet kicking them! Which one is the "real" chair and the
"real" foot? And what is the actual objective location of that chair?
If this is not IN-direct, what would be?
Brook >
Au contraire! Marr's vision model is entirely representational. From the
moment the image registers on the (synthetic) retina, all computational
processing of that image occurs inside the computer, or inside a brain.
Nothing gets projected out into the world again! There is *NO* "reverse
inference" going on here, but a feed-forward progression of *forward*
inference, from input stimulus to internal mental model. The computer never
has access to *any* information that is not explicitly represented in the
machine. That is an indirect representational algorithm!
Steven Lehar
Yet I could nowhere find a clear statement of what Gibson's "profound
epistemological error" is presumed to be.
For sure, Gibson avoided questions of mechanism. But that hardly seems
to be an epistemological error.
I am quite sure that Gibson understood the physics and biology, and
recognized that stimulation of retinal cells and the transmission of
signals on the optic nerve were part of the causal processes involved
in that direct perception.
Rickert >
For a summary of the epistemological issue and the various alternatives, see
Direct perception embodies a profound epistemological error.
Rickert >
Actually, Gibson made his views on the role of the retina perfectly clear,
and they were very much at odds with the consensus view on it.
Gibson, J. J. (1966) The Senses Considered as Perceptual Systems. Boston:
Houghton Mifflin.
p. 263:
The reason why Gibson denied that the retina records an image and
transmits it to the brain, is that to even allow this much
representationalism in the visual process is to acknowledge that the
principle behind representationalism is perfectly feasable, and that
the first stage of visual processing is apparently representational.
Steve Lehar said (among much else):
I respond:
Try another way:
Steve:
Me:
Steve:
Me:
Andrew
Ok, we've been round and round the direct perception v.s.
representationalism debate enough times to see that nobody is about to
change their minds, no matter HOW eloquent or persuasive the arguments
are on either side. Why is this so?
This is a sure sign of a *paradigm* debate!
As Kuhn explained, (Kuhn 1970 The Structure of Scientific Revolutions)
there is a profound difference between theories and paradigms, and how
to debate them. The problem is that what we are debating is not a
question of theory, like the question of whether the representation in
the brain is digital or analog, or whether time-to-collision
information is available from the optic array, which would be resolved
by the normal rules of logic and evidence. But in this case what we
are debating are the *foundational assumptions* with which we come to
the debate in the first place. Debates between paradigms tend to go
round and round in futile circles, because the participants are
debating from different foundational assumptions. We take our
foundational assumptions as a *given*, as obviously self-evident
*fact*, and from that perspective the opposing paradigm appears
patently absurd, it makes us wonder how intelligent, educated people
could possibly defend such an absurd and indefensible view.
But if alternative paradigms are to be fairly evaluated, it is
necessary to temporarily and provisionally suspend one's own
paradigmatic assumptions, (a feat that many find impossible to do) and
accept the assumptions of the alternative paradigm **as if they could
actually be true**. Only then can the competing paradigms be fairly
compared, not on the basis of the perceived incredibility of their
initial assumptions, but on the overall coherence and self-consistency
of the world view that they implicate in total.
In the case of our debate, we hear some state that it is "obvious"
that we experience things directly, while others state that it is
"obvious" that perception is indirect. If we begin with either of
those assumptions, we are sure never to reach agreement. In the case
of the historical debate between an earth-centered or sun-centered
cosmos, the earth-centered people argued that the idea of the whole
earth with all its mountains and forests and oceans spinning and
flying through space is so absurd and incredible on the face of it,
that it does not matter what evidence you might cite, they would never
be convinced!
There is a parallel with the current discussion, because as in that
ancient debate, there is a certain asymmetry in the two views: one
alternative is the "naive" view, in the sense that that is the view
that we adopt by default, even before giving the question any serious
thought, while the other view appears initially to be patently absurd.
In paradigmatic debates, the argument that one view "seems incredible" is
no valid argument. Many of the greatest discoveries of science seemed
initially to be so incredible that it took decades or even centuries before
they were generally accepted. But accepted they were, eventually. And the
reason why they were accepted was not because they had become any less
incredible. Facts such as the immensity of the universe, and its
cataclysmic genesis from a singularity in space and time, as well as the
smallness of the atom, or the bizarre properties of quantum phenomena, are
just as incredible today as they were when they were first discovered. And
yet all of these incredible theories have taken their place in the realm of
accepted scientific knowledge, not because they have become any less
incredible since they were first proposed, but because the evidence for
them has been irrefutable. In science, irrefutable evidence triumphs over
incredibility, and this is exactly what gives science the power to discover
unexpected or incredible truth.
There is an asymmetry in this debate: all representationalists were once
naive realists, whereas most direct perceptionists have never been
representationalists. I can tell you that I find representationalism **just
as incredible** as any of you do. It is absolutely incredible that my
physical skull should be larger than the dome of the sky. And it is
incredible that the brain can construct and maintain a real-time volumetric
moving image of the world with the rich detail and fidelity of the world I
see around me. Given current knowledge of neurophysiology, that appears to
be absolutely incredible!
But before we deploy Occam's Razor, we must first balance the scales and
take a full accounting of the alternatives under consideration. For the
alternative is that we can somehow become aware of objects and surfaces in
the external world *without* the mediation of sensory processing and
internal representations. That, in my view, is not just incredible, it is
incoherent, as demonstrated by the fact that nobody has ever, or could ever
possibly build a robot that can demonstrate the *principle* behind direct
perception in a simple model.
I am therefore suspicious of direct perceptionists who focus exclusively on
the incredible aspects of representationalism, without also acknowledging
the incredible aspects of direct perception. I can accept someone who
argues that both views appear incredible, but that they consider this view
to be somewhat less incredible than that one. That is a valid and
reasonable position. But anyone who does not see the profound problems in
direct perception (as I see the profound problems in representationalism)
is suspect of being a paradigmatic partisan, that they accept one view as
plainly obvious, thus requiring no further proof, while the other appears
patently absurd no matter what the evidence. One suspects of such people
that they never really understood the alternative position enough to give
it any serious consideration.
To those people I implore that they entertain the *possibility* that they
may perhaps be mistaken.
-William Shakespeare, Hamlet, Act I, Scene 5.
Steve Lehar
Steven Lehar
Steven perhaps sees this as a problem. I don't. While I tilt toward
the direct perception side, I am not all all concerned that Steven and
Alex are strongly committed to representationalism. We don't need to
put all of our eggs in the one basket. Science is best served when a
problem is studied from several different perspectives. And may the
best perspective (whichever that is) win.
I don't think I agree with Steven's assessment.
No doubt naive realism is pretty much the received view. But direct
perception is not the same thing as naive realism. I suspect that if
people with an ordinary science education were asked to choose
between Gibson's account of vision and Marr's account of vision, most
would prefer Marr's account. So the default view would be closer to
that of the representationalists than to that of the direct
perceptionists. There are subtleties to Gibson's theory that make it
a little difficult to appreciate.
However, that is not the alternative. It seems that Steven can only
see one side of the paradigm shift.
The alternative, or one alternative, is that we can become aware of
objects and surfaces *with* the mediation of sensory processing, but
*without* the mediation of internal representations.
Incidently, I don't doubt that some of our interactions with the
world are mediated by internal representations.
-NWR
Reply to Neil Rickert:
Rickert>
Are you serious? It does not bother you in the least that people believe
two mutually contradictory theories of perception? Surely the goal of
science is to discover which of those two views is right, and which is
wrong. I am passionately interested in that question!
slehar >>
Rickert >
And exactly *HOW* would this be implemented in a simple robotic model? How
can a robot "become aware" (aquire immediate parallel access) to
environmental information *without* the mediation of internal
representations?
We hear these *words* loud and clear, but the *concept* behind the words
remains as clear as mud. Again, until you can demonstrate the *principle*
behind direct perception in a simple model, this concept is so vague as to
be virtually meaningless.
Given my comments at the launching of this "Theories v.s. Paradigms"
thread, isn't it perfectly clear what is happening here?
Isn't it perfectly clear that Rickert himself does not have a clear idea of
what he means by direct perception, or at least not clear enough to tell us
how to build the robot to demonstrate the principle. And yet he is
strangely blind to this gaping hole in his concept of perception. He
appears to be strangely blind to the profound problems inherent in the
concept of direct perception, he does not even acknowledge that any kind of
problem exists.
Indeed, Rickert exhibits all the signs of a paradigmatic partisan. He is
arguing from the initial assumption that perception is direct, but he is
unable to question that initial assumption itself. To him that assumption
appears so manifestly obvious that it requires no explanation.
I believe that the whole notion of direct perception is an elaborate
rationalization to try to make some kind of logical sense out of the
profoundly paradoxical observation that experience is outside of our head,
as if in the world itself, and yet vision is clearly representational, from
the retina on in to the brain.
But at a very deep level there is one small part of Rickert's mind that is
aware of this paradox, although his conscious mind is in denial over that
recognition. So, like Gibson, his response is to get more dogmatic and
emphatic about his certainty that his view is right, even though he cannot
explain it to us in any kind of detail, all he can do is to emphasize again
and again that you can perceive the world directly without representations.
Until Rickert can find it in himself to acknowledge the *problem* that we
are discussing, that is, the difficulty of actually implementing direct
perception as he conceives it, further debate is simply useless, because we
are arguing from different foundational assumptions. Rickert BEGINS with
the assumption that perception is direct, so he cannot contemplate the
possibility that it might not be.
