Cognitive constraints on the dative alternation across space and registers Melanie R¨othlisberger KU Leuven Quantitative Lexicology and Variational Linguistics
ICLC 13 (Newcastle), July 25, 2015
Introduction
A new project ▶
▶
“Exploring probabilistic grammar(s) in varieties of English around the world” (5-year project, 2013-2018) main goal: understand the plasticity of probabilistic knowledge of English grammar, on the part of language users with diverse regional and cultural backgrounds (see Szmrecsanyi et al. to appear)
▶
Project members: ▶ ▶ ▶ ▶
▶
PI: Benedikt Szmrecsanyi Particle placement: Jason Grafmiller Genitive: Benedikt Heller Dative: Melanie R¨othlisberger
supervisors: Benedikt Szmrecsanyi, Marianne Hundt, Jason Grafmiller
Introduction ▶
Main goal: understand the plasticity of probabilistic knowledge of English grammar, on the part of language users with diverse regional and cultural backgrounds
(see Geeraerts 2010; Kristiansen and Geeraerts 2013; Szmrecsanyi 2010; Geeraerts 2005; Gries 2013 and others)
Introduction ▶
▶
Main goal: understand the plasticity of probabilistic knowledge of English grammar, on the part of language users with diverse regional and cultural backgrounds Cognitive sociolinguistics:
(see Geeraerts 2010; Kristiansen and Geeraerts 2013; Szmrecsanyi 2010; Geeraerts 2005; Gries 2013 and others)
Introduction ▶
▶
Main goal: understand the plasticity of probabilistic knowledge of English grammar, on the part of language users with diverse regional and cultural backgrounds Cognitive sociolinguistics:
“Cognitive Linguistics must embrace society and contexts of usage” (Schmid 2015, this conference) (see Geeraerts 2010; Kristiansen and Geeraerts 2013; Szmrecsanyi 2010; Geeraerts 2005; Gries 2013 and others)
Introduction
Aim:
Introduction
Aim: ▶
How do linguistic internal (cognitive) and external features (sociolinguistic) shape speakers‘ cognitive grammar?
Introduction
Aim: ▶
▶
How do linguistic internal (cognitive) and external features (sociolinguistic) shape speakers‘ cognitive grammar? Do cultural differences exist in syntactic alternations where the choice of variant is heavily influenced by cognitive processes?
The dative alternation
The dative alternation
(1) a. He gives [Mary]recipient [a present]theme (ditransitive dative) b. He gives [a present]theme to [Mary]recipient (prepositional dative)
The dative alternation
▶ ▶
semantic similarity rich discussion (see Goldberg 2002; Green 1974; Oehrle 1976; Rappaport Hovav and Levin 2001; Gropen et al. 1989)
▶
alternation-based generalizations in the mental grammar → “cause someone to receive”(Perek 2012)
Previous research
Conditioning factors: ▶ semantic/pragmatic: animacy, thematicity ▶ processing-related: discourse accessibility, pronominality, complexity/end weight, definiteness ▶ economy: TTR, . . . ▶ external variables: register, style, geographic location ▶ ...
