Transcript Слайд 1 - Institute of Linguistics, Russian Academy

Andrej A. Kibrik
Olga B. Markus
Local discourse structure
in Upper Kuskokwim Athabaskan
SSILA Conference
Berkeley, July 2009
1
2
Basic information about Upper
Kuskokwim Athabaskan (UKA)
 About 30 speakers left out of the population of





about 200
Most speakers reside in the village of Nikolai
Actual use of UKA – in two or three households
Prior work – Collins and Petruska 1979
Kibrik’s field trips in 1997 and 2001
As in other Athabaskan:
 polysynthesis
 highly complex verb morphology and morphophonemics
3
4
Data
 Natural discourse recordings (transcribed)




Folk stories
Personal stories
Conversation (pre-arranged)
Interview at school
 In all – 3 hours 20 minutes of talk
5
Lena Petruska, the oldest speaker
6
Theory
 Local discourse structure:


Elementary discourse units (EDUs)
EDUs are elementary behavioral acts of
discourse processing
EDUs are identified on the basis of a cluster of
prosodic features:





Tonal contour
Central accent
Tempo pattern
Loudness pattern
Pausing
7
Example (1): tonal contours
a
b
c
d
e
f
8
Example (1b): tempo pattern
a


b
c
d
sighwdlaɁ 720ms / 3 = 240 ms per syllable
todoltsitł’ ts'eɁ 1800 ms / 4 = 450 ms per syllable
e
f
9
Example (1): pausing
a
b
c
d
e
f
10
Properties of EDUs
 Prosodically identified EDUs display
interesting content-related properties
 Cognitively: manifest a focus of
consciousness (Chafe)
 Semantically: typically report event/state
 Grammatically: often coincide with clauses
11
EDUs and clauses
 Clausal EDUs
 Short EDUs (less than one canonical
clause)
 Long EDUs (more than one canonical
clause)
12
EDU types in example (1)
 Clausal: b, c, f
 Short:
a
d
e
regulatory (discourse marker)
subclausal (topic)
fragmentary (false start)
13
Quantitative data:
an overview
 965 EDUs in the data set
 Clausal EDUs – 70.8%
 Short EDUs – 14.8%
 Long EDUs – 14.4%
14
Clausal EDUs (683 = 100%)
 Headed by a lexical verb – 84%
 Headed by a verb of being – 6%
 Non-verbal – 10%
(1b, c)
(1f)
(2)
15
Non-verbal clausal EDU
(2) ‘(There was) also lots of marten skins’
16
Short EDUs (143 = 100%)
 Regulatory – 13%
 Fragmentary – 20%
 Nominalized – 7%
 Subclausal – 50%
(1a)
(1e)
 Prospective – 42%
 Retrospective – 18%
(1d)
(3)
17
Retrospective subclausal
EDUs
Increment:
(3) ‘That is why that happened to me then,
because of the icon’
18
Long EDUs (139 = 100%)
 Concatenation – 19%
 Adverbial – 0%
 Relative clause + main clause – 2%
 Non-quotative complement clause
+ main clause – 42%
 Quotative clause + main clause – 37%
(4)
(5)
19
Concatenation
(4) ‘He went inside and lay down’
 danaɁediyo
 naztanh
150 ms
385 ms
20
Quotative clause +
main clause
(5) ‘You should also come slide with me, I
told her’
21
EDUs and clauses in a
typological perspective
Language
English (Chafe 1994)
Percentage
of clausal EDUs
60%
Mandarin (Iwasaki and Tao 1993)
39.8%
Sasak (Wouk 2008)
51.7%
Japanese (Matsumoto 2000)
68%
Russian (Kibrik and Podlesskaya 2009)
68.6%
Upper Kuskokwim
70.8%
22
A possible explanation
 Percentage of clausal EDUs is correlated
with the degree of a language’s:
 degree of morphological complexity
 grammatically marked distinction of inflected
verbs from other predicate types
 Probably the languages overtly marking
verbs as dedicated predicative elements
more strongly correlate clauses with EDUs
23
Conclusions
 EDUs as universal building blocks of local



discourse structure are perfectly well identifiable
in a polysynthetic language
EDUs display a high correlation with clauses
Short and long EDU types, as known in other
languages, are also found in Upper Kuskokwim
An account of EDUs and their types is a
necessary component of a grammatical
description of any language, less studied and
endangered languages not excluded
24
Some directions for further
research
 Different intonation contours – their
discourse semantics
 Interaction of discourse prosody with
lexical tone, vestigially present in some
idiolects
25
TsenɁan!
 Thanks to all speakers of Upper Kuskokwim, both

mentioned and unmentioned above
Thanks to many individuals and organizations that
helped to collect and process the data, in
chronological order:










Michael Krauss
James Kari
Raymond Collins
Alaska Native Language Center
Fulbright Program
Endangered Language Fund
Bernard Comrie
MPI for Evolutionary Anthropology, Leipzig
Russian Foundation for the Humanities
National Science Foundation
26
27
28
29
Welcome to Nikolai
30
31
32
33
34
35
36
37