Science and Technology news... :: View topic

Wolf Kirchmeir · Guest

On 6 Aug 2003 11:12:26 -0700, Eray Ozkural exa wrote:

[quote]If you believe by mind I mean something like an eternal soul you do
not understand much about philosophy of mind.
[/quote]
Oh, I understand quite a bit about philosophy of mind. At one time, I was
well on the way toward a philosophy degree, in fact. It>s your philosophy I
don>t understand, since it seems to consist mostly of strenuously denying
that behaviourism has any validity or relevance. Considering that you want to
build a machine that exhibits "mind", that>s inconsistent, to put it gently.

HTH

--
Best Wishes,
Wolf Kirchmeir, Blind River ON
"Not that brains are everything --
you>ll also need a skull to put them in." (Nancy Franklin, 1997)
<just one w and plain ca for correct address>

Fred Mailhot · Guest

On 8/6/03 11:12 AM, "Eray Ozkural exa" <erayo@bilkent.edu.tr> wrote:
[quote]
I would tell you the exact relation between external behavior and the
mind, but am not quite sure you can understand, therefore I might
spare you from a real philosophical discussion. :)

[/quote]
Please Eray...enlighten us...I know that I, for one, would love to know the
"EXACT relation between external behaviour and the mind"...moreover, I>m
quite certain that I can understand anything that will come out of your
mouth (unless it>s in Turkish, of course *grin*)...

Fred.

Eray Ozkural exa · Guest

Fred Mailhot <fred.mailhot@videotron.ca> wrote in message news:<BB56C2F1.6C7C%fred.mailhot@videotron.ca>...
[quote]
Please Eray...enlighten us...I know that I, for one, would love to know the
"EXACT relation between external behaviour and the mind"...moreover, I>m
quite certain that I can understand anything that will come out of your
mouth (unless it>s in Turkish, of course *grin*)...
[/quote]
Gladly, since you so sincerely seek enlightenment. Something you will
never find in Chomsky>s school. [+]

It is exactly the same relation between the I/O history of a
computational device and a complex non-linear function the device
realizes. It is usually not possible to decide what the function is by
analyzing any interval of I/O. Even if you had the whole history
(which includes the future and therefore practically impossible), you
would not be able to derive the function for those above a certain
complexity (which is not too high). Typically all mammal minds are
functions above that threshold. Many devices we build are like that,
too complex for behavioral analysis to unveil the mathematical nature
of the underlying function. [*]

That is the uninteresting relation.

The above trivial observation also lies at the heart of Turing>s
semi-philosophical 1950 paper, which was meant for laymen
(non-mathematicians). A necessary aspect in the imitation game was the
computer performing equivalent to humans *consistently*. That means a
mathematical limit of number of games towards infinity, which is
non-sense in practice.

Furthermore, at that time they did not know much about algorithmic
information theory and Turing>s paper has not made any observations on
how the complexity might affect the practicality. In particular Turing
made a huge mistake in assuming that one could use natural language to
effectively probe *all* mental facilities. That is wrong with respect
to my above explanation. Maybe I should write a rebuttal to Turing>s
paper which can analyze the situation in light of recent results. We
can be more realistic about what can be done with language.

I do not intend to write in excess of what is strictly necessary. To a
person equipped with a graduate level understanding of theory of
computation and complexity the above argument should be sufficient.

Regards,

[+] I>m hoping this inappropriate tone of yours is not related to my
debunking of Chomsky>s dogmatic assumptions such as "innateness
assumption". Anyway, I will ignore your humor and explain to you what
I meant by that remark.

[*] If you are unclear about this let me explain it to you in a more
"material" way. A physicist friend of mine asked me, cannot we simply
measure the sensory input and effector output of the brain and
henceforth simulate it on the grounds of data obtained? I answered
"No. One cannot derive a simulation of a complex non-linear system in
that fashion for many operational patterns will not be present on the
data you have collected"

PS: I know that I have just claimed that Turing was wrong in the paper
in a sense that has not been proposed before. I think my observation
is finishing what French>s and Davidson>s analysis of the matter
started. If you speak of this idea, I urge you to credit it to me
afterwards. I believe it underlines the limit of behavioral analysis
in a succinct way.

