LAWS@SRI-AI.ARPA (12/08/84)
From: AIList Moderator Kenneth Laws <AIList-REQUEST@SRI-AI>
AIList Digest Saturday, 8 Dec 1984 Volume 2 : Issue 173
Today's Topics:
Journals - LISP Papers & Computational Intelligence,
Brain Theory - Caenorhabditis Elegans,
Cognition - Infant Amnesia & PBS and The Brain,
Seminars - Speech & Language & Memory & Math Representation (CSLI),
Conference - Intelligent Systems and Machines
----------------------------------------------------------------------
Date: Fri 7 Dec 84 15:16:31-PST
From: Michael Georgeff <georgeff@SRI-AI.ARPA>
Subject: Journals for LISP papers
I wish to submit a paper on a new and efficient method for implementing
funargs in LISP (currently an SRI Tech Note) to an appropriate
journal. Anyone know of any GOOD journal that publishes papers
on programming languages and implementations??
Michael Georgeff.
------------------------------
Date: Thu 6 Dec 84 16:58:08-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Computational Intelligence--New Journal
Computational Intelligence/Intelligence Informatique is a new journal which
will publish in English/French high-quality original theoretical or
experimental research in computational (aritificial) intelligence. The
editors are Nick Cercone/Simon Fraser University and Gordon McCalla/Univer.
of Saskatchewan. Editorial board includes L. Bolc, A. Mackworth, A. Ortony,
R. Perrault, E. Sandewall, A. Sloman, n. Sridharan, D. Wilkins etc.
Subscription rates: U.S. $85 institutional, $47 personal. It will be
a quarterly with the first issue to be available February 1985.
It will be published by the National Research Council of Canada and
sponsored by the Canadian Society for Computational studies of intelligence.
For more information: Distribution, R-88 (Computational Intelligence),
National Research Council of Canada, Ottawa, Ontario, Canada, K1A OR6.
Special rates for members of Canadian Societies. Manuscripts should be
addressed to the editors, Computational Intelligence, Computing Science
Department, Simon Fraser University, Burnaby, British Columbia, Canada,
V5A 1S6.
I will be ordering this title for the Math/CS Library. [...]
Harry Llull
------------------------------
Date: Thu 6 Dec 84 12:01:07-CST
From: ICS.DEKEN@UTEXAS-20.ARPA
Subject: brains, kludges, and elegance
The most substantial evidence of brain kludgery or lack thereof
curiously resides in the structure (completely mapped) of a nematode
named ... "elegant."
Caenorhabditis elegans has 302 neurons of 118 different types, which
make about 8000 synapses in total (each process synapses with about
50% of its neighbors). The lineage of every one of these cells is
known, and the process by which neurological structures are formed may
well seem, to a computer scientist, a kludge. Bilateral symmetry is
not produced, for example, in the "logical" mirror-image development
of a single precursor. The word "kludge," though, carries a pejorative
connotation which seems inappropriate - there are multiple forces and
priorities at work. (One might similarly feel that "kludge" is not
the right word to describe democracy relative to totalitarianism.)
A better word, which may mean something to biologists and others
outside the hacker's ken, might be "fossiliferous," used to describe
any system (program or biological organism) which carries along the
baggage of its own trial-and-error evolution. As Sulston, White,
Thomson, and Schierenberg put it:
"... the perverse assignments, the cell deaths, the long-range
migrations - all the features which could, it seems, be
eliminated from a more efficient design - are so many
developmental fossils."
(There is a three-part series on C. elegans in Science of 22 Jun,
6 Jul, and 13 Jul).
------------------------------
Date: Thu, 6 Dec 1984 01:59 EST
From: MINSKY%MIT-OZ@MIT-MC.ARPA
Subject: Infant Amnesia V2 #165
The general evidence for infant memories is pretty poor. For one
thing, as Ken Laws points out, there is something mighty suspicious
about those handfuls of memories each person claims. In psychiatry
some of these are called "screen memories". A very common feature is
to remember some scene, sort of "eidetically" -- but on questions, the
subject very often sees itself right in the center of the stage!
Since this is impossible, obviously, the conclusion is that the memory
is a construct.
