[Prev][Next][Index][Thread]
CSLI Calendar, 22 October 1997, vol. 13:6
C S L I C A L E N D A R O F P U B L I C E V E N T S
______________________________________________________________________
22 October 1997 Stanford Vol. 13, No. 6
______________________________________________________________________
A weekly publication of the
Center for the Study of Language and Information (CSLI)
Stanford University, Ventura Hall, Stanford, CA 94305-4115
____________
ACTIVITIES DURING 22 OCTOBER - 31 OCTOBER 1997
WEDNESDAY, 22 OCTOBER
12:15pm Developmental Brownbag
Jordan Hall 420:286
Infants' Responses to Emotional Signals at the End of the
First Year
Donna Mumme
Stanford University
4:15pm EE380: Computer Systems Laboratory Colloquium
Gates B03 (NEC Auditorium)
Java SmartCards
Eduard De Jong
Integrity Arts
Abstract below
THURSDAY, 23 OCTOBER
12 noon CSLI CogLunch
Cordura Hall, Room 100
Lexical Knowledge and Semantic Heterogeneity
Geoff Nunberg
Xerox PARC/Stanford Linguistics
Abstract below
4:00pm Xerox PARC Forum
George Pake Auditorium, Xerox PARC
Machine Translation and Human Translation
Douglas Hofstadter
Indiana University
Abstract below
4:15pm Seminar on Computational Learning and Adaptation (SCLA)
Gates 100
Bump Hunting in High-Dimensional Data
Jerome Friedman
Department of Statistics, Stanford.
Abstract below
7:30pm Phonology Workshop
Margaret Jacks Hall 460:146
The Cyclic Effect of Suffix Ordering in the Bantu Verb
Stem (with special reference to applicativized
causatives)
Larry Hyman
Department of Linguistics, Berkeley
Abstract below
FRIDAY, 24 OCTOBER
12 noon Logic Lunch
Room 380:383N
"Reducing" Mathematical Objects and Relations to Abstract
Objects and Relations
Edward N. Zalta
CSLI/Philosophy
Abstract below
12:30pm Seminar on People, Computers, and Design
Gates B01 (HP classroom)
Supporting Cooperative and Personal Surfing with a
Desktop Assistant
Hannes Marais and Krishna Bharat
DEC Systems Research Center
Abstract below
3:15pm Cognitive Seminar
Jordan Hall 420:100
A Test of Conditional Independence and Inter-rater
Agreement
Ewart Thomas
Stanford Psychology
3:15pm Philosophy Department Colloquium
Philosophy 90:92Q
Semantic Inferentialism and Logical Expressivism
Robert Brandom
University of Pittsburgh
WEDNESDAY, 29 OCTOBER
12:15pm Developmental Brownbag
Jordan Hall 420:286
Infant Speech Processing
John Pinto
Stanford University
4:15pm Computer Musings
Gates B01
35 Years of (Linear) Probing
Donald Knuth
Computer Science
Abstract below
4:15pm EE380: Computer Systems Laboratory Colloquium
Gates B03 (NEC Auditorium)
BDTI Processors for DSP-intensive Applications
Jeff Bier
Berkeley Design Technology Inc.
[http://www.stanford.edu/class/ee380/contents.html]
THURSDAY, 30 OCTOBER
12 noon CSLI CogLunch
Cordura Hall, Room 100
Two Models of Dialogue: What's the Score?
Phillip Staines
University of New South Wales
Abstract below
4:00pm Xerox PARC Forum
George Pake Auditorium, Xerox PARC
Failing and Succeeding at Real-World Artificial
Intelligence: Experiences in Three Decades
Peter Hart
Ricoh Silicon Valley, Inc.
FRIDAY, 31 OCTOBER
12:30pm Seminar on People, Computers, and Design
Gates B01 (HP classroom)
Video Rewrite and the Magic Morphin' Mirror.
Trevor Darrell, Chris Bregler, Michele Covell
and Malcolm Slaney
Interval Research
3:15pm Cognitive Seminar
Jordan Hall 420:100
Decision Making in the N-Person Prisoners' Dilemma
Paul Whitmore
3:15pm Philosophy Department Colloquium
Bldg. 90:92Q
"Physician-Assisted Suicide and Public Policy"
Gerald Dworkin
UC Davis
____________
ANNOUNCEMENT
I'm moving the mailing lists that I use for sending the short and the
long form of the calendar to a listserver. New rules for
unsubscribing and subscribing should appear at the end of this
message. I expect a few glitches this first time so please bear with
me, but, I hope in the long run that this will make managing the
mailing lists much easier.
