The Epistemology of Direct Realism
Michael Schmitz
raised are epistemological ones, and I'm trying to distinguish the different
aspects of Direct Realism (DR), I title the post accordingly.
[MS, previously]
> Now, if you want to hold that objects of seeing are mental objects, I
think
> you need a very good argument. If you simply declared to ordinary people
> that as a matter of fact you have seen mental objects, I think people will
> find that a bit strange. My point was that the so-called argument from
> illusion could not be such an argument, because this argument really
> *presupposes* the existence of intentional contents directed at external
> objects, because it is supposed to start from such facts as that a person
> hallucinates that there is a dagger in front of him.
>
> Cathy:
> I agree that it sounds odd to talk about seeing mental objects, but that's
> because natural language simply doesn't distinguish regularly between
mental
> and physical objects. It simply assumes (or rather, the conceptual
> framework which underlies natural language assumes) that some mental
objects
> are physical objects, and the result of this is the naive realism implicit
> in natural language.
It seems to me that it is rather you who simply assumes, apparently without
argument, that some physical objects are really mental objects. We agree
that naive or direct realism (terminologically I prefer "direct", for
reasons that I will soon post about) is implict in natural language, but I
think that also means that the onus is on you to give reasons against
realism. Terms like "seeing", "hearing" are implicitly defined through their
occurrences in such contexts as "I heard the bells, saw the man running
towards me" etc., and if you want to challenge our usual way of speaking you
have to show that there is some tension in it that can only be resolved by
abandoning DR, or that there are facts unknown to common sense which force
us to abandon it. It is not enough to simply assert its falsity and say we
tend to believe because it is implicit in language!
> But it's possible, with a bit of practice, to perceive the contents of
> conscious experience as mental objects, by deconstructing what we usually
> perceive to be physical objects into their experiential components.
One can become *aware*, with practice and after a lot of reflection, of the
fact that one has visual experiences with certain representational contents,
but that is an awareness in thought. It does not follow that one sees or
otherwise perceives visual experiences. Indeed there is no sense to the idea
of seeing a visual experience.
>
>
> MS:
> If by that you mean that "see" is sometimes used in a sense in which the
> existence of a visual experience is sufficient for seeing (I), and
sometimes
> in a sense in which the representational success of that experience and
thus
> a corresponding object is required (II), I quite agree. So to avoid
> confusion all we need to do is to resolve that ambiguity, and again, it
does
> not matter which interpretation of "seeing" we chose. Once the ambiguity
is
> dissolved, the baselesness of the argument from illusion becomes apparent,
> for it only seems plausible through a fallacy of equivocation. For the
> premise that something is seen when a dagger is hallucinated is only
> plausible when "seeing" is taken in sense I. But the conclusion that some
> (special mental) *object* is seen will only follow when "seeing" is taken
in
> sense II which requires a corresponding object.
>
> Cathy:
> The problem is that natural language does not distinguish between (I) and
> (II), but routinely concatenates them into a single perceptual package.
So
> if you want to resolve the ambiguity you are already doing something that
> natural language does not do, and you are no longer using natural
language.
So, are you suggesting I did not speak English anymore -;? Well, I suppose
you are saying I'm going beyond common sense in making this distinction. I
agree, but so what? If we never went beyond what is common sense now, there
would never be any progress. You are, as you are saying yourself, not only
going beyond common sense, but you are flatly *contradicting* it on the
issue of DR, and I was pointing out that the best-known argument for your
position only seems compelling if one equivocates on the different senses of
seeing. So I don't see what your point is.
>
> I also think your version of events, that regards a perception as
veridical
> if it relates to some external object, gets the epistemic situation
exactly
> the wrong way round. What actually happens is, we assume some external
> object must be present because we regard our perceptions as veridical.
No, I think you misunderstand me. So far I have not really said anything
about epistemology and about how we *judge* whether some experience is
veridical or not. I was talking about what *constitutes* veridicality and
this is, trivially enough, the presence of the corresponding state of
affairs. And from my point of view the distinction you say I get the wrong
way round does not really make sense. To ascribe veridicality is *ipso
facto* to ascribe the presence of a corresponding object. I don't see how
there can be an issue of epistemological priority at this point. However,
the issue presumably is that you think that epistemologically we start off
in the way Descartes supposed, that is, we only know that we have all these
states of consciousness and the question then is whether we can trust them,
specifically the deliverances of our senses, and you sense correctly that I
do not accept this picture and want to turn it upside down. So let me
conclude with some brief remarks on this.
1) I think we develop the sort of self-consciousness necessary for
self-ascribing states of consciousness and perceptual experiences much later
than the capacity to perceive external objects, so that in fact we are never
in the Cartesian situation. Accordingly, perceiving and self-ascribing
perceptions are clearly distinct. Animals and small children for example,
perceive all the time without knowing that they do.
2) Nor does it make sense even as a thought experiment to doubt all things
at the same time. To doubt some things, other beliefs need to be kept
constant.
3) It is not the case that we generally need reasons to trust our perceptual
experiences. Only in very special situations do we need such reasons - for
example, if we know that in certain contexts we are usually deceived.
Normally *we need reasons to doubt* the deliverances of our senses.
Especially point 3), I think, is the central fact that you and Descartes and
an entire tradition get the wrong way round, and like many contemporary
philosophers I'm trying to turn them up right again.
Best,
Michael
Andrew Brook
will not copy here). However, I would put one central point even more strongly.
Those who try to deny that ordinary perception is direct perception of the world
around cannot even do so successfully. Why? Because there is no stateable notion
of 'direct perception' by contrast to which what we have in normal perception is
not direct perception.
If anyone thinks I am taking too strong a stand here ('the opposition is subtly
incoherent'), there is a very easy way to show this: articulate the missing
concept of direct perception -- a non-mysterious, scientifically-compatible
notion according to which the perception that I have the computer screen in
front of me is something less than direct perception of that screen. I challenge
you. Nay, I defy you. Articulate such a notion and I will shut up. Don't and
Michael and I will have won. Simple. (Hey, it is late and I get feisty at night.)
Andrew
--
Andrew Brook, Professor of Philosophy
Director, Institute of Cognitive Science
Member, Canadian Psychoanalytic Society
2217 Dunton Tower, Carleton University
Ottawa ON, Canada K1S 5B6
Ph: 613 520-3597
Fax: 613 520-3985
Web: www.carleton.ca/~abrook
Cathy Reason
>It seems to me that it is rather you who simply assumes, apparently without
>argument, that some physical objects are really mental objects. We agree
>that naive or direct realism (terminologically I prefer "direct", for
>reasons that I will soon post about) is implict in natural language,
Actually I don't think direct realism is implicit in natural language. What
is implicit in natural language is naive realism, which is not a coherent
philosophical position to begin with.
And nowhere did I say that physical objects were really mental objects. I
said that the subjective experience which makes up our perception of
physical objects can itself be perceived (or "directly attended to", if you
don't like this use of the word "perceived") as mental objects.
>but I
>think that also means that the onus is on you to give reasons against
>realism. Terms like "seeing", "hearing" are implicitly defined through
their
>occurrences in such contexts as "I heard the bells, saw the man running
>towards me" etc., and if you want to challenge our usual way of speaking
you
>have to show that there is some tension in it that can only be resolved by
>abandoning DR, or that there are facts unknown to common sense which force
>us to abandon it. It is not enough to simply assert its falsity and say we
>tend to believe because it is implicit in language!
As a matter of fact I didn't say that we should abandon direct realism.
What I said was, direct realism is trivial, becasue it is really just a
question of epistemological perspective. It's trivial because it doesn't
actually tell us anything at all about consciousness, or how it works, or
what the relationship is between subjective and objective perception.
>One can become *aware*, with practice and after a lot of reflection, of the
>fact that one has visual experiences with certain representational
contents,
>but that is an awareness in thought. It does not follow that one sees or
>otherwise perceives visual experiences. Indeed there is no sense to the
idea
>of seeing a visual experience.
