LINGUIST List 3.312

Sun 05 Apr 1992

Confs: Mid-American, Prosody, Cog Sci

Editor for this issue: <>

Directory

1. Frances Ingemann, Update: Mid-American
2. Cynthia McLemore, Prosody Workshop
3. Peter Cole, Yael Zic Cog Sci Talk -Clarification

The Mid-American Linguistics Conference and the Conference on Siouan
and Caddoan Languages will be held on October 18-19, 1992, at the Uni-
versity of Missouri-Columbia. Abstracts should be sent to Donald Lance,
Linguistics Program, 107 Tate, University of Missouri, Columbia, MO 65211
(FAX 314-882-5785; e-mail engdlumcvmb.missouri.edu) or Louanna Furbee,
Anthropology, 200 Swallow Hall, University of Missouri, Columbia, MO
65211. Abstract deadline: September 1, 1992.

 Call for Papers

 Workshop on Prosody in Natural Speech Data

 5 August - 12 August 1992 (Wednesday-Wednesday)

 University of Pennsylvania
 Philadelphia, PA, USA

Prosody has become increasingly important in many areas of language-
related research -- natural language processing, phonology, sociolinguistics,
pragmatics, semantics, and language acquisition. Nevertheless, researchers
trying to describe and understand prosodic phenomena are faced with the
basic problem of reconciling actual patterns of pitch and time in naturally
occurring speech with the categories and functions hypothesized by various
theories. Thus definitions of notions such as "prosodic phrase," "pitch
accent," "focus," "contrastive stress," etc., usually are not adequate for
an individual researcher to use in describing a stretch of talk, much less
for different researchers to agree on. This extensional uncertainty does
not necessarily invalidate the concepts, which may be motivated by certain
clear cases; however, both theory and applications would undoubtedly benefit
from a clarification of how such concepts apply to ordinary speech. In
addition, there are identifiable and recurrent prosodic phenomena in
ordinary speech that don't have an agreed-on linguistic description.

This workshop is aimed at developing an approach to prosodic analysis
that is grounded in a wide range of natural speech data, spanning
different languages as well as different discourse types. All aspects
of prosody will be considered, including pitch range, tune form, text-tune
mapping, stress/prominence, and phrasing; the relationship of these
features to textual, situational, and cultural context; and their role
in poetic structuring and rhetorical functioning in discourse. We will
focus primarily on comparative data analysis, with concurrent evaluation
of proposed theories.

Researchers working on prosodic form and function in natural speech of
any kind are encouraged to submit abstracts. Analyses should be of either
a single body of prosodically transcribed data, or the distribution of
prosodic forms across instances of use. In addition to the initial
submission of abstracts, participants will be asked to submit work-in-
progress before the workshop, including the speech material on which
their analysis depends, so that a large body of analyzed material will be
available for common consideration. Subsequent theoretical discussions
can then refer to the group's collective experience of prosodic analysis,
grounded in a common body of analyzed data.

Specific topics to be addressed include:

 (1) the identification, transcription and measurement of recurrent forms
 (2) the relation between prosody and other linguistic forms
 (3) the role of prosody in poetics and rhetoric
 (4) the relation of prosody to aspects of sociocultural context
 (5) the nature of evidence for positing formal and functional categories

Facilities will be provided so that group analyses can be conducted
interactively from the computer console.

Deadlines: 11 May 1992 1-2 page abstracts due
 25 May 1992 Notification of acceptance
 10 July 1992 Papers due

Limited funds will be available for participant stipends.

Submissions and inquiries should be sent to either co-sponsor:

Mark Liberman			Cynthia McLemore
University of Pennsylvania	University of Pennsylvania
Linguistics Department		Inst. for Research in Cognitive Science
619 Williams Hall		3401 Walnut St.
Philadelphia, PA 19104	Philadelphia, PA 19104
mylunagi.cis.upenn.edu		camunagi.cis.upenn.edu

The workshop is funded by the Institute for Research in Cognitive
Science, University of Pennsylvania.

*****************************************************************************
Clarification: The talk is at 10:10. There will be open discussion with
Dr. Ziv at 3:35. Call the Dept. of Linguistics, U Delaware for driving
directions - 302-831-6806, linguisticsudel.edu.
*****************************************************************************

 UNIVERSITY OF DELAWARE
 COGNITIVE SCIENCE COLLOQUIUM


 YAEL ZIV
 THE HEBREW UNIVERSITY OF JERUSALEM

"Pragmatic Constraints on Anaphora in Left and Right Dislocations"

 ABSTRACT

 A pragmatic account is proposed of the possible anaphoricity
evident in so-called left and right dislocation constructions, as
in the following:
(1a) Hei is here, Jimi (1a') I don't like themi, the copsi.
(2a) Billyi I met himi in town last night. (2b)* He/Himi I met
Billyi in town last night. It is shown that L(eft) D(islocations)
(2a above) do not display properties associated with bound
variables (evidence from VP ellipsis and Quantifiers), but rather
exemplify pragmatically determined coreferentiality.

 In addition, C-Command is argued to be inapplicable since LDs
are analyzed as E(expressions) and not as S(entential entities).
The evidence consists in the peculiar properties of the initial NP
with respect to fronting (e.g. Adverb preposing) and
morphosyntactic status as well as the restrictions on embeddability
evident in LD's. An account of the anaphoric options evident in
LDs will be shown to involve considerations of discourse anaphora.
The pragmatic approach adopted will be based on linearity in
conjunction with the differential properties of the various
referring expressions and the specific discourse functions
associated with LDs. It will be shown that the possible
anaphoricity patterns follow from the capacity of the various
referring expressions to access and activate the relevant discourse
entities.

 Unlike LD constructions, R(ight) D(islocation) (1a, 1a' above)
will be shown to display grammatical properties associated with
S(entential units) (e.g. complying with the constraints on
subjacency and showing a predictable morphosyntactic dependence
between the two coreferential NP's expressed overtly in languages
like Modern Hebrew where the relevant cases are morphologically
distinct). It will, however, be argued that the available
anaphora options in these constructions are insightfully accounted
for in discourse functional terms. The relevant account will
utilize considerations of the type evident in LD anaphora above.
Similar linearity factors together with an appropriate
characterization of the differential accessing and activating
capacity of the various referring expressions coupled by the
specific discourse function associated with RD will be shown to
make the correct predictions and bring out the otherwise
unaccounted for anaphora parallelisms between LDs and RDs. The
above pragmatic characterization will be shown to predict the
impossibility of identity statements RDs and the well-formedness
of the corresponding attributive readings.



 MONDAY, APRIL 13, 1992
 10:10 A.M.
 325 PURNELL HALL
 UNIVERSITY OF DELAWARE
 NEWARK, DELAWARE

 AT 3:35 DR. ZIV WILL TALK IN ROOM 316B WOLF HALL (OPEN DISCUSSION).