Voice choice in Totoli discourseSonja Riesberg, Kurt Malcher, Maria Bardají i Farré & Nikolaus P. Himmelmann
Totoli
2
Austronesian symmetrical voice
(1) a. i Ali nambalung deuk itu ACTOR VOICEi Ali non‐mbalung deuk itu (subject = Ali)HON PN AV.RLS‐throw.at dog DIST
b. deuk itu nimbalung i Ali UNDERGOER VOICEdeuk itu ni‐mbalung i Ali (subject = dog)dog DIST UV.RLS‐throw.at HON PN
‘Ali threw (stones) at the dog.’
3
Austronesian symmetrical voice
A language is a symmetrical voice language if
a. it has more than one basic transitive construction,b. the non‐subject arguments behave equally in all voices, andc. the verb is morphologically equally marked in all different voices.
(cf. Riesberg 2014: 11)
all this makes symmetrical voice alternations very different from active‐passive alternations
4
Austronesian symmetrical voice
% AV % UV Trans. clauses Reference
Balinese 62 38 1851 (+ 8% PASS) Pastika (1999: 62)
Besemah 56 44 899 McDonnell (2016)
Pendau 45 55 443 Quick (2005: 235)
Tondano 30 70 314 Brickell (2017)
Totoli 28 72 962 this study
ClassicalMalay 27 73 115 Cumming (1991: 85)
5
Research question
How do speakers of symmetrical voice languages choose one transitive construction over the other?
• Subject choice? Argument‐related properties?
• Functional difference between symmetrical and asymmetrical voice alternations? New factors? Differing choices?
6
Previous studies
What determines voice choice?• definiteness/specificity (Adams & Manaster‐Ramer 1988)
• topicality (Givón 1983, Cooreman 1983)• referential distance (Wouk 1996, Pastika 1999, Quick 2005)• topic persistence (Dryer 1994, Payne 1994, Pastika 1999, Wouk 1999)
• discourse transitivity/eventiveness (Hopper 1983, Cumming 1991, Wouk 1996, Pastika 1999, McDonnell 2016)
• (syntactic constraints)
McDonnell 2016
7
Totoli study: corpus
• 27 texts (02 hours 50 minutes)
• 53 speakers (26 female, 27 male)
• 3152 clauses, 6.745 intonation units
• 962 transitive clauses (excluded: RC, CC, LV, Indonesian)
• Variety of interactivity and text types
8
Totoli study: corpus
Dialogue Texts Hours IUs tr. clauses Monologue Texts Hours IUs tr. clauses
everydayconversation 5 00:22:27 1188 119 folk tales 6 00:35:25 1336 193
personalexperience 1 00:16:12 486 97 personal
experience 2 00:11:34 435 25
procedural 2 00:14:32 878 127 procedural 2 00:20:12 578 147
task‐oriented 4 00:32:53 1297 135 retelling of Pear Film 5 00:16:49 547 119
TOTALS 12 01:26:04 3849 478 15 01:24:00 2896 484
9
Totoli study: annotation
• GRAID (Haig & Schnell 2014): grammatical relations, morphological forms, animacy features
• RefIND (Schiborr et al. 2018): tracking of discourse referents• RefLex (Baumann & Riester 2012) (simplified): information status of referents
(3) tau magala kayu
tau mog‐ ala kayu
person AV‐ take wood
GRAID: np.h:a_av v:pred_AV np:p_av
RefIND: 10 11
RefLex: *** new
‘the boys picked up wood‘ [king_frog.059]10
Totoli study: factors
• Argument related factors• Argument inherent properties
• Animacy• Humanness
• Discourse related properties• Topicality• Activation state and argument realization• Tracking use• Generalizability
• Verb/clause related factors• Grounding
• Discourse transitivity and subordination• Valency increasing morphology
• Collostruction strength• Inherent orientation of root
• Structural priming• Prime distance• Root identity• Speaker identity• Function continuity
• Interactivity and text type• Dialogue vs. monologue• Text types
11
12
Totoli study: Cramér‘s Vvariable matrix size Cramér’sV
Argument realization (2x4) 0.284Priming (2x3) 0.241Activation state (2x4) 0.183 (data subset n = 432)Referential distance (2x3) 0.16 (data subset n = 258)Tracking (2x4) 0.148Generalizability (2x4) 0.14Humanness (2x4) 0.139Animacy (2x4) 0.125Text type (2x6) 0.12Valency incr. morph. (2x2) 0.119
No statistically significant association:
Topic persistence (2x3) 0.07Mood (2x2) 0.012Clause type (2x2) 0.006Interactivity (2x2) 0
Effect size (Cohen (1988: 222))0.1 ≤ V < 0.3 ‐ small0.3 ≤ V < 0.5 ‐ moderateV > 0.5 ‐ strong
13
Totoli study: argument realization
Some preferences.