But my point was that we can become aware of our subjective experiences in a
way which is at least as direct as the way we believe we become aware of
physical objects. If you don't want to call that direct awareness "seeing"
or "perceiving", that's ok with me, but I think we need a word to describe
this sort of direct awareness of subjective experience. It is not just " an
awareness in thought", if by that you mean an indirect awareness.
And whatever this particular process of direct awareness is, I shall
maintain that the same process is at work regardless of whether we are
looking at mental or physical objects.
>So, are you suggesting I did not speak English anymore -;? Well, I suppose
>you are saying I'm going beyond common sense in making this distinction. I
>agree, but so what? If we never went beyond what is common sense now, there
>would never be any progress. You are, as you are saying yourself, not only
>going beyond common sense, but you are flatly *contradicting* it on the
>issue of DR, and I was pointing out that the best-known argument for your
>position only seems compelling if one equivocates on the different senses
of
>seeing. So I don't see what your point is.
I absolutely agree that we need to go beyond common sense. But you can
hardly appeal to common sense to support your notion of "seeing", since you
admit you are going beyond common sense.
>No, I think you misunderstand me. So far I have not really said anything
>about epistemology and about how we *judge* whether some experience is
>veridical or not. I was talking about what *constitutes* veridicality and
>this is, trivially enough, the presence of the corresponding state of
>affairs. And from my point of view the distinction you say I get the wrong
>way round does not really make sense. To ascribe veridicality is *ipso
>facto* to ascribe the presence of a corresponding object.
The keyword here is "ascribe". The point is, it is irrelevant to this
process whether some external object is really present or not. The
awareness we have when we perceive something is not dependent on the
presence of some external object.
>I don't see how
>there can be an issue of epistemological priority at this point. However,
>the issue presumably is that you think that epistemologically we start off
>in the way Descartes supposed, that is, we only know that we have all these
>states of consciousness and the question then is whether we can trust them,
> specifically the deliverances of our senses, and you sense correctly that
I
>do not accept this picture and want to turn it upside down. So let me
>conclude with some brief remarks on this.
What Descartes said was that we can never prove the existence of the
external world. He was right. We don't prove it, we just assume it -- as
you have done here.
>1) I think we develop the sort of self-consciousness necessary for
>self-ascribing states of consciousness and perceptual experiences much
later
>than the capacity to perceive external objects, so that in fact we are
never
>in the Cartesian situation. Accordingly, perceiving and self-ascribing
>perceptions are clearly distinct. Animals and small children for example,
>perceive all the time without knowing that they do.
The reason most of us are never in the "Cartesian situation" is that most of
us, most of the time, never feel the need to prove our assumptions about the
external world. We're quite happy just to assume them.
>2) Nor does it make sense even as a thought experiment to doubt all things
>at the same time. To doubt some things, other beliefs need to be kept
>constant.
I agree, otherwise we lapse into empirical incoherence.
>3) It is not the case that we generally need reasons to trust our
perceptual
>experiences. Only in very special situations do we need such reasons - for
>example, if we know that in certain contexts we are usually deceived.
>Normally *we need reasons to doubt* the deliverances of our senses.
I agree again.
>Especially point 3), I think, is the central fact that you and Descartes
and
>an entire tradition get the wrong way round, and like many contemporary
>philosophers I'm trying to turn them up right again.
Actually the whole point of "brain in a vat" stories is that we don't need
to distrust our perceptual experiences -- they work irrespective of whether
the external world exists or not.
Most of us accept that some sort of external world exists, since our
perceptions are not random. But all we can really say about that external
world is that it takes the form of some agency which orchestrates our
conscious perceptions. I have a feeling that beneath these "philosophy of
mind" discussions is a metaphysic which dates back to nineteenth century
materialism (which is itself really just Cartesian dualism with one half
thrown away). And that metaphysic seems to me to have a long way to go
before it even catches up with twenty-first century physics.
Cathy
[Catherine Reason]
Steven Lehar
there is no stateable notion of 'direct perception' by contrast to which
what we have in normal perception is not direct perception.
articulate the missing concept of direct perception -- a non-mysterious,
Articulate such a notion and I will shut up. Don't and Michael and I will
have won. Simple. (Hey, it is late and I get feisty at night.)
Ok, I guess I can step in here. (There's nothing better than a good feisty
debate, although my response will be strictly emotion-neutral!)
Someone earlier raised the analogy of a telescope *through* which one views
the world. Contrast that with a video viewfinder that consists of a
television camera and video display, instead of a lensing system. The
telescope is direct perception, because you cannot see anything unless
there is something there to be seen. The telescope cannot hallucinate. It
can have internal reflections, and dirt on the lens, and so forth, but
those too are directly seen in the telescope (rather than *through* it). If
you were asked what you are seeing, you can truly say that you are seeing
the objects in the field of view, directly, unmediated. What you see is
what's out there. There is no internal register where features have to be
constructed, processed, or stored.
The video viewfinder on the other hand *is* representational. What you are
looking at is glowing phosphor on a glass screen. It is a pattern of
luminence which is actually constructed by the mechanism to represent
something else beyond the lens of the camera. The video viewfinder *is*
capable of true hallucinations. You can see video noise and other visual
artifacts that have no objective reality anywhere beyond the lens of the
camera.
You are right that there is some ambiguity in terminology when viewing a
real object *through* either a telescope or the video viewfinder, you can
argue that you are seeing the object itself, through the medium of your
telescope or the screen. But all ambiguity disappears in the case of
hallucinations. A telescope cannot construct visual hallucinations, because
it never constructs any image at all.
The phenomena of dreams and hallucinations prove that human vision is
representational, because the system *is* capable of constructing complete
visual scenes that have no objective reality beyond the brain, that is
something that only a representationalist mechanism is capable of.
...oOo...
As for articulating a non-mysterious scientifically-compatible notion of
the concept of direct perception, we can put it thus: Say you are having a
visual experience in the form of a specific spatial pattern of color and
shape in a specific portion of your visual field. You can "see" that
experience **in the sense** of having direct access to that experience,
that is, you could (if requested) report on its shape, color, and location.
That information is available to you immediately and in parallel, in the
form of a spatial image, that you could copy onto a page as a picture (if
requested).
According to direct perception, it is possible to have immediate parallel
access to the information content of a visual experience *WITHOUT* having
that information explicitly encoded anywhere in your brain.
Direct perception is having access to information about the external world
which is not explicitly encoded in the physical brain or sensory system.
...oOo...
Of course when you put it in such stark terms, there are very few who would
defend direct perception of that form. As we hear in this forum, there are
countless intermediate forms of semi-direct perception that finesse the
issue of what is represented and where. They acknowledge that some parts of
experience are represented (e.g. after-images are registered in the retina)
but insist that other aspects are perceived directly. There are as many
variants on direct perception as there are proponents of it, and very few
take the stark extreme stance of Gibson and O'Regan that denies *any*
representation whatsoever in perception.
But the point is that as long as you allow that *any* portion of your
experience is direct or unmediated by representations in your brain, you
run into the same profound paradox as the more obviously indefensible
position of Gibson and O'Regan.
This is why I raise the issue of a robotic demonstration. How could you
possibly make a robot respond to "environmental" information that is *NOT*
explicitly encoded in its computer brain? A robot might scan around with
active exploratory saccades, as demanded by Gibson, but unless the
information gathered in those saccades is stored or registered in some kind
of spatial database, the "experience" of the robot (i.e. the information to
which it has direct and immediate access) will be the same as the
experience of a visual agnosia patient, i.e. it will see a sequence of
disconnected glimpses of the world that never cohere into a globally
organized spatial structure.
Steve Lehar
Arnold Trehub
from a conflation of the expressions "direct perception" and "direct realism".
In my simple lexicon, a legitimate example of *direct perception* would be
the visual experience resulting from looking at a real object directly, as
opposed to looking at the object as it is displayed on a video screen (to use
Steve Lehar's example). On the other hand, (again in my lexicon) *direct
realism* implies that the visual experience resulting from directly looking
at a real object always directly matches, in all salient respects, the true
properties of the real object of regard. I think this is untenable.