Steve
Steve Lahar represents direct realism, a position I hold, this way:
But that is not my view at all!!!! As I have said over and over and
.... OVER! (The caps may get me censored.) Nor, given that it is an
utterly implausible view, is it the view of most other direct
realists. A rough approximation of my view would be:
Furthermore, I don't think the difference between us is a paradigm
shift difference, I think it is a difference induced by an implacable
belief, totally untouchable by evidence or argument, that *if* there
is a medium of perception, *then* the resulting perceptions cannot be
direct. To which I respond:
You can define the word *direct* this way if you want, but if you
define it as it is usually defined, that we are aware of objects
around us, not just end-products of sensory processing in our brains,
then the inference needs argument and evidence to support it. And we
have seen not a single bit that the direct realists in the crowd have
not been able to deal with handily.
Andrew
Lehar >>
Brook >
Well in that case you are not the target of that criticism.
Brook >
And exactly *HOW* would that occur??? If you cannot explain how this
concept might be implemented in an artificial robot, then the idea is
so vague as to be *meaningless*! Simply stating "over and over and
.... OVER!" that perception is direct is just not going to cut it!
Steve
Lehar:
Sizemore:
I agree with Glen as against Steve on which view comes first. Though I
agree with Steve that some kind of inchoate realism would seem true to
most people who have never studied cognition, I was thoroughly
indoctrinated in indirect representationalism in school as though it was
the obvious and even the only conceivable point of view (Steve?). It
then took me years to battle my way out of that bewitchment and back to
a view that allowed me to accept what seems to me obviously true -- that
I see, touch, taste, feel, and smell things other than myself,
especially including other people. One good name for my view (he said
being deliberately provocative) would be: direct representationalism.
Which comes first? When we are children we begin with the most naive of
naive realism--that the world of experience is the world itself, viewed
directly where it lies.
Then we learn about the eye with retina and optic nerve that project to the
brain, an obviously representational system. At this point our mental image
of the problem goes into a bistable state. We understand the causal chain
of vision, but at the same time we observe that the experience, from the
end of the causal chain, jumps back out of your head again and appears back
out in the world, "like a stretched rope when it snaps", as Bertrand
Russell expressed it. This is the state in which direct realists become
permanently stuck.
Andrew Brook>
A very apt name for a befuddled epistemology! Experience is both at the end
of the causal chain, and it is also back out at its beginning. Vision is
obviously representational, at least as far as the retina (except for
extremists like Gibson and Sizemore) and yet at the same time perception is
direct, as if bypassing the retina and viewing the world directly again.
Paradoxically, both appear to be true simultaneously. "Direct
representationalism" indeed!
It is only by really taking representationalism seriously that the
epistemological paradox is resolved, although that comes at the cost of an
almost *incredible* neurophysiological hypothesis, that the brain is
capable of constructing a three-dimensional volumetric real-time moving
model of the external world with as much rich spatial detail as you see in
the world around you.
Even if you continue to find that hypothesis too incredible to swallow,
will you not at least admit the profound paradox inherent in the direct
perception view? As I said in the "Theories v.s. Paradigms" thread, if you
cannot even see that there is a paradox at all, then further debate is
really a waste of time.
We've beaten this direct perception v.s. representationalism debate nearly
to death, without much progress. And yet one view is right, and the other
is wrong. One day the conclusive evidence will come in that will finally
settle the issue once and for all. Maybe even in our own lifetimes! That
raises the question:
What would it take to convince you? What kind of evidence can you possibly
imagine that would settle the issue finally and conclusively? Or are we all
so dogmatic that we will go to our graves in obstinate denial no matter
what the evidence?
Just for the fun of it, let me hypotheticalize two alternative future
scenarios that once and for all *PROVE* the correctness of
representationalism, and of direct perception, respectively. Would the
paradigmatic partizans among us be convinced by these? I suspect the
responses might be illuminating.
H1: REPRESENTATIONALISM IS PROVEN!
It is discovered in the future that the brain is one giant resonator, just
humming with one global resonance through all its tissues, right down to
the spinal cord. This sets up volumetric spatial standing waves in the
various cortical areas, in patterns like three-dimensional Chladni figures.
Although the different cortical areas are connected, they are also
independent resonators, each one sustaining its own spatial standing wave,
as in a separate Chladni plate. And yet they are also coupled, so that the
pattern in one cortical area is coupled to the patterns in all the other
areas. If you modulate the resonance in one area, it has an influence on
the pattern in all the other areas simultaneously. This immediate parallel
coupling between resonating areas turns out to be the solution to
the "binding problem". If you *were* the standing wave in your brain (which
in fact you *are*), there would be no way for you to tell that your
resonance is distributed over different resonators, because the pattern of
your resonance would be unified.
Careful neurophysiological recordings reveal that the experience of
volumetric substance and void, solid matter surrounded by empty space, is
encoded by the *phase* of the vibration: positive phase within perceived
objects, and negative phase in the surrounding void. Experienced colors
correlate with a cylic phase representation (as in the NTSC color
television standard, for those who are familiar) which finally explains why
phenomenal color defines a circular space (the color circle) even though
the spectrum is linear. When a subject views a three-dimensional scene,
neurophysiologists can sample that volumetric experience in various parts
of the visual cortex, and actually "read" what the subject is experiencing
in any portion of his visual space by the resonance in that part of his
cortex.
When a person turns their head, or moves about in the world, the spatial
image in each of the cortical areas rotates and translates in synchrony,
like the images in an array of television sets in a shop display that are
all tuned to the same channel.
If the mapping of experience in the brain were decoded so that *every
aspect* of experience could be deciphered from the outside with appropriate
probes, which revealed a complete world of experience all encoded inside
your brain, and if neurophysiologists could transmit signals into the
visual cortex and predictably cause the subject to experience a red square,
or a blue circle, or whatever, by using the right Fourier code, would THAT
be sufficient to finally convince the doubting direct perceptionists out
there?
I suspect not.
Ok, then lets switch to hypotheticalization #2
H2: DIRECT PERCEPTION IS PROVEN!
Uh, here I have a little more difficulty imagining the "experimentum
crucis" (Oops! My paradigmatic partizanship is showing!) But lets give it
a try regardless.
In the future it is discovered that there are in fact NO representations in
the brain! Even the image on the retina is not really an image as such,
transmitted from the eye to the brain, but instead, the eye is a tool for
active exploration of the environment that detects environmental
invariances OUT IN THE WORLD where the objects of perception reside.
This bizarre notion is finally proven beyond a shadow of doubt with the
invention of the "experience meter", a device tuned to some quantum-
mechanical cat-in-the-box state of existence in the world, so that it can
detect when a material object is being experienced by someone. If you point
the experience meter at an object in front of a person, the meter lights up
when that person's eyes are open, and blinks out when their eyes are closed.
The meter has a screen that shows the information content of the subject's
experience--monochrome for people who are color blind, full color for
people with normal vision. When a subject with visual agnosia views an
object, the experience meter displays a chaotic jumble of ever shifting
visual features instead of a coherent image. There is a peculiar void, or
loss of signal from the rear faces of objects that are not exposed to the
subject's direct view, as well as for background objects that are occluded
by foreground objects. This finally proves that experience is not some
mysterious non-existent entity, but a real physically measurable quantity
that exists out in the physical world, not in a person's brain.
The experience meter also records *affordances*. When a subject views a
chair, there is a measurable "sit-onable" affordance detected that appears
superimposed on the chair, and that affordance has the mysterious property
that it has a causal influence on the subject's motor system, so as to tend
(when circumstances are right) to make the subject actually move toward the
chair and attempt to sit on it. Furniture manufacturers use the experience
meter to objectively measure the "sit-onable" appeal of various furniture
designs. Neurophysiologists map out in exquisite detail the relation
between the measured affordances, and the spatial influence that they exert
on the various muscles of the body, by some kind of mysterious action-at-a-
distance from the affordance out in the world, to the muscles inside the
body.
Would THAT be sufficient to finally convince the doubting
representationalists out there?
I gotta say, *I* would be convinced!
Would you?
Steve
Steve, absolutely the right approach, ingeniously carried out. Trouble
is, it does not address the issue between, for example, you and me. I
can't imagine anything that would 'prove' direct perception *as you
describe it*, anything compatible with what we now know about the brain,
anyway. But within representationalism, there are two houses.
There is the house of those who think 'If a representational medium is
present, the results can only be indirect perception/consciousness.'
And there is the house of those who think, 'The right kind of
representational medium lets us go right through it all the way to the
world itself, so that the result is direct perception/consciousness.'
For me, this debate between direct and indirect representationalists,
which is within representationalism, is the interesting one.
Andrew
Andrew Brook >
[1]:
[2]:
Well then can you imagine the "experimentum crucis" that would select
between THOSE two alternatives?
If not, then what does the second alternative actually *mean*?
Lehar has conflated conceptual and empirical issues. The only slight
difference is that he considers future experiments, whereas most
cognitivists claim that representationalism has already been
proven. Once again, the notion that psychological phenomena require
current representations and stored and retrieved representations (as
well as a host of related notions like expectation, unconscious
inference, unconscious rule-following, etc. etc.) for their
explanation is an assumption, as is the view that they do not.
As long as these views remain assumptions, and it is not clear that
they can ever be anything else, philosophical debate and conceptual
analysis is the only avenue via which they can be compared. Did Lehar
not, in fact, argue this himself when he invoked Kuhn a couple of days
ago?
