[Prev][Next][Index][Thread]
Sesquipedalian #26
-
To: ling-local, ling-others
-
Subject: Sesquipedalian #26
-
From: Alistair Murray <alistair@csli.stanford.edu>
-
Date: Fri, 29 Apr 94 10:54:23 PDT
-
Cc: gopher-quip
-
Flags: 000000000000
The SESQUIPEDALIAN WEEKLY HERALD Volume IV, Number 26
\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-
April 29, 1994
-\-\-\ LOOK WHO'S TALKING /-/-/-
The semantics interest group will present the following talk on April
29th (Friday) 1pm at Cordura 100. All are welcome.
"What Do Reciprocals Mean?"
Mary Dalrymple, Xerox PARC
Makoto Kanazawa, Stanford University
Sam Mchombo, University of California-Berkeley
Stanley Peters, Stanford University
Abstract
Research on reciprocals has uncovered a variety of semantic
contributions that the reciprocal can make, creating problems for
proposals that the reciprocal unambiguously means something weak
(e.g., Langendoen 1978). However, there is no real evidence that
reciprocals are ambiguous, despite previous claims to the contrary
(e.g., Fiengo and Lasnik 1973). First, we classify the apparently
heterogeneous list of meanings proposed in previous research into a
natural taxonomy, showing how they arise from a small stock of logical
operations and predicates. Second, we exhibit a partial ordering of
the various reciprocal meanings according to logical strength, which
we make crucial use of in determining what reciprocals mean in each
specific context where they appear. Third, we claim that a reciprocal
statement expresses the strongest candidate meaning that is consistent
with the fundamental logical properties of the relation expressed by
the scope of the reciprocal. This claim is supported by a large
collection of examples we have gathered from various corpora.
************************************************************************
-\-\-\ LINGUISTICS COLLOQUIUM /-/-/-
AN OPTIMALITY-THEORETIC APPROACH TO TEXTSETTING IN FOLK SONGS
Bruce Hayes
Department of Linguistics
UCLA
Friday, April 29th at 3:30 pm
CSLI, Cordura Hall, Room 100
Happy hour will follow.
********************************************************************
Abstract
Published versions of folk songs usually present the first stanza of
a song with its syllables aligned to the musical notes, but print
the subsequent stanzas at the bottom of the page. The singer is
assumed to be able to align these later stanzas with the music,
despite the fact that corresponding lines of different stanzas often
have different numbers of syllables. Recent work at UCLA suggests
that even musically untrained subjects have little difficulty in
performing a similar task, namely that of arriving at appropriate
rhythmic patterns for novel lines of text.
A recent paper by John Halle and Fred Lerdahl cogently addresses the
question of how people do this, and proposes as an answer a first-
pass account of the rhythmic textsetting algorithm. This paper
reports the results obtained so far in a second attack on the
problem. The basic ingredients of the project are as follows:
(a) A corpus of 479 folk song lines, annotated for stress
pattern, prosodic domains (at all levels), and the text
setting employed in the source material.
(b) A grammar, formulated under Optimality Theory
(Prince and Smolensky 1993) and consisting of 130 constraints.
The grammar defines for any input a tableau, which ranks the
output scansions in order of preference and lists the
constraints they violate. Accuracy of tableaux is insured by
a simple computer program that checks all the logically
possible scansions for every line. The rules rely heavily on
earlier work in musical rhythm and metrics, notably work
of Lerdahl and Jackendoff (1983) and of Kiparsky (1977).
(c) A set of judgments from a group of native speakers,
specifying how each would scan the lines of the corpus.
The project is still in progress, but some tentative results seem
worth reporting now.
First, text-setting seems fairly clearly to be an optimality
problem, in the sense defined by Prince and Smolensky (1993). In
particular, the scansion of certain lines is often regulated by
constraints that are violated in other lines, where this is
necessary to avoid violating other, stricter constraints.
Second, the data largely back up the initially counterintuitive
claim of Optimality Theory that constraints are ranked in
lexicographic order; i.e. no number of violations of laxer
constraints can override a single violation of a stricter
constraint. Apparent exceptions are few and probably reanalyzable.
Third, principles of English metrics originally worked out by
Kiparsky and others for art verse (the "beginnings free, endings
strict" principle, and the bounding of metrical constraints within
prosodic domains) are crucial to an adequate account of folk
metrics.
Fourth, the talk will address a possible role for extensive
inductive learning in linguistics, in particular the possibility of
relying on induction in trimming back the amount of hypothesized
innate knowledge, as well as the suitability of Optimality Theory as
the basis for induction. The talk will discuss the performance of
one inductive learning algorithm (a slightly modified version of
Tesar and Smolensky 1993) in learning the patterns of English
textsetting.
**************************************************************************
University of California-Berkeley
Phonology Laboratory Colloquium
................................