Michael Schmitz wrote that he believed the following:
>
> a) that there are visual (auditory, tactile) experiences with
> representational contents,
> b) that these representational contents are directed at non-mental objects
> different from the experience itself (this might be my gloss on "external";
> note that this need not mean external to the perceiver's body) and
> c) that these representational contents are at least sometimes, and in fact
> normally, satisfied.
>
Then Andrew Brook wrote:
>
> I agree with everything that Michael said in his aptly-titled message ...
>
So it appears to me that while Andrew endorses *direct perception* he does
not *really* endorse *direct realism*. Is this an argument over the
definition of *direct realism*?
Arnold Trehub
==========================================================================
Alfredo Pereira Jr
> The phenomena of dreams and hallucinations prove that human vision is
> representational, because the system *is* capable of constructing compl
ete
> visual scenes that have no objective reality beyond the brain, that is
> something that only a representationalist mechanism is capable of.
APJ - What do dreams and hallucinations represent? Your argument is good,
but works for constructivism, not for representationalism.
> ...This is why I raise the issue of a robotic demonstration. How could
you
> possibly make a robot respond to "environmental" information that is *N
OT*
> explicitly encoded in its computer brain? A robot might scan around wit
h
> active exploratory saccades, as demanded by Gibson, but unless the
> information gathered in those saccades is stored or registered in some
kind
> of spatial database, the "experience" of the robot (i.e. the informatio
n
to
> which it has direct and immediate access) will be the same as the
> experience of a visual agnosia patient, ...
APJ - I agree that:
a) once a cognitive system A is conscious of a pattern B, then it has an
"internal
encoding" (not necessarily a representation) of B; and
b) in mammals most of the encoding is made in the brain.
However, the brain is an open, interactive system, therefore I don“t ag
ree
that:
c) the conscious content is determined by the brain.
E.g., if a cat appears in my visual field it is not the case that it is t
he
result of
a theater played inside the brain, having as actors a cat-representation
and
a myself-representation. The most plausible interpretation is that the br
ain
interacted physically with an external being, and based on the being“s
sensory
patterns the brain constructed an episode where it organized the availabl
e
information according to pragmatic constraints.
I understand that *all* forms of Realism as well as Idealism are
mistaken, since they try to explain cognition as a result of only one sys
tem
(brain/mind or external reality), while we know from contemporary physics
and biology that the brain (the support of the mind) is an open, adaptive
,
plastic, interactive system.
Any convincing explanation of cognition must consider that from the start
there is an interaction of the cognitive agent with the world, and that s
uch
an interaction changes both the agent and the world. Cognition is a
construction that occurs in the context of such an interactive process.
In the history of philosophy, it is well known that Husserl (at the end
of his intellectual trajectory), Heidegger and Merleau-Ponty reached a
similar conclusion (following a different path; Husserl based his reanoni
ng
on the crucial role played by intersubjective interactions).
Best Regards,
Alfredo Pereira Jr.
Andrew Brook
I think my confusion about the terminology at issue in this discussion
stems from a conflation of the expressions "direct perception" and
"direct realism". In my simple lexicon, a legitimate example of *direct
perception* would be the visual experience resulting from looking at a
real object directly, as opposed to looking at the object as it is
displayed on a video screen (to use Steve Lehar's example). On the other
hand, (again in my lexicon) *direct realism* implies that the visual
experience resulting from directly looking at a real object always
directly matches, in all salient respects, the true properties of the
real object of regard. I think this is untenable.
I respond:
Arnold, I agree with you that one of the problems in our conversation is
great unclarity about what different people mean by terms. That said, I
think that your 'direct realism' is a view that nobody has ever held --
certainly that nobody ever should hold. What I mean by 'direct realism'
and 'direct perception' is nothing more than that when I perceive, it is
objects in the world I perceive, not some intermediary from which I
infer the existence of objects. I am seeing the screen in front of me
right not, I am not aware of the screen by inference from something in
my head which is what I am really immediately conscious of. Just that,
nothing more. Then the interesting question is: what machinery allows me
to directly perceive and be aware of things in the world around me?
Note: my view does not commit me to any particular view about *how much'
of the objects around me I perceive, or in what way. My own view is that
I perceive a few middle-sized properties of these objects: shape,
colour, contours, sometimes function and so on and so forth, and do not
perceive, directly or otherwise, a great many others: chemical
properties, quantum properties, gravitational properties, and so on and
so forth.
Harwood Fisher
>APJ - I agree that:
a) once a cognitive system A is conscious of a pattern B, then it has an
"internal encoding" (not necessarily a representation) of B; <
Hi Alfredo,
The idea that a construction process is different from (and in some way,
mutually exclusive of) representations
would appear a route to some concept of direct perception. However, the
concepts of 'coding' and 'representation' would have to be de-coupled.
The problem seems to me to be the different coding capabilities and the
different 'codes' for different subsystems. Thus, is some part of the
transmission of information not via neural signaling?
Even assume that a percept can be a complex of different sorts of codes,
and that these are 'constructed' right at the point of initial reception.
I would think that the interaction of the different code systems would call
for some retrieval before we can call the experience a percept. For
instance, in the gestalt experiments, the size of objects is determined by
extrinsic meanings. These meanings have all sorts of 'encoding.' No? To
have a strictly constructionist view, these meanings would have to be
immediately present in order to interact with various codes that determine
the 'direct percept' of an object or form. But that state of affairs would
call for immediate availability of relevant meanings in each case of a
percept--and each case would be drawn from a great array of different
objects.
The alternative of course is a lightning fast transmission from the point
of perception to some central station at which there is a construction. In
this view, the transmission bothers me, because whatever is transmitted
would appear to be a transformation of a code or if you want an 'encoding.'
But I'll come to the transformation issue in a moment.
Could you comment more on the difference between an 'internal encoding' and
a 'representation'? If there is more than one coding system, then I can't
see how the percept is merely constructed. Tectonically? Somewhere in
such a process there would seem to be a transformation of one code into
another. Even if one assumes that code A is a cause of event (a) and code
B is a cause of event (b); such that [(a)(b)] becomes the percept, the
relationship between (a)(b) vis a vis the relationship of code A and code B
is an example of a transformation. So, in a word, I can't see how
'representations' can be out of the picture.
Harwood Fisher
Steven Lehar
The phenomena of dreams and hallucinations prove that human vision is
representational, because the system *is* capable of constructing complet
e
visual scenes
Pereira >
What do dreams and hallucinations represent? Your argument is good,
but works for constructivism, not for representationalism.
< Pereira
A hallucinated cat represents a cat. A hallucinated colored surface
represents a colored surface. We have yet to discover what it is that is
doing the representation, whether synchronous oscillations or neural
spiking frequencies, but as to what is represented appears to us directly
,
even if only as a color or shape.
Pereira >
However, the brain is an open, interactive system, therefore I don´t
agree
that: c) the conscious content is determined by the brain.
E.g., if a cat appears in my visual field it is not the case that it is t
he
result of a theater played inside the brain, having as actors a cat-
representation and a myself-representation.
< Pereira
But then what *is* the conscious content determined by if not by the brai
n?
In the case of veridical perception you can say it is determined by the
real cat out there. But in the case of a hallucinated cat, there's got to
be *something* that defines the shape and color of that cat, what else ca
n
it be but something in the brain? And whatever that thing is in the brain
that makes you think you are seeing a cat, is obviously representational,
because you obviously don't have a *real* cat in your brain to represent
itself.
Pereira >
The most plausible interpretation is that the brain interacted physically
with an external being, and based on the being´s sensory patterns th
e brain
constructed an episode where it organized the available information
according to pragmatic constraints.
< Pereira
Plausible enough for a real cat. But what about the hallucinated cat? See
n
during a trance during which the body is immobile? What is the *ontology*
of the hallucinated cat? What is its substance? What is the medium that
exhibits the shape and color of our experience? If these are not the
subjective manifestation of states of our brain, what else could they
possibly be?
I think this debate is going round and round in circles. See my new threa
d
on Theories v.s. Paradigms.
Steve
Neil W Rickert
to get to some of the issues being discussed with respect to direct
or representational theories of perception. I am discussing DSL to
illustrate the practical issues involved with picking up
information.