But:• no clear cut association of the two “mixed” constellations in the middle columns and one of the two voices
14
Totoli study: discussion
Despite statistical significance, factors not suited for explaining the differences in usage conditions.
• Distributions rarely deviate strongly from the average distribution of voices in the corpus.
• None of the factor groups investigated involves a constellation where one voice is used close to 100% of the time.
15
16
McDonnell (2016: 223)17
Totoli study: summary & outlook
• Study confirms that symmetrical voice alternations differ both structurally and functionally from asymmetrical voice alternations.
• Parameters relevant to asymmetrical voice alternations are not relevant for voice choice in symmetrical alternations in the same way.
• Combinations of factors should be modelled.
• Other factors have to be considered.
18
Pilot‐study: episode boundaries
• Corpus: 5 Pear Movie re‐tellings from 5 different Totoli speakers 143 transitive clauses (29 AV vs. 113 UV)
• Divided into episodes• Correlation between episode boundaries and voice:
Episode beginning
AV UV
15%
85%
Pear Story corpus
AV UV
20%
80%
Tentative Small numbers
19
To be continued...
Thank you for your attention!
Thanks to:
Max Hörl Marc Heinrich
20
ReferencesBaumann, Stefan and Arndt Riester. 2012. Referential and lexical givenness: Semantic,
prosodic and cognitive aspects.” In Gorka Elordieta and Pilar Prieto (eds), Prosody and Meaning, 119–162. Berlin & New York: Mouton de Gruyter.
Chafe, W. (ed.). 1980 The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production. Norwood, New Jersey: Ablex.
Cooreman, Ann. 1983. Topic continuity and the voicing system of an ergative language: Chamorro. In Talmy Givón (ed), Topic continuity in discourse. A quantitative cross‐linguistic study, 425–489. Amsterdam & Philadelphia: John Benjamins Publishing Company.
Cumming, Susanna. 1991. Functional change: The case of Malay constituent order. Berlin: De Gruyter.
Dryer, Matthew S. 1994. The Discourse Function of the Kutenai Inverse. In Talmy Givón(ed), Voice and Inversion, 65–99. Amsterdam & Philadelphia: John Benjamins.
Givón, Talmy. 1983. Introduction. In Talmy Givón (ed), Topic continuity in discourse. A quantitative cross‐linguistic study, 3–41. Amsterdam & Philadelphia: John BenjaminsPublishing Company.
Haig, Geoffrey & Schnell, Stefan. 2014. Annotations using GRAID (Grammatical Relations and Animacy in Discourse).
Hopper, Paul J. 1983. Ergative, passive, and active in Malay narrative. In Flora Klein‐Andreu (ed.), Discourse perspectives on syntax, 67–88. New York, et al.: Academic Press.
21
McDonnell, Bradley. 2016. Symmetrical voice constructions in Besemah: A usagebasedapproach. Santa Barbara: University of California, Santa Barbara dissertation.
Riesberg, Sonja. 2014. Symmetrical Voice and Linking in Western Austronesian Languages (Pacific Linguistics 646). Boston/Berlin: Mouton de Gruyter.
Pastika, I Wayan. 1999. Voice Selection in Balinese Discourse. Canberra: Australian National University dissertation.
Payne, Thomas E. 1994. The Pragmatics of Voice in a Philippine Language. In Talmy Givón(ed), Voice and Inversion, 318–64. Amsterdam & Philadelphia: John Benjamins.
Schiborr, Nils N. & Schnell, Stefan & Thiele, Hanna. 2018. RefIND — Referent Indexing in Natural‐language Discourse: Annotation guidelines. Version 1.1. University of Bamberg. (https://multicast.aspra.uni‐bamberg.de/#annotations)
Quick, Phil. 2005. Topic continuity, voice and word order in Pendau. In I Wayan Arka and Malcolm Ross (eds), The many faces of Austronesian voice systems: some new empirical studies, 221–242. Canberra: Pacific Linguistics.
Wouk, Fay. 1999. Sasak is different : A discourse perspective on voice. Oceanic Linguistics 38.1, 91–114.
Wouk, Fay. 1996. Voice in Indonesian discourse. Studies in Language 20.2, 361–410.