What's worse, with careful questioning, one usually finds that the
memory has indeed been rehearsed, as Ken remarks, perhaps
periodically. Presumably it has been reconstructed in the process,
too -- and can hardly be called a memory, but rather, an elaborated
theory or fantasy.
Finally, even more careful questioning is revealing: how do you know
that this was when you were three years old? Oh, I'm sure of it. It
was the day my dog was run over. An innocent clue like that most
likely points to an incident of Freudian magnitude; a loss or death,
itself rehearsed perhaps for months, and then, unconsciously, for all
the rest of one's life.
In any case it is silly to haggle over the sharpness of the cutoff of
infantile amnesia. I like theories like this: our experience is first
encoded in rather stupid ways; a square is seen as a line attached to
another line attached to another line, etc. Like an early
assembly-language. Later, a square is represented as "closed path of
equal lines" and, later, orthognal pairs of parallels, etc. -- going
to Fortrams to Pascals to LOGs to SMALLTalks to who-knows-what. The
representatins and their interpreters grow more sophisticated, and
those first machine-languages of infancy just can't be always
upwards-compatible. So, even if those early memories were not, in
fact, entirely ever lost, they're doomed to become
unintelligible, eventually.
------------------------------
Date: 1 Dec 1984 20:03:35 EST
From: HARTUNG@USC-ISI.ARPA
Subject: PBS & The Brain
Hello,
I too have been watching the PBS series on the brain. And while I find
it to be remarkably up to date, I do have a concern about it. This is a
concern not just for this series but for many physiological explanations
of experiential phenomena presented to lay audiences. When statements are
made that such and such an area of the brain is responsible for some known
effect, or that damage to location X results in some new and peculiar
observed behavior, these statements are (I fear) taken in a way they are
not meant to be.
The lay audience has a different frame of reference than a psychologist or
neurophysiologist. Scientists studying brain functions view their subjects
as complex models involving the interaction of a variety of known components:
neurotransmiters; ganglia; axon projections; structures, etc. The majority
of the audiance has only limited exposure to these objects and concepts and
not enough time to really develop a similar framework to view all this new
knowledge in. Instead I believe they do what people usually do when under-
standing new material and that is relate it to what they already know. What
people already know is that their brain is responsible for their subjective
awareness of the world. And as a result of the attempt to integrate knowledge
about the brain with the fact that it is the seat of subjective experience
there is a strong possibility that people will believe that these explanations
of brain functioning are in fact explanations of how it is that they have
an experiential component to their lives.
Such physiological explanations will probably never supply the answer to
the question of how it is that we have the kind of experience of things that
we do. For good argument on this point I refer you to Nagel's article in the
Oct. '74 Philsophical Review "What is it like to be a bat?". But, lay
audiences are rarely if ever informed of this.
Another point frequently skipped in presenting brain physiology to lay
audiences is the great importance of subjective experience in the functioning
of cognition. (See Natsoulas, T. Residual Subjectivity. American Psychologist
March 1978.) Indeed subjectivity is so inseparable from cognition that it
raises serious questions about the capacity of digital machines to perform
the full range of human abilities, given that such digital machines may not
be able to achieve a subjective perspective (Searle, J. Minds, Brains, and
Programs. Behavioral and Brain Sciences, Vol. 3, No. 3). Arguments concerning
the mind brain problem have even come to doubt the capacity of present
scientific approaches to the study of mental phenomena and their relationship
to physical phenomena to have any success (Fodor, J.A. Methodological solipsism
considered as a research strategy in cognitive psychology. Behavioral and
Brain Sciences Vol. 3 No. 3).
I assume that the AI-LIST audience is aware of details of these arguments.
The television audience mostly is not. I understand the reluctance of
television producers to include arguments as abstract and difficult as those
on the mind-brain problem. Not to mention the fact that certain religious
groups find them upsetting. However, I feel it is important for us who are
scientists, to encourage people consulting us about presentations they would
make to lay people, to provide the broadest possible context for our arguments,
and always to remember who our audience is and how different their perspective
may be.
Michael A. Moran
Lockheed Advanced Software Laboratory
address HARTUNG@USC-ISI
------------------------------
Date: Wed 5 Dec 84 21:28:17-PST
From: Dikran Karagueuzian <DIKRAN@SU-CSLI.ARPA>
Subject: Seminars - Speech & Language & Memory & Math Representation
(CSLI)
[Excerpted from the CSLI Newsletter by Laws@SRI-AI.]
