[Prev][Next][Index][Thread]
Sesquipedalian #30
the SESQUIPEDALIAN Volume VII, No. 30
\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/
Fall of Constantinople (1453) May 29, 1997
-/-\-/ LOOK WHO'S TALKING /-\-/-
-- Elizabeth Traugott gave a plenary talk entitled: "Subjectification
as externalization: A study of the development of discourse markers"
and participated in a panel on semantic change at the Third Conference
on Conceptual Structure, Discourse, and Language, University of
Colorado at Boulder, May 24-27.
-- LLC6: The Sixth Annual Workshop on Logic, Language and Computation
takes place this weekend at CSLI (see below). Don't miss it!
Department affilliates who will be presenting talks include Jerry
Hobbs, John Fry, and Ann Copestake.
-/-\-/ LINGUISTICS COLLOQUIUM /-\-/-
Friday, May 30, 3:30pm
Margaret Jacks Hall (460), Room 146
Shuichi Yatabe
Tokyo University and Stanford
Particle ellipsis and focus projection in Japanese
In this talk, I will examine the properties of particle ellipsis in
Japanese in hopes of discovering some general principles governing the
distribution of focus.
The nominative case particle in Japanese (_ga_) is omissible under
certain circumstances, as in the sentence _Taiyoo (ga) detekita_ 'The
sun came out'. The descriptive goal of this talk is to determine
exactly under what circumstances the particle is allowed to be absent.
Authors such as Tsutsui, Tateishi, and Kageyama have made claims such
as the following:
-The particle is harder to drop when it marks an agentive NP.
-The particle is harder to drop when it marks the subject of
an individual-level predicate.
-The particle is harder to drop when the particle is not
adjacent to a predicate.
These seem to be fairly adequate descriptions of observed tendencies,
but there are exceptions to each of these generalizations, as has been
pointed out in the previous literature.
In this talk, I will first argue that particle ellipsis in Japanese is
subject to what I will call the Antifocus Constraint, a constraint
which prohibits ellipsis of a particle when either the NP marked by
that particle or its head noun is focused. (An expression is said to
be _focused_ when it is interpreted as contrasting with some other
object of the same type.) Then, making use of some intonational facts
noted by Selkirk, Tateishi, Kori, and others, I will argue that the
Antifocus Constraint alone captures all the facts that the three
generalizations listed above are intended to capture. In other words,
it will be my claim that the distribution of focus within a sentence
(and not just the distribution of prosodic prominence) is affected by
the internal semantic makeup of the sentence.
------------------
Reception follows.
For directions and a complete list of colloquia, see
http://www-linguistics.stanford.edu/Linguistics/colloq/colloq.html
-/-\-/ PHONOLOGY WORKSHOP /-\-/-
Rm 146, Bldg 460
Thursday, 5/29/97, 7:30 pm
CALIBRATING PHONETIC DISTANCE
John Nerbonne
(reporting on collaboration with Wilbert Heeringa)
Alfa-Informatica, BCN, Rijksuniversiteit Groningen
nerbonne@let.rug.nl
Kessler (1995) analyzed dialect data using Levenshtein distance,
which has also been applied to sequence comparison problems in
spelling, protein and DNA sequences, bird songs and speech recognition
(Kruskal 1983). To apply Levenshtein distance, one imagines string
mapping operations associated with costs. The basic operations are
insertion, deletion and replacement, and they always apply to single
characters. Levenshtein distance between two strings is the cost of
the least costly set of operations mapping one string to another.
Nerbonne, Heeringa, van den Hout, van der Kooi, Otten and van de Vis
(1996) applied this method to 100 words from 20 Dutch dialects. The
dialect data from Blacquart et al. (1925/1982) was represented in
phonetic transcription. The distance between varieties was taken to
be the average of the 100 word-pair differences, and clustering based
on least-error was applied to the matrix of resultant distances. This
results in dendograms in which the main dialect groups are found in
the highest level groupings, in this case, Flemish, Franconian,
Frisian and Lower Saxon.