In a couple of earlier message, I have attempted to explain Gibson
and clarify what I thought was a misunderstanding of his position. I
want to be clear that in this current message I make no pretense that
I explaining Gibson. I am presenting my own views (my current
views), and it would not surprise me if Gibsonians disagreed.
Here is an outline of my home network:
The phone line is connected to a DSL modem.
The output of the DSL modem is connected to an ethernet cable,
which in turn connects to a home router.
My computer is connected to the router via another ethernet cable.
We might say that information comes in on the phone line, is
converted to a different form by the modem for sending to the router
via the ethernet cable, and then resent from the router to the computer.
We could perhaps say that the ethernet signals, the contents of the
router memory buffer, the signals on the next ethernet, and the
contents of the memory buffer in the computer are all
representations.
As it happens, I live some distance from the telephone company
Central Office (or CO) through which I am connected to a DSL
service. Because of that distance, there is a 56db attenuation on
the telephone line. That is, the signal arriving on the phone
company is 1/400000 of the signal that was sent at the CO. Most of
the energy has been absorbed en route. Worse still, various sources
of noise have polluted the incoming signal.
The critical step, in receiving the DSL signal, is at the modem. For that
is where very weak signals have to be detected, and useful information
obtained from them.
Because of the weakness of the signal, I have already run a direct wire
from my modem, to where the phone line enters the house. This gets me
a better signal quality entering the modem. I have seriously considered
moving the modem to the basement, where it will be closer yet to the
incoming phone line.
By contrast, the ethernet cables and router don't matter much. The
signals on those components are strong, and won't be overwhelmed by
noise even if I use longer ethernet cables, and even if I run those
cables through an electrically noisy area.
Where should we identify the important step here? I say that the
important step is at the modem. The other steps are mainly
implementation details, copying representations into the buffers where
it is convenient to use them. But the modem is the step that deals
with the external signal and creates a usable representation in the
first place.
Let's revisit perception. I agree with Steven, that we use
representations. But the question is in where perception occurs. I
say that the critical step is where the sensory cells detect the
information coming from the external world. I say that we should
identify the act of perception with the initial generation of useful
representations. That makes perception direct. There may well be
indirect stages involved in the later processing of these
representations. But perception is what happens at the initial stage
where the valuable information is separated out from the sea of noise
in which it is embedded.
Representationalism appears to be based on the assumption that
forming representations is easy, and that the hard work goes into
analysing the representation. By contrast, I am saying that creating
*useful* representations is difficult. But, if we have done a good
job of creating useful representations, then much of the analysis has
already been done in the initial generation of those
representations.
-NWR
Michael Schmitz
MS previously
>It seems to me that it is rather you who simply assumes, apparently without
>argument, that some physical objects are really mental objects. We agree
>that naive or direct realism (terminologically I prefer "direct", for
>reasons that I will soon post about) is implict in natural language,
CR
Actually I don't think direct realism is implicit in natural language. What
is implicit in natural language is naive realism, which is not a coherent
philosophical position to begin with.
And nowhere did I say that physical objects were really mental objects. I
said that the subjective experience which makes up our perception of
physical objects can itself be perceived (or "directly attended to", if you
don't like this use of the word "perceived") as mental objects.
MS
Well, you wrote:
"Cathy:
I agree that it sounds odd to talk about seeing mental objects, but that's
because natural language simply doesn't distinguish regularly between mental
and physical objects. It simply assumes (or rather, the conceptual
framework which underlies natural language assumes) that some mental objects
are physical objects, and the result of this is the naive realism implicit
in natural language."
So that's why I thought that you said that common sense ('natural language')
mistakes *some* (I never said all) mental objects for physical ones. If
that's not what you are saying, I have to say I really don't know what you
are saying. Also, I have to complain about the fact that you always talk
about naive realism as implicit in natural language and as distinct from
direct realism, but you never say what naive realism actually amounts to. I
already asked you this two posts back (16/04):
[CR previously]
> Naive realism, as I understand it, asserts (or assumes) that the
phenomenal
> properties of consciousness are themselves properties of the external
world,
[MS previously]
I'm not sure what you mean by that. Do you mean that naive realism is the
"non-representational" view? (View 1) in Elizabeth Schier's taxonomy). So
defined, I reject it, too. Or do you mean that the naive realist also needs
[to be]
a naive realist about colors, sounds, and other so-called manifest
properties? Or what?
but you did not answer my question.
CR
But my point was that we can become aware of our subjective experiences in a
way which is at least as direct as the way we believe we become aware of
physical objects. If you don't want to call that direct awareness "seeing"
or "perceiving", that's ok with me, but I think we need a word to describe
this sort of direct awareness of subjective experience. It is not just " an
awareness in thought", if by that you mean an indirect awareness.
MS
No, I don't mean to suggest that this awareness is 'indirect'. I think there
needs to be a lot of intellectual background in order for this awareness to
be possible, but I don't think that makes it 'indirect'.
CR
I absolutely agree that we need to go beyond common sense. But you can
hardly appeal to common sense to support your notion of "seeing", since you
admit you are going beyond common sense.
MS
I think I can appeal to common sense in going beyond common sense. It is an
argument for my proposal that it can preserve virtually all or even all
things we usually want to say, such as that what we see are the familiar
objects we usually take ourselves to be seeing, but that sometimes certain
things only seem to be case, that we have hallucinations etc. I go beyond
common sense in being more precise at certain points, by introducing
technical terminology to explicitly mark certain distinctions etc., but I do
not contradict common sense, and so I can appeal to common sense when
arguing against a different proposal that does.
MS previously
>No, I think you misunderstand me. So far I have not really said anything
>about epistemology and about how we *judge* whether some experience is
>veridical or not. I was talking about what *constitutes* veridicality and
>this is, trivially enough, the presence of the corresponding state of
>affairs. And from my point of view the distinction you say I get the wrong
>way round does not really make sense. To ascribe veridicality is *ipso
>facto* to ascribe the presence of a corresponding object.
CR
The keyword here is "ascribe". The point is, it is irrelevant to this
process whether some external object is really present or not. The
awareness we have when we perceive something is not dependent on the
presence of some external object.
MS
Do you deny that ascribing veridicality ipso facto *means* ascribing the
presence of a corresponding object? I thought you said that you don't want
to deny DR, but only naive realism. But here you clearly deny it.
CR
What Descartes said was that we can never prove the existence of the
external world. He was right. We don't prove it, we just assume it -- as
you have done here.
MS
No, I don't merely assume, say, that this is a keyboard on which I'm typing.
I can *see* it, I can *feel* it, and so I *know* that this is a keyboard and
not part of me and thus external. And so I know that there is an external
world. It's not a question of faith or assumption. I don't think it matters
much whether this is a proof of the external world. In some sense, it may be
considered a (Moore-style) proof (see G.E. Moore "Proof of the External
World"), in some sense perhaps not. But either way, no more proof is needed,
just a critique of Cartesianism.
MS previously
>3) It is not the case that we generally need reasons to trust our
perceptual
>experiences. Only in very special situations do we need such reasons - for
>example, if we know that in certain contexts we are usually deceived.
>Normally *we need reasons to doubt* the deliverances of our senses.
CR
I agree again.
MS
It's good that you agree for a change, and even again :-). On the other
hand, even though it is reasonable that you agree with these points, I think
in the end it shows how idle this whole Cartesian subjectivism is. Because,
as you say:
CR
Actually the whole point of "brain in a vat" stories is that we don't need
to distrust our perceptual experiences -- they work irrespective of whether
the external world exists or not.
MS
but that just shows how pointless the whole exercise is. You will continue
to act like everbody else, make statements in the way everbody does, treat
evidence in the way everbody does, and so on. You will presumably even
distinguish belief and knowledge in the way everbody does. Only at a certain
point you will comment all of this by saying something like:
CR
But all we can really say about that external
world is that it takes the form of some agency which orchestrates our
conscious perceptions.