ABSTRACT OF NEXT WEEK'S SEMINAR
``A Generalized Framework for Speech Recognition''
This talk will describe a framework for speaker-independent,
large-vocabulary and/or continuous speech recognition being developed at
Schlumberger (Fairchild). The framework consists of three components:
1) a finite-state pronunciation network which models relevant
acoustic-phonetic events in the recognition vocabulary;
2) a set of generalized acoustic pattern matchers; and
3) an optimal search strategy based on a dynamic programming algorithm.
The framework is designed to accommodate a variety of (typically disparate)
approaches to the speech recognition problem, including spectral template
matching, acoustic-phonetic feature extraction and lexical pruning based
on broad-category segmentation. A working system developed within this
framework and tailored to the digits vocabulary will also be described. The
system achieves high recognition accuracy on a corpus spoken by
approximately 250 talkers from 22 ``dialect groups'' within the continental
United States.
---Marcia Bush
____________
ABSTRACT OF NEXT WEEK'S COLLOQUIUM
``Data Semantics''
Abstract: There is a growing agreement of opinion that several semantic
phenomena can only be adequately dealt with in a theory which takes
partiality seriously, a theory of partial objects. There is no agreement
about what these partial objects are; for instance, whether they represent
``pieces of the world'' or ``states of partial information about the world.''
Yet, the choice of the perspective determines in large part the potential
of the theory. I will discuss various aspects of Data Semantics, a theory
being developed by Frank Veltman and me, which takes the second
perspective as basic: the semantic behavior of several types of expressions
can best be understood if we take them to relate to our lack of information,
and regard them as patterns on how information can grow. I will argue that
problems concerning quantification and equality force us to distinguish
between different kinds of partial objects.
---Fred Landman
F1 (AND F3) PROJECT MEETING
Title: Self-propagating Search of Memory
Speaker: Pentti Kanerva
Time/Date: Tuesday, December 11, 3:15 p.m.
Place: Ventura Seminar Room
Abstract: Human memory has been compared to a film library that is indexed
by the contents of the film strips stored in it. How might one construct
a computer memory that would allow the computer (a robot) to recognize
patterns and to recall sequences the way humans do? The model presented
is a simple generalization of the conventional random-access memory of a
computer. However, it differs from it in that (1) the address space is very
large (e.g., 1,000-bit addresses), (2) only a small number of physical
locations are needed to realize the memory, (3) a pattern is stored by
adding it into a SET of locations, and (4) a pattern is retrieved by POOLING
the contents of a set of locations. Patterns (e.g., of 1,000 bits) are
stored in the memory (the memory locations are 1,000 bits wide) and they
are also used to address the memory. From such a memory it is possible to
retrieve previously stored patterns by approximate retrieval cues--thus,
the memory is sensitive to similarities. By storing a sequence of patterns
as a linked list, it is possible to index into any part of any "film strip"
and to follow the strip from that point on (recalling a sequence).
____________
AREA C MEETING
Topic: Theories of variable types for mathematical practice,
with computational interpretations
Speaker: Solomon Feferman, Depts. of Mathematics and Philosophy
Time/Date: 1:30-3:30 p.m., Wednesday, December 12
Place: Conference Room, Ventura Hall
Abstract: A new class of formal systems is set up with the following
characteristics:
1) Significant portions of current mathematical practice (such as in
algebra and analysis) can be formalized naturally within them.
2) The systems have standard set-theoretical interpretations.
3) They also have direct computational interpretations, in which all
functions are partial recursive.
4) The proof-theoretical strengths of these systems are surprisingly
weak (e.g. one is of strength Peano arithmetic).
Roughly speaking, these are axiomatic theories of partial functions and
classes. The latter serve as types for elements and functions, but they
may be variable (or "abstract") as well as constant. In addition, an element
may fall under many types ("polymorphism"). Nevertheless, a form of typed
lambda calculus can be set up to define functions.
The result 3) gets around some of the problems that have been met with
the interpretation of the polymorphic lambda calculus in recent literature
on abstract data types. Its proof requires a new generalization of the
First Recursion Theorem, which may have independent interest.
The result 4) is of philosophical interest, since it undermines
arguments for impredicative principles on the grounds of necessity for
mathematics (and, in turn, for physics).