____________
EE380 COMPUTER SYSTEMS COLLOQUIUM
on Wednesday, 22 October 1997, 4:15pm to 5:30pm
NEC Auditorium (B03), Gates Computer Science Building
[http://www.stanford.edu/class/ee380/]
Programming SmartCards with JAVA
Eduard de Jong
Chief Scientist
Integrity Arts
The traditional approach to programming for smart cards does not allow
the creation of loadable executable code and requires programmers with
experience in programming in Assembly. This does not allow card
manufacturers to quickly respond to changes in the market and limits
the flexibility of what the smart card provides to end-users. Object
Oriented programming provides a more dynamic approach to card
applications. Objects give good support for security both for the
programmer and run-time implementation. Java is an excellent object
oriented development environment for programmers to create
applications that are used in smart cards
This Seminar will examine the following topics:
* The Next Generation of Smart Cards: Technology Trends
* Overview of Object-Oriented Programming
* Comparison of Java and Java in a Card
* Card Software Architecture
About the speaker: Born in Amsterdam, Eduard de Jong is Chief
Scientist and co-founder of Integrity Arts. Dr. de Jong directs the
research and development of Java-based smart card applications. His
professional interests include computer security and architecture,
hardware design and system implementation, functional programming, and
the impact of technology upon society. Prior to his current position,
he was co-founder, co-director and Chief Technical Officer of QC
Technology BV in Zaandam, The Netherlands. He has also done
breakthrough work on electronic purse standardization and digital
cash.
A short list of organizations that have published Dr. de Jong's work
includes Future Generation Computer Systems, CARDIS 1996, CARDIS 1994,
UNIFORUM '96, and the University of Amsterdam. His educational
background includes a Ph.D. in Computer Science and a Master of
Science in Physics, both from from the University of Amsterdam in The
Netherlands.
Integrity Arts, a recent acquisition of Sun Microsystems and based in
San Mateo, Calif., was founded in early 1995 to bring advanced
software development concepts, tools, and methods to the smart card
industry. The company develops programming tools as well as object and
source code modules to create smart card and terminal operating
systems and applications. All IA products use an open architecture,
modular approach to software development.
____________
CSLI COGLUNCH
on Thursday, 23 October 1997, 12 noon
Cordura Hall, Room 100
[http://www-csli.stanford.edu/csli/Coglunch/]
Lexical Knowledge and Semantic Heterogeneity
Geoff Nunberg
Xerox PARC and Stanford Linguistics
Modern linguistic theory has been committed to a kind of solipsistic
individualism, where we take notions like "grammar" and the like
purely as the constructions of an individual mind confronted with a
set of sentences, without giving any consideration to what other
people might have in mind. What makes this maneuver possible is an
idealization to an understanding of the object of linguistic theory as
the individual representation of the language of a homogeneous
speech-community whose rules are matters of perfect knowledge to all
its members. But this idealization carries various costs --
particularly, as is usually the case, when we fail to acknowledge it
as such: we may be tempted to misinterpret the phenomenal reflexes of
linguistic heterogeneity as the properties of individual
representations of the language. The problems are particularly acute
in connection with knowledge of lexical meanings, where heterogeneity
is an inevitable consequence of social differentiation.
A good example of this displacement involves the distinction between
lexical and encyclopedic information. I'll suggest that we can make
this distinction intelligible only if we interpret it in the light of
the problem of coordinating linguistic practices in a heterogeneous
community, taking "the language" as something out there, rather than
in here.
____________
XEROX PARC FORUM
on Thursday, 23 October 1997, 4:00pm - 5:00pm
George Pake Auditorium, Xerox
[http://www.parc.xerox.com/forum]
Machine Translation and Human Translation
Douglas Hofstadter
Center for Research on Concepts and Cognition
Indiana University, Bloomington
Using a short poem by Clement Marot, a French poet of the sixteenth
century, I'll look at various manners of translating, including
content-only, form-only, and content-plus-form in various mixtures,
and compare the kinds of results that they yield. I'll also consider
Searle's "Chinese Room" thought experiment in the context of
translation.
Biography: Douglas Hofstadter is College Professor of computer science
and cognitive science, director of the Center for Research on Concepts
and Cognition, and adjunct professor of philosophy, psychology,
comparative literature, and the history and philosophy of science at
Indiana University. His Pulitzer-prize-winning book "Godel, Escher,
Bach: an Eternal Golden Braid" (1979) has had considerable impact on
people in many disciplines, ranging from philosophy to mathematics to
artificial intelligence to music, and beyond. He has written four
other books, including the recent "Le Ton Beau de Marot", as well as
numerous articles, and for a number of years wrote a column for
Scientific American. He is spending the current academic year at the
Stanford Humanities Center.
Suggested reading: "Le Ton beau de Marot", Douglas Hofstadter, 1997.
Refreshments will be served from 3:45 to 4:00.