MS
It seems to me such a statement is, to use one of Wittgenstein's favorite
metaphors, a wheel that does not turn, does no work and is only attached to
the cognitive machinery for purely decorative purposes.
Michael
Alex Green
Best Wishes
Alex Green
-NWR
Send instant messages to your online friends http://uk.messenger.yahoo.com
Cathy Reason
formatting :-)
Michael Schmitz wrote:
Well, you wrote:
"Cathy:
I agree that it sounds odd to talk about seeing mental objects, but that's
because natural language simply doesn't distinguish regularly between mental
and physical objects. It simply assumes (or rather, the conceptual
framework which underlies natural language assumes) that some mental objects
are physical objects, and the result of this is the naive realism implicit
in natural language."
So that's why I thought that you said that common sense ('natural language')
mistakes *some* (I never said all) mental objects for physical ones. If
that's not what you are saying, I have to say I really don't know what you
are saying.
Cathy again:
Well, that is different from your earlier paraphrase of my remarks, which
was:
"It seems to me that it is rather you who simply assumes, apparently without
argument, that some physical objects are really mental objects"
What I am saying is that naive realism assumes the conscious representation
of the world to be, in fact, the external world. As to whether this is a
mistake or not, that really depends on what you expect our perceptual
systems to do. If you want them to answer ontological questions, then it's
clearly a mistake. But if you want them to perceive the world, then it's
probably a good thing to do.
MS:
Also, I have to complain about the fact that you always talk
about naive realism as implicit in natural language and as distinct from
direct realism, but you never say what naive realism actually amounts to. I
already asked you this two posts back (16/04):
[MS previously]
I'm not sure what you mean by that. Do you mean that naive realism is the
"non-representational" view? (View 1) in Elizabeth Schier's taxonomy). So
defined, I reject it, too. Or do you mean that the naive realist also needs
[to be]
a naive realist about colors, sounds, and other so-called manifest
properties? Or what?
but you did not answer my question.
Cathy again:
Sorry, but I thought I'd already defined what I meant by naive realism. If
you want a different definition, how about the one I gave above?
As to whether this is the same as Elizabeth Schier's View #1 -- no, I don't
think it is. Mainly because View #1 appears to me not to distinguish
between naive realism and certain forms of direct perception (those which
relate exclusively to objective perception and don't attempt to deal with
the subjective aspects).
<snip>
MS
I think I can appeal to common sense in going beyond common sense. It is an
argument for my proposal that it can preserve virtually all or even all
things we usually want to say, such as that what we see are the familiar
objects we usually take ourselves to be seeing, but that sometimes certain
things only seem to be case, that we have hallucinations etc. I go beyond
common sense in being more precise at certain points, by introducing
technical terminology to explicitly mark certain distinctions etc., but I do
not contradict common sense, and so I can appeal to common sense when
arguing against a different proposal that does.
Cathy again:
I'm far less certain than you that it is possible to "go beyond common
sense" without contradicting it in some way. But notwithstanding that, I
don't accept the principle that common sense should always be assumed a
priori to be correct.
CR
The keyword here is "ascribe". The point is, it is irrelevant to this
process whether some external object is really present or not. The
awareness we have when we perceive something is not dependent on the
presence of some external object.
MS
Do you deny that ascribing veridicality ipso facto *means* ascribing the
presence of a corresponding object? I thought you said that you don't want
to deny DR, but only naive realism. But here you clearly deny it.
Cathy again:
Could you say what it is about that statement that "clearly denies " direct
realism?
MS
No, I don't merely assume, say, that this is a keyboard on which I'm typing.
I can *see* it, I can *feel* it, and so I *know* that this is a keyboard and
not part of me and thus external. And so I know that there is an external
world. It's not a question of faith or assumption. I don't think it matters
much whether this is a proof of the external world. In some sense, it may be
considered a (Moore-style) proof (see G.E. Moore "Proof of the External
World"), in some sense perhaps not. But either way, no more proof is needed,
just a critique of Cartesianism.
Cathy again:
Ok, if you want to put it like that. But my point is still that your
"seeing" and "feeling" and "knowing" that you are typing on a keyboard is
not dependent on some external object actually being present. If you want
to assert in addition that some external object is present, then you have to
make an assumption.
CR
Actually the whole point of "brain in a vat" stories is that we don't need
to distrust our perceptual experiences -- they work irrespective of whether
the external world exists or not.
MS
but that just shows how pointless the whole exercise is. You will continue
to act like everbody else, make statements in the way everbody does, treat
evidence in the way everbody does, and so on.
Cathy again:
That indeed is exactly my point. The correct working of our perceptual
systems is largely independent of these sorts of ontological questions.
MS:
You will presumably even
distinguish belief and knowledge in the way everbody does. Only at a certain
point you will comment all of this by saying something like:
CR
But all we can really say about that external
world is that it takes the form of some agency which orchestrates our
conscious perceptions.
MS
It seems to me such a statement is, to use one of Wittgenstein's favorite
metaphors, a wheel that does not turn, does no work and is only attached to
the cognitive machinery for purely decorative purposes.
Cathy again:
Well, Wittgenstein is always good for a bit of philosophical cabaret. But
your response here is singularly lacking in any kind of argument. So I'm
not even sure what you're trying to say here, but let's see if I can tease
it out.
Do you mean that my view is so self-evident as to be trivial? I think an
analysis of the metaphysics implicit on modern physicalism would contradict
that. Or do you mean that my view is incorrect in some way? In that case I
think you need to say in what way it's incorrect.
Or do you just mean the question is not worth discussing? In that case I'd
say it's worth discussing to the extent that we are actually discussing it.
(Although I do think it has some implications for the issue of causal
completeness.)
Cathy
[Catherine Reason]
Arnold Trehub
>
> What I mean by 'direct realism' and 'direct perception' is nothing more
> than that when I perceive, it is objects in the world I perceive, not
> some intermediary from which I infer the existence of objects.
>
I fully agree that our phenomenal experience *as such* is not an inference
from something else.
>
> I am seeing the screen in front of me right now, I am not aware of the
> [computer] screen by inference from something in my head which is
> what I am really immediately conscious of. Just that, nothing more.
>
It seems to me that what is described here is a perspectival experience;
i.e., an immediate phenomenal experience of the computer screen in a
geometric relationship (some distance in front of) to the *self* in an
egocentric space. Andrew, can we agree, for the purposes of this discussion,
that "direct realism", "direct perception", and "perspectival experience"
are equivalent terms?
>
> Then the interesting question is: what machinery allows me
> to directly perceive and be aware of things in the world around me?
>
Fine! Now we cut to the chase. My claim is that the machinery you are
looking for is the *retinoid system*. This is a putative system of neuronal
brain mechanisms which represent 3D egocentric space and the layout of
objects within this space in perspectival relationships to the *self locus*,
the fixed perspectival origin (home position) of the brain's phenomenal "I".
The neuronal structure and dynamics of the retinoid system are detailed
and discussed in my book *The Cognitive Brain* (MIT Press, 1991).
A large body of empirical evidence in support of the retinoid model is
also presented in the book.
Arnold Trehub
==========================================================================
Bill Adams
The difference between viewing the world via a telescope and via a video
monitor is only about the relative difficulty of exploration in each case.
Gibson often said that the single most difficult problem for any theory of
perception is to explain error. His explanation was extremely simple. If
an animal is restricted in its ability to explore its environment to the
fullest extent of which it is capable, error can result. Conversely, since
animals alive today have adapted successfully to their environments, given
enough time and freedom to explore, they always will arrive at veridical
perception.
A smudge on a telescope lens is easily dismissed because it is easy to look
a the lens, see a smudge, and wipe it off. Lens flare in a telescopic view
is less obvious to discount, unless you know something about refractive
aberration. That's a matter of exploring your telescope thoroughly and
completely. Static noise on a video monitor can likewise be ignored if you
understand (have thoroughly explored) your equipment.
In a dream, you may have non-veridical, quasi-perceptual experience, but
that's because you are unable to adequately explore. You're in bed; asleep;
it's dark; your muscles are inoperable, etc. You are restricted from free
exploration of your environment so it is not surprising that error results.
