Science and Technology news... :: View topic

Lester Zick · Guest

On Fri, 29 Aug 2003 21:00:23 -0400, Joe Legris <jalegris@xympatico.ca>
wrote:

[quote]Lester Zick wrote:
On Thu, 28 Aug 2003 23:29:58 -0400, Joe Legris <jalegris@xympatico.ca
wrote:

[. . .]

Representationalism is certainly the direction I>m leaning toward but it
means different things to different people. You seem to take it in the
most literal sense possible - representations are little copies of real
objects in the head. I take an information theoretic view -
representations are internal patterns of neural structure or activity.
Some of these representations are internal models. By definition, a
model is not a copy. Models predict only limited aspects of the world,
precisely because it is impossible to capture the entire reality.

I>m not sure why a model cannot be a copy. Certainly as a matter of
definition. I think a copy would be a model in any event just as the
original is a model for the original. What needs to be decided is what
makes a copy a copy and what makes something not a copy. Then the
original question of whether a model is a copy can be decided
accordingly. Either way you>re going to have a model but not
necessarily a copy.

If you define a model as a simplified representation of a system then it
is not a copy by definition. End of argument.
[/quote]
Why not just define representations we use to judge things as a not
copies? Why rationalize them as copies or not copies according to
simplicity or complexity? For that matter why define them as models or
not models?

None of this has any bearing on whether or not models are copies or
not. You asked the right question before. Now you>re just defining
away what you can>t explain.
[quote]
Models in the brain are distillations of the aspects of reality that are
relevant to the organism. For example, a brain may contain a ballistic
model that allows one to predict where a thrown rock will land. The
relevant aspect of such a model is the relationship between perceived
heft of an object and it>s expected trajectory when thrown with a
particular effort. It is conceivable that the brain actually solves the
differential equation of motion using a neural network that has been
tuned over years of practice. Such a model says nothing and need say
nothing about any other aspect of projectile throwing. Through neural
connections to rest of the brain it receives inputs about the mass of
the rock and the intention of the thrower and produces outputs about the
expected range. Of course, we cannot yet demonstrate the existence of
such a model, but it is at least plausible and reasonably parsimonious.

Compare this with a copy of a rock and a copy of the space surrounding
the organism. It is patently ridiculous. Why so much detail? Is it
necessary to include every little stone on the ground as well? How about
the individual leaves on the trees? If you are fully invested in the
copy theory you would probably want to do so just in case the organism
decides to use a different rock, or chooses another target. Of course,
we cannot yet demonstrate the existence of such a model, and we
certainly never will. It is implausible, wasteful of neural resources,
inflexible (what if it>s an indoor game of volleyball rather than
outdoor rock throwing) and completely unparsimonious.

So much word salad. It looks like you have a very specific idea where[/quote]
you want to get to and not a clue as to how to get there. Everything
you say here is just a series of arbitrary suppositions and
definitions.

Using such an approach anyone can justify anything and claim they>ve
gotten anywhere they want to go. You could claim the ballistic model
you refer to is really an analog computer algorithm. This is just so
much speculation. You still don>t have any idea whether what you>d
like to deal with is a copy or not. End of argument.

Regards - Lester

remove DEL in address for email

Efecto Mariposa · Guest

I>m not so sure, Wolf. Arturo Rosenbluth once said that the best model
of a cat was another cat. Preferably, the same cat.

Wolf Kirchmeir wrote:
[quote]On Thu, 28 Aug 2003 22:07:44 GMT, Lester Zick wrote:

The brain apparently does contain representations that model certain
aspects of reality, but models are not copies.

Well I>m not really sure. What ensures or prevents models from being
copies?

If you>ve ever made model airplanes, you should know the difference.

[/quote]

John · Guest

Is it even necessary for a brain to produce a model of copy of the
environment? Perhaps all it has to do is identify the features relevant to
producing an appropriate response, which negates a whole lot of work in
forming a model or whatever?

--
johnYYYcoe@tpg.com.au

remove YYY in reply
"Glen M. Sizemore" <gmsizemore2@yahoo.com> wrote in message
news:d0f24d303114a0cbb84b45b2e39586ac@news.teranews.com...
[quote]Lester Zick wrote:

If copy theory is an inadequate basis on which to explain how the mind
produces what appear to be copies of external phenomena - which I
agree that it is - I would like to ask the following question:

In the absence of copy theory how is it that some external sensation A
is processed or operated on by brain B so as to produce some C as a
non copy of A yet with the ability to operate on C to produce some
apparent copy of A?