The present talk focuses on extensions to this work which aim at
calibration as well as at refinements. The calibration involves
developing a more exact method of comparing the results achieved here
with those in dialectology, and the refinements involve the phonetic
representation in features, the use of an information-gain weighting,
and the analysis of diphthongs.
References:
Blacquart, E., et al. (1925/1982). Reeks nederlandse dialectatlassen.
Kessler, Brett (1995). Computational dialectology in Irish Gaelic. In
Proc. of the European ACL, Dublin, pp. 60-67.
Kruskal, Joseph (1983). An overview of sequence comparison. In
D. Sankoff & J, Kruskal, eds, Time Warps, String Edits and
Macromolecules: The Theory and Practice of Sequence Comparison.
Addison-Wesley, Reading Mass., pp. 1-44.
Nerbonne, John, Wilbert Heeringa, Erik van der Hout, Peter van der
Kooi, Simone Otten & Willem van de Vis (1995). Phonetic distance
between Dutch dialects. In G. Durieux, W. Daelemans & S. Gillis,
eds., Proc. of CLIN '95, Antwerpen, pp. 185-202. Also available as
http://grid.let.rug.nl/~nerbonne/papers/dialects.ps
*****
All are welcome for pre-workshop pizza (7:00 pm). RSVP.
-/-\-/ SYMBOLIC SYSTEMS FORUM /-\-/-
Special Event
Honors Thesis Presentations
Thursday, May 29, 4:15 pm
Cordura 100 (CSLI)
-=> NOTE NEW LOCATION! <=-
Five SSP seniors will present their Honors thesis projects which
they have developed over the academic year. The Topics are:
VISUAL MUSIC (Max Klee), NATURAL LANGUAGE INTERFACE (Marissa Mayer),
AN INEXPENSEIVE RANGE FINDER (a visual system for a robotic arm)
(Thomas Willeke), COOPERATIVE PROBLEM SOLVING (Michael Ebstyne),
and THE EXPLANATORY LIMITS OF SCIENCE (Edwing Ong).
There will be demos of some of these cool projects. Don't miss
the opportunity to learn about your peers/friends/students Honors
theses.
Refreshments will be served. Bring your friends!
****************************************************************************
Abstracts
Marissa Mayer
Title: A Natural Language Interface for the Advisable Planner
In order to aid user interactions with AI planning technologies, I created an
interface that accepts written or spoken advice from the user, translates into
an acceptable advice representation, and then passes it to a system
(the Advisable Planner) that is capable of influencing the plan generated.
I'll discuss how this interface operates in the domain it was created for and
how to extend it to new domains.
Max Klee
Title: VISUAL MUSIC:
A new expressive medium through which graphics animate to music.
This C++ application synchronizes the animation of images on the screen
with frequency information extracted in real time from the audio
port.
Thomas Willeke
title: An inexpensive range finder
The range finder I developed is part of an ongoing research effort
to build a robotic arm capable of assisting people with disabilities.
Such an arm must be both affordable, and have the ability to find the
distance and shape of objects to be handled.
Michael Ebstyne
Cooperative Problem Solving
This study investigates the quantitative impact of cooperative
searches. At a general level, it is an exploration of information
technologies in an information age. The two search technologies
explored are genetic algorithms and standard agent simulations. These
technologies were applied to the following constraint satisfaction
problems: cryptoarithmetic and algorithmic musical composition.
Cooperative searches demonstrate a super-linear improvement in performance
over non-cooperative searches. My talk will focus on the standard agent
simulations which were used to explore cryptoarithmetic problems.
-\-/-\ LLC-6 \-/-\-
THE SIXTH CSLI WORKSHOP ON LOGIC, LANGUAGE, AND COMPUTATION
Dates: May 30-June 1 1997
Location: Cordura Hall Conference Room, Stanford University
This workshop brings together scholars having an interest in
logic - philosophers, linguists and computer scientists - with the
overall aim of facilitating interdisciplinary interaction.
