----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: HGCALLAWAY[…]aol.com
Date: Wed, 4 Dec 2002 08:38:59 EST
X-Message-Number: 1

Creath Thorn wrote:

----quote----------
If the chess-playing machine is completely dyadic, then it exists
perpetually in secondness, a world of strict reaction. How, then, can we
account for our perception that the machine is predictable, does have
habits, does even at times, as Kasparov said, appear to act with
intelligence? If a machine's behavior can only be described in terms of
thirdness (and how else could a satisfactory chess move be so described?),
then how can it be "completely dyadic"?
----end quote-----

In a way, this comment brings us back to the unsatisfactory state of the
distinction between "thirdness" and "secondness," in recent discussions. In a
sense, the ques-tion is "Could a machine, e.g., the chess-playing computer,
come to exemplify thirdness?"

But there is perhaps another interesting way to approach this nest of issues
and questions. We see something of this in the concept of "artificial
intelligence." Is there a such thing as artificial intelligence? Well, it
seems that the study of intel-ligence, human intelligence in particular,
properly belongs to psychology. But does the subject-matter of artificial
intelligence belong to psychology too? Well, if the chess-playing computer
takes up habits, exhibits intelligence, and can also be said to follow
various strategies, perhaps, then, it is starting to look like something that
properly belongs to the subject-matter of psychology. Again, psychologists
might attempt to understand properly human intelligence by means of the
attempt to simulate it with programs and computers. If such an approach
actually casts some light on human intelligence, then the ultimate results of
such research, including its artifacts, might come to belong to the proper
subject-matter of psychology.

All of this is to suggest that the question of whether there is a single
subject-matter including both natural and artificial intelligence is
something that depends on the concrete results of inquiry and research in
particular specified fields, and if so, then it is not a kind of question we
could hope to answer a priori or independent of those results and possible
results. By the same token, what is to count as genuine thirdness remains
unclear in many cases, and depends on the results of inquiry to provide less
indefinite accounts of thirdness. Right?

Howard

H.G. Callaway
(hgcallaway[…]aol.com)

----------------------------------------------------------------------

Subject: Re: Hermeneutic Equivalence Classes
From: Jon Awbrey <jawbrey[…]oakland.edu>
Date: Wed, 04 Dec 2002 10:20:27 -0500
X-Message-Number: 2

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

HEC. Note 11

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

| Leibniz, "Elements of a Calculus" (cont.)
|
| 16. Hence we can also know by symbolic numbers which term does not
| contain another; for we have only to test whether the number
| of the latter can divide exactly the number of the former.
|
| For example, if the symbolic number of man is assumed to be 6, and
| that of ape to be 10, it is evident that neither does the concept
| of ape contain the concept of man, nor does the converse hold,
| since 10 cannot be exactly divided by 6, nor 6 by 10.
|
| If, therefore, it is asked whether the concept of the wise man is
| contained in the concept of the just man, i.e. if nothing more is
| required for wisdom than what is already contained in justice, we
| have only to examine whether the symbolic number of the just man
| can be exactly divided by the symbolic number of the wise man.
| If the division cannot be made, it is evident that something
| else is required for wisdom which is not required in the just
| man. (This "something else" is a knowledge of reasons; for
| someone can be just by custom or habit, even if he cannot give
| a reason for the things he does.) I will state later how this
| minimum which is still required, or, is to be supplied, can also
| be found by symbolic numbers.
|
| Leibniz, 'Logical Papers', p. 22.
|
| Leibniz, G.W., "Elements of a Calculus" (April, 1679),
| G.H.R. Parkinson (ed.), 'Leibniz: Logical Papers', pp. 17-24,
| Oxford University Press, London, UK, 1966. (Couturat, 49-57).

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: Peter Skagestad <Peter_Skagestad[…]uml.edu>
Date: Wed, 04 Dec 2002 11:58:48 -0500
X-Message-Number: 3

> Joe & List,

I have been meaning to address the intriguing question Joe raised in his
paper as to the origin of the formal distinction between AI and IA. I
certainly never thought I had originated the distinction. I more or less
implicitly assumed that I had found it in Engelbart's writings where it
is, as Joe notes, implicit from 1962 on, but perhaps not made explicit.
In 1960 J.C.R. Licklider distinguished AI from what he called
"Man-computer symbiosis", which he himself certainly regarded as
complementary projects - briefly, Licklider embraced AI as a long-term
objective while advocating symbiosis as an intermediate strategy. But
this is still not exactly intelligence augmentation; for a while, it
seems, Licklider and Engelbart regarded the two programs as identical,
but they gradually came to see the relationship between the two as
problematic; in the simplest terms, Lickliuder's program prioritized
user-friendliness, which was never a priority for Engelbart.