If you take a fully awake person and put a bag over their head, they also
will fail to perceive the environment correctly. Most cognitive psychology
experiments create this problem. They artificially restrict the person's
ability to explore the environment, then act surprised to get unusual
results.
But I admit that Gibson's theory breaks down when it comes to the physical
body that actually performs the exploration. Gibson just assumed the
existence of an intact, exploratory, physically embodied animal. In his
1966 book, "The Senses Considered As Perceptual Systems," he argued for a
holistic, Gestalty view of perception based on the coordination of the
senses, but not constructed from the senses. He was adamantly not an
associationist, and often spoke of "Lockean poisoning" and "Kantian
poisoning", in reference to how associationist and constructionist ideas of
perception continue to distort our thinking.
Conveniently however, that meant he could talk about the whole body as "the
perceptual organ," and he did not have to deal with anything like energy
transduction or brain representations. Inevitably, that meant that
Gibsonian direct perception, which begins with the indisputable fact of
perceptual experience, does not make explanatory contact with the physical
world. I believe this has been the source of Steve's frustration with it and
the reason why he continues to call for a robotic demonstration. Of course
there can be no robotic demonstration of direct perception, since direct
perception cannot explain how holistic perceptual experience is instantiated
in the common sense, realist, physical world.
Therefore, as a proponent of "stark" Gibsonian, non-representational direct
perception, I must temporarily concede to the persistent arguments of Steve
Lehar. Common sense realism, especially as it presupposes the physicality of
the exploratory body, cannot be denied or ignored forever, as Gibsonian
direct perception attempts to do.
Representationalism enjoys the advantage of the presupposition of common
sense realism, and from that, attempts to construct or derive perceptual
experience. At best, it can only hope to construct a functional brain/body
system that would be devoid of experience (because experience is not
physical). But if representationalism is headed to a dead-end, Gibsonian
direct perception stands at the brink of a visual cliff, unable to take the
next step into common sense realism.
Stalemate then? For now. What needs to happen, it seems to me, is for
phenomenologists to explain how the physical body can be understood as a
mental object, a projection of the mind, rather than as a biological,
physical, material object.
There is some basis for undertaking an inquiry into the possibility that the
body is a mental projection, such as the work of Merleau-Ponty, and more
recently, Judith Butler, and demonstrations that one's conceptualization of
one's body is amazingly elastic (e.g., Ramachandanran & Hirstein, 1997). I
cannot possibly explain in a posting how it might be plausible to say that
the mind could project a real physical object, one that would pass every
common sense and scientific test of physical realism. I attempt to spell
out the causal sequence in a forthcoming book.
That idea is "absolutely incredible" for the presupppositions of common
sense realism, but as good detectives know, when everything reasonable has
been eliminated, we are left with the incredible.
I have one closing comment: I don't agree that this discussion has only
gone in circles. I have learned a great deal about the representationalist
point of view, and also about my own assumptions. I want to thank all the
participants, but especially Steve Lehar for that.
Bill Adams
References
Adams, W.A. (forthcoming, 2005). What Does It All Mean? A Humanistic
Account of Human Experience. Exeter, U.K.: Imprint Academic.
Butler, J. (1993). Bodies that Matter: On the Discursive Limits of 'Sex'.
NY: Routledge.
Ramachandandran, V.S., and Hirstein, W. (1997). Three laws of qualia: what
neurology tells us about the biological functions of consciousness. Journal
of Consciousness Studies, 4, 5/6, 429-57.
bill....@bainbridge.net
http://members.bainbridge.net/~bill.adams
Alfredo Pereira Jr
> The idea that a construction process is different from (and in some way
,
> mutually exclusive of) representations
> would appear a route to some concept of direct perception. However, th
e
> concepts of 'coding' and 'representation' would have to be de-coupled.
APJ - Sure.
> The problem seems to me to be the different coding capabilities and the
> different 'codes' for different subsystems. Thus, is some part of the
> transmission of information not via neural signaling?
APJ - There are different kinds of neural signaling with different codes.
Spikes ("nerve pulses") have approsimately the same amplitude, and so
the information from sensors to brain is (assumed to be) encoded in the
frequency of spiking and maybe in the timing (variation of time intervals
between spikes). But when these pulses reach a synapse, a richer code
is mobilized. There are dozens of transmitters and membrane receptors,
and also voltage-gated ion channels to transmit the message to the
post-synaptic
neuron. Each dendrite of the post-synaptic neuron also has synapses with
several other neurons. In a few words, what happens is that the
incoming signal is received by a system with a code that is more complex
(it has more components and consequently more combinatorial possibilities
)
than the system that transmited it from the peripheral sensor. Therefore,
we are enabled to say that neurons *intepret* the incoming signal, instea
d
of representing the message that the signal carries.
> Even assume that a percept can be a complex of different sorts of codes
,
> and that these are 'constructed' right at the point of initial receptio
n.
APJ - They are constructed from such a complex, but the codes are integra
ted
at the end of the process by means of a super-code, in the context of a
natural (n-valued) quantum computing.
As far as I can see there is no integration at the points of initial
reception.
> I would think that the interaction of the different code systems would
call
> for some retrieval before we can call the experience a percept. For
> instance, in the gestalt experiments, the size of objects is determined
by
> extrinsic meanings. These meanings have all sorts of 'encoding.' No?
APJ - Yes!
>To have a strictly constructionist view, these meanings would have to be
> immediately present in order to interact with various codes that determ
ine
> the 'direct percept' of an object or form. But that state of affairs
would
> call for immediate availability of relevant meanings in each case of a
> percept--and each case would be drawn from a great array of different
> objects.
APJ - Therefore constructionism is not compatible with "directionisms"...
But it is not indirect either. Both 'direct' and 'indirect' ideas are
compromised
with representationalism and/or realism. A construction is neither direct
nor
indirect!
> The alternative of course is a lightning fast transmission from the poi
nt
> of perception to some central station at which there is a construction.
In
> this view, the transmission bothers me, because whatever is transmitted
> would appear to be a transformation of a code or if you want an
'encoding.'
> But I'll come to the transformation issue in a moment.
APJ - Yes, there is a transformation and then there is no "directness".
> Could you comment more on the difference between an 'internal encoding'
and
> a 'representation'?
APJ - Yes, thanks for asking. This is a central issue. Representation is
a
kind of
encoding that preserves a linear mapping between what is represented and
the
final product (this is approximately Allan Newell´s concept of
representation).
But there are other kinds encoding operations that don´t preserve such
a
linear
mapping. This is approximately the concept of "homeomorphism" (a good
discussion
of this notion was recently made by Juergen Fell, in the paper "Identifyi
ng
Neural
Correlates of Consciousness: the State Spece Approach", published at
Consciousness and Cognition 13, 2004, p. 709-729).
In the brain, it is well known that only the encodings at the primary
sensory cortex preserve a linear mapping (and only for some aspects of th
e
stimulus).
> If there is more than one coding system, then I can't
> see how the percept is merely constructed. Tectonically?
APJ - As I said before I believe something like quantum computing is
responsible for the magic. Of course this is highly speculative at this
moment.
I also believe that a coherent system with some degree of entanglement ha
s
aspects of self-similarity (a fractal geometry) and then quantum informat
ion
reflects
macroscopic content (e.g. the general state of the body and environment).
Like Leibiniz´ monads (that reflect the whole universe), each part of t
he
coherent
system has information about the other parts of the system.
> Somewhere in
> such a process there would seem to be a transformation of one code into
> another. Even if one assumes that code A is a cause of event (a) and c
ode
> B is a cause of event (b); such that [(a)(b)] becomes the percept, the
> relationship between (a)(b) vis a vis the relationship of code A and co
de
B
> is an example of a transformation. So, in a word, I can't see how
> 'representations' can be out of the picture.
APJ - I argued that encodings are not necessarily representations, but I
don´t
deny that in some cases (the exception, not the rule) they are
representations.
Andrew Brook
What I am saying is that naive realism assumes the conscious
representation of the world to be, in fact, the external world.