Previous research ▶
Previous probabilistic models: “harmonic alignment” (ease of processing): ▶ ▶ ▶ ▶ ▶
given >new animate >inanimate definite >indefinite pron >non-pron short >long → end-weight
recipient >theme → ditransitive theme >recipient → prepositional (see Bresnan et al. 2007; Bresnan and Ford 2010; Bock 1982, 1986; Bock et al. 1992; McDonald et al. 1993, and others)
Previous research
▶
Probabilistic constraints shared across South Asian varieties of English → general processing principles in all varieties of English (Bernaisch et al. 2014)
▶
Varietal differences with regard to cultural (e.g. animacy) and processing-related factors (end-weight) (Bresnan and Hay 2008; Bresnan and Ford 2010)
Data & Methodology
Data I tap into 9 different varieties of English (as sampled in the International Corpus of English (ICE)): ▶ 1 million words per variety ▶ 12 different subregisters
Retrieving dative tokens 1. extract dative tokens using verb list 2. define choice context (incl. pronouns), leave out, e.g.: ▶ ▶ ▶
fixed and idiomatic expressions (e.g. bring it to the boil) spatial goals (e.g. send their daughter to school) beneficiaries (e.g. We get them uh typed photo copies)
3. annotate tokens for sociolinguistic and cognitive factors 4. restrict dataset: RecWordLth = 18 (DO), ThemeWordLth = 23 (PD) (see Bresnan et al. 2007; Bresnan and Hay 2008; Bresnan and Ford 2010; Bernaisch et al. 2014; De Cuypere and Verbeke 2013)
Conditioning factors ▶
Semantic/Pragmatic factors: ▶ ▶
▶
Processing-related factors: ▶ ▶ ▶ ▶
▶
▶
Animacy: animate vs inanimate Theme concreteness: concrete vs non-concrete Pronominality: pron versus non-pron Discourse accessibility: given versus new Definiteness: definite versus indefinite Complexity: log value of weight ratio; complex versus simple Recipient person: local vs non-local
Sociolinguistic factors ▶ ▶
register: spoken vs. written / informal versus formal country: GB, CAN, NZ, IRE, SIN, IND, HK, JA, PHI
Variable ‘Country’
Dative proportions across all nine ICE corpora, N=8549
Variable ‘Register’
Proportion of dative tokens across register
Generalized mixed-effect logistic regression
▶ ▶
Model predicts: prepositional dative random effects included: ▶ ▶ ▶
▶ ▶
Verb sense nested in Verb heads of recipient and theme text category & file
deviation coding for ‘country’ and ‘register’ function glmer() in lme4 package in R
Results
Effects of predictors
Predicted outcome: PD; C -value: 0.98; Accuracy: 93.6% (baseline: 69 %)
Main effects Ditransitive: give [Mary]recipient [a present]theme Prepositional: give [a present]theme to [Mary]recipient
Main effects
▶
all predictors influence the choice of construction as predicted: ▶ ▶ ▶ ▶ ▶
given >new animate >inanimate definite >indefinite pron >non-pron short >long
recipient >theme → ditransitive theme >recipient → prepositional
The dative alternation across space “give it to the most appropriate or the person that fits the criteria best” (JA:S1A-060) “Giving all those worthless women thousands of children” (JA:W2F-012)
The dative alternation across register
The dative alternation across space & register ▶
effect directions follow the overall pattern...: ▶ ▶ ▶
▶
pron >non-pron short >long ...
Variety-specific patterns:
The dative alternation across space & register ▶
effect directions follow the overall pattern...: ▶ ▶ ▶
▶
pron >non-pron short >long ...
Variety-specific patterns: ▶
effect of end-weight (short before long) is weaker in Indian English and stronger in Jamaican English
The dative alternation across space & register ▶
effect directions follow the overall pattern...: ▶ ▶ ▶
▶
pron >non-pron short >long ...
Variety-specific patterns: ▶
▶
effect of end-weight (short before long) is weaker in Indian English and stronger in Jamaican English recipient pronominality is more important for speakers of Indian and Canadian English than for speakers of Jamaican English
The dative alternation across space & register ▶
effect directions follow the overall pattern...: ▶ ▶ ▶
▶
pron >non-pron short >long ...