By the way, I should acknowledge that Joe is right that I would accept
that AI and IA, considered simply as research projects, are
complementary, rather than inherently rival. (Of course rivalry always
arises when funding priorities are to be set. For instance, user
interface development (IA) has benefitted enormousy from the development
of optical character recognition, image processing, and speech
processing and recognition, developmentsd which were originally
motivated by the AI project. But Joe is also right that my own emphasis
has been on the distinctness of IA, rather than on the complementarity.

But who spelled out the distinction in explicit terms? The credit,
apparently, belongs to the computer scientist Frederick Brooks at UNC
Chapel Hill. Howard Rheingold, in his book Virtual Reality (1991, p.
37), quotes Brooks as follows:

"I believe the use of computer systems for intelligence amplification is
much more powerful today, and will be at any given point in the future,
than the use of computers for artificial intelligence (AI).... In the AI
community the objective is to replace the human mind by the machine and
its program and its database. In the IA community, the objective is to
build systems that amplify the human mind by providing it with
computer-based auxiliaries that do the things that the mind has trouble
doing." So far Brooks.

As a matter of personal biography, I was originally sensitized to the
distinction through my professional involvement with machine translation
(MT), an offshoot of AI research, and computer-aided translation (CAT),
my favorite example of IA, although everything I have actually written
about it so far has ended up in my wastebasket. ("For what I have
published I can only hope to be forgiven; for what I have burned I
deserve to be praised." Alexander Pope) Let me just say that the most
popular CAT tool is the Translation Memory Tool where the translator
works with an electronic source file and, as s/he translates, the
program builds up a translation memory which matches each
source-language sentence with the corresponding target-language
sentence. Then, if the translator encounters a sentence already
translated, the program will automatically retrieve the translation from
its memory and let the translator accept it or modify it. This is
labor-saving when translating e.g. user manuals, which can be extremely
repetitive, and it also helps the trtanslator maintain consistency of
terminology. But the chief value of memory tools is what is calle "fuzzy
matching": if a sentence comes up that matches a previously translated
sentince 85% or more, the program will flag it as a fuzzy match and
bring up the previous translation. These programs are enormously popular
in the translator community, where MT is almost universally detested.

As a manager in a translation company, I frequently need to prepare bids
on contracts and for that purpose determine the most cost-effective way
of carrying out a project. Sometimes the best way is with a memory tool,
especially if there is a lot of repetition, sometimes it will be MT with
human post-editing, if there is very little repetition, but simple
language, limited terminology, etc., and the client is not especially
concerned with style. ("Post-editing, by the way, may sound redundant,
but it is a term of art, contrasted with the "pre-editing" of source
texts to make them easy for the MT program to process.) I have gone into
this to illustrate that one may recognize the value of both AI and IA
approaches to the same problems, while nonetheless having to choose
between them in particular cases.

But I have gone on for long enough.

Cheers,
Peter




----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: Kenneth Ketner <b9oky[…]TTACS.TTU.EDU>
Date: Wed, 04 Dec 2002 11:13:21 -0600
X-Message-Number: 4

I wouldn't say that contemporary computers which only run dyadic
programs manipulate signs; even 'manipulate' is anthropomorphic
ascription. They just run programs. If you or I see some things
happening on a computer - say pixels on a screen - and these have
potential for participation in the triadic sign relation (which is the
root meaning of 'sign'), which some interpretant can also participate in
on a suitable occasion, then a sign process (semeiosis) can involve some
aspect of a computer activity (dynamic action, or dynamis ??is that a
proper greco-roman ending??) - an example is someone reading these
pixels (a dynamis of computers which are running programs). But pencils,
paper, rocks, sand, tree bark, and so on also can function in the same
manner.

.
--
Kenneth L. Ketner
Paul Whitfield Horn Professor
Institute for Studies in Pragmaticism
Texas Tech University
Charles Sanders Peirce Interdisciplinary Professor
School of Nursing
Texas Tech University Health Sciences Center
Lubbock, TX 79409-0002
806 742 3128
Office email: b9oky[…]ttacs.ttu.edu
Home email: ketner[…]arisbeassociates.com
Office website: http://www.pragmaticism.net
Personal website: http://www.wyttynys.net




----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: Kenneth Ketner <b9oky[…]TTACS.TTU.EDU>
Date: Wed, 04 Dec 2002 11:17:46 -0600
X-Message-Number: 5

Jon, could you give me a reference on machines with irreducibly
unbounded memories?