There are simple extensions of these theories, not meeting condition 2),
in which there is a type of all types, so that operations on types appear
simply as special kinds of functions.
NL1 MEETING
Topic: ``Association with Focus''
Speaker: Mats Rooth
Time/Date: 2 p.m., Friday, December 7
Place: Trailer Seminar Room
Note: The content will overlap with but be non-identical to the
presentation the speaker gave in the intonation seminar.
Abstract: In the context of adverbs of quantification, conditionals, and
``only,'' focus can have truth conditional significance. Suppose Mary
introduced Bill and Tom to Sue and performed no other introductions. Then
``Mary only introduced Bill to SUE'' is true, while ``Mary only introduced
BILL to Sue'' is false. Similarly, ``MARY always takes Sue to the movies''
and ``Mary always takes SUE to the movies'' have different truth conditions.
My general claim is that focus influences truth conditions indirectly: the
semantics of the constructions in question involve contextual parameters,
typically unspecified domains of quantification, which are fixed by a
focus-influenced component of meaning. This idea is executed in a Montague
grammar framework.
------------------------------
Date: Fri, 7 Dec 84 10:36:07 EST
From: Morton A Hirschberg <mort@BRL-BMD.ARPA>
Subject: Conference - Intelligent Systems and Machines
CALL FOR PAPERS
1985 Conference on Intelligent Systems and Machines
Dates: April 23-24, 1985
Place: Oakland University
Rochester, Michigan
Technical papers reflecting both advances and applications in all aspects of
intelligent systems and machines will be considered. Suggested topics include,
but are not restricted to:
Intelligent Robotics, Machine Intelligence, C3I, Adaptive Control and
Estimation, Visual Perception and Computer Vision, Pattern Recognition
and Image Processing, Artificial Intelligence for Engineering Design,
Intelligent Simulation Tools, Computer-Integrated Manufacturing Systems,
Knowledge Representation, Expert Systems, Game Theory and Military
Strategy, Interpretation of Multisensor Information, Automatic Message
Understanding, Natural Language and Automatic Programming.
Authors are requested to submit a 300-500 word abstract by January 31, 1985 to:
Professor Nan K. Loh, Conference Chairman,
(313)377-2222
Professor Christian Wagner, Technical Review Committee Chairman
(313)377-2215
Center for Robotics and Advanced Automation
School for Engineering and Computer Science
Oakland University
Rochester, Michigan 48063
The conference will be preceded by tutorials on AI and Robotics held 22 April.
------------------------------
End of AIList Digest
********************bwm@ccice2.UUCP (Brad Miller) (12/10/84)
> Date: Thu, 6 Dec 1984 01:59 EST > From: MINSKY%MIT-OZ@MIT-MC.ARPA > Subject: Infant Amnesia V2 #165 > > ... I like theories like this: our experience is first > encoded in rather stupid ways; a square is seen as a line attached to > another line attached to another line, etc. Like an early > assembly-language. Later, a square is represented as "closed path of > equal lines" and, later, orthognal pairs of parallels, etc. -- going > to Fortrams to Pascals to LOGs to SMALLTalks to who-knows-what. The > representatins and their interpreters grow more sophisticated, and > those first machine-languages of infancy just can't be always > upwards-compatible. So, even if those early memories were not, in > fact, entirely ever lost, they're doomed to become > unintelligible, eventually. This doesn't explain how someone with an edictic or photographic memory can examine a scene after being exposed to it and discover things about it. The theory I like says that memories of scenes or situations are stored as holograms. We simply garbage-collect our earlier memories due to lack of access, that is, we haven't stored them as instances of anything, so there are no pointers to them. Brad Miller -- ...[rochester, cbrma, rlgvax, ritcv]!ccice5!ccice2!bwm
kay@flame.UUCP (Kay Dekker) (12/15/84)
[my, that was a yummy bug-fix!] Brad Miller in <532@ccice2.UUCP> says: >This doesn't explain how someone with an edictic or photographic memory >can examine a scene after being exposed to it and discover things about it. Sorry, but I wasn't aware that eidetic memeory had been shown to exist. Am I out-of-date, people? Kay -- "But what we need to know is, do people want nasally-insertable computers?" ... mcvax!ukc!flame!kay