Variety-specific patterns: ▶
▶
▶
effect of end-weight (short before long) is weaker in Indian English and stronger in Jamaican English recipient pronominality is more important for speakers of Indian and Canadian English than for speakers of Jamaican English speakers of New Zealand E, Irish E, Jamaican E and Hong Kong E are sensitive to the communicative setting (style)
Findings of the model ▶
all predictors influence the choice of construction following ease of processing (cf. Bernaisch et al. 2014)
Findings of the model ▶
▶
all predictors influence the choice of construction following ease of processing (cf. Bernaisch et al. 2014) Probabilistic differences across the varieties can be observed with regard to processing-related factors (end-weight, pronominality, theme definiteness), semantics (theme concreteness) and style (register)
Findings of the model ▶
▶
all predictors influence the choice of construction following ease of processing (cf. Bernaisch et al. 2014) Probabilistic differences across the varieties can be observed with regard to processing-related factors (end-weight, pronominality, theme definiteness), semantics (theme concreteness) and style (register) → “probabilistic indigenization” (Szmrecsanyi et al. to appear)
Findings of the model ▶
▶
▶
all predictors influence the choice of construction following ease of processing (cf. Bernaisch et al. 2014) Probabilistic differences across the varieties can be observed with regard to processing-related factors (end-weight, pronominality, theme definiteness), semantics (theme concreteness) and style (register) → “probabilistic indigenization” (Szmrecsanyi et al. to appear) Overall constructional preferences by variety: HK, JA, SIN, PHI and IND → prepositional dative GB, IRE, NZ and CAN → ditransitive dative
Conclusion
Conclusion I
▶
▶ ▶
processing-related intralinguistic variation across varieties of English (Bresnan & Ford 2010) genitives: possessum length * variety processing-related factors are socially-sensitive
Conclusion I
▶
▶ ▶ ▶
processing-related intralinguistic variation across varieties of English (Bresnan & Ford 2010) genitives: possessum length * variety processing-related factors are socially-sensitive exposure? substrate influence?
Conclusion II ▶
Do cultural differences exist in syntactic alternations where the choice of variant is heavily influenced by cognitive processes?
Conclusion II ▶
Do cultural differences exist in syntactic alternations where the choice of variant is heavily influenced by cognitive processes? ▶
there is a shared cognitive grammar
Conclusion II ▶
Do cultural differences exist in syntactic alternations where the choice of variant is heavily influenced by cognitive processes? ▶ ▶
there is a shared cognitive grammar prevalence of cognitive-related differences rather than social ones
Conclusion II ▶
Do cultural differences exist in syntactic alternations where the choice of variant is heavily influenced by cognitive processes? ▶ ▶
▶
there is a shared cognitive grammar prevalence of cognitive-related differences rather than social ones → cognitive systems evolve not universally but depend on the speech community?
Conclusion II ▶
Do cultural differences exist in syntactic alternations where the choice of variant is heavily influenced by cognitive processes? ▶ ▶
▶
▶
there is a shared cognitive grammar prevalence of cognitive-related differences rather than social ones → cognitive systems evolve not universally but depend on the speech community? → connection between sociolinguistic and cognitive factors? (see Gries 2013:8 on the correlation between register and processing)
Next steps ▶
annotation ▶
▶
▶
▶
▶ ▶
implement the more fine-grained distinction of text types in ICE add a more fine-grained level for complexity (e.g. relative clauses, PPs, etc.) include persistence/syntactic priming and other factors
include data from web-based corpora for the same varieties: Corpus of Global Web-based English (GloWbE) explore by-variety effects explore language contact & substrate influence (WALS)
Thank you!
[email protected] http://wwwling.arts.kuleuven.be/qlvl/ProbGrammarEnglish.html
This presentation is based upon work supported by an Odysseus grant of the Research Foundation Flanders (FWO) (grant no. G.0C59.13N).
References I Bernaisch, T., Gries, S. T., and Mukherjee, J. (2014). The dative alternation in South Asian English(es): Modelling predictors and predicting prototypes. English World-Wide, 35(1):7–31. Bock, J. K. (1982). Toward a Cognitive Psychology of Syntax: Information Processing Contributions to Sentence Formulation. Psychological Review, 89(1):1–47. Bock, J. K. (1986). Syntactic Persistence in Language Production. Cognitive Psychology, 18:355–387. Bock, J. K., Loebell, H., and Morey, R. (1992). From conceptual roles to structural relations: bridging the syntactic cleft. Psychological review, 99(1):150–71. Bresnan, J., Cueni, A., Nikitina, T., and Harald, B. (2007). Predicting the Dative Alternation. In Boume, G., Kraemer, I., and Zwarts, J., editors, Cognitive Foundations of Interpretation, pages 69–94. Royal Netherlands Academy of Science, Amsterdam. Bresnan, J. and Ford, M. (2010). Predicting Syntax: Processing dative constructions in American and Australian Varieties of English. Language, 86(1):168–213. Bresnan, J. and Hay, J. (2008). Gradient grammar: An effect of animacy on the syntax of give in New Zealand and American English. Lingua, 118(2):245–259. De Cuypere, L. and Verbeke, S. (2013). Dative alternation in Indian English: A corpus-based analysis. World Englishes, 32(2):169–184. Geeraerts, D. (2005). Lectal variation and empirical data in Cognitive Linguistics. In Ruiz de Mendoza Ib´ an ˜ez, F. and Pe˜ na Cervel, S., editors, Cognitive Linguistics. Internal Dynamics and Interdisciplinary Interactions, Cognitive Linguistics Research [CLR], pages 163–189. Mouton de Gruyter, Berlin/New York. Geeraerts, D. (2010). La r´ eception de la linguistique cognitive dans la linguistique fran¸caise. M´ emoires de la Soci´ ete de Linguistique de Paris, 18:185–192. Goldberg, A. E. (2002). Surface generalizations: An alternative to alternations. Cognitive Linguistics, 13-14:327–356. Green, G. (1974). Semantics and Syntactic regularity. Indiana University Press, Bloomington. Gries, S. T. (2013). Sources of variability relevant to the cognitive sociolinguist, and corpus- as well as psycholinguistic methods and notions to handle them. Journal of Pragmatics, 52:5–16.