There are really two parts to this question:

How does B operate on A to produce some C as a non copy of A?

Such that B can also operate on C to produce some apparently realistic
image of A?

Regards - Lester

"Copy theory" is a straw-man. The only reference I can find to such a
thing is here:

Chander, M.J., & Boyes M. (1982). Social Cognitive Development. In B. B.
Wolman (Ed), Handbook of developmental psychology (pp. 387-402).
Englewood Cliffs, NJ: Prentice Hall.

As they describe it, a child>s (age 2-3) ability to judge the contents
of other people>s minds is likened to a recording device that copies the
world without interpretation. Because of this young children fail the
false belief task, where they are incapable of conceiving that someone
else may believe something that they know is false.

This has nothing to do with the general nature of consciousness and the
perception of reality. No one seriously believes that the brain contains
a copy of reality.

The brain apparently does contain representations that model certain
aspects of reality, but models are not copies.

When the brain (B) operates on sensed external variables (A) to produce
a model (C), it uses information and internal structures developed
over>>the life and phylogenic history of the organism. Presumably, the
brain
has a reentrant structure (i.e. internal feedback) so that the resulting
model can also be taken as another input variable.

All perception is the result of brain processes. If some perceptions
appear more "realistic" than others I would expect that the perceiver
has had some practice making such discriminations and has found that
certain attributes, such as visual clarity, a sense of wakefulness, and
correspondence to intuitive physical properties of the world, have
generally led to satisfactory outcomes. Or at least realistic ones.

--
Joe Legris

Absolute nonsense, Joe. This is the hallmark of representationalism.
In the
case of the "cognitive map" it is quite explicit, but it is usually only
thinly disguised. What the hell is the "prototype" theory for
Chrissakes.
Honestly, Joe, it is not like you>re a drooling crack head, but
sometimes
you say the......well......dumbest things.

JL: Representationalism is certainly the direction I>m leaning toward but
it
means different things to different people. You seem to take it in the
most literal sense possible - representations are little copies of real
objects in the head. I take an information theoretic view -
representations are internal patterns of neural structure or activity.
Some of these representations are internal models. By definition, a
model is not a copy. Models predict only limited aspects of the world,
precisely because it is impossible to capture the entire reality.

GS: But this doesn>t get around the problem of homunculism. Model, copy,
representation......they still are the thing that is seen, heard, felt,
tasted etc. etc. etc.

JL: Internal models cannot be observed directly (yet) because they are
part
of the brain>s overall process - outside the context of the brain of
which they are a part, internal models are effectively indecipherable
patterns.

GS: But they are what the person observes, right?

JL: No one has ever observed a copy of anything in any brain. That is
because there are no copies.

GS: Oh......right.......there are only models. Big difference there, Joe.

JL: We>ve been through this before: if there
were actual copies of anything in the brain then a "brain equivalent"
would be required to make sense of it, which leads to infinite regress.

GS: The same is true of "models" in the brain.

"Joe Legris" <jalegris@xympatico.ca> wrote in message
news:3F4EC8B6.1000103@xympatico.ca...
Glen M. Sizemore wrote:

"Joe Legris" <jalegris@xympatico.ca> wrote in message
news:3F4E7634.6010006@xympatico.ca...

Lester Zick wrote:

If copy theory is an inadequate basis on which to explain how the
mind
produces what appear to be copies of external phenomena - which I
agree that it is - I would like to ask the following question:

In the absence of copy theory how is it that some external sensation
A
is processed or operated on by brain B so as to produce some C as a
non copy of A yet with the ability to operate on C to produce some
apparent copy of A?

There are really two parts to this question:

How does B operate on A to produce some C as a non copy of A?

Such that B can also operate on C to produce some apparently
realistic
image of A?

Regards - Lester

"Copy theory" is a straw-man. The only reference I can find to such a
thing is here:

Chander, M.J., & Boyes M. (1982). Social Cognitive Development. In B.
B.
Wolman (Ed), Handbook of developmental psychology (pp. 387-402).
Englewood Cliffs, NJ: Prentice Hall.