__
Eray Ozkural
PhD student
CS Dept., Bilkent Univ.

Fred Mailhot · Guest

Hi Eray...

On 8/6/03 10:19 PM, "Eray Ozkural exa" <erayo@bilkent.edu.tr> wrote:
[quote]
Gladly, since you so sincerely seek enlightenment. Something you will
never find in Chomsky>s school. [+]
[/quote]
*shrug*...I>ll take my enlightenment where I can find it, and I>ve managed
to find some in generative linguistics...

[quote]It is exactly the same relation between the I/O history of a
computational device and a complex non-linear function the device
realizes.
[/quote]
To be clear, this is what I hear/read you saying:

"The exact relation between external behaviour and mind is exactly the same
relation as that between the I/O history of a computational device and
whatever complex non-linear function that device realizes."

So are you saying that:

1) mind is "the cmplx nonlin fxn"

or

2) mind is WHAT COMPUTES "the cmplx nonlin fxn"

Or am I beside the mark ?? I think you>re saying that mind IS the
function...

We>ll take it on faith that I>ve understood you correctly...

Well for starters, the functions (read "dynamical system") emulated by
computational devices need not be complex OR nonlinear, so I>m not really
sure what you>re getting at...or are you talking about the function (dyn.
sys.) realized at the physical level ?? (I assume you know that emulation
and realization are not the same thing)

[quote]It is usually not possible to decide what the function is by
analyzing any interval of I/O.
[/quote]
Seems to me that this is exactly what reverse engineering tries to do...and
while I agree that it>s pretty touch and go, I don>t think that it>s
impossible to ever find out what the true function is...

[quote]Even if you had the whole history
(which includes the future and therefore practically impossible), you
would not be able to derive the function for those above a certain
complexity (which is not too high).
[/quote]
"those" what above a certain complexity ?? I/O sequences ?? Do you have a
ref. for this hardness result ?? I>d like to know more about that...

[quote]Typically all mammal minds are functions above that threshold.
[/quote]
Not to pick at nits, Eray, but this is clearly an assumption on your part...
a very plausible one, to be sure, but an assumption nonetheless...

And for clarity>s sake, I>ll assume whenever you say "mind" from now on that
you mean "function that takes (something) as input and outputs external
behaviour"....

(is it me, or does that sound behaviourist, with the word "mind" thrown in?)

[quote]Many devices we build are like that, too complex for behavioral analysis to
unveil the mathematical nature of the underlying function. [*]

That is the uninteresting relation.

The above trivial observation also lies at the heart of Turing>s
semi-philosophical 1950 paper, which was meant for laymen
(non-mathematicians).
[/quote]
I>ve read the paper, thanks...and since it appeared in Mind, I don>t know if
either of "semi-philosophical" or "for laymen" are proper
characterizations...

[quote]A necessary aspect in the imitation game was the
computer performing equivalent to humans *consistently*. That means a
mathematical limit of number of games towards infinity, which is
non-sense in practice.
[/quote]
Stop telling me that infinite sequences are unrealizable in practice...I>m
not a dimwit, regardless of what you may think.

[quote]Furthermore, at that time they did not know much about algorithmic
information theory and Turing>s paper has not made any observations on
how the complexity might affect the practicality.
[/quote]
The complexity of what ??

Turing>s paper was meant as an instigator of ideas...I>m quite certain that
questions of "practicality" were pretty unimportant to him when he wrote
it...

[quote]In particular Turing made a huge mistake in assuming that one could use
natural language to effectively probe *all* mental facilities.
[/quote]
Turing>s paper wasn>t about probing mental faculties, Eray...all he claimed
was (don>t have the paper handy for a direct quotation, David) something to
the effect of "if a computer could consistently exhibit behaviour X, then we
would be foolish if we insisted on saying that it>s not intelligent"...

Furthermore, I think natural language is PLENTY strong enough as a probe for
intelligence. Have you read Dan Dennett>s "Can Machines Think" or any of Bob
French>s several papers on the topic ??