No one has this stuff worked out by any conceivable stretch of the
imagination, but, quite bluntly, the representationalists think that
they do and this faith is, I assert, largely a product of the
ubiquitous, literal, usage of real representations such as photographs
and paintings etc. It is bewitchment by metaphor.
But even paradigmatic issues can be determined by experiment, at least in
principle. The earth-centered cosmos has now been conclusively rejected by
the "experiment" of flying around the back side of the moon without
colliding with the crystal sphere that supposedly supports it in space.
Although this experiment was technologically not feasable in Ptolomy's
time, Ptolomy and Copernicus could have predicted the different outcomes
of this "experimentum crucis" based on the two theories.
Any "theory" that is *NOT* testable by any experiment, even in principle,
is not a scientific theory at all, but a pure *belief*. For example the
existence of immaterial souls, if they are by definition undetectable by
physical means.
If you cannot describe an experiement that could prove direct perception at
least *in principle* with some kind of future technology, then direct
perception is a *belief* not a theory, since it predicts nothing different
than the alternative representationalist hypothesis.
But the truth is that the principle of representationalism *is*
demonstrable in a simple robotic system, while the principle of direct
perception remains as mysterious as the immaterial soul! After all these
rounds of debate, we still have no idea how such a system could possibly be
built in a real physical system.
Steve
Lehar >
...If the mapping of experience in the brain were decoded so that *every
aspect* of experience could be deciphered from the outside with appropriate
probes, which revealed a complete world of experience all encoded inside
your brain ... would THAT be sufficient to finally convince the doubting
direct perceptionists out there?
I suspect not.
You are right. That would not convince me.
Lehar >
[some of the details snipped]
This bizarre notion is finally proven beyond a shadow of doubt with the
invention of the "experience meter", a device tuned to some quantum-
mechanical cat-in-the-box state of existence in the world, so that it can
detect when a material object is being experienced by someone. If you point
the experience meter at an object in front of a person, the meter lights up
when that person's eyes are open, and blinks out when their eyes are closed.
This would convince me that I should consult
the Amazing Randi, so
that he can uncover and debunk the trickery involved in the
demonstration.
The trouble with Steven's two examples is that they are both
quite implausible. They look like sleight of hand parlor tricks.
A scientist needs better evidence than that.
To convince me, I need a detailed account of the relevant processes.
This should, preferably, be at the level of information processing.
It should plausibly account for all of the unresolved questions. And
it should be supported by empirical evidence.
-NWR
Neil Rickert >
Well then is there *ANY* possible future experiment that would prove it to
your satisfaction? If not, then you are a hopeless paradigmatic partisan.
Rickert >
What *kind* of evidence are you talking about? Describe the experiment!
Rickert >
Notice that the direct perceptionists are always long on high sounding
verbiage and vague concepts, but very short on specific mechanisms and
mechanical details.
They cannot describe a simple robotic system that would demonstrate the
*principle* of direct perception in an actual physical system, and they
cannot describe a future experiment that would prove it one way or another.
All this just confirms my original suspicion that the concept of direct
perception is no *theory* at all, but merely a series of vague
rationalizations used to justify their naive realist intuitions. It is
impossible to pin them down on *any* kind of specifics.
Steve
Lehar:
Sizemore:
Lehar:
Sizemore:
Lehar:
Sizemore: Yes. Whether or not there are representations that are
stored and retrieved etc. etc. are [not] beliefs, they are
assumptions.
Lehar:
Sizemore:
An alternative view is that perception is behavior, and that
behavioral function is mediated by physiology. Unless all "mediation"
is "representation" (and I argue it is not) this simple statement
constitutes the beginning of a scientific approach. You are correct
that "we still have no idea how such a system could possibly be built
in a real physical system," and we have representationalism to thank
for the wild-goose chase that constitutes much of "cognitive
neuroscience." We understand, in some complete sense, behavior at
about the level of habituation of the gill-withdrawal reflex in
Aplysia, and maybe some classical conditioning of the system (and
despite the language that Kandel uses, there is little to be called
"representation" here, unless everything orderly is
representation). Little wonder that "we still have no idea how such a
system could possibly be built in a real physical system," if the
system in question is the behavior we call seeing and hearing etc.,
and the other behavior of which it is a part. We know some of the
behavioral observations that need to be explained, at least to the
extent that some portions of psychology actually demonstrate
behavioral regularities in individual subjects, but we do not yet know
how physiology mediates such behavior. But to depend on the obviously
category-error-ridden conceptual muddle that constitutes
representationalism is no solution.
The question whether psychological phenomena require internal
representations is not just a conceptual question, as Sizemore
suggests, it is also an empirical question that will one day be
confirmed one way or the other experimentally, as soon as we figure
out the code of the brain. Because the theory of direct perception
states that there is no need for internal representations, or in the
softer version defended by Andrew Brook, the internal representations
do not need to encode ALL of the information of experience, because
SOME of that information can be perceived directly *through* the
representation (whatever that means). In any case, if it is discovered
that the brain actually *does* explicitly encode *ALL* of the
information in our experience, that will pretty much prove direct
perception to be false.
Of course there will still be dogmatic defenders of direct perception even
after that discovery, such as Glen Sizemore who tells us that he would not
be convinced even if representations *are* found in the brain, because we
would still not understand how we *see* those representations. (But then
neither does Direct Perception tell us how we can *see* the world *without*
representations, so the mystery of experience remains regardless.)
But you can never convince *everyone*. After all, there are still people
who believe in God and intelligent design, as opposed to evolution, despite
the overwhelming evidence to the contrary.
The *only* kind of paradigm that remains purely conceptual, in Sizemore's
usage, are theories that make no testable predictions whatsoever. For
example a version of direct perception that posits that there *are*
complete and explicit representations in the brain that encode *all* the
information of experience, but that perception is still direct, and does
not actually *use* those representations. Or the theory that God did design
the world, but that he used evolution has his mechanism of creation. Those
kinds of theories are indeed purely conceptual, not at all empirical,
because the experimental evidence is identical for both alternatives. But
such theories are not theories at all, they are *beliefs*, and thus fall
outside the realm of science.
Steve
The question of Conceptual v.s. Empirical issues raises a question about
Andrew Brook's softer concept of direct perception *through*
representations.
Question to Andrew Brooks:
Is there *any* information in our experience that is *not* explicitly
represented in the brain? Like the information about the world that is
experienced *through* the representation instead of *in* it? If so, then
your concept of direct perception *through* representations could be tested
in principle by seeing whether that information is in fact encoded in the
brain or not.
If on the other hand you posit that *all* of the information in our
experience *is* explicitly encoded in the brain, but we still view the
world "directly" *through* that representation, then we are in perfect
agreement, that in veridical (non-illusory) perception we are viewing the
world "directly" in that sense. But in that case your theory becomes
indistinguishable from represntationalism. (Viewing "directly" *through* a
representation strikes me as a contradiction in terms.)
Steve
Maybe this will help situate my position. I agree with Glen Sizemore that what
is between us is mainly conceptual, not empirical. Another way to put this is
that we are not disagreeing about the facts of human information processing and
behaviour, we are disagreeing about how these facts should be interpreted, what
they entail for theory.
It is hard to tell whether your "direct representationist" position really
is conceptual, or whether it has empirical implications. It is very
slippery the way you describe it. But in any case, if it *is* purely
conceptual, and makes no predictions, then it is entirely vacuous. It is
like saying that "the representation is so good, it really feels as if I am
seeing the world directly". Well we all agree with that! But if your theory
does not make any predictions, then it is no theory at all! It is a belief,
and a rather vague and ill-definable one at that.
Glen Sizemore calls them "assumptions" rather than "beliefs". But the issue
is what are they assumptions about? If they are *untestable* assumptions,
such as the existence of immaterial souls, then they are not scientific
assumptions at all. If they *are* scientific assumptions they *must* make
some kind of predictions, even if they are only testable in principle.
In fact there is a very profound empirical issue that is wrapped up in this
debate, and that is the question of how one would construct a robot that
operates by the same principle as our own brain. Either it needs to be
equipped with a representation of the world, or it needs to be designed to
extract that information from the world directly. And the debate also has
profound implications for neuroscience, that is, should we even be *looking
for* representations in the brain, and should we expect those
representations to encode *all* of our experience, or only part of it?
It is exactly the profound implications of this debate for artificial
intelligence, philosophy, and neuroscience, that make the epistemological
question interesting in the first place!
Steve,
if your theory/interpretation/whatever and mine are both
compatible with all the facts, they are *equally* untestable, at least
at the bar of the facts. As I said yesterday, I suspect that that is the
case.
Andrew
*IF* our theories are indeed both identical with respect to
predictions and future discovery of facts, then they are identical,
because you would claim, as I do, that every aspect of experience must
necessarily be explicitly encoded in the brain, and thus your theory
is indistinguishable from representationalism. Welcome to the most
reasonable explanation!
But then your rhetoric about seeing things directly, *through* the
representation amounts to no more than that the representation is so
damned convincing that it really *seems* as if we are viewing the
world directly, although you acknowledge that you are seeing it
"through" the representation, which is tantamount to saying that you
are seeing the representation, while having the vivid impression that
you see the world itself, as when watching through a television
monitor.
If on the other hand you continue to insist that seeing *through* the
representation is something different than observing the state of your
brain, while believing that you are seeing the world directly, then
you will have to explain what that actually *means* in real
down-to-earth terms with sufficient specificity that we would know how
to build a robot that sees directly by that same principle, or that we
could describe the experiment that would distinguish between the
two--for example that there is information encoded in the
representationalist view that is not encoded in the direct
representatioanlist view. We still do not know what you actually mean
by this term.