It is organized by Johan van Benthem, Henriette de Swart
Grigori Mints, Rob van Glabbeek, and Martina Faller
This year's program reflects the usual lively mix of topics and
interests that CSLI was designed to bring together. It includes
recent developments in dynamic processing of linguistic and
non-linguistic information, as well as key techniques from
mathematical logic that underlie both formal proof and computation.
The mix also includes contributions by established researchers and by
newcomers to the field, in line with a long-standing tradition.
Finally, some contributions this year reflect the growing interactions
with a broader world, witness talks on logic teaching, as well as
presentations by representatives of industry and high performance
computing. We hope to broaden our 'circle of debate'.
See the LLC6 webpage for more information on speakers, abstracts and
schedule: http://www-linguistics.stanford.edu/Linguistics/llc6
or mailto:faller@csli.stanford.edu
-\-/-\ JOB ANNOUNCEMENTS \-/-\-
(REDUNDANCY NOTICE: For fuller listings of these and other jobs, don't
forget to check the Jobs binder in the Greenberg Room, and the file
'jobslist.txt' on the CSLI directory /user/linguistics.)
-- SENSORY SYSTEMS: Sensory, Inc., Sunnyvale, is searching for a voice
recordings coordinator. The positoin will have the following duties:
arrange and manage voice recording sessions, involving both adult and
child subjects. Some travel is involved. Maintain contacts for
future recordins. Join Linguistics team in processing speech
synthesis and voice recoginition data, using Sensory software. Assist
with product testing. This is a full-time, entry-level position.
Qualifications: Excellent interpersonal skills for working with both
adults and children. Some linguistics background (speech, phonetics)
is helpful. Must have own transportation and be able to work flexible
hours. Familiarity with computers a must. Japanese, Korean,
Cantonese, Mandarin or French language fluency a definite plus.
Please mail or fax resume to
Sensory, Inc.
Attn: Paul Geiner, Linguistics Manager
521 East Weddell Drive
Sunnyvale, CA 94089
Fax: 408 744 1299
mailto:paul@sensoryinc.com
-- Dragon System, Inc is an innovative speech and recognition company
with approximately 192 employees in Newton, Massachusetts. Currently
we have several openings in our Research Department for Research
Engineers. Attached is a brief description of our current vacancies.
Please do not hesitate to contact the director of Human Resources,
Tamah Rosker, at 617-965-5200 x584 or
mailto:Tamahr@dragonsys.com
Imaginative Research Scientist needed to work on exciting speech
recognition research projects. Background in speech recognition
desirable but not essential. Candidate must have:
(I) strong academic/research background (Ph.D. or M.S. in computer
science, electrical engineering, mathematics, or other quantitative
specialty);
(II) advanced C or C++ programming skills
(REDUNDANCY NOTICE: For fuller listings of these and other jobs, don't
forget to check the Jobs binder in the Greenberg Room, and the file
'jobslist.txt' on the CSLI directory /user/linguistics.)
/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\
-\-/-\ CONSERVE DISK SPACE \-/-\-
So you may delete your copy after you've read it (or better yet,
before you've read it), the Sesquipedalian Weekly Herald is stored
online at Stanford (in directory /user/linguistics/Sesquip/), and at
Berkeley (in the directory /usr/pub.), or on the Linguistics
Department home page (http://www-linguistics.stanford.edu/). The most
current issue of the Herald can be found by typing 'help quip'.
Neither Stanford University nor the Linguistics Department, nor any of
their employees, makes any warranty, whatsoever, implied, or assumes
any legal liability or responsibility regarding any information,
disclosed, in this publication, or represents that its use would not
infringe privately owned rights. No specific reference constitutes or
implies endorsement, recommendation, or favoring by Stanford
University or the Linguistics Department, or their employees. Any
similarity to actual linguists, living or dead, is purely
coincidental. The views and opinions expressed herein do not
necessarily reflect those of Stanford University or the Linguistics
Department, or their employees, and shall not be used for advertising
or product endorsement purposes. No user-serviceable parts inside.
WARNING: Contains a substantial amount of non-tobacco products. It is
a violation of federal law to use this product in a manner
inconsistent with its labelling.
All rights revoked
\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/-\-/