Time is of the Essence:
Temporal Factors in Speech Perception
Steven Greenberg
Department of Linguistics, UC-Berkeley
International Computer Science Institute
Monday, May 2, 1994
12:00 - 1:00 p.m.
46 Dwinelle Hall
SUMMARY
Models of speech perception have traditionally focused on spectral
properties (i.e., formant patterns) of the speech signal as the
physical basis of phonetic classification. So pervasive is the
spectral approach that it may properly be considered one of the
central tenets of speech science and associated disciplines.
And yet there are facets of speech theory that have never been
adequately explained on the basis of a purely spectral
characterization of the speech signal, most notably the perceptual
invariance of phonetic elements associated with widely divergent
formant patterns and the influence of visual input on the acoustic
percept (e.g., lip reading, the "McGurk" effect).
This presentation will focus on the "neglected" dimension of time,
as it pertains to phonetic classification and the macro-structure of
speech. What are the important temporal parameters governing speech
perception? How do these interact with the acoustic spectrum? And what
are the underlying neurological mechanisms that account for the
spectro- temporal properties of speech?
**************************************************************************
PH.D. STUDENTS ENTERING AUTUMN '94:
--------------------------------
BRAD DAVIDSON received his BA (magna cum laude) in International Relations in
May, 1990. Currently in the MA program here, he would like to study the impact
and outcomes of language planning in minority communities, looking at issues
of cultural rebirth and increased separatism. He is also interested in viewing
the relation between language and ethnic identity, and language and social
groupings, from a historical perspective, looking at how languages evolved and
continue to evolve, and under what social pressures, and how these languages
have been shaped into their present states.
JAMES FLECK received his BA (Summa cum laude) in English, with a minor
in Linguistics, at the University of Oregon in March, 1993. His
undergraduate linguistics studies focused on (functionalist) syntax
and semantics, phonetics, phonology, historical and comparative
linguistics, and sociolinguistics. His career goal is to earn a Ph.D.
and to secure a university faculty position as a professor of
Linguistics and researcher working in the areas of syntactic and
semantic theory, discourse analysis, pragmatics, and sociolinguistics,
with a background in the history of the English language.
JOHN FRY received his BA in Linguistics and Computer Science at UCLA in
June, 1990. He spent the summer of his junior year working as an
intern at Digital Equipment Corporation in Evry, France. After
graduating he moved to Yokohama, Japan, where he spent three years as
a C programmer, developing video games for the Nintendo and Sega home
entertainment systems. It was in 1988, taking a syntax course, that he
thought about applying parallel processing to linguistics. He reasoned
vthat phonetic, syntactic, and semantic analysis, which are treated as
distinct subfields in linguistics, in fact occur in parallel in the
human brain. He concluded that linguistics would be an ideal target
for one or more methods of parallel processing.
GARRET LUTES received his BA (magna cum laude) in Physics at Yale in May, 1986.
He is re-enrolling in the Ph.D. program which he left in Fall 1988. At that
time he moved to New York where he began a new and exciting life as a member
of an experimental dance company. During his time there he continued his
linguistic pursuits as a hobby, re-reading some phonetics and phonology
materials from his time at Stanford, and learning Polish (he lived in the
middle of a Polish neighborhood in Brooklyn). Thus, his interests in
linguistics have become much more clearly defined. He is currently working on
finishing a paper on Scandanavian accent, which he presented in April, 1993
in a Phonology workshop at Stanford.
ARMAN MAGHBOULEH received his BA (High distinction) in Engineering, with a
minor in Psychology, at Harvey Mudd College in May, 1986. He received his MS
in Computer Science at Yale University in May, 1989. Currently in the
MA program here, it is Arman's expectation that by completing the linguistics
Ph.D. program here, he will become more fully conversant in speech-related
fields, gain research experience and be ready to perform speech research in
the coming decades.
YUKIKO MORIMOTO received her BA in English at Mesa State College, Colorado
in May, 1992. She will receive her MA in Linguistics at Cal State University,
Fresno in May, 1994. Her MA program has provided her with a strong background
in general linguistics through a sequence of undergraduate and graduate
courses in phonology, syntax, and historical linguistics. Her purpose in
applying for the Ph.D. program is to expand her theoretical knowledge of the
discipline, to pursue her research interests in syntax and sociolinguistics,
and to prepare herself for a research and teaching career in a university
setting.
QING ZHANG received her BA in English at Nankai University, Tianjin, China in
July 1991. She will receive her MA in Linguistics at Nankai University in July,
1994. Through her graduate study, her interest falls in the area of
sociolinguistics, particularly, the cross-linguistic analysis of speech acts
and all the aspects related to their realizations. This interest comes mainly
>From her reading in sociolinguistics, discourse analysis, and pragmatics. It
has acquainted her with the speech act theory and the state of the art. She
hopes to investigate Chinese people's speech behavior as not much has been
reported on the study of speech acts in Chinese, unlike the bulk of research
done in the west. Her professional goal is to become an independent researcher
and a teacher of linguistics.