Doesn't it bother you to defend a view so adamantly and with such
conviction, while being unable to specify exactly what you mean by it? It
sure has me confused!
But then so do all the other attempts at rationalizing direct perception!
Steve
Steve,
1. Stop the aggressive bombast. I am not intimidated.
2. I have put a lot of effort into specifying my position in detail,
much more than you have put into specifying yours. That you either do
not understand what I am saying or do not believe that I could mean is
another matter.
Andrew
Brook >
I am sorry, the intent was not to intimidate, nor to belittle your
contribution. You have done a masterful job of articulating a very slippery
issue, and you have my deepest respect for your willingness to explain your
thinking very honestly and in excruciating detail. I am really very
interested to hear what you have to say.
My point was that I am truly baffled at how you perceive your own position.
You freely admit that you cannot describe a robot that operates by direct
perception, and that you cannot conceive of an experiment that would
differentiate between our two views. And now in the most recent exchange
you raise the level of ambiguity of your concept of direct perception one
more level by saying that you **don't know** whether your view suggests
that the representation in the brain encodes *all* of the information in
your experience or not. Given all that, can you understand my puzzlement
about exactly what it is you are proposing?
Contrast that with the powerful conviction and certainty with which you
argue your position, and it makes me wonder just what it is that you are so
powerfully certain *of*? As far as I can tell, you acknowledge *everything*
about representationalism, except for the one single statement that you
know for a fact that your experience is *direct*, *not* an experience of a
representation. Can you not see a contradiction here? The concept
of "direct representationalism" is a contradiction in terms, because
representationalism is by definition indirect, by way of a representation.
In all our long debate, I have never once heard you acknowledge this as any
kind of paradox. It need not be fatal; you might still argue that you
consider that paradox more tolerable than the incredible notion that the
whole world is represented in your head. That would be a reasonable and
understandable position. But do you not even see what we
representationalists see as a problem with your view?
And how can you maintain your view with such supreme conviction when you
cannot describe an experiment that would differentiate our views? Isn't it
in the very nature of scientific theories that they make predictions?
Shouldn't this at least dampen the magnitude of your conviction, or at
least your expectation that others be persuaded to join your faith? As I
explained earlier, I can see the profound problems in the
representationalist view, but in my view they are outweighed by the deeper
paradoxes of the direct perception view. But do you even see where we see a
problem in your position? Why do we not hear you at least acknowledge that
much?
Your arguments just confirm my initial suspicion that the theory of direct
perception is logically indefensible, but that people hold it with great
conviction simply because it *SEEMS* so obviously to be true, and the whole
theory of direct perception is just an elaborate rationalization of that
initial assumption. Can you disabuse me of that suspicion?
Brooks >
Really? Are you counting THIS?
and THIS?
and THIS?
I have invited you to read Chapter 1 of my book (available on-line) but I
have not hear your commentary on it. I would be interested to hear your
reaction to it. What do you make of the "introspective retrogression"? I
would really like to hear what you have to say.
Steve
Lehar:
Sizemore:
Lehar:
Sizemore >
Sizemore >
Whether or not something is a representation depends on whether somebody or
something uses it as a representation. A picture in a newspaper is only
representational when viewed as a depiction of something else, not when the
newspaper is used to line your bird cage. Voltages in a computer memory, or
patterns of activation in a retina, are not representations unless or until
some process further downstream interprets them as such, at which point
they become representations only to that process.
Sizemore >
The retina simply responds to light, it does not experience any kind of
representation, just its own state. Lower brain functions interpret the
signal from the retina as a pattern of light, like the meaningless patterns
in an abstract hallucination. The next higher brain functions interpret
those patterns of light as meaningless objects in an illuminated scene,
like the view of an abstract sculpture. The next higher levels interpret
the pattern of objects as a meaningful scene, for example a view of a face.
And the highest levels of brain function interpret the face as someone you
know, and call up the appropriate response such as nodding or greeting the
experienced person. Only the last stage involves the *whole* visual brain,
although not necessarily the auditory, olfactory, or limbic functions,
which may or may not engage in any particular experience. The lower level
functions do not involve the whole brain, and in fact there are an array of
visual deficits caused by failures of various regions of the brain (e.g.
visual agnosia) that clearly demonstrate that the whole brain does not
always have to be involved in every kind of experience.
And likewise in synthetic vision. An image on a photodiode array is just a
pattern of voltages, it does not represent anything until the video camera
is plugged into a computer that interprets that signal as a pattern of
light, or an image. Further algorithms work on that data on the assumption
that it is a pattern of light, to detect presumed features in the scene,
and then further algorithms work on that feature data to extract presumed
objects from that presumed scene. Those are all representations, whether or
not there is any correspondence between two domains. If the lens cover is
on, then there is no pattern of light, although the rest of the algorithm
mistakenly interprets the signal as an image of light nonetheless.
Sizemore >
The word "see" implies a viewer, and that does not apply inside the brain,
because the viewer is the whole brain. Instead, we experience the states of
our own brain, and when the whole brain directs its attention to an
external object (by way of its internal representations) then we call that
process "seeing". It is a fallacy to insist that experience necessarily
involves eyes and a separate viewer. But we've been over this ground once
before already.
Sizemore >
True enough, but it *is* a *testable* assumption, at least in principle,
and a good way to test it is by demonstration with a simple model system.
Representationalism is easy enough to demonstrate. Direct perception is
more difficult to demonstrate, because nobody has ever articulated the
concept with enough specificity to either build a model, or to make
predictions.
Ok, lets try the *ontological* approach to the Direct Perception
v.s. Representationalist debate. What is the ontology of experience?
First let us agree what we mean by experience. For example visual
experience is the colored three-dimensional volumetric world you see
around you when you open your eyes. This is distinct from the
objective external world in the fact that when you close your eyes,
the visual world disappears, or rather, it is transformed into a foggy
brownish space of indefinite extent, while the real world continues to
exist unaffected by the blinking of your eyes. Whether you are a
direct or indirect perception advocate, we can agree on the definition
of visual experience, even though representationalists believe it to
be located inside the brain, while direct perceptionists locate
experience out in the world beyond the retina.
In either case, visual experience takes the form of modulations of
color qualia across a volumetric space. For example a checkerboard
pattern is experienced as an alternating modulation of black and white
squares. What is the *ontology* of those alternating qualia? What is
it that flips from black to white and back again? We know that it is
an experience, and that experience is spatially extended in a
pictorial fashion, but is there anything in the external physical
world that corresponds to that experienced alternation? What is its
substance? Does it even have one?
The representationalist answer is that qualia are different states of
the physical brain, and thus they are located inside the brain. In
other words the brain must posess some continuous spatial medium
across which extends a pattern of alternating states. Whether these
states correleate with voltages, spiking frequencies, or some standing
wave representation, remains an open question at this point. But
experience has physical presence in the physical universe known to
science.
But what is the direct perceptionist's answer? What is the "stuff"
that changes color across space that you experience? And where is it
located? And is it in principle detectable by scientific means at that
location?
If my experience of a chess board is out there where the chessboard
exists, is the black and white pattern I experience the alternation of
the pigment in the paint on its surface, perceived directly? Or is it
the reflectivity of the surface experienced directly? Or is it the
intensity of reflected light experienced directly? Or is it a pattern
of activation in my retina? What is its ontology?
I think that direct perceptionists are uncertain about the ontology of
experience, they perceive it in a bistable manner, as being both an
external objective, and an internal subjective entity, and their
answers, when probed, flip back and forth between these two as if a
pattern in your brain could somehow be also outside of your head. The
whole concept of direct percption is founded on a profound
epistemological error, that we can in principle be conscious of things
which are not explicitly represented in our brain. Direct perception
states the *problem* of experience, it does not offer a *solution* to
it that can either guide the construction of a model of the concept,
or even an experiment to test the concept. The theory of direct
perception is every bit as mysterious as the property of consciousness
that it is supposed to explain.
The representationalist position is more coherent because it posits a
single ontology, that the modulations of the qualia of visual
experience are modulations of the physical state of your brain across
some spatial representational medium. And it makes the testable
prediction that that medium and its modulations will one day be
discovered and decoded in the brain. My vote is for a standing wave
representation using a Fourier code to produce moving volumetric
holographic images in the brain, and that those images correspond
directly to our experience.
Glen Sizemore will complain that there remains the problem of
experience, and why it is we have it when our brain is in certain
states. But if you accept that mind is a physical process taking place
in the physical mechanism of the brain, and you acknowledge that the
brain is conscious, then that already is an admission that a physical
process taking place in a physical system can under certain conditions
be conscious.
Besides, the mystery of experience, or why consciousness exists in the
brain, is by no means unique to representationalism. Direct perception
cannot resolve that one either. Representationalism at least offers an
account of the functional aspects of experience that can be expressed
in actual models and make testable predictions. Direct perception does
not even offer that much.
Steve
It's an ingenious argument. It seems that you could use that method
to prove that we don't eat food, we eat representations of food.
It makes you wonder how we get our nutrients.
You seems to be assuming the kind of Cartesian theater that Dennett
criticised. I don't agree with that. But even assuming a Cartesian
theater, you are misusing "experience". For example, my experience
in watching a movie includes my emotions and thought. It isn't just
what was played on the screen. Contrary to your assertion, I don't
locate experience out in the world. I don't consider it a thing. It
seems to me that treating experience as a thing is a category mistake.