Jon Awbrey wrote:

>o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
>
>KK = Kenneth Ketner
>
>KK: Machines (by which is meant devices that run algorithms,
> Universal Turing machines, contemporary publicly purchasable
> computers) function in an exclusively dyadic manner by means
> of chains of dyadic relations. Source: many, but see
> "Peirce and Turing" in SEMIOTICA 68-1/2 (1988), 33-61.
>
>This is not so. One can say that finite state machines are 2-adic,
>but machines with irreducibly unbounded memories are not like this,
>and they have to be considered as operating in the medium of 3-adic
>relations, proceeding from at least two independent arguments, say,
>(1) the unpredetermined external input, and (2) the finite plus the
>infinite parts of the internal state, to generate (3) a contingently
>terminating intermediary result at each step. The difference between
>what finite state machines can do and higher classes of automata like
>turing machines can do is basically what the debate between Chomsky
>and Skinner was all about.
>
>Jon Awbrey
>
>o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
>
>---
>Message from peirce-l forum to subscriber b9oky[…]ttacs.ttu.edu
>To unsubscribe send a blank email to: leave-peirce-l-3181N[…]lyris.ttu.edu
>
>
>

--
Kenneth L. Ketner
Paul Whitfield Horn Professor
Institute for Studies in Pragmaticism
Texas Tech University
Charles Sanders Peirce Interdisciplinary Professor
School of Nursing
Texas Tech University Health Sciences Center
Lubbock, TX 79409-0002
806 742 3128
Office email: b9oky[…]ttacs.ttu.edu
Home email: ketner[…]arisbeassociates.com
Office website: http://www.pragmaticism.net
Personal website: http://www.wyttynys.net




----------------------------------------------------------------------

Subject: Re: Thinking with a Manichaean Bent?
From: Charles F Rudder <cf_rudder[…]juno.com>
Date: Wed, 4 Dec 2002 11:35:21 -0600
X-Message-Number: 6



HOWARD, JOE, WILLIAM, LIST

I have not followed developments in AI or cognitive science in several
years and had not seen the AI -- IA distinction before reading Joe's
manuscript. Apropos of thinking with a Manichaen bent and less directly
mechanical augmentations of intelligence, does anything I say in the
attached manuscript that I wrote three or four years ago as part of a
correspondence with a friend and colleague make any sense to anyone? I
acknowledge the "arm chair" economics.

Charles

----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: Jon Awbrey <jawbrey[…]oakland.edu>
Date: Wed, 04 Dec 2002 13:22:53 -0500
X-Message-Number: 7

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

PSI. Note 21

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

KK: Jon, could you give me a reference on machines
with irreducibly unbounded memories?

Ken,

Here is a selection of good references on
computability theory and related subjects:

| Arbib, M.A.,
|'Brains, Machines, Mathematics', 1st ed. 1964,
| 2nd ed., Springer-Verlag, New York, NY, 1987.

| Boolos, G.S. & Jeffrey, R.C.,
|'Computability & Logic', 2nd ed.,
| Cambridge University Press, Cambridge, UK, 1980.

| Cutland, Nigel J.,
|'Computability: An Introduction to Recursive Function Theory',
| Cambridge University Press, Cambridge, UK, 1980.

| Davis, M.,
|'Computability & Unsolvability', 1st ed. 1958,
| enlarged ed., Dover Pubs., New York, NY, 1982.

| Denning, P.J., Dennis, J.B., Qualitz, J.E.,
|'Machines, Languages, and Computation',
| Prentice-Hall, Englewood Cliffs, NJ, 1978.

| Rogers, Hartley, Jr.,
|'Theory of Recursive Functions and Effective Computability',
| McGraw-Hill, New York, NY, 1967.

Jon

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: Kenneth Ketner <b9oky[…]TTACS.TTU.EDU>
Date: Wed, 04 Dec 2002 13:49:09 -0600
X-Message-Number: 8

Which of those would be the one on machines with irreducibly unbounded
memories?


--
Kenneth L. Ketner
Paul Whitfield Horn Professor
Institute for Studies in Pragmaticism
Texas Tech University
Charles Sanders Peirce Interdisciplinary Professor
School of Nursing
Texas Tech University Health Sciences Center
Lubbock, TX 79409-0002
806 742 3128
Office email: b9oky[…]ttacs.ttu.edu
Home email: ketner[…]arisbeassociates.com
Office website: http://www.pragmaticism.net
Personal website: http://www.wyttynys.net




----------------------------------------------------------------------

Subject: Re: Hermeneutic Equivalence Classes
From: Jon Awbrey <jawbrey[…]oakland.edu>
Date: Wed, 04 Dec 2002 14:52:12 -0500
X-Message-Number: 9