**************************************************************************
-\-\-\ CONGRATULATIONS TO..... /-/-/-
.....Scott A. Schwenter - for his publication:- The grammaticalization
of an anterior in progress: Evidence from a peninsular Spanish
dialect. Studies in Language 18:71-111 (1994).
**************************************************************************
-\-\-\ LINGUISTIC CONFERENCES /-/-/-
The Twelfth International Conference on Historical Linguistics
13-18 August 1995 at Hulme Hall University of Manchester
Plenary speakers
Elizabeth Traugott
Theo Vennemann
Anthony Kroch
Susan Herring
Paul Kiparsky
Alice Harris
Barry Blake
Aditi Lahiri
Ian Roberts
Workshops will include:
The lexicon and semantic change;
Changes in numeral systems;
Cross-linguistic evidence for syntactic change;
The influence of the Hansa and Low German on European languages.
For further information contact:
ICHL, Department of Linguistics, University of Manchester, Manchester
M13 9PL, UK. Tel/Fax: +44 61-275 3187 e-mail: ICHL1995@man.ac.uk
If you are interested and would like to be put on the mailing list
please fill in this form and give it to Nigel Vincent or Kersti
Borjars, or mail it to the above address
Name:
Address:
Institution:
e-mail:
**************************************************************************
INTERNATIONAL COLLOQUIUM
ON
PROCESSES OF LINGUISTIC COMMUNICATION
Linguistics Department
University of Wales, Bangor
23-25 June, 1994
#75.00 including full board
Speakers include:
* Ruth Kempson
* Lesley Milroy
* Dan Sperber
* Deirdre Wilson
* Malcah Yaeger
Over the last 15 years, theoretical and empirical research has
extended considerably our knowledge of what goes on in linguistic
communication and our ability to explain it. There have been a number
of perspectives, most notably Relevance Theory and Communication
Accommodation Theory. Hitherto, there has been little contact among
perspectives. The aim of the Colloquium is to bring together
approaches to mutual benefit.
The Colloquium will begin at lunchtime on Thursday 23rd June and end
at lunchtime on Saturday 25th June 1994. Accommodation will be
provided in one of the College's halls of residence.
The Colloquium is being organised in the Linguistics Department,
School of English and Linguistics, University of Wales, Bangor,
Gwynedd, LL57 2DGI UK. Fax: 0248/382928.
The organisers are:
Robert Borsley (Tel. 0248/382270. E-mail: elsO03@uk.ac.bangor),
Peter Garrett (Tel. 0248/382269. E-mail: elsO06@uk.ac.bangor).
**************************************************************************
-\-\-\ JOB OPENINGS /-/-/-
RESEARCH ASSOCIATE/FELLOW IN COMPUTATIONAL LINGUISTICS
THE UNIVERSITY OF BIRMINGHAM
SCHOOL OF ENGLISH
Applications are invited from suitably qualified and experienced
graduates for the above post. The job entails handling most of the
computing side of the development of software tools for studying large
samples of natural language, principally but not exclusively English.
Candidates should be familiar with UNIX and the C programming
language, as well as having all-round skills in computing. They
should have a flexible approach to research and should be interested
in helping to find practical solutions to problems that arise in
research. The samples of language run to many millions of words, and
although the responsibility for the corpora is shared among a number
of groups, the English School needs to maintain its own independent
facility.
The appointment is from as soon as possible until 31st July 1995.
Salary in the range of 12,828 to 25,107 pounds per annum.
Application forms (returnable by 20th May 1994) and further
particulars available from:
The Director of Staffing Services,
The University of Birmingham,
Edgbaston,
BIRMINGHAM B15 2TT
Telephone +44 - (0) 21-414-6483 (24 hours)
Please quote reference no. A 14763/94.
Working towards equal opportunities.
**************************************************************************
-\-\-\ CALL FOR PAPERS /-/-/-
Functionalism/Formalism
23rd Annual UWM Linguistics Symposium
University of Wisconsin-Milwaulkee, April 6-8, 1995
Featured Speakers [pending funding]:
WERNER ABRAHAM, STEPHEN ANDERSON, JOAN BYBEE, ALICE DAVISON, J0HN
DUBOIS, TALMY GIVON, KENNETH HALE, MICHAEL HAMMOND, MARTIN HASPELMATH,
BRUCE HAYES, HOWARD LASNIK, ALEC MARANTZ, EDITH MORAVCSIK, GEOFFREY
NATHAN, MICHAEL NOONAN, FREDERICK NEWMEYER (CO-ORGANIZER), DORIS
PAYNE, DAVID PESETSKY, JANET PIERREHUMBERT
We are seeking papers that:
- speak to the relationship between linguistic functionalism and
formalism; or
- highlight the advantages or drawbacks of some functional or
formal approach; or
- provide analyses of the same data from multiple perspectives; or
- explore the basic assumptions about language and cognition that
underlie the two approaches; or
- trace the history of one or both approaches; or
- offer general discussions of the formalist functionalist
dichotomy and its implications; or
- otherwise throw light on the similarities and differences between
the two approaches and their assessment.