The argument about blinking the eyes is interesting, because I think
that actually argues for direct perception. If there is some sort of
volumetric representation, then you would think your visual
experience would persist during a blink, perhaps slowly fading out.
May I suggest that the "foggy brownish space of indefinite extent" is
closer to what is represented.
SL >
There is no such stuff. You have confused the issue by your misuse
of "experience".
...
I seriously doubt that there is enough DNA in the human genome to
encode the hardware specifications that would be required for the
proposed system.
In another message ("Conceptual vs. Empirical Issues"), Steven wrote:
SL >
There is nothing explanatory about representationalism. Most
representationalists admit that they are unable to explain conscious
experience. The argument about an infinite regression of homunculuses
keeps coming up precisely because representationalism explains
nothing.
Whether a system uses direct perception, or is based on
representations, in an implementation issue, not an explanatory
issue. So lets stop the arguing, and wait for until there is enough
empirical evidence to answer questions about implementation details
in homo sapiens.
Rickert >
No, just that we experience internal representations of the
external "nouminal" food that nourishes us.
Rickert >
It seems there are a lot of these "category mistakes" in direct perception.
I see a spatial structure that is my experience, it is distinct from the
world itself, and yet we are not permitted to think of that spatial
structure as anything we can talk about. It seems that a lot of direct
perception involves prohibitions against certain concepts, as in
behaviorism, that fobade discussion of conscious experience. Curiously,
all of these forbidden concepts are the very things that reveal the
incoherency of direct perception.
Rickert >
I see. Because closing the eyelids makes the real world out there cease to
exist momentarily. Hmmmm...
Rickert >
No, you would expect it to blink out immediately when the input data stream
is blocked, like the image on a photodiode array when you put the lens
cover on.
Rickert >
There I believe you put your finger on what I believe is the principal
reason why representationalism is generally not given serious consideration.
Rickert >
Well it does explain the *functional* aspect of vision, that is, how the
information of the world gets in to the computational hardware of the
brain. Direct perception does not even explain that much. And direct
perception does not explain experience either, it merely prohibits
discussion of it.
Rickert >
I think that is wise. We are not making any further progress in
understanding each other. With this last ontological argument I have
expended my last big arrow from my quiver. I will provide a summary
overview of all the arguments we have covered in these various threads
under the Subject:
Summary: Direct Perception v.s. Representationalism
Steve
We have had a full and informative exchange on the question of
Direct Perception v.s. Representationalism, and it seems we have
pretty much exhausted that topic by now, as we are now just repeating
the same arguments over and over again. So I will bow out of the
discussion in a major way at this point, although I will answer any
residual questions that people might have. I have no more major
arguments to make on the subject. Most sincere thanks to all who
have contributed, and to all those lurkers out there who have found
this debate worth following. If nothing else, it is in my view a most
fascinating topic, and resolving this issue once and for all would be
of the greatest significance for philosophy, psychology, and
neuroscience. I hope we have advanced that cause if only by a small
notch.
I will take this opportunity to summarize the whole debate as I see it
from the representationalist perspective. The origin of direct
perception is naive realism. It is almost impossible to shake the very
vivid impression that what we see in experience is the world
itself. But this concept is profoundly at odds with the
neurophysiology of perception, with sense organs that transmit
information to the brain, an obviously representational system. One of
these views must be right, and the other is wrong, unless as Andrew
Brook claims, the truth lies somewhere in between.
It turns out however that the direct perception view is incoherent,
whether in its pure or partial form, because it involves the organism
having knowledge of things which are not explicitly represented in its
brain. Proponents of direct perception, from Gibson onwards, must
sense some kind of a problem at least subconsciously, as seen in their
supreme confidence that their view is right, even though they cannot
articulate their position with sufficient specificity as to explain
how an artificial robotic system could possibly be built that operates
by direct perception. Even more surprising, they cannot seem to
formulate any possible future experiment that would resolve the issue
definitively one way or the other. And strangest of all, they appear
to have a peculiar blindness to this gaping hole in their view of
perception, as if they simply cannot understand the objections of the
representationalists. It would be one thing if they acknowledged that
there is some kind of paradox involved, but that they consider that
paradox more tolerable than the incredible representationalist
view. But direct perceptionists cannot seem to even bring themselves
to acknowledge that there is any kind of problem at all.
The more extreme form of direct perception espoused by Gibson is like
behaviorism: it is full of prohibited concepts and so-called "category
errors" that forbid one to acknowledge experience as a "thing", or to
recognize our experience as pictorial, or that it has spatial extent,
when those facts are plain for all to see as soon as they just open
their eyes. "Seeing is behavior", Glen Sizemore tells us, and there is
supposedly no meaning to words like "information" and "representation"
and "processing", concepts which have become part of our everyday
lexicon since the arrival of computer technology in our lives. One
gets the sense that direct perception is a religion rather than a
scientific hypothesis, as seen in Gibson's stark refusal to even
discuss sensory processing at all.
Andrew Brook's more moderate concept of "direct representationalism"
sounds at first more reasonable, because he allows for representations
in the brain, but paradoxically, he insists that we view the world
"directly" *through* the representations in our brain, a concept that
strikes me as a contradiction in terms. If our brain uses
representations, then of course our experience consists of those
representations, and we cannot view the world directly except by way
of them. When questioned more closely, Brook wriggles and squirms
until his explanation becomes almost identical to a
representationalist thesis, with the sole exception that he continues
to insist that our experience of the world is direct, although he
cannot even explain exactly what he means by that term.
Until the *principle* of direct perception can be demonstrated in a
simple robot model, the concept is so vague and incoherent as to be
essentially meaningless.
Conscious experience has an information content, and information
cannot exist without a physical medium or carrier to carry that
information. That medium can only be the brain, where the sensory
nerves terminate, and from whence the motor nerves originate.
A theory that makes no predictions about possible future discoveries
in the brain, is no scientific theory at all, but is more like an
article of faith.
Direct perception cannot explain the *ontology* of the vivid spatial
structure of visual experience, those volumetric objects bounded by
continuous colored surfaces that disappear when we close our
eyes. These are obviously a product of the brain, and yet they appear
out in the world. They can only be states of our own brain, and thus be
located in our brain.
Like Behaviorism, direct perception only survives by simply
prohibiting discussion of concepts like the information,
representation, and processing in the brain, and by prohibiting
discussion of the manifest properties of the vivid spatial structure
of experience.
The chief argument raised against representationalism is the question
of experience, and why we have it when our brain processes sensory
information. This is indeed a deep philosophical quandary. However we
know for a fact that experience does exist, and that the brain is the
organ of conscious experience. The only reasonable location for
experience is inside the brain as a representation.
We do not need "internal eyes" to "see" the representations in our
brain, we simply experience the structure of certain patterns of
energy in our brain directly. While this concept may seem deeply
troubling to some, it is nowhere near as troubling as the concept of
awareness of the world out beyond the brain, where there is no
computational or representational hardware to do the experiencing. Direct
perception does not offer any better explanation for the question of
experience, except by prohibiting discussion of
experience as a "category error".
O'Regans concept of probing the external environment as if it were an
internal memory is demonstrably false, because the three-dimensional
spatial information of the external world is by no means immediately
available from glimpses of the world, but requires the most
sophisticated and as-yet undiscovered algorithm to decipher that
spatial information from the retinal input.
The absurdity of O'Regan's concept is highlighted by the condition of
visual agnosia, a visual integration failure, because the condition of
apperceptive agnosia is the failure of a visual function whose
existence O'Regan effectively denies.
The question of direct perception v.s. representationalism is not a
conceptual issue that is beyond the reach of science, it is a very
significant empirical issue with profound implications for the nature
of perceptual representation in the brain, and for our attempts to
replicate the principle of perception in artificial robot models.
The phenomena of dreams and hallucinations clearly demonstrate that
the brain is capable of constructing vivid spatial experiences in the
absence of an external world available for direct inspection. Perception
is a guided hallucination, constrained by sensory input.
In the face of this overwhelming array of indisputable evidence, how
can intelligent, educated people continue to insist that our
experience of the world is direct? The answer is the very vivid
impression that our experience simply *appears* direct. Some people
just cannot bring themselves to accept the view towards which all of
the evidence inevitably points. This issue is the ultimate example of
a paradigm debate, because seeing things from the representationalist
perspective requires that one inverts one's entire epistemology to
recognize that the world which appears outside is actually inside
one's head. It is admittedly a difficult concept to swallow.
I have prepared an exerpted summary of this whole debate on-line, which
can be found at:
http://cns-alumni.bu.edu/~slehar/cartoonepist/EpistDebate.html
Steve Lehar
Steve Lehar has presented a strong defense of representation in his
summary. Unfortunately, as he himself has acknowledged previously, it
is unlikely to cause too many minds to change. Many people have
suggested that maybe there is a way to reconcile the opposing views of
direct vs indirect perception, and this seems desirable since the
absence of empirical evidence that would clearly determine the correct
position suggests that there is a possibility that both approaches
could be wrong, or both could be right. I submit that both could be
right.
...
In sum, yes there is a picture, but it is not a copy of the world. It
is, for all intents and purposes, the world.
Carreno >
The only form of direct perception that might be right is one that is
experimentally indistinguishable from representationalism, and thus makes
no testable predictions to distinguish between the two. That form of direct
perception however is indistinguishable from representationalism, because
it posits that every aspect of experience is necessarily replicated in the
brain.
