LINGUIST List 25.2854
Mon
Jul 07 2014
Diss: Swahili;
Computational Ling, Pragmatics, Text/Corpus Ling:
Mwamzandi: 'Swahili Word Order
Choices...'
Editor for this issue:
Danuta Allen <danutalinguistlist.org>
Date: 07-Jul-2014
From: Mohamed Mwamzandi
<mohamed.mwamzandi
mavs.uta.edu>
Subject: Swahili Word Order
Choices: Insights from Information
Structure
E-mail this message to a
friend
Institution: University of Texas at
Arlington
Program: PhD in Linguistics
Dissertation Status: Completed
Degree Date: 2014
Author: Mohamed Yusuf Mwamzandi
Dissertation Title: Swahili Word Order Choices:
Insights from Information Structure
Linguistic Field(s): Computational
Linguistics
Pragmatics
Text/Corpus
Linguistics
Subject Language(s):
Swahili (swh)
Dissertation Director:
Laurel Smith Stvan
Joseph Sabbagh
Jeffrey Witzel
Jason Kandybowicz
Dissertation Abstract:
In pragmatics, cross-linguistics studies have
shown non-canonical
word order can often be explained if
information structure is taken into
consideration. In this dissertation, I explore
the role of information
structure on word order variation in Swahili,
an SVO language
belonging to the Bantu branch of the
Niger-Congo languages. In
particular, I explore how the notion of topic
may explain word order
variation in adnominal demonstratives phrases
and reciprocal
constructions, as informed by a corpus-based
analysis. The term
‘adnominal demonstrative’ is used to
distinguish pronominal
demonstratives such as huyu ‘this’ from
demonstratives that co-occur
with nouns such as huyu mtu ‘this person’.
Aside from a brief mentions
(Ashton 1944; Givon 1976; Leonardo 1985, 1987;
Wilt 1987; Carstens
1991, 1998), the pragmatic function and
syntactic position of
adnominal demonstratives have not yet been
investigated in Swahili via
corpus analysis. Another instance where word
order variation rests on
information structure is that of Swahili
Discontinuous Reciprocal (DR),
and the Simple Reciprocal (SR). While other
studies have discussed
the DR as a syntactic derivative of the SR
(Vitale 1981); or the DR as a
syntactic strategy to resolve unbalanced
coordination (Mchombo and
Ngalande 1980, Mchombo 1993); or the DR and SR
as distinct
structures that are underivable from each other
(Seidl & Dimitriadis
2002), I argue that the variation is motivated
by the principle of
givenness which requires familiar information
to come first in a
sentence before new information. Though these
constructions present
word order variation that warrants an
explanation, they have received
little attention in information structure
studies.
Class 1 (animate nouns) adnominal
demonstratives from the Helsinki
Corpus of Swahili are examined in the two
attested word orders:
NP+DEM and DEM+NP. Statistical analysis of the
dataset indicated
that the NP+DEM order was more frequently used
if the topic was
active (used in previous sentence), p<0.001,
while the DEM+NP order
was more frequently used if the topic was
semiactive or inactive
(referents within the utterance situation or
after topic shift in texts),
p<0.00.
Corpus examples from two distinct verb
categories, namely
conversation and “marry” verbs (Levin 1993),
were analyzed to
investigate the effect of givenness and verb
category on reciprocal
variation. To complement the tokens found in
the corpus, I also set
about eliciting native speaker judgments on
controlled sets of
grammatical constructions using questionnaires
administered via the
DMDX software. The corpus and accessibility
ratings results indicated
that givenness, rather than verb category, is
the main predictor of
Swahili reciprocal variation.
As mentioned above this study addressed Swahili
reciprocal and
adnominal demonstrative word order variation,
which to the best of my
knowledge have not been analyzed under the
auspices of information
structure, or approached via corpus analysis.
This research will be of
interest to those studying information
structure, deixis, reciprocals, and
syntax. Further, the results of this study will
help those interested in
Swahili grammar such as Swahili language
instructors and Swahili
second language learners understand the
different pragmatic value of
these constructions.
Page Updated: 07-Jul-2014