(btw, Bob French was the translator of GEB into French)

French>s stuff is ALL recommended reading...running the gamut from
foundational issues to technical neural-net related stuff...his page is
here:

http://www.ulg.ac.be/cogsci/rfrench.html

[quote]That is wrong with respect to my above explanation.
[/quote]
You mean w.r.t. your claim that looking at I/O isn>t enough to determine the
"complex non-linear function" that is the "mind" ??

[quote]Maybe I should write a rebuttal to Turing>s
paper which can analyze the situation in light of recent results. We
can be more realistic about what can be done with language.
[/quote]
Once again, have a look at Dennett>s paper, or a French>s stuff...we can do
a LOT more with language than most people suspect. And more to the point,
we don>t need more than language to get the result that Turing was talking
about.

The point is that the Turing Test is, if anything, TOO strong...it>s
unlikely that any computer could EVER pass it...which only makes it more
likely that if one DID pass it, we>d be stupid not to call it intelligent...

[quote]I do not intend to write in excess of what is strictly necessary.
[/quote]
You rarely do....*grin*...

[quote]To a person equipped with a graduate level understanding of theory of
computation and complexity the above argument should be sufficient.
[/quote]
Well, your "argument" is certainly _not_ sufficient, since I remain
skeptical...and moreover:

1) Your above statements do NOT require "a graduate level understanding" of
anything, let alone theory of computation and complexity, since you don>t
seem to have made any substantive arguments. You used the words "complex"
and "non-linear" in a couple of places, but I can point you to recent papers
in semiotics that do the same thing...

2) You>re a little too elitist sometimes, Eray...try to get off your
educational pedestal and understand that some people are incredibly good at
understanding things without needing a formal eduction...

[quote][+] I>m hoping this inappropriate tone of yours is not related to my
debunking of Chomsky>s dogmatic assumptions such as "innateness
assumption". Anyway, I will ignore your humor and explain to you what
I meant by that remark.
[/quote]
I>ll make no apologies for my irreverence...I reserve it for people who
claim to have "the answer" to anything...

And I don>t recall having seen you debunk anything...if you>re referring to
the "Universal Driver" argument, I>m not convinced and am slowly working on
a reply...

Moreover, Chomsky clearly states in several different places that the
nativist stance is a hypothesis, not an assumption...he even says "it may be
wrong at the end of the day, but I think there are good reasons to believe
it"...

[quote]
[*] If you are unclear about this let me explain it to you in a more
"material" way. A physicist friend of mine asked me, cannot we simply
measure the sensory input and effector output of the brain and
henceforth simulate it on the grounds of data obtained? I answered
"No. One cannot derive a simulation of a complex non-linear system in
that fashion for many operational patterns will not be present on the
data you have collected"
[/quote]
OK...before I say that that>s not actually an explanation, but simply a
repetition of what you said before...

explain to me what an "operational pattern" is...

And as for simulating nonlinear systems...have a look at the following by
Sam Roweis and Zoubin Gharamani...

http://www.cs.toronto.edu/~roweis/papers/nllds_preprint.pdf

[quote]PS: I know that I have just claimed that Turing was wrong in the paper
in a sense that has not been proposed before.
[/quote]
Well, I have no way of knowing whether or not the way in which you claim
Turing was "wrong" has been claimed before, but I do feel compelled to point
out that your claim is very much beside the point that Turing made in his
paper.

[quote]I think my observation is finishing what French>s and Davidson>s analysis
of the matter started.
[/quote]
Then you haven>t read Turing or French correctly...I don>t know what
Davidson had to say on the matter....

[quote]If you speak of this idea, I urge you to credit it to me
afterwards.
[/quote]
ROTFLMFAO...*giggle*...oh...oh that>s funny...are you being serious ??

In that case, I _urge_ you to write it up and submit it for publication...

[quote]I believe it underlines the limit of behavioral analysis
in a succinct way.
[/quote]
And I don>t....*shrug*...

[quote]Eray Ozkural
PhD student
CS Dept., Bilkent Univ.
[/quote]
You should update your webpage, since it appears to say that you>re
expecting an MS, and not a PhD...

And why do you have your Linguistics and Semantics courses under
"Programming Languages" ?? And why are they separated ?? Do you not think
that Semantics is part of Linguistics ?? Or do you mean denotational
semantics ??