The more extreme version of direct perception that prohibits
representations of any sort, cannot be right if representationalism is also
right.
Carreno >
But there are two aspects of the "picture", one that disappears when you
close your eyes, and the other that continues to exist unchanged. That
clearly places one world on the outside beyond your eyelids, and the other
one inside on this side of your eyelids. The failure to distinguish these
fundamentally different worlds is exactly the theory of direct perception.
Carreno >
It is very easy to elucidate. If "taking it at face value" means making a
sensory image of it and sending it to the brain for processing, then that
is representationalism, and "direct perception" would be wrong. If "taking
it at face value" means behavior and experience *as if* we had a copy of
the world in our brain, but we actually *don't*, then the causal loop
between perception and behavior is critically broken, and perceptual motor
function remains forever a deep dark mystery.
Steve
Steven indeed gave the representationalist perspective. It was
perhaps more a polemic than a summary.
SL>Until the *principle* of direct perception can be demonstrated in a
We should put this in perspective. There is no robot model that
demonstrates representational perception. There is no robot model
that credibly demonstrates any kind of perception.
-NWR
Reply to Neil Rickert:
Rickert >
Obviously it was a view from the representationalist perspective; feel free
to compose a summary from the "direct perception" perspective. And there
was an overall message to that polemic, which is that direct perception may
seem plausible enough when debated issue by issue, but it falls apart when
viewed in the aggregate, when one takes a "big picture" view as revealed in
the summary. Advocates of direct perception cannot even agree on what their
theory states, whether there are any representations in the brain or not,
or if there are, whether they encode all of experience or only some of it.
They cannot make testable predictions of future experiments that would
resolve the matter one way or the other, nor can they build a functioning
robot to demonstrate the *principle* behind the concept.
The motivation behind direct percepton on the other hand is perfectly
clear, it is motivated by the very vivid naive impression that what we
experience is the world itself, rather than an internal representation. In
that sense it is very much like the "animism" of the turn of the last
century whose advocates insisted that life is "something more" than just
chemical reactions, although they were unable to define what the added
ingredient might be, or how it would be detected in principle, or how it
could be implemented in a simple model.
Lehar >>
Rickert >
Any robot model with a camera, computer, and servos, demonstrates the
*principle* behind representationalism, and thus clarifies concepts such
as "information", "representation", and "processing" in terms that are
perfectly clear to anyone who has used a computer. There is *no* such
simple demonstration of the *principle* behind direct perception, nor is
there even consensus among advocates what that term actually means, or what
it says with any specificity about the perceptual process, or how it would
be implemented in a robot model. It is quite extraordinary how dogmatically
and with such supreme confidence the advocates defend what is a pretty
vague concept.
Steve
Direct perception is entirely consistent with the neurophysiology
perception. It is inconsistent merely with a certain ideology of what
follows from these facts. Steve, saying what you do here shows as
serious an unwillingness to come to grips with what we are actually
saying as your earlier statement that we believe that representation
is outside the head. We believe that representation can be of or about
thing outside the head. But your statement is about ... the nature of
the vehicle, mine is about .... what it is about, what
it makes us conscious of. If you refuse to even try to grasp this
distinction, I hypothesis that you are letting ideology substitute for
evidence and argument.
Lehar >
[The clause in asterisks] No, it does not ... and I have said why over
and over and over. What does it take to get you at least to
acknowledge that we believe what we say we believe? Go back over my
messages. I have addressed this point at least half a dozen
times. Again, I suspect you are letting unshakeable ideology go proxy
for looking at the arguments.
Brook>
You have yet to tell us whether the representations in the brain
necessarily encode *all* of experience or just *some* of it. This may
seem like an irrelevant detail to you, but for me that issue is
*everything*. Because if you acknowledge that some experience is not
explicitly represented, then you actually *ARE* claiming that part of
your experience is out in the world, not in your head, and I challenge
you to demonstrate *that* in a robot model. So are you herewith now
denying that *any* aspect of experience can exist without being
explicitly represented in your brain?
Lehar >>
Brook >
[Original Message]
Date: Thu, 7 Apr 2005 07:09:55 -0700
From: Glen Sizemore
Subject: Re: The geometry of conscious experience
In fact there is no need for an internal observer of the scene, since
the internal representation is simply a data structure like any other
data in a computer, except that this data is expressed in spatial form
(Earle 1998, Singh & Hoffman 1998). For if a picture in the head
required a homunculus to view it, then the same argument would hold
for any other form of information in the brain, which would also
require a homunculus to read or interpret that information. In fact
any information encoded in the brain needs only to be available to
other internal processes rather than to a miniature copy of the whole
brain. The fact that the brain does go to the trouble of constructing
a full spatial analog of the external environment merely suggests that
it has ways to make use of this spatial data.
I do not think that the absurdities of representationalism
go away so easily. It is becoming widely known that saying we see a
representation immediately raises the issue of how seriously to take
it. If it is a metaphor, then it is useless (unless we can show that
it is likely to somehow be literal), but if it is taken literally,
then it raises the issue of infinite regress.
[Original Message]
Date: Fri, 8 Apr 2005 09:42:17 -0500
From: Steven Lehar
Subject: Re: The geometry of conscious experience
[Original Message]
Date: Sat, 9 Apr 2005 06:03:02 -0700
From: Glen Sizemore
Subject: Re: The geometry of conscious experience
The key issue is not the word "see" or "use", to describe how the
brain makes use of internally represented information, it is the
question of whether the internal representation is processed by other
internal processes in the brain, or whether it requires a copy of the
*whole brain* in order to "see" that represented data. Only the latter
formulation leads to the infinite regress.
< Lehar
I disagree. The key issue is whether or not what
has to be explained is simply placed inside the head
(as a category error). We ask "What is 'seeing'"? and
representationalism answers that "it is the 'seeing'
of a representation." It doesn't matter if the answer
is "it is one part of the brain 'seeing' the
representation located in another part." Or: "it is
one part of the brain 'making use of' a representation
in another." Or: "it is one part of the brain
"processing" the representation in another part."
If Sizemore claims that a picture in your brain requires a miniature
copy of your whole brain to "see" that picture, then surely this
objection would apply to *any* information represented in the brain,
including verbal, linguistic, and cognitive knowledge, all of which
would require a miniature copy of the whole brain to interpret or
process that cognitive information.
< Lehar
I do not only raise the issue with respect to "pictoral data." I am
critical of all metaphorical uses of 'representatation.' I put no
stock in 'information' being 'represented' in the brain.
< Glen Sizemore
[Original Message]
Date: Sat, 9 Apr 2005 11:12:10 -0500
From: Steven Lehar
Subject: Re: The geometry of conscious experience
The key issue is not the word "see" or "use", to describe how the brain
makes use of internally represented information...
<< Lehar
I disagree. The key issue is whether or not what
has to be explained is simply placed inside the head...
< Sizemore
I do not only raise the issue with respect to "pictoral data." I am
critical of all metaphorical uses of "representatation." I put no stock
in "information" being "represented" in the brain.
< Sizemore
I claim that explanatory fictions like "representations," and
earlier versions like "beliefs" and "knowledge," shed
no light on what is going on in the nervous system
... What does it mean for "information" to be "processed?" ...
< Sizemore
[Original Message]
Date: Mon, 11 Apr 2005 08:11:06 -0500
From: Steven Lehar
Subject: Direct Perception v.s. Representationalism
[Original Message]
Date: Mon, 11 Apr 2005 11:13:32 -0400
From: Andrew Brook
Subject: Re: Direct Perception v.s. Representationalism
[Original Message]
Date: Sat, 9 Apr 2005 15:37:46 -0700
From: Augustin Carreno [stincarr@GMAIL.COM]
Subject: Re: The geometry of conscious experience
[Original Message]
Date: Tue, 12 Apr 2005 09:08:14 -0500
From: Steven Lehar
Subject: Re: Direct Perception v.s. Representationalism
Steve Lehar's message on anti-representationalism seems to me to be
exactly right. Theorists who present themselves as denying the
existence of representations almost always, on closer inspection, turn
out to be denying that certain kinds of representations exist
< Andrew Brook
I think the argument against representation should be made based on
physics. ... can the brain really afford to hold
a copy of the entire universe together with whatever extra brain power
is needed to analyze it? Judging by nature's preference for parsimony,
the answer is no.
< Carreno
[Original Message]
Date: Tue, 12 Apr 2005 12:16:30 -0400
From: Andrew Brook
Subject: Re: Direct Perception v.s. Representationalism
But before
deploying Occam's razor we must first balance the scales, and take a full
accounting of the alternatives under consideration. For the alternative is
that we experience the world directly, as if bypassing the causal chain of
sensory processing. As incredible as representationalism might seem, the
alternative is even more incredible.
< Lehar
[Original Message]
Date: Tue, 12 Apr 2005 14:46:32 -0500
From: Steven Lehar
Subject: Re: Direct Perception v.s. Representationalism
How we can do this is a wonderful mystery but that we do it
seems, to me at least, pretty much beyond question.
< Brook
[Original Message]
Date: Tue, 12 Apr 2005 21:27:14 -0400
From: Andrew Brook
Subject: Re: Direct Perception v.s. Representationalism
[Original Message]
Date: Wed, 13 Apr 2005 10:19:40 -0500
From: Steven Lehar
Subject: Re: Direct Perception v.s. Representationalism
I'm not sure what's going on here. We should be on the same side.