GOLDEN EAR AWARD: The forum will begin promptly at 4:00. This week,
the forum will feature a Thematic Hofstadter Medley of four different
pieces, played by Richard Kade, each with connections to "Godel,
Escher, Bach" or "Le Ton beau de Marot". This "pre-program" will be
10 minutes long, and it will begin in the Auditorium at 3:50. The
person who identifies the most compositions will be presented with the
Golden Ear Award at the end of the forum.
____________
SEMINAR ON COMPUTATIONAL LEARNING AND ADAPTATION (SCLA)
on Thursday, 23 October 1997, 4:15pm to 5:30pm
Gates 100
[http://www-csli.stanford.edu/cll/scla.html]
Bump Hunting in High-Dimensional Data
Jerome H. Friedman
Department of Statistics
Stanford University
mailto:jhf@stat.stanford.edu
Many data analytic questions can be formulated as (noisy) optimization
problems. They explicitly or implicitly involve finding simultaneous
combinations of values for a set of ("input") variables that imply
unusually large (or small) values of another designated ("output")
variable. Specifically, one seeks a set of subregions of the input
variable space within which the value of the output variable is
considerably larger (or smaller) than its average value over the
entire input domain. In addition it is usually desired that these
regions be describable in an interpretable form involving simple
statements ("rules") concerning the input values. This talk describes
a new procedure directed towards this goal based on the notion of
"patient" rule induction. This patient strategy is contrasted with the
greedy ones used by most rule induction methods, and semi-greedy ones
used by some partitioning tree techniques such as CART. Applications
involving scientific and commercial data bases are presented.
____________
STANFORD PHONOLOGY WORKSHOP
on Thursday, 23 October 1997, 7:30pm
Margaret Jacks Hall 460:146
The Cyclic Effect of Suffix Ordering in the Bantu Verb Stem
(with special reference to applicativized causatives)
Larry Hyman
Department of Linguistics
University of California, Berkeley
In this paper I deal with some problems in phonological cyclicity that
result directly from surface suffix orderings that are conditioned
strictly by morphotactic constraints (often in violation of
compositionality/the mirror principle). I argue that cyclicity,
contrary to common assumption, need not be motivated by the tendency
for morphologically complex forms to reflect phonological properties
of corresponding base forms. As evidence I cite Bantu cases of
cyclicity in which the effect of the cycle is to obscure the
similarity among root allomorphs of related words. The Bantu case
involves a pervasive conflict arising between the applicative -il- and
causative -i- suffixes: Both a causativized applicative
[ [ [ root ] appl ] caus ]
and an applicativized causative
[ [ [ root ] caus ] appl ]
can only be realized in the order -il-i-. Since the caus-appl
structure is more frequent, it is constantly contradicted by the
surface linear order appl-caus of the morphs. In Hyman (1994),
semantic, morphological and phonological arguments were presented that
the sequence should be derived cyclically via "interfixation":
root --> root-i- --> root-il-i-.
Languages like Bemba exhibit cyclic "frication" conditioned by -i-,
e.g. Bemba luf-is-i- 'lose for/at' (from lub-il-i-). This paper
expands the earlier work to consider languages which "undo" inner
frication with a "replacive" C, e.g. /k/ in Matumbi, Yao. I argue that
the unifying property between the two resolutions of the Bantu
conflict is the transparency of derivation as one goes from CVC- to
CVC-i- to CVC-ili-, not "correspondence".
The organizers are updating the pinterest email list; please let them
know if you wish to be added to this list.
Hee-Sun Kim mailto:heesun@csli.stanford.edu
or
Shiao-Wei Tham mailto:tham@csli.stanford.edu
____________
LOGIC LUNCH
on Friday, 24 October 1997, 12 noon
Math Corner 380:383N
"Reducing" Mathematical Objects and Relations to
Abstract Objects and Relations
Edward N. Zalta
[http://mally.stanford.edu/zalta.html]
CSLI/Philosophy Department, Stanford University
In this talk I describe a method of interpreting an arbitrary
mathematical theory T in such a way that the constants of T denote
abstract objects and the predicates of T denote abstract relations.
The instrument of this "Platonist" reduction is the theory of abstract
objects and relations (described in my previous talk at the Logic
Lunch series this quarter). Once this theory is defined, more
familiar and specific (kinds of) abstract objects (such as possible
worlds, times, concepts, monads, fictions, sets, numbers, etc.) and
abstract relations (such as membership, successor, etc.) become
"reducible" to the entities axiomatized by the theory. However, there
are two distinct kinds of reduction: classical and "philosophical".
In the classical reductions, the basic principles of a target theory T
become derivable simply by adding definitions of the primitive notions
of T to the axiomatic theory of abstract objects. However, theories
stronger than Peano Number Theory are not reducible in this way.
Instead, arbitrary mathematical theories are subject to
"philosophical" reduction---the object and relation terms of
mathematical theory T become uniquely interpreted as abstract objects
and relations, respectively, once certain accepted (constitutive)
facts about T have been added to the background theory.