Papers will be 20 minutes long, with a 10 minute discussion period to
follow. Please send 8 copies of an anonymous abstract and a 3x5 card
containing the title of the paper and your name, affiliation, and
address. The abstract may be up to one typed page, with figures and
references allowed on a second page. Since we need a camera-ready copy
for reproduction in the meeting handbook if accepted, we prefer
regular mail over email or fax.
A selection of the conference papers, supplemented with some invited
contributions, will be published by John Benjamins in a set of volumes
edited by Michael Darnell, Edith Moravcsik, Frederick Newmeyer, and
Michael Noonan.
Send your abstract to:
'95 UWM Symposium Committee
Department of Linguistics
University of Wisconsin-Milwaukee
Milwaukee, WI 53201-0413
USA
ABSTRACT DEADLINE: NOVEMBER 1, 1994
For further information:
Email: Edith Moravcsik (edith@convex.csd.uwm.edu) or Michael Noonan
(noonan@convex.csd.uwm.edu). Phone: Noonan at (414) 220-4539,
Moravcsik at (414) 229-6794, or leave messages at (414) 229-4285.
Fax: (414) 229-6258.
Snailmail: write to Noonan or Moravcsik at the postal address above.
===========================================================================
-\-\-\ SECOND CALL FOR PAPERS /-/-/-
for the
Third International Conference on
The COGNITIVE SCIENCE of NATURAL LANGUAGE PROCESSING
Dublin City University, 7-8 July 1994
Subject Areas: This is a non-exclusive list of subjects which fall
within the scope of CSNLP. It is intended as a guide only.
* Corpus-based NLP
* Connectionist NLP
* Statistical and knowledge-based MT
* Linguistic knowledge representation
* Cognitive linguistics
* Declarative approaches to NLP
* NLG and NLU
* Dialogue and discourse
* Human language processing
* Text linguistics
* Evaluation of NLP
* Hybrid approaches to NLP
Submissions may deal with theoretical issues, applications, databases
or other aspects of CSNLP, but the importance of cognitive aspects
should be borne in mind. Papers should report original substantive
research.
Theme: Corpus-Based Approaches
-----------------------
Since the conference follows on the heels of the ACM SIGIR meeting, we
have decided to emphasise the use of corpora in NLP. Papers dealing
with corpus- based approaches (advantages, disadvantages,
applications, etc.) will be preferred. Text and speech corpora are
equally welcome.
Invited Speakers:
The following speakers have been invited to give keynote talks:
Roger Garside, University of Lancaster;
Hans Kamp, Universitaet Stuttgart;
Cathy Sotillo, University of Edinburgh.
Not all are confirmed as yet.
Registration and Accommodation: The registration fee will be IR#40,
and will include proceedings, lunches and one evening meal.
Accommodation can be reserved in the campus residences at DCU.
Accommodation will be "First come, first served": there is a heavy
demand for campus rooms in the summer.
To register, contact Alex Monaghan at the addresses given below.
Payment in advance is possible but not obligatory.
This conference immediately follows SIGIR, a major Information
Retrieval conference, also at DCU. There is a limited amount of
funding available under the EC Human Mobility initiative, for
participants who are under 35 and citizens of one of the 12 EC member
states.
Submission of Abstracts:
Those wishing to present a paper should submit a 400-word abstract to
arrive not later than 13/5/94. Abstracts should give the author's full
name and address, with Email address if possible, and should be sent
to:
CSNLP
Alex Monaghan
School of Computer Applications
Dublin City University
Dublin 9
Ireland
Email submissions are also acceptable, plain ASCII text please to:
alex@compapp.dcu.ie (internet)
Completed papers should be around 8 pages long, although longer papers
will be considered if requested. Camera-ready copy must be submitted
to arrive in Dublin by 27/6/94. No particular conference style will be
imposed, but papers should be legible (12pt laser printed) and
well-structured.
Deadlines:
13th May --- abstracts to arrive in Dublin
1st June --- notification of authors
27th June --- camera-ready copy to arrive in Dublin
1st July --- final date for registration, accommodation, meals etc.
*************************************************************************
-\-\-\ 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 both at Stanford (in directory /user/linguistics/Sesquip), and
at Berkeley (in the directory /usr/pub.) 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. 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.
This journal printed on 100% recycled electrons