Ahh...it says you haven>t updated since 2001...ok...

Anyway...I await your reply with baited breath, Eray...

Fred. :0)

Glen M. Sizemore · Guest

Erape: 1) Naive behaviorism a la Skinner, David, Glen-animal that *asserts*
mind is identical to external behavior.

GS: No, Skinnerians do not hold this. First of all, they hold that the
meanings of terms are to be found in the variables that control the speaker'
s verbal behavior. Often, some of these variables may be inferred from
examining "usage of the terms." Virtually all of us are "taught" to use
"mental" terms in the

third-person when we are in the presence of someone behaving in particular
ways*. That is, such responses are reinforced in the presence of a person
behaving in a particular way, and such behavior becomes a discriminative
stimulus that controls the speaker>s verbal behavior. For this reason, I do
not typically talk about "folk-psychology" because it implies that people
"have" a "theory of mind" in the same sense that scientists "have theories."
My son, for example, does not display his "theory of mind," when he
correctly "uses" mental terms, he is simply emitting operants under the
discriminative control of the behavior of others (as well as under control
of his own behavior). Later, no doubt, he will be exposed to true theories
of mind, as are virtually all people in the modern world. Even when we have
actual theories of mind, we continue to respond verbally in the sense that I
described earlier, it is just that we now have another, rather extensive,
verbal repertoire that reflects our indoctrination into the ranks of
philosophers and "scientists"

Erape: This is a foolish doctrine
that is discarded (somewhat similar to naive functionalism) and cannot
explain such things as mad-man pain or dreams, beliefs, feelings, etc.

GS: Erape will have to tell me what "mad-man pain" is, but I have a great
deal to say about "pain." Dreams appear to be behavior that is under minimal
control of the current environment. Since we do not really know much about
how brain activity relates to behavior, there is little to be said, with
certainty, about neurobiological mechanisms. We are, I think, hundreds of
years away from having anything resembling a complete description of how the
brain mediates the effects of the simplest of contingencies investigated in
the laboratory. Sadly, it is "theories of mind" that increasingly stand in
the way of any real progress, despite the self-congratulatory ranting of
many neurobiologists. The meaning of "beliefs," like other controversial
terms, must be assessed by examining both colloquial and "technical" usage,
but I have already discussed the "use of mental terms" in a general way
above. The issue of "feelings" gives rise to a discussion of a topic that
shows, again, that your characterization of radical behaviorism is
incorrect. That issue involves how it is we come to talk about events to
which we alone have access. This is the very issue that Wittgenstein
struggled with in his discussion of the impossibility of a private language,
and he came, essentially, to the same conclusions as Skinner. Skinner>s
version is much clearer, however, because he elucidates clearly the
behavioral processes involved and more fully examines the implications. His
position is this: We "become aware of" features of the world in the sense of
"responding to them" as our behavior comes under discriminative control of
these features. Some of our responses are verbal, and the verbal community,
thus, plays a huge role in shaping the events of which we are aware. This
holds for verbal responses under discriminative control of our own behavior,
both public and private. So when we introspect, we observe and talk about
our own behavior or closely associated states of our bodies. Some of our
self-descriptive repertoire is "accurate" because it is established when the
behavior serving a discriminative function is occurring overtly and
descriptions of such behavior remains intact when the it becomes greatly
reduced in intensity. Other aspects of our behavior are not observed or
described accurately and we are frequently unable to "describe our
feelings."

So you are wrong in that behaviorists are well aware of the importance of
the topic of feelings, dreams, imagining etc. in the development of
institutionalized mentalism. So, in the instances in which "mind" is "used
as a synonym" for subjectivity, behaviorists do not assert that the
"referent" is "external behavior" (I prefer "public behavior").

*Indeed, if one examines the etymology of mental terms, one finds that they
were once frank references to behavior.

"Eray Ozkural exa" <erayo@bilkent.edu.tr> wrote in message
news:fa69ae35.0308070703.668d11bc@posting.google.com...
[quote]Hi Wolf,

So, why should you imply that I believe in a funky version of
Descartes>s dualism? I never said that.