Me:
Not this naive realist. What I say is that conscious representation is
*of* the external world, specifically, makes me aware of the external
world. I don't believe and certainly don't assume that conscious
representation *is* the external world. In fact, I don't know what Cathy
means by this statement. How could I believe that a representation,
something in my head, is something external to me?
Cathy, do you mean that I assume that what is really consciousness of my
own representations is consciousness of the external world? If that is
what you mean, I don't assume that, either. I deny that consciousness of
the world is really consciousness of states in my brain. Now, I could be
wrong about this -- that is what I'd like those who reject direct
realism to try and show, so that I'd know what the arguments are. But
those who reject direct realism do not present argument or evidence for
their view, they just assume that they have to be right and the opposite
point of view has to be wrong. Sorry, maybe we are wrong, but it is
certainly not obvious that we are.
Cathy has said what I just quoted a number of times and I have made the
same response a number of times. If she doesn't mean what I
reconstructed in the previous para., I don't understand what she means
-- and probably she would say the same about me.
Michael Schmitz
Well, at this point the most challenging aspect of this discussion is the
formatting :-)
M
Sorry, I'm not sure I can do better than I have, but I will try again - and
also keep working on the formatting -;
M prev
So that's why I thought that you said that common sense ('natural language')
mistakes *some* (I never said all) mental objects for physical ones. If
that's not what you are saying, I have to say I really don't know what you
are saying.
Cathy again:
Well, that is different from your earlier paraphrase of my remarks, which
was:
"It seems to me that it is rather you who simply assumes, apparently without
argument, that some physical objects are really mental objects"
M
But that's the same state of affairs described from the point of view of
common sense (DR) which I support. Since you say that common sense mistakes
mental objects for physical ones, I can say, from the point of view of
common sense (DR), that you claim that some physical objects are really
mental objects, right?
C
What I am saying is that naive realism assumes the conscious representation
of the world to be, in fact, the external world. [snip]
M
That's supposed to be something that ordinary folks believe? No, I don't
think that's correct. Very small children do not have a concept of seeing
and consciousness at all, such that there is a sense in which they only know
the 'external' world, but that's not the same as identifying conscious
representation and external world (Failure to differentiate is not the same
as identification.)
Around 3, 4 years of age, I think, children are beginning to develop a
theory of mind including a concept of seeing. (In fact, a concept of seeing
is acquired prior to a full-blown theory of mind.) Certainly adults do
distinguish the state of seeing from its objects, have the concept of
illusion, hallucination, of seeming etc. Of course, most people have not
thought much about the questions that this raises, and nobody has thought of
all of them, but so what? Do you seriously suppose that adults are not able
to distinguish episodes of seeing from their objects?
Cathy again:
I don't accept the principle that common sense should always be assumed a
priori to be correct.
M
Neither do I. Does anybody? But other things being equal, it is still better
if common sense is respected as much as possible.
CR prev
The keyword here is "ascribe". The point is, it is irrelevant to this
process whether some external object is really present or not. The
awareness we have when we perceive something is not dependent on the
presence of some external object.
MS prev
Do you deny that ascribing veridicality ipso facto *means* ascribing the
presence of a corresponding object? I thought you said that you don't want
to deny DR, but only naive realism. But here you clearly deny it.
Cathy again:
Could you say what it is about that statement that "clearly denies " direct
realism?
M
DR says that the veridicality of a perception consists in the presence of a
corresponding object, and that at least some perceptions are veridical. So
how can you say that the presence of this object is irrelevant to the
ascription of veridicality? Well, if try to answer this question myself, I
now suppose what you mean is that we cannot ever *really know* whether a
perception is veridical, right? But that thesis, I believe, you would better
express by saying that we cannot *really ascribe* veridicality at all.
M prev
No, I don't merely assume, say, that this is a keyboard on which I'm typing.
I can *see* it, I can *feel* it, and so I *know* that this is a keyboard and
not part of me and thus external. And so I know that there is an external
world. It's not a question of faith or assumption. I don't think it matters
much whether this is a proof of the external world. In some sense, it may be
considered a (Moore-style) proof (see G.E. Moore "Proof of the External
World"), in some sense perhaps not. But either way, no more proof is needed,
just a critique of Cartesianism.
Cathy again:
Ok, if you want to put it like that. But my point is still that your
"seeing" and "feeling" and "knowing" that you are typing on a keyboard is
not dependent on some external object actually being present. If you want
to assert in addition that some external object is present, then you have to
make an assumption.
M
I take it that you don't want to deny that in the ordinary sense of
"seeing", "feeling" and "knowing" I see and feel my keyboard and know that I
am typing on it. For example, if we were not discussing epistemology, but
would be talking about this in some everyday context, I assume you would be
ready to accept these things - if you did not have some special reason to
think I was hallucinating, lying, trying to deceive or some such. However,
you think you can accept this in the ordinary sort of way, and still claim
that there is deep epistemological question that is yet to be answered -
even though it then turns out it cannot possibly be answered. Let me put
this by saying that while you accept that I see and feel the keyboard and
know that I am typing on it, you think I still do not *really see* and
*really feel* that *external object* and do not *really know* about the
presence of this *external object.* Now my point is best put by saying that
your notions of *really feeling*, *really knowing* and *external object*,
according to which we can never *really know*, *really see* and *really
feel* the presence of *external objects*, are useless and meaningless
because, as you say yourself, they are just totally irrelevant to the sort
of questions we normally ask, like whether I know whether what I am feeling
and seeing here is a keyboard. I'm not saying I can answer your further
question, I'm saying there really is no such question, because a question
that could not possibly be answered is not really a proper question at all.
And this is also the point of the Wittgensteinian parable from my last post.
I think our previous exchange confirms this interpretation of your position:
C prev
Actually the whole point of "brain in a vat" stories is that we don't need
to distrust our perceptual experiences -- they work irrespective of whether
the external world exists or not.
M prev
but that just shows how pointless the whole exercise is. You will continue
to act like everbody else, make statements in the way everbody does, treat
evidence in the way everbody does, and so on.
Cathy again:
That indeed is exactly my point. The correct working of our perceptual
systems is largely independent of these sorts of ontological questions.
M
Largely? But not entirely? In which sense not entirely?
C prev
But all we can really say about that external
world is that it takes the form of some agency which orchestrates our
conscious perceptions.
M prev
It seems to me such a statement is, to use one of Wittgenstein's favorite
metaphors, a wheel that does not turn, does no work and is only attached to
the cognitive machinery for purely decorative purposes.
Cathy again:
Well, Wittgenstein is always good for a bit of philosophical cabaret.
M
If Wittgenstein is good for cabaret, what is Descartes good for? Ghost
trains perhaps?
C
But your response here is singularly lacking in any kind of argument. So
I'm
not even sure what you're trying to say here, but let's see if I can tease
it out.
M
OK. Let's see if you can be my Socratic midwife -;
C
Do you mean that my view is so self-evident as to be trivial?
M
Not quite.
C
[snip]
Or do you mean that my view is incorrect in some way?
M
That's getting closer, though I might prefer to say it does not even get to
the point where it is a genuine *view* of anything which might be incorrect.
It's more that it is incorrect in being a pseudo view or pseudo theory which
does not even qualify for other forms of incorrectness such as falsity.
C
Or do you just mean the question is not worth discussing? In that case I'd
say it's worth discussing to the extent that we are actually discussing it.
M
That's yet closer, but I think the real trouble is that it is impossible to
discuss your question. We can perhaps, for a brief stretch, discuss whether
we can discuss it, but we cannot discuss itself. And the reason is that you
have defined the 'issue' of whether we *really perceive* anything in such a
way that no possible evidence could be relevant to it.
Michael
Cathy Reason
But that's the same state of affairs described from the point of view of
common sense (DR) which I support. Since you say that common sense mistakes
mental objects for physical ones, I can say, from the point of view of
common sense (DR), that you claim that some physical objects are really
mental objects, right?
Cathy:
Well I'd understood that you weren't saying that representations are
physical objects, just that they are the means by which one perceives
physical objects. So I'm not sure what these physical objects are that I'm
supposed to be claiming are mental objects.