As they describe it, a child>s (age 2-3) ability to judge the contents
of other people>s minds is likened to a recording device that copies
the
world without interpretation. Because of this young children fail the
false belief task, where they are incapable of conceiving that someone
else may believe something that they know is false.

This has nothing to do with the general nature of consciousness and
the
perception of reality. No one seriously believes that the brain
contains
a copy of reality.

The brain apparently does contain representations that model certain
aspects of reality, but models are not copies.

When the brain (B) operates on sensed external variables (A) to
produce
a model (C), it uses information and internal structures developed
over
the life and phylogenic history of the organism. Presumably, the brain
has a reentrant structure (i.e. internal feedback) so that the
resulting
model can also be taken as another input variable.

All perception is the result of brain processes. If some perceptions
appear more "realistic" than others I would expect that the perceiver
has had some practice making such discriminations and has found that
certain attributes, such as visual clarity, a sense of wakefulness,
and
correspondence to intuitive physical properties of the world, have
generally led to satisfactory outcomes. Or at least realistic ones.

--
Joe Legris

Absolute nonsense, Joe. This is the hallmark of representationalism.
In the
case of the "cognitive map" it is quite explicit, but it is usually
only
thinly disguised. What the hell is the "prototype" theory for
Chrissakes.
Honestly, Joe, it is not like you>re a drooling crack head, but
sometimes
you say the......well......dumbest things.

Representationalism is certainly the direction I>m leaning toward but it
means different things to different people. You seem to take it in the
most literal sense possible - representations are little copies of real
objects in the head. I take an information theoretic view -
representations are internal patterns of neural structure or activity.
Some of these representations are internal models. By definition, a
model is not a copy. Models predict only limited aspects of the world,
precisely because it is impossible to capture the entire reality.

Internal models cannot be observed directly (yet) because they are part
of the brain>s overall process - outside the context of the brain of
which they are a part, internal models are effectively indecipherable
patterns.

No one has ever observed a copy of anything in any brain. That is
because there are no copies. We>ve been through this before: if there
were actual copies of anything in the brain then a "brain equivalent"
would be required to make sense of it, which leads to infinite regress.

--
Joe Legris

[/quote]

Glen M. Sizemore · Guest

JL: Models in the brain are distillations of the aspects of reality that are
relevant to the organism. For example, a brain may contain a ballistic
model that allows one to predict where a thrown rock will land. The
relevant aspect of such a model is the relationship between perceived
heft of an object and it>s expected trajectory when thrown with a
particular effort. It is conceivable that the brain actually solves the
differential equation of motion using a neural network that has been
tuned over years of practice.

GS: Right! And when the homunculus is steering the body to a point at the
end of the trajectory it consults this integral. That is, if it isn>t too
busy checking grammar rule-book so it can cause the guy to yell, "I got it!"

JL: Such a model says nothing and need say
nothing about any other aspect of projectile throwing. Through neural
connections to rest of the brain it receives inputs about the mass of
the rock and the intention of the thrower and produces outputs about the
expected range. Of course, we cannot yet demonstrate the existence of
such a model, but it is at least plausible and reasonably parsimonious.

GS: Yeah right. Postulating a differential equation-solving entity in the
head is real parsimonious. But I guess, after all, if the Moon can solve a
differential equation, why not our homunculus? I mean "brain."

JL: Compare this with a copy of a rock and a copy of the space surrounding
the organism. It is patently ridiculous. Why so much detail? Is it
necessary to include every little stone on the ground as well? How about
the individual leaves on the trees? If you are fully invested in the
copy theory you would probably want to do so just in case the organism
decides to use a different rock, or chooses another target. Of course,
we cannot yet demonstrate the existence of such a model, and we
certainly never will. It is implausible, wasteful of neural resources,
inflexible (what if it>s an indoor game of volleyball rather than
outdoor rock throwing) and completely unparsimonious.

GS: None of this matters. It doesn>t eliminate the reasons that both models
and copies are ludicrous.

"Joe Legris" <jalegris@xympatico.ca> wrote in message
news:3F4FF727.4080605@xympatico.ca...
[quote]Lester Zick wrote:
On Thu, 28 Aug 2003 23:29:58 -0400, Joe Legris <jalegris@xympatico.ca
wrote:

[. . .]