Behaviorism (as a theory) is as irrelevant as identity theory of mind
(or dualism), that is what I am saying.

It has relevance only as far as showing "Look people thought of this
theory before, but it couldn>t solve this and that therefore it was
abandoned". Just like identity theory. Besides behaviorism is not
simply the theory that mind is identical to mind, but something else
in addition: a doctrine. It *asserts* the truth of three outrageous
claims. [1] A philosophical theory doesn>t *assert* things, it argues
for them. Still, of course you can think of it as a theory, ignoring
the full doctrine, as it is viewed in philosophy of mind textbooks.

Behavioral analysis surely has benefits in scientific study, in those
cases we cannot reason at all about a system we will resort to
simplistic external observations and "guess" what>s inside the "black
box". However, such explanations have a very limited scope. Think
making "anatomy" without cutting apart or scanning any being.

However, we could talk about two versions of behaviorism
1) Naive behaviorism a la Skinner, David, Glen-animal that *asserts*
mind is identical to external behavior. This is a foolish doctrine
that is discarded (somewhat similar to naive functionalism) and cannot
explain such things as mad-man pain or dreams, beliefs, feelings, etc.
2) Your sort of behaviorism which tries to classify brain operation as
"internal behavior". Note that all computational systems have a
message passing facet and an operational facet which could be viewed
as a manifest of small communicative behaviors and switching
behaviors. However, the correct analysis of such a system is always
computational. That is because one needs a powerful computational
model (like a TM) to characterize the "internal behavior" of a
computer one way or another, ie a finite state machine or a markov
model will not do. In that manner, when we take it this far your
version of brain behaviorism becomes quite similar to computational
view of mind.

Now, if you have such sophistication in philosophy I challenge you to
respond to this post point by point and drop those annoying HTHs. You
are not helping with awkward arguments. I>m cross-posting to
comp.ai.philosophy because the distinction I am drawing is
philosophical as well as it concerns cognitive scientists. (Typically
"behaviorism" as a doctrine is incompatible with all cognitive
sciences)

Regards,

__
Eray Ozkural

Quoting from http://plato.stanford.edu/entries/behaviorism/
(1) Psychology is the science of behavior. Psychology is not the
science of mind.

(2) Behavior can be described and explained without making reference
to mental events or to internal psychological processes. The sources
of behavior are external (in the environment), not internal (in the
mind).

(3) In the course of theory development in psychology, if, somehow,
mental terms or concepts are deployed in describing or explaining
behavior, then either (a) these terms or concepts should be eliminated
and replaced by behavioral terms or (b) they can and should be
translated or paraphrased into behavioral concepts.

"Wolf Kirchmeir" <wwolfkir@sympatico.can> wrote in message
news:<jbysxveflzcngvpbpna.hj86bq2.pminews@news1.sympatico.ca>...
On 6 Aug 2003 11:12:26 -0700, Eray Ozkural exa wrote:

If you believe by mind I mean something like an eternal soul you do
not understand much about philosophy of mind.

Oh, I understand quite a bit about philosophy of mind. At one time, I
was
well on the way toward a philosophy degree, in fact. It>s your
philosophy I
don>t understand, since it seems to consist mostly of strenuously
denying
that behaviourism has any validity or relevance. Considering that you
want to
build a machine that exhibits "mind", that>s inconsistent, to put it
gently.

HTH[/quote]

Eray Ozkural exa · Guest

erayo@bilkent.edu.tr (Eray Ozkural exa) wrote in message news:<fa69ae35.0308070703.668d11bc@posting.google.com>...
[quote]abandoned". Just like identity theory. Besides behaviorism is not
simply the theory that mind is identical to mind, but something else
[/quote]
correction: the theory that mind is identical to behavior. sorry.


www.GetXFactor.com Leading Technology, Science, Agriculture News and information	Part of the Identityscape.com network... getxfactor.com jmoodmusic.com smartbusinesschoices.com mintdepot.com lowfaresalways.com evangelicalview.com shoppingpodder.com soproudlywehail.com webnews.ws currenthumor.com

Science and Technology news... Forum Index -> Cognitive Science Forum	Goto page Previous 1, 2
Page 2 of 2