C
What I am saying is that naive realism assumes the conscious representation
of the world to be, in fact, the external world. [snip]
M
That's supposed to be something that ordinary folks believe? No, I don't
think that's correct. Very small children do not have a concept of seeing
and consciousness at all, such that there is a sense in which they only know
the 'external' world, but that's not the same as identifying conscious
representation and external world (Failure to differentiate is not the same
as identification.)
Cathy:
I agree, identification was the wrong word to use.
M
Around 3, 4 years of age, I think, children are beginning to develop a
theory of mind including a concept of seeing. (In fact, a concept of seeing
is acquired prior to a full-blown theory of mind.) Certainly adults do
distinguish the state of seeing from its objects, have the concept of
illusion, hallucination, of seeming etc. Of course, most people have not
thought much about the questions that this raises, and nobody has thought of
all of them, but so what? Do you seriously suppose that adults are not able
to distinguish episodes of seeing from their objects?
Cathy:
No, just that the conscious representation of the world is not distinguished
from the world it represents.
C
I don't accept the principle that common sense should always be assumed a
priori to be correct.
M
Neither do I. Does anybody? But other things being equal, it is still better
if common sense is respected as much as possible.
Cathy:
Well, I'm not sure if I agree with this either. Lewis Wolpert once said
that in science, the common sense explanation can be reliably regarded as
the one most likely to be wrong :-)
But I guess the real problem I have with this is that common sense
inevitably depends on a large number of asumptions, and I really don't see
that it's possible to question these assumptions without appearing to
contradict common sense.
M
DR says that the veridicality of a perception consists in the presence of a
corresponding object, and that at least some perceptions are veridical. So
how can you say that the presence of this object is irrelevant to the
ascription of veridicality? Well, if try to answer this question myself, I
now suppose what you mean is that we cannot ever *really know* whether a
perception is veridical, right? But that thesis, I believe, you would better
express by saying that we cannot *really ascribe* veridicality at all.
Cathy:
By all means consider it expressed that way if you wish. However this means
that the definition of veridicality in DR, as you describe it, depends on a
state of affairs which is completely impossible to determine. And by your
Wittgenstein criterion, we would have to assume that your notion of
veridicality was a wheel which didn't turn, decorative ornamentation
attached to the cognitive machinery, etc etc (although I'm not actually sure
what bit of cognitive machinery it would be attached to). Since your notion
of veridicality depends on the answer to a question which you admit is
completely unanswerable, in fact pointless, we would by this criterion have
to say that your notion of veridicality is itself rather pointless.
However I'm not really interested in pursuing that point, because I don't
agree with it. I think I might actually find your definition of
veridicality rather useful.
M
I take it that you don't want to deny that in the ordinary sense of
"seeing", "feeling" and "knowing" I see and feel my keyboard and know that I
am typing on it. For example, if we were not discussing epistemology, but
would be talking about this in some everyday context, I assume you would be
ready to accept these things - if you did not have some special reason to
think I was hallucinating, lying, trying to deceive or some such. However,
you think you can accept this in the ordinary sort of way, and still claim
that there is deep epistemological question that is yet to be answered -
even though it then turns out it cannot possibly be answered. Let me put
this by saying that while you accept that I see and feel the keyboard and
know that I am typing on it, you think I still do not *really see* and
*really feel* that *external object* and do not *really know* about the
presence of this *external object.* Now my point is best put by saying that
your notions of *really feeling*, *really knowing* and *external object*,
according to which we can never *really know*, *really see* and *really
feel* the presence of *external objects*, are useless and meaningless
because, as you say yourself, they are just totally irrelevant to the sort
of questions we normally ask, like whether I know whether what I am feeling
and seeing here is a keyboard.
Cathy:
I'm afraid all this stuff about "really knowing", etc, is an invention of
your own and not something I would endorse at all. And since you yourself
describe these terms as useless and meaningless, I'm really not sure why you
conjured them into being in the first place.
M
I'm not saying I can answer your further
question, I'm saying there really is no such question, because a question
that could not possibly be answered is not really a proper question at all.
Cathy:
And yet your notion of veridicality is completely dependent on the answer to
this pointless, unanswerable, non-existent question.
M
And this is also the point of the Wittgensteinian parable from my last post.
Cathy:
I would not want to lay down injunctions about how other people should use
language, but I feel entitled to point out that in the common sense use of
language, it's not actually a requirement that all questions must be
answerable. To that extent, your Wittgensteinian parable appears to
contradict common sense, which is something I take it you do not wish to do.
M
I think our previous exchange confirms this interpretation of your position:
<previous exchange snipped>
Cathy:
You're correct of course, but nowhere did I say the question was answerable.
Actually I said the reverse -- the question is completely unaswerable, and
this is what Descartes originally said as well.
C
That indeed is exactly my point. The correct working of our perceptual
systems is largely independent of these sorts of ontological questions.
M
Largely? But not entirely? In which sense not entirely?
Cathy:
To the extent that we want them to answer ontological questions, our
perceptual systems leave something to be desired. They are particularly
unsatisfactory when it comes to answering pointless, unanswerable,
non-existent questions.
C
Well, Wittgenstein is always good for a bit of philosophical cabaret.
M
If Wittgenstein is good for cabaret, what is Descartes good for? Ghost
trains perhaps?
Cathy:
I've never heard Descartes' steam train impression but I understand
Wittgenstein tells better jokes.
M
OK. Let's see if you can be my Socratic midwife -;
Cathy:
I have to tell you that I'm not authorized to dispense pain-killing
medication.
<snip>
C
[snip]
Or do you mean that my view is incorrect in some way?
M
That's getting closer, though I might prefer to say it does not even get to
the point where it is a genuine *view* of anything which might be incorrect.
It's more that it is incorrect in being a pseudo view or pseudo theory which
does not even qualify for other forms of incorrectness such as falsity.
Cathy:
I didn't really see it as being a theory at all (pseudo or otherwise) but a
bit of pure pragmatism. We can't answer questions about the existence of
the external world, but we can make judgements. I made such a judgement
when I said that some sort of external world probably does exist because our
perceptions are not completely random. However, given that one postulates
the existence of an external world, one should assume the absolute minimum
about it (you know, entities not being multiplied unnecessarily and all
that). And it seems to me that the absolute minimum one can say about the
external world is that it is some sort of agency which orchestrates our
conscious perceptions. It seems to me however, that most forms of
physicalism or materialism assume rather more than this.
You might like to regard this as a sort of positivism, but I suspect you
won't.
C
Or do you just mean the question is not worth discussing? In that case I'd
say it's worth discussing to the extent that we are actually discussing it.
M
That's yet closer, but I think the real trouble is that it is impossible to
discuss your question. We can perhaps, for a brief stretch, discuss whether
we can discuss it, but we cannot discuss itself. And the reason is that you
have defined the 'issue' of whether we *really perceive* anything in such a
way that no possible evidence could be relevant to it.
Cathy:
Well actually I didn't. You were the one who introduced the notion that
veridical perception depended on the presence of some external object. I
pointed out that we do not, in fact, determine veridicality on this basis.
But I agree with you -- the question is unanswerable. But we have had quite
a few rounds of this discussion so far, and only now do I get you to admit
that the existence of the external world is not only impossible to determine
but largely irrelevant to the workings of perception.
I press this point because it seems to me an example of a more general one.
Any process which involves a correlation of states is wholly dependent on
the reliability of the mechanism which does the correlating. And if the
process which is used to test the reliability of this correlating mechanism
itself depends on a correlation of states, the same applies, and we have an
infinite regress. And as just about any form of objective perception, and
just about any sort of computation (even non-Turing computation) seems to
depend on a correlation of states, it makes me wonder what sort objective
perception, or what sort of computation, or what sort of anything, could
ever
determine any proposition with absolute certainty. And that would probably
be a pointless, unanswerable, non-existent question if it weren't for the
fact that there is one proposition we can establish with absolute
certainty -- the proposition that we have subjective experiences. Just what
sort of objective mechanism can do this? Can such a mechanism even exist?
Cathy
[Catherine Reason]