Representationalism is certainly the direction I>m leaning toward but it
means different things to different people. You seem to take it in the
most literal sense possible - representations are little copies of real
objects in the head. I take an information theoretic view -
representations are internal patterns of neural structure or activity.
Some of these representations are internal models. By definition, a
model is not a copy. Models predict only limited aspects of the world,
precisely because it is impossible to capture the entire reality.

I>m not sure why a model cannot be a copy. Certainly as a matter of
definition. I think a copy would be a model in any event just as the
original is a model for the original. What needs to be decided is what
makes a copy a copy and what makes something not a copy. Then the
original question of whether a model is a copy can be decided
accordingly. Either way you>re going to have a model but not
necessarily a copy.

If you define a model as a simplified representation of a system then it
is not a copy by definition. End of argument.

Models in the brain are distillations of the aspects of reality that are
relevant to the organism. For example, a brain may contain a ballistic
model that allows one to predict where a thrown rock will land. The
relevant aspect of such a model is the relationship between perceived
heft of an object and it>s expected trajectory when thrown with a
particular effort. It is conceivable that the brain actually solves the
differential equation of motion using a neural network that has been
tuned over years of practice. Such a model says nothing and need say
nothing about any other aspect of projectile throwing. Through neural
connections to rest of the brain it receives inputs about the mass of
the rock and the intention of the thrower and produces outputs about the
expected range. Of course, we cannot yet demonstrate the existence of
such a model, but it is at least plausible and reasonably parsimonious.

Compare this with a copy of a rock and a copy of the space surrounding
the organism. It is patently ridiculous. Why so much detail? Is it
necessary to include every little stone on the ground as well? How about
the individual leaves on the trees? If you are fully invested in the
copy theory you would probably want to do so just in case the organism
decides to use a different rock, or chooses another target. Of course,
we cannot yet demonstrate the existence of such a model, and we
certainly never will. It is implausible, wasteful of neural resources,
inflexible (what if it>s an indoor game of volleyball rather than
outdoor rock throwing) and completely unparsimonious.

--
Joe Legris
[/quote]

Wolf Kirchmeir · Guest

On Sat, 30 Aug 2003 13:07:26 GMT, Glen M. Sizemore wrote:

[quote]GS: Right! And when the homunculus is steering the body to a point at the
end of the trajectory it consults this integral. That is, if it isn>t too
busy checking grammar rule-book so it can cause the guy to yell, "I got it!"
[/quote]
There>s no homunculus in Joe>s description, anymore than there is one in a
computer programmed to solve differential equations.

--
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>

Wolf Kirchmeir · Guest

On Fri, 29 Aug 2003 21:25:39 GMT, Glen M. Sizemore wrote:

[quote]GS: But this doesn>t get around the problem of homunculism. Model, copy,
representation......they still are the thing that is seen, heard, felt,
tasted etc. etc. etc.
[/quote]
No, they>re not. The "thing that is seen., etc" is outside the system. In an
information processing system, it>s the process itself that>s the model.
Anyone who has a program in which data is "passed" to a subroutine which then
"returns" data to the routine that "called" the subroutine knows that "model"
is very abstract concept indeed. "Homunculus" is so irrelevant to this
issue, it isn>t even wrong. However, it makes sense to say that the "program
solves a problem." If some people think there>s a little man inside the
computer that does the thinking, well, that shows that they don>t understand
what "solving a problem" really is. Even when they do it themselves. Even
when they are playing catch with their kid.

--
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>

Wolf Kirchmeir · Guest

On Fri, 29 Aug 2003 23:57:11 -0500, Efecto Mariposa wrote:

[quote]I>m not so sure, Wolf. Arturo Rosenbluth once said that the best model
of a cat was another cat. Preferably, the same cat.
[/quote]
I like that!

--
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>

Joe Legris · Guest

Glen M. Sizemore wrote:
[quote]JL: Models in the brain are distillations of the aspects of reality that are
relevant to the organism. For example, a brain may contain a ballistic
model that allows one to predict where a thrown rock will land. The
relevant aspect of such a model is the relationship between perceived
heft of an object and it>s expected trajectory when thrown with a
particular effort. It is conceivable that the brain actually solves the
differential equation of motion using a neural network that has been
tuned over years of practice.