< Brook
Here is why I think that some form of direct realism is virtually
unrejectable. When I open my eyes, I see a chair just as directly as when I
kick it, I kick a chair. In both cases, it is the chair that I am in
contact with. In neither case am I in ... contact with any intermediary.
What could be more direct than seeing or kicking an object in plain view in
front of me? If this is not direct, what would be?
< Brook
1. Simple robot model? I sketched how at the end of my message. In the same
way that vision systems since Marr's have been able to extract three
dimensional objects from two dimensional arrays, our vision system not only
extracts three dimensional objects but allows us 'reverse infer' down the
causal chain to be directly aware of them. What more do you want?
< Brook
[Original Message]
Date: Thu, 14 Apr 2005 14:20:55 -0500
From: Neil W Rickert
Subject: Re: Direct Perception v.s. Representationalism
But Gibson's profound epistemological error backed him into a corner which
is ultimately indefensible, which made him get defensive and dogmatic in
his later years, as often happens to those who commit themselves to
defending the indefensible.
[Original Message]
Date: Sun, 17 Apr 2005 09:37:47 -0500
From: Steven Lehar
Subject: Re: Direct Perception v.s. Representationalism
Yet I could nowhere find a clear statement of what Gibson's "profound
epistemological error" is presumed to be.
< Rickert
But I am quite sure that Gibson understood the physics and biology, and
recognized that stimulation of retinal cells and the transmission of
signals on the optic nerve were part of the causal processes involved in
that direct perception.
< Rickert
"The very idea of a retinal pattern sensation that can be impressed on the
neural tissue of the brain is a misconception, for the neural pattern never
even existed in the retinal mosaic. There can be no anatomical engram in
the brain if there was no anatomical image in the retina. The retina jerks
about. It has a rapid tremor. It even has a gap in it (the blind spot). It
is a scintillation, not an image. ... The whole idea stems from the
persistent myth that there has to be something in the brain that is
visible, and from Johannes Müller's assumption that the nerves telegraph
messages to the brain."
[Original Message]
Date: Tue, 19 Apr 2005 15:33:56 -0400
From: Andrew Brook
Subject: Re: Direct Perception v.s. Representationalism
Ok, stand in front of a chair, and before you kick it, touch your finger
to one eyeball (through the eye lid) and push it gently to one side,
until you see a double image. Now KICK! Now you see TWO chairs, and TWO
feet kicking them! Which one is the "real" chair and the "real" foot?
And what is the actual objective location of that chair? If this is not
IN-direct, what would be?
Well, it is still the chair that we see -- and we have extremely
reliable means for sorting out when the perceptual medium is working
well and when it is distorting things. All I want to insist on is that
when we see, we see the world (most of the time). We see *via* various
media, but we not aware of these media (except when things go wrong or
we otherwise pay attention to them), we are aware of the things in the
world that kick them into action. That's all.
of course there is intermediate *machinery" but that
does not mean that the resulting *perception* or *consciousness* is
indirect, i.e., by inference from something in the head.
Unless, that is, you mistake our experience for a direct view of the
world itself, bypassing the chain of sensory processing, which in
impossible in principle!
Why should a 'direct view of the world' have to bypass sensory
processing? It proceeds via, though, sensory processing. What makes it
direct is that it makes us aware of the world, not the processing. How
it does this is, at this point in history, mostly wide-open.
It is extremely difficult to banish the last vestiges of naive realism
from our philosophy.
I don't know what naive realism is but I certainly do not want to banish
realism because it is true! (IMHO, of course.)
[Original Message]
Date: Tue, 19 Apr 2005 10:09:43 -0500
From: Steven Lehar
Subject: Theories v.s. Paradigms
"There are more things in heaven and earth, Horatio, than are dreamt of in
your philosophy."
[Original Message]
Date: Tue, 19 Apr 2005 15:03:40 -0500
From: Neil W Rickert
Subject: Re: Theories v.s. Paradigms
Ok, we've been round and round the direct perception v.s.
representationalism debate enough times to see that nobody is about to
change their minds, no matter HOW eloquent or persuasive the arguments are
on either side. Why is this so?
There is an asymmetry in this debate: all representationalists were once
naive realists, whereas most direct perceptionists have never been
representationalists.
But before we deploy Occam's Razor, we must first balance the scales and
take a full accounting of the alternatives under consideration. For the
alternative is that we can somehow become aware of objects and surfaces in
the external world *without* the mediation of sensory processing and
internal representations.
[Original Message]
Date: Wed, 20 Apr 2005 09:52:59 -0500
From: Steven Lehar
Subject: Re: Theories v.s. Paradigms
Steven perhaps sees this as a problem. I don't. While I tilt toward
the direct perception side, I am not all all concerned that Steven
and Alex are strongly committed to representationalism. We don't
need to put all of our eggs in the one basket.
< Rickert
But before we deploy Occam's Razor, we must first balance the scales and
take a full accounting of the alternatives under consideration. For the
alternative is that we can somehow become aware of objects and surfaces in
the external world *without* the mediation of sensory processing and
internal representations.
<< slehar
However, that is not the alternative. It seems that Steven can only see
one side of the paradigm shift.
The alternative, or one alternative, is that we can become aware of
objects and surfaces *with* the mediation of sensory processing, but
*without* the mediation of internal representations.
< Rickert
[Original Message]
Date: Tue, 19 Apr 2005 20:47:36 -0400
From: Andrew Brook
Subject: Re: Theories v.s. Paradigms
"We can somehow become aware of objects and surfaces in the external world
*without* the mediation of sensory processing"
"we can somehow become directly, i.e., noninferentially, aware of
objects and surfaces in the external world *through* the medium of
sensory processing"
[Not released on PSYCHE-D (censored?)]
Date: Wed, 20 Apr 2005
From: Steven Lehar
Subject: Re: Theories v.s. Paradigms
"We can somehow become aware of objects and surfaces in the external world
*without* the mediation of sensory processing"
<< Lehar
But that is not my view at all!!!! As I have said over and over and
.... OVER! (The caps may get me censored.)
< Brook
we are aware of objects around us, not just end-products of sensory
processing in our brains
< Brook
[Original Message]
Date: Wed, 20 Apr 2005 11:17:11 -0700
From: Glen Sizemore
Subject: Re: Theories v.s. Paradigms
There is an asymmetry in this debate: all
representationalists were once naive realists, whereas
most direct perceptionists have never been
representationalists.
Obviously, you are asserting that "representationalism" is a view
battling for acceptance, but nothing could be further from the
truth. The fact is that it is one of the first academic positions
concerning human behavior to be taken seriously. It wasn't really
until the 20th century with behaviorism (and I include later
Wittgenstein here) that any serious challenge was mounted. Mental
"theories" of representation simply morphed seamlessly into
neurobiological "theories".
[Original Message]
Date: Wed, 20 Apr 2005 15:32:56 -0400
From: Andrew Brook
Subject: Which came first
[Original Message]
Date: Fri, 22 Apr 2005 10:27:21 -0500
From: Steven Lehar
Subject: Re: Which came first
One good name for my view ... would be: direct representationalism.
< Andrew Brook
[Original Message]
Date: Fri, 22 Apr 2005 05:51:30 -0500
From: Steven Lehar
Subject: What would it take to convince you?
[Original Message]
Date: Fri, 22 Apr 2005 13:39:50 -0400
From: Andrew Brook
Subject: Re: What would it take to convince you?
[Original Message]
Date: Fri, 22 Apr 2005 15:30:11 -0500
From: Steven Lehar
Subject: Re: What would it take to convince you?
Steve, absolutely the right approach, ingeniously carried out. Trouble
is, it does not address the issue between, for example, you and me. I
can't imagine anything that would 'prove' direct perception *as you
describe it*, ... But within representationalism, there are two houses.
There is the house of those who think 'If a representational medium is
present, the results can only be indirect perception/consciousness.'
And there is the house of those who think, 'The right kind of
representational medium lets us go right through it all the way to the
world itself, so that the result is direct perception/consciousness.'
< Andrew Brook
[Original Message]
Date: Fri, 22 Apr 2005 12:02:14 -0700
From: Glen Sizemore
Subject: Re: What would it take to convince you?
[Original Message]
Date: Fri, 22 Apr 2005 16:10:43 -0500
From: Steven Lehar
Subject: Re: What would it take to convince you?
[Original Message]
Date: Fri, 22 Apr 2005 14:12:59 -0500
From: Neil W Rickert
Subject: Re: What would it take to convince you?
H1: REPRESENTATIONALISM IS PROVEN!
< Lehar
H2: DIRECT PERCEPTION IS PROVEN!
< Lehar
[Not published on PSYCHE-D. Censored?]
From: Steve Lehar
Date: 4/22/2005
Subject: What would it take to convince you?
You are right. That would not convince me.
< Neil Rickert
The trouble with Steven's two examples is that they are both
quite implausible. They look like sleight of hand parlor tricks.
A scientist needs better evidence than that.
< Rickert
To convince me, I need a detailed account of the relevant processes.
This should, preferably, be at the level of information processing.
It should plausibly account for all of the unresolved questions. And
it should be supported by empirical evidence.
< Rickert
[Original Message]
Date: Sat, 23 Apr 2005 08:33:27 -0400
From: "Glen M. Sizemore"
Subject: Conceptual vs. Empirical Issues
But even paradigmatic issues can be determined by experiment,
at least in principle. The earth-centered cosmos has now been
conclusively rejected by the "experiment" of flying around the back
side of the moon without colliding with the crystal sphere
This only shows that "paradigmatic issues" are not
necessarily "conceptual issues." Besides, you yourself emphasized the
incommensurability of "opposing paradigms."