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

HEC. Note 12

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

| Leibniz, "Elements of a Calculus" (cont.)
|
| 17. From this, therefore, we can know whether some
| universal affirmative proposition is true. For
| in this proposition the concept of the subject,
| taken absolutely and indefinitely, and in general
| regarded in itself, always contains the concept of
| the predicate.
|
| For example, all gold is metal; that is, the concept of metal is
| contained in the general concept of gold regarded in itself, so that
| whatever is assumed to be gold is by that very fact assumed to be metal.
| This is because all the requisites of metal (such as being homogeneous
| to the senses, liquid when fire is applied in a certain degree, and then
| not wetting things of another genus immersed in it) are contained in the
| requisites of gold, as we explained at length in article 7 above. So if
| we want to know whether all gold is metal (for it can be doubted whether,
| for example, fulminating gold is still a metal, since it is in the form of
| a powder and explodes rather than liquefies when fire is applied to it in
| a certain degree) we shall only investigate whether the definition of metal
| is in it. That is, by a very simple procedure (once we have our symbolic
| numbers) we shall investigate whether the symbolic number of gold can be
| divided by the symbolic number of metal.
|
| Leibniz, 'Logical Papers', pp. 22-23.
|
| Leibniz, G.W., "Elements of a Calculus" (April, 1679),
| G.H.R. Parkinson (ed.), 'Leibniz: Logical Papers', pp. 17-24,
| Oxford University Press, London, UK, 1966. (Couturat, 49-57).

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

----------------------------------------------------------------------

Subject: Re: Peircean Semiotic & Intelligence Augmentation
From: Jon Awbrey <jawbrey[…]oakland.edu>
Date: Wed, 04 Dec 2002 15:26:22 -0500
X-Message-Number: 10

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

PSI. Note 22

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

KK: Which of those would be the one on machines
with irreducibly unbounded memories?

Any machine that is not finite state is of this order.
Working within the Chomsky-Schutzenberger hierarchy as
a standard of comparison, we have the following orders:

1. Finite State Automata & Regular Languages.
2. Push-Down Automata & Context-Free Languages.
3. Linear-Bounded Automata & Context-Sensitive Languages.
4. Turing Automata & General Recursive Languages.

Anything properly higher order than FSA's would
be said to have in principle unbounded memories.

Perhaps you are tripping up on the word "irreducibly",
which I use here merely because the word "ineliminably"
trips up my tongue, even when speaking purely subvocally,
and I judged that the mathematical usage of "essentially"
might cause more trouble in this context than it's worth.

My rough intuition is that anything with some kind of ineliminable indeterminacy
or irreducible infinity about it has probably got some "thirdness" in the works.
It is not for naught that Peirce used the trefoil knot as a symbol of infinity.
But these are not proofs -- our intuitions tend to desert us when we get beyond
the levels of the finite state and the primitive recursive -- and if we want to
do anything more than sit around on our rules of thumb, we will have to put our
personal guesses and our popular sentiments aside, and see what we can see by
undertaking the rigors of proof.

Jon Awbrey

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

----------------------------------------------------------------------

Subject: Re: Hermeneutic Equivalence Classes
From: Jon Awbrey <jawbrey[…]oakland.edu>
Date: Wed, 04 Dec 2002 17:30:04 -0500
X-Message-Number: 11

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

HEC. Note 13

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

| Leibniz, "Elements of a Calculus" (cont.)
|
| 18. But in the particular affirmative proposition it is not necessary that
| the predicate should be in the subject regarded in itself and absolutely;
| i.e. that the concept of the subject should in itself contain the concept
| of the predicate; it is enough that the predicate should be contained in
| some species of the subject, i.e. that 'the concept of some instance or
| species of the subject should contain the concept of the predicate',
| even though it is not stated expressly what the species is.
|
| Consequently, if you say, "Some experienced man is prudent", it is
| not said that the concept of the prudent man is contained in the
| concept of the experienced man regarded in itself. Nor, again,
| is this denied; it is enough for our purpose that some species
| of experienced man has a concept which contains the concept of
| the prudent man, even though it is not stated expressly just
| what that species is. For even if it is not said expressly
| here that the experienced man is a prudent man who also has
| natural judgement, it is enough that it is understood that
| some species of experienced man involves prudence.
|
| Leibniz, 'Logical Papers', p. 23.
|
| Leibniz, G.W., "Elements of a Calculus" (April, 1679),
| G.H.R. Parkinson (ed.), 'Leibniz: Logical Papers', pp. 17-24,
| Oxford University Press, London, UK, 1966. (Couturat, 49-57).

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

----------------------------------------------------------------------

Subject: Dating Service
From: Jon Awbrey <jawbrey[…]oakland.edu>
Date: Wed, 04 Dec 2002 23:16:26 -0500
X-Message-Number: 12

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

| The purpose of this web project is to disassemble
| the Collected Papers of Charles Saunders Peirce
| [edited by Charles Hartshorne and Paul Weiss,
| Cambridge: Harvard University Press, 1931-1958]
| and to reassemble them in chronological order.

http://www.princeton.edu/~batke/peirce/

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o



---

END OF DIGEST 12-04-02

.