References II
Gropen, J., Pinker, S., Hollander, M., Goldberg, R., and Wilson, R. (1989). The learnability and acquisition of the dative alternation in English. Language, 65(2):203–257. Kristiansen, G. and Geeraerts, D. (2013). Contexts and usage in Cognitive Sociolinguistics. Journal of Pragmatics, 52:1–4. McDonald, J. L., Bock, J. K., and Kelly, M. H. (1993). Word and world order: Semantic, phonological, and metrical determinants of serial position. Cognitive Psychology, 25:188–230. Oehrle, R. T. (1976). The grammatical status of the English dative alternation. Doctoral dissertation, MIT. Perek, F. (2012). Alternation-based generalizations are stored in the mental grammar: Evidence from a sorting task experiment. Cognitive Linguistics, 23(3):601–635. Rappaport Hovav, M. and Levin, B. (2001). Polysemy, monosemy, and the dative alternation. Szmrecsanyi, B. (2010). The English genitive alternation in a cognitive sociolinguistics perspective. In Geeraerts, D., Kristiansen, G., and Peirsman, Y., editors, Advances in Cognitive Sociolinguistics, pages 141–166. Mouton de Gruyter, Berlin/New York. Szmrecsanyi, B., Grafmiller, J., Heller, B., and R¨ othlisberger, M. Around the world in three alternations: modeling syntactic variation in global varieties of English. English World-Wide.
Manual coding of register
register by variety
Mixed-effect model Predictor
Odds ratio
Pr(>—z—)
Significance
REC COMPLEXITY: complex THEME COMPLEXITY: complex REC PERSON: non-local REC ACCESSIBILITY: new REC ANIMACY: inanimate THEME PRON: non-pron REC PRON: non-pron REC DEFINITENESS: indefinite THEME DEFINITENESS: indefinite WEIGHT (rec/theme) THEME CONCRETENESS: non-concrete VARIETY: hk VARIETYire * REGISTERspokform VARIETYire * REGISTERspokinf VARIETYhk * REGISTERspokinf VARIETYhk * REGISTERwrittenform VARIETYjam * REGISTERspokinf VARIETYjam * REGISTERwrittenform VARIETYnz * REGISTERwrittenform VARIETYcan * REC PRONnon-pron VARIETYind * REC PRONnon-pron VARIETYjam * REC PRONnon-pron VARIETYcan * THEME CONCRETENESSnon-concrete VARIETYind * WEIGHT VARIETYjam * WEIGHT
2.4130885 0.5027482 2.4224678 1.4493789 2.6755396 0.2041128 6.9203502 1.7564450 0.4886220 20.5422340 1.5951425 2.0023395 1.9885652 0.5414433 1.9632923 0.3980724 0.5003841 2.3421414 1.9634870 2.5012822 3.1125789 0.2906772 0.2811699 0.3279156 7.3228191
1.55e-05 2.59e-05 4.37e-07 0.00415 1.81e-12 0.00071