Any "theory" that is *NOT* testable by any experiment, even
in principle, is not a scientific theory at all, but a pure
*belief*.
But, as I have pointed out, the issues in question are
NOT theoretical issues. They are conceptual issues. They concern the
ASSUMPTIONS that underlie theory. And the empirical side of things is
frequently distinct from both theory and concepts
If you cannot describe an experiement that could prove
direct perception at least *in principle* with some kind of future
technology, then direct perception is a *belief* not a theory, since
it predicts nothing different than the alternative representationalist
hypothesis.
But the truth is that the principle of representationalism *is*
demonstrable in a simple robotic system, while the principle of direct
perception remains as mysterious as the immaterial soul! After all
these rounds of debate, we still have no idea how such a system could
possibly be built in a real physical system.
I don't think that representationalism is "demonstrable
in a simple robotic system". ... showing orderly relations between
"the world" and a "pattern of neural activity" is not "proof" of
representation. Representation, like "computation," is defined by its
functional relation to the behavior of an animal, human or
otherwise. To invoke "representation" is to invoke behavioral
characteristics of whole animals, hence the charge of homunculism.
[Original Message]
Date: Sat, 23 Apr 2005 15:25:20 -0500
From: Steven Lehar
Subject: Re: Conceptual vs. Empirical Issues
[Original Message]
Date: Sat, 23 Apr 2005 15:35:20 -0500
From: Steven Lehar
Subject: Re: Conceptual vs. Empirical Issues
[Original Message]
Date: Sun, 24 Apr 2005 15:59:44 -0400
From: Andrew Brook
Subject: Re: Conceptual vs. Empirical IssuesE
[Original Message]
Date: Mon, 25 Apr 2005 09:44:18 -0500
From: Steven Lehar
Subject: Re: Conceptual vs. Empirical Issues
[Original Message]
Date: Mon, 25 Apr 2005 15:56:21 -0400
From: Andrew Brook
Subject: Re: Conceptual vs. Empirical Issues
[Original Message]
Date: Mon, 25 Apr 2005 16:13:12 -0500
From: Steven Lehar
Subject: Re: Conceptual vs. Empirical Issues
[Original Message]
Date: Tue, 26 Apr 2005 17:10:23 -0400
From: Andrew Brook
Subject: Re: Conceptual vs. Empirical Issues
[Original Message}
Date: Wed, 27 Apr 2005 09:58:33 -0500
From: Steven Lehar
Subject: Re: Conceptual vs. Empirical Issues
1. Stop the aggressive bombast. I am not intimidated.
< Brook
2. I have put a lot of effort into specifying my position in detail,
much more than you have put into specifying yours.
< Brooks
A Cartoon Epistemology
Gestalt Isomorphism and the Primacy of Subjective Conscious Experience
The World In Your Head: A Gestalt view of the mechanism of conscious experience
[Original Message]
Date: Mon, 25 Apr 2005 18:24:46 -0700
From: Glen Sizemore
Subject: Re: Conceptual vs. Empirical Issues
Because the theory of direct perception states that
there is no need for internal representations, or in
the softer version defended by Andrew Brook, the
internal representations do not need to encode ALL of
the information of experience, because SOME of that
information can be perceived directly *through* the
representation (whatever that means).
Here is where Lehar ignores a point I have made several times. We
cannot identify "representation" with some observation of a
correspondence between two domains. If we did, then all
correspondences would be a matter of representation. That would make
all of science representation. ... Whole persons look at
representations and they "use" their eyes and their brain. How does a
"part of the brain" do this, and how do we justify "our" belief that
this could possibly make sense? ... Lehar may say that we don't see
the representation ... but the problem is that that is what we do with
literal visual representations, we see them. If the things we say
about literal representations do not apply to "brain representations,"
what does that say about "our" cavalier use of language?
The *only* kind of paradigm that remains purely conceptual, in
Sidemore's usage, are theories that make no testable predictions
whatsoever. ... But such theories are not theories at all, they are
*beliefs*, and thus fall outside the realm of science.
As long as these views remain assumptions, and it is not clear that
they can ever be anything else, philosophical debate and conceptual
analysis is the only avenue via which they can be compared. Did Lehar
not, in fact, argue this himself when he invoked Kuhn a couple of days
ago?
< Sizemore
[Original Message]
Date: Wed, 27 Apr 2005 09:05:28 -0500
From: Steven Lehar
Subject: Re: Conceptual vs. Empirical Issues
Here is where Lehar ignores a point I have made several times. We cannot
identify "representation" with some observation of a correspondence between
two domains. If we did, then all correspondences would be a matter of
representation. That would make all of science representation.
< Sizemore
Whole persons look at representations and they "use" their eyes and their
brain. How does a "part of the brain" do this, and how do we justify "our"
belief that this could possibly make sense?
< Sizemore
Lehar may say that we don't see the representation ... but the problem is
that that is what we do with literal visual representations, we see them.
If the things we say about literal representations do not apply to "brain
representations," what does that say about "our" cavalier use of language?
< Sizemore
Some questions are empirical questions and some are conceptual. It may be
that conceptual questions may become empirical, but the fact of the matter
is that the notion that "representations" are necessary to explain
perception is simply an assumption.
< Sizemore
[Original Message]
Date: Mon, 25 Apr 2005 10:06:44 -0500
From: Steven Lehar
Subject: Ontology of Experience
[Original Message]
Date: Tue, 26 Apr 2005 16:45:36 -0500
From: Neil W Rickert
Subject: Re: Ontology of Experience
But what is the direct perceptionist's answer? What is the "stuff"
that changes color across space that you experience?
< SL
For you direct perception offers the "simplicity" of being intuitively
believable, whereas for me representationalism offers the "simplicity" of
being causally explanatory.
< SL
[Original Message]
Date: Wed, 27 Apr 2005 13:27:42 -0500
From: Steven Lehar
Subject: Re: Ontology of Experience
It's an ingenious argument. It seems that you could use that method
to prove that we don't eat food, we eat representations of food.
It makes you wonder how we get our nutrients.
< Rickert
you are misusing "experience". ... I don't consider it a thing. It
seems to me that treating experience as a thing is a category mistake.
Rickert
The argument about blinking the eyes is interesting, because I think
that actually argues for direct perception.
< Rickert
If there is some sort of volumetric representation, then you would think
your visual experience would persist during a blink, perhaps slowly fading
out.
< Rickert
I seriously doubt that there is enough DNA in the human genome to
encode the hardware specifications that would be required for the
proposed system.
< Rickert
There is nothing explanatory about representationalism. Most
representationalists admit that they are unable to explain conscious
experience. The argument about an infinite regression of
homunculuses keeps coming up precisely because representationalism
explains nothing.
< Rickert
So lets stop the arguing, and wait for until there is enough
empirical evidence to answer questions about implementation details
in homo sapiens.
< Rickert
[Original Message]
Date: Thu, 28 Apr 2005 08:13:24 -0500
From: Steven Lehar
Subject: Summary: Direct Perception v.s. Representationalism
[Original Message]
Date: Thu, 28 Apr 2005 12:19:14 -0700
From: Augustin Carreno
Subject: Re: Summary: Direct Perception v.s. Representationalism
[Original Message]
Date: Thu, 5 May 2005 11:32:41 -0500
From: Steven Lehar
Subject: Re: Summary: Direct Perception v.s. Representationalism
there is a possibility that both approaches could be wrong, or both
could be right. I submit that both could be right.
< Carreno
In sum, yes there is a picture, but it is not a copy of the world. It
is, for all intents and purposes, the world.
< Carreno
if we stipulate that the world is indeed already there, then the advantage
of making a copy of it over taking it at face value is difficult to
elucidate.
< Carreno
[Original Message]
Date: Thu, 28 Apr 2005 15:26:48 -0500
From: Neil W Rickert
Subject: Re: Summary: Direct Perception v.s. Representationalism
SL>simple robot model, the concept is so vague and incoherent as to be
SL>essentially meaningless.
[Original Message]
Date: Thu, 5 May 2005 12:10:25 -0500
From: Steven Lehar
Subject: Re: Summary: Direct Perception v.s. Representationalism
Steven indeed gave the representationalist perspective. It was
perhaps more a polemic than a summary.
< Rickert
Until the *principle* of direct perception can be demonstrated in a
simple robot model...
<< Lehar
We should put this in perspective. There is no robot model that
demonstrates representational perception. There is no robot model
that credibly demonstrates any kind of perception.
< Rickert
[Original Message]
Date: Thu, 28 Apr 2005 16:40:50 -0400
From: Andrew Brook
Subject: Re: Summary: Direct Perception v.s. Representationalism
It turns out however that the direct perception view is incoherent,
whether in its pure or partial form, **because it involves the organism
having knowledge of things which are not explicitly represented in its
brain.**
< Lehar
[Original Message]
Date: Thu, 5 May 2005 12:29:28 -0500
From: Steven Lehar
Subject: Re: Summary: Direct Perception v.s. Representationalism
Steve, saying what you do here shows as serious an unwillingness to
come to grips with what we are actually saying as your earlier
statement that we believe that representation is outside the head. If
you refuse to even try to grasp this distinction, I hypothesis that
you are letting ideology substitute for evidence and argument.
< Brook
**because it involves the organism having knowledge of things which are not
explicitly represented in its brain.**
<< Lehar
[The clause in asterisks] No, it does not ... and I have said why over and
over and over. What d