While experimental studies can by now be judged as a standard technique in the study of individual grammars, their application in cross-linguistic and typological studies is a more recent development which has created a new paradigm of cross-linguistic research. This paradigm promises various advancements in our understanding of Between-Languages differences: (a) it increases the Reliability of the compared observations across languages; (b) it allows for precise estimations of Gradient Phenomena, such as word order preferences or effects of context.
This seminar focuses on the investigation of argument structure with a focus on psych-verb grammar: our current knowledge relies on fine-grained intuitions about word order preferences, binding possibilities, compatibilities with indicators of aspectual categories, etc. These phenomena involve gradience that cannot be precisely estimated without the use of exact quantitative methods. Moreover, a part of the assumed variation between languages is certainly due to the variation between grammarians, since this type of judgments frequently varies between authors (also within the same language). Hence, the use of exact data-gathering methods and the development of experimental and corpus designs is crucial for understanding the effect of argument asymmetries on the grammatical phenomena relevant in this field. The present course discusses empirical evidence from three different methodological paradigms on the same grammatical phenomenon, thereby highlighting the potential contribution of each method to language comparison.
Lecture 1. Controlled speech production
Controlled speech production is a means to observe speakers’ behavior under identical conditions. A sentence formation study on German, Greek, Chinese, and Turkish shows that only the two former languages display psych effects on the linearization with transitive experiencer-object verbs, while the two latter languages do not provide evidence for an exceptional behavior at the syntactic level (Verhoeven 2014). Methodological issues: sampling for cross-linguistic experiments; statistical comparisons between languages with generalized mixed-linear models.
Suggestions for Reading
Jaeger, F.T. & Norcliffe, E. J. 2009. The cross-linguistic study of sentence production. Language and Linguistics Compass 3:866-887. (http://onlinelibrary.wiley.com/doi/10.1111/j.1749-818X.2009.00147.x/epdf)
Verhoeven, E. 2014. Thematic prominence and animacy asymmetries. Evidence from a cross-linguistic production study. Lingua 143, 129-161.
Lecture 2. Cross-linguistic corpus studies
The findings in controlled speech production will be validated by a large-scale cross-linguistic corpus study (10 000 tokens per language). This study examines preferences in the choice of word order and the choice of subject in German, Greek, Turkish, and Chinese and the influence of factors that license not canonical word orders, i.e., disharmonic alignment of thematic roles with the animacy hierarchy and the referentiality scale (see Verhoeven 2015 on German; Verhoeven 2016 on Greek, Turkish, Chinese). The results reveal several differences between the languages that are traced back to their word order properties, pro-drop and the morphological strategies for passive and (anti-)causative formation. Methodological issues: sampling for cross-linguistic corpus studies, corpus design with verbs as random factors.
Suggestions for Reading
Bresnan, J. & Ford, M. 2010. Predicting syntax: Processing dative constructions in American and Australian varieties of English. Language 86(1):186-213. (https://web.stanford.edu/~bresnan/86.1.bresnan.pdf)
Haig, G. & Schnell, S. 2016. The discourse basis of ergativity revisited. Language 92(3): 591-618. (https://www.academia.edu/12395366/The_discourse_basis_of_ergativity_revisited._Haig_Geoffrey_and_Stefan_Schnell._2016._Language_92_3_pp._591-618._)
Verhoeven, E. 2015. Thematic asymmetries do matter! A corpus study of German word order. Journal of Germanic Linguistics 27.1, 45-104.
Lecture 3. Speakers’ intuitions
A crucial part of the relevant generalizations for argument structure is gained through fine-grained intuitions about thematic and event-structural properties of particular verbs, i.e. evidence that cannot be detected with preferences in speech production. In this lecture, we will discuss a cross-linguistic acceptability study that investigates agentivity and stativity as determinants of psych verb grammar in German, Greek, Turkish, Chinese and Yucatec Maya (Verhoeven 2010). Furthermore, by means of judgments of contextual felicity we will compare accusative fronting and dative fronting in German, Greek, Hungarian, and Korean (Temme & Verhoeven 2016). This study shows that the licensing conditions for arguments with the same morphological case differ across languages, which correlates with assumptions about the role of the case in the constituent structure of the languages at issue. Methodological issues: developing stimuli for cross-linguistic acceptability studies; control of lexicalizations and intervening grammatical factors.
Suggestions for Reading
Simík, R. & Wierzba, M. 2017. Expression of information structure in West Slavic: Modeling the impact of prosodic and word order factors. to appear in Language (https://ling.auf.net/lingbuzz/003304).
Temme, A., Verhoeven, E. 2016. Verb class, case, and order: A crosslinguistic experiment on non-nominative experiencers. Linguistics 54.4.
Verhoeven, E. 2010. Agentivity and stativity in experiencer verbs: Implications for a typology of verb classes. Linguistic Typology 14, 213-251.
Lecture 4. Comparison between methods
The methodologically interesting questions are: (a) whether the measures that we get with different empirical approaches (judgments, production) correlate and (b) how we estimate the mapping between different types of data. We will present evidence that the gradience in the intuition of agentivity of particular verbs correlates with the word order frequencies of these verbs (German and Greek). The mapping between different data types will be examined with models based on logistic regression.
Suggestions for Reading
Bresnan, J. 2007. Is syntactic knowledge probabilistic? Experiments with the English dative alternation. In S. Featherston & W. Sternefeld, eds., Roots: Linguistics in search of its evidential base. Mouton de Gruyter, Berlin: 75-96. (https://web.stanford.edu/~bresnan/Roots_05_Bresnan.pdf)
Featherston, S. 2005. The Decathlon Model of empirical syntax. In S. Kepser & M. Reis, eds., Linguistic evidence: empirical, theoretical, and computational perspectives. De Gruyter, Berlin: 187-208. (http://www.el.uni-tuebingen.de/sam/papers/LingEvid.pdf)
Verhoeven, E. & Temme, A. 2017. Word order acceptability and word order choice. Featherston et al. eds. Linguistic Evidence 2016 Online Proceedings. Tübingen: Universität Tübingen.
Verhoeven, E. 2017. Features or scales in word meaning? Verb classes as predictors of syntactic behaviour. Belgian Journal of Linguistics, Special Issue NWASV2.