GS: Right! And when the homunculus is steering the body to a point at the
end of the trajectory it consults this integral. That is, if it isn>t too
busy checking grammar rule-book so it can cause the guy to yell, "I got it!"

JL: Such a model says nothing and need say
nothing about any other aspect of projectile throwing. Through neural
connections to rest of the brain it receives inputs about the mass of
the rock and the intention of the thrower and produces outputs about the
expected range. Of course, we cannot yet demonstrate the existence of
such a model, but it is at least plausible and reasonably parsimonious.

GS: Yeah right. Postulating a differential equation-solving entity in the
head is real parsimonious. But I guess, after all, if the Moon can solve a
differential equation, why not our homunculus? I mean "brain."

JL: Compare this with a copy of a rock and a copy of the space surrounding
the organism. It is patently ridiculous. Why so much detail? Is it
necessary to include every little stone on the ground as well? How about
the individual leaves on the trees? If you are fully invested in the
copy theory you would probably want to do so just in case the organism
decides to use a different rock, or chooses another target. Of course,
we cannot yet demonstrate the existence of such a model, and we
certainly never will. It is implausible, wasteful of neural resources,
inflexible (what if it>s an indoor game of volleyball rather than
outdoor rock throwing) and completely unparsimonious.

GS: None of this matters. It doesn>t eliminate the reasons that both models
and copies are ludicrous.

[/quote]

But there are models of reality in other parts of the body, why not the
brain? For example, the spectral response of our eyes is a model of the
illumination available from the sun. The historical distribution of skin
and hair pigmentation is a model of the distribution of sunlight. The
ability of our livers to break down some toxins and not others is a
model of the sort of chemical threats that existed in the pre-industrial
era. The structure of our teeth is a model of the diet of of our
ancestors. And so on...

Even the probability of a response can be seen as a model of the
contingencies of reinforcement. That probability must come from
somewhere and it>s a safe bet that it>s in the brain.

--
Joe Legris

Glen M. Sizemore · Guest

[quote]GS: But this doesn>t get around the problem of homunculism. Model, copy,
representation......they still are the thing that is seen, heard, felt,
tasted etc. etc. etc.
[/quote]
WK: No, they>re not. The "thing that is seen., etc" is outside the system.

GS: According to me it is.

WK: In an
information processing system, it>s the process itself that>s the model.

GS: And just what does this mean?

WK: Anyone who has a program in which data is "passed" to a subroutine which
then
"returns" data to the routine that "called" the subroutine knows that
"model"
is very abstract concept indeed. "Homunculus" is so irrelevant to this
issue, it isn>t even wrong.

GS: You>re a victim of your arrogance again, Wolf. That hasn>t happened for
a while. I say the notion is relevant, and nothing you have said here
changes that.

WK: However, it makes sense to say that the "program
solves a problem."

GS: Yeah....metaphorical sense. A person who "explodes in anger" is not
really an explosive. You should think on that one for a while, Wolf. It
makes sense to say that a person "exploded" but that doesn>t make it a
scientifically useful notion. Really, Wolf, I wish you would or could think
more carefully..

Wolf: If some people think there>s a little man inside the
computer that does the thinking, well, that shows that they don>t understand
what "solving a problem" really is. Even when they do it themselves. Even
when they are playing catch with their kid.

GS: I just don>t have the taste for metaphor and tautology that you do Wolf.

"Wolf Kirchmeir" <wwolfkir@sympatico.can> wrote in message
news:jbysxveflzcngvpbpna.hkg7kx3.pminews@news1.sympatico.ca...
[quote]On Fri, 29 Aug 2003 21:25:39 GMT, Glen M. Sizemore wrote:

GS: But this doesn>t get around the problem of homunculism. Model, copy,
representation......they still are the thing that is seen, heard, felt,
tasted etc. etc. etc.

No, they>re not. The "thing that is seen., etc" is outside the system. In
an
information processing system, it>s the process itself that>s the model.
Anyone who has a program in which data is "passed" to a subroutine which
then
"returns" data to the routine that "called" the subroutine knows that
"model"
is very abstract concept indeed. "Homunculus" is so irrelevant to this
issue, it isn>t even wrong. However, it makes sense to say that the
"program
solves a problem." If some people think there>s a little man inside the
computer that does the thinking, well, that shows that they don>t
understand
what "solving a problem" really is. Even when they do it themselves. Even
when they are playing catch with their kid.