____________
SEMINAR ON PEOPLE, COMPUTER, AND DESIGN
on Friday, 24 October 1997, 12:30-2:00pm
Gates B01 (HP Classroom)
[http://www-pcd.stanford.edu/seminar/]
(SITN Channel E2)
Supporting Cooperative and
Personal Surfing with a Desktop Assistant
Hannes Marais and Krishna Bharat
DEC Systems Research Center
[http://www.research.digital.com/SRC/personal/Johannes_Marais/home.html]
[http://www.research.digital.com/SRC/staff/bharat/bio.html]
We present the use of desktop assistants in the context of web surfing
and show how such a tool may be used to support activities in both
cooperative and personal surfing. By cooperative surfing we mean
surfing by a community of users who choose to cooperatively and
asynchronously build up knowledge structures relevant to their
group. Specifically, we describe the design of an assistant called
Vistabar, which lives on the Windows desktop and operates on the
currently active web browser. Vistabar instances working for
individual users support the authoring of annotations and shared
bookmark hierarchies, and work with community interest profiles to
make findings highly available. Thus, they support a form of community
memory. Vistabar also serves as a form of personal memory by indexing
pages the user sees to assist in recall. We present rationale for the
assistant's design, show some roles it could play to support surfing
and discuss implementation strategies and ongoing extensions.
Biographies: Hannes Marais has been a member of the research staff at
Digital's Systems Research Center in Palo Alto since April 1996. His
background is in user interfaces with a recent focus on the web.
Krishna Bharat is a member of the research staff at Digital's Systems
Research Center in Palo Alto. He received his PhD at the Graphics,
Visualization, and Usability Center of the College of Computing at the
Georgia Institute of Technology last year with a thesis topic of
"Supporting the Construction of Distributed, Interoperative, User
Interface Applications".
____________
COMPUTER MUSINGS
on Wednesday, 29 October 1997, 4:15pm
Gates B01 (HP Classroom)
[http://www-cs-faculty.stanford.edu/~knuth/musings.html]
35 Years of (Linear) Probing
Donald Knuth
Many computer programs access large tables of data by using a
classical variant of hashing called "linear probing," also known to
children as the game of "musical chairs." When the author first
studied the characteristics of this simple method in 1962, he came to
understand that the mathematical analysis of algorithms was a rich
subject suitable for a lifetime of study. This talk surveys the
mathematical analysis of algorithms by focusing on advances that have
been made in the analysis of linear probing from 1962 to 1997, and by
noting surprising relations between this algorithm and other important
algorithms and combinatorial problems, including the study of random
graphs.
____________
CSLI COGLUNCH
on Thursday, 30 October 1997, 12 noon
Cordura Hall, Room 100
[http://www-csli.stanford.edu/csli/Coglunch/]
Two Models of Dialogue: What's the Score?
Phillip Staines
University of New South Wales
[http://www.arts.unsw.edu.au/philosophy/pstaines.htm]
In a recent important work, "Making it Explicit", Robert Brandom sets
up a model of dialogue within a deontic pragmatics. As he says "A
crucial measure according to which a theory of speech acts ought to be
assessed is its treatment of what one is doing in producing an
assertion." In this book he takes the effect of assertion to be
changing the conversational score by altering participants'
commitments and entitlements. This paper compares Brandom's approach
with work in (Formal) Dialectic which also takes commitment as its
central notion, but has no explicit role for entitlement.
____________
END MATERIAL
The CSLI Calendar appears weekly on Wednesdays throughout the academic
year. Announcements, abstracts, and other information to appear in
the Calendar should be submitted to the editor, who reserves the right
to decide what does or does not go in the calendar
[mailto:incalendar@csli.stanford.edu].
Requests to be added to the mailing list should be sent to
[mailto:majordomo@csli.stanford.edu]. With the lines in the body of
the text of either
subscribe csli-calendar
for the long form or
subscribe csli-short-calendar
for the short form. Problems with subscribing or unsubscribing should
be sent to owner-csli-calendar@csli.stanford.edu.
The full current issue is at
[http://www-csli.stanford.edu/csli/Archive/calendar/1997-98/current.html]
and the archives at
[http://www-csli.stanford.edu/csli/Archive/calendar/].
People on most of the CSLI computers can type 'help csli-calendar' to
see the current issue. They can also type 'help quip' to see the
Linguistics Department newsletter, the Sesquipedalian,
[http://www-linguistics.stanford.edu/Linguistics/].
The CSLI Calendar is also posted each week to
[news://nntp-csli.stanford.edu/csli.bboard].
Information about CSLI's research program is available at
[http://www-csli.stanford.edu/csli/].
For maps to the Stanford University campus see
[http://www.stanford.edu/home/visitors/maps.html].
____________