--
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

[/quote]

Glen M. Sizemore · Guest

[quote]GS: Right! And when the homunculus is steering the body to a point at the
end of the trajectory it consults this integral. That is, if it isn>t too
busy checking grammar rule-book so it can cause the guy to yell, "I got
it!"[/quote]

WK: There>s no homunculus in Joe>s description, anymore than there is one in
a
computer programmed to solve differential equations.

GS: Jesus, Wolf. The homunculus is not "in" the "computation," it is "in"
whoever "uses" the "computation." A whole person uses a computation. That>s
precisely why the notion is so fucking stupid. To explain human behavior in
terms of computations and their "use" is to imply a user of the computation.
The term"use" implies a human being. Why don>t you think about that for, for
a change, Wolf. Once again, WHOLE HUMANS use the results of integrating
differential equations.

I do not share your taste for metaphor and tautology.

"Wolf Kirchmeir" <wwolfkir@sympatico.can> wrote in message
news:jbysxveflzcngvpbpna.hkg7082.pminews@news1.sympatico.ca...
[quote]On Sat, 30 Aug 2003 13:07:26 GMT, Glen M. Sizemore wrote:

GS: Right! And when the homunculus is steering the body to a point at the
end of the trajectory it consults this integral. That is, if it isn>t too
busy checking grammar rule-book so it can cause the guy to yell, "I got
it!"

There>s no homunculus in Joe>s description, anymore than there is one in a
computer programmed to solve differential equations.

--
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

[/quote]

Joe Legris · Guest

Glen M. Sizemore wrote:
[quote]GS: Right! And when the homunculus is steering the body to a point at the
end of the trajectory it consults this integral. That is, if it isn>t too
busy checking grammar rule-book so it can cause the guy to yell, "I got

it!"

WK: There>s no homunculus in Joe>s description, anymore than there is one in
a
computer programmed to solve differential equations.

GS: Jesus, Wolf. The homunculus is not "in" the "computation," it is "in"
whoever "uses" the "computation." A whole person uses a computation. That>s
precisely why the notion is so fucking stupid. To explain human behavior in
terms of computations and their "use" is to imply a user of the computation.
The term"use" implies a human being. Why don>t you think about that for, for
a change, Wolf. Once again, WHOLE HUMANS use the results of integrating
differential equations.

I do not share your taste for metaphor and tautology.

[/quote]

Your vitalism is showing again. Why is it that a homunculus inevitably
resides in an information processing description of people but not in
that of computers? Apparently "behaviour" has special qualities that no
machine could ever exhibit.

--
Joe Legris

Wolf Kirchmeir · Guest

On Sat, 30 Aug 2003 20:42:45 GMT, Glen M. Sizemore wrote:

[quote]WK: In an
information processing system, it>s the process itself that>s the model.

GS: And just what does this mean?
[/quote]
Exactly what it says.

And it ain>t metaphor. It>s mathematics.

--
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>

Wolf Kirchmeir · Guest

On Sun, 31 Aug 2003 15:38:48 GMT, Lester Zick wrote:

[quote]I understand what solving a problem is well enough to know the machine
doesn>t do it. There is certainly a little man somewhere around the
machine who designed and wrote the general solution that solves the
specific problem.

[quoting me:]
This is Creationism sneaking in by the back door...

This is an interesting characterization that I don>t quite understand.
By all means if this is what I>m suggesting then let>s get it in the
front door instead of the back.

I don>t think of it that way. Can you explain what creationism has to
do with my remark and why you take it to be so. As far as I can tell
it>s just an accurate description of the current state of the art.
[/quote]
It>s a version of the argument from design: "if there is a programmed
machine, there must be a programmer" <--> "if there is a universe, there must
be a creator." Paley, who is the best known purveyor of this argument, used a
watch as the first element of his analogy. You>re using a programmed machine.
Dawkins titled his book about evolution "The Blind Watchmaker" in reference
to the argument from design, which he rejects.

I reject it, but more because I reject the notion that a verb ("action word)
implies an actor. Whether a verb implies an actor depends entirely on whether
the verb refers to an action, and that' an empirical question, not a
philosophical one. Many verbs don>t refer to actions at all. The verbs we use
to describe the universe almost all refer to processes, ie, changes in state.
We can induce a change in state, but that doesn>t mean we "do" the change in
state.

(Sidebar: The argument with Glen about homunculi arises from his belief that
"compute" is an action, whereas I believe it>s a process. The "computer"
changes state at least once -- in a complex computation it changes state many
times -- and that change in state is the computation. The final state is the
"solution" or "result of the computation." The fact we humans have
symbolised the process, and use the symbolisation to "perform computations"
(and to build machines that perform the computation) doesn>t mean that every
computation requires that there be a person that "performs computations.")

--
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>

Lester Zick · Guest

On Sun, 31 Aug 2003 00:57:48 -0400 (EDT), "Wolf Kirchmeir"
<wwolfkir@sympatico.can> wrote:

[quote]On Sun, 31 Aug 2003 00:38:56 GMT, Lester Zick wrote:

I understand what solving a problem is well enough to know the machine
doesn>t do it. There is certainly a little man somewhere around the
machine who designed and wrote the general solution that solves the
specific problem.

This is Creationism sneaking in by the back door...
[/quote]
This is an interesting characterization that I don>t quite understand.
By all means if this is what I>m suggesting then let>s get it in the
front door instead of the back.

I don>t think of it that way. Can you explain what creationism has to
do with my remark and why you take it to be so. As far as I can tell
it>s just an accurate description of the current state of the art.
[quote]
If the machine doesn>t do it, why build it?

For the same reason you build and use pencil and paper.[/quote]

Regards - Lester

remove DEL in address for email

dan michaels · Guest

Joe Legris <jalegris@xympatico.ca> wrote in message news:<3F4FF727.4080605@xympatico.ca>...

[quote]
If you define a model as a simplified representation of a system then it
is not a copy by definition. End of argument.

Models in the brain are distillations of the aspects of reality that are
relevant to the organism. For example, a brain may contain a ballistic
model that allows one to predict where a thrown rock will land. The
relevant aspect of such a model is the relationship between perceived
heft of an object and it>s expected trajectory when thrown with a
particular effort. It is conceivable that the brain actually solves the
differential equation of motion using a neural network that has been
tuned over years of practice.
[/quote]

Much of what you say above makes sense. However, one surmises that the
brain does not solve differential equations, per se, in the
computational or mathematical sense. Actions are too inaccurate. A
better proposition might be that the case of stone throwing involves a
big adaptive feedback loop - with an external loop [involving the
target] and a number of internal [to the brain and organism] loops.
The key is that feedback loops adjust their outputs by making
comparisons.

All together, the stone-throwing problem involves real-time visual
perception [of the target], tactile input [re stone weight, size,
etc], long-term memory [ie, remembrance of earlier and similar
experiences], short-term memory [ie, successive comparisons], and
requisite motor actions [ie, stone throwing]. Motor actions alone
involve several internal feedback loops, and the other aspects factor
into the larger feedback loop.

No one simply picks up a stone and throws it dead on-target. Even
professional baseball pitchers and football quarterbacks have to warm
up and get their systems essentially "calibrated" every time anew.
What happens is these guys throw the ball, and then use the results of
the toss to adapt their next throws. The major factor is feedback
comparisons stored in STM from one toss to the next. The loop and
motor actions are adapted each time, depending upon whether the
previous toss went long or short, etc. Even so, you never get perfect
repetitions, because the processes involve "estimation" rather than
hard calculation.

So, each stone toss involves real-time visual perception, comparison
to stored experiences, comparison to previous tosses, and active
adaptive adjustments of motor actions. This is not quite the same as
getting a solution using differential equations. Feedback and
comparisons in STM are the key factors.

You can probably make a case that each new action is almost an
open-loop or ballistic action, based upon "internal adjustments"
involving "comparisons" with the short-term memory of the previous
action [as stored in all relevant subsystems in the brain] - more so
than a direct calculation each time. The first throw is more or less a
guess, successive throws are mainly feedback adjustments. In addition,
each new throw is not quite an open-loop action, because you can also
sense how your muscles and joints are moving during the action [due to
the proprioceptive feedback loops, etc], and adjust the action even
while it>s proceeding - eg, finish strong, finish weak, twist, halt,
etc - because all actions take time to complete.

Feedback loops and real-time comparisons. What do you think?


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, 3 Next
Page 2 of 3