Processing and representation of clitic doubling and scrambling (working title)

Language is not only a structural system of items and rules but also a social phenomenon of communicative interaction deeply embedded in human cognition. As such, it is a complex instance of human behaviour and therefore ultimately processed by the brain. This issue appears even more intriguing if we take into account the vast number of different languages – estimates range between 6.000 to 15.000 distinct languages – that are currently spoken all over the world. Despite this large variability, linguistic research has disclosed certain shared principles underlying the structural and functional organization of different languages. For instance, the encoding and decoding of argument structure (“Who did what to whom?”) is heavily influenced by information structural principles (“Which communicative information is shared by the interlocutors? In which order is the information presented by the speaker?”) and linguistic prominence principles (e.g., referentiality hierarchies; semantic role hierarchies etc.). Neurolinguistics – especially in a neuro-typological framework – has started to investigate links between these shared principles operative in linguistic systems and mechanisms of the general cognitive system such as attention, working memory and prediction. These insights will contribute to our understanding of the roots of language processing in human cognition. The principles at work in shaping argument structure induce certain language-specific ways of encoding following extensive preferences (e.g., canonical word order, agent > patient). However, due to special communicative needs the principles may be in conflict and require atypical encoding strategies. To facilitate the correct interpretation of a sentence, all language systems exhibit specific formal means for indicating such deviances from the hierarchy-based preference (e.g., differential object marking for animacy asymmetries; cleft sentences for focus marking of an object). From a processing perspective, it is assumed that the parser constantly makes feature-based predictions with regard of the elements that are expected to turn up next. There is a growing body of evidence that at least some predictions can be correlated to the principles mentioned above and that atypical patterns cause prediction errors indicated by surprise and costly reanalysis effects similar to structural violations.

In my project, I investigate clitic doubling (CD) and clitic left dislocation (CLLD) in Bulgarian as an example of a special coding mechanisms for marking object arguments. In CD/CLLD, an object argument is represented by both, a full noun (or pronoun) and a short clitic pronoun:

1. Izborite gi spečeli opozicijata.                                                                           Clitic left dislocation

elections-ART-PL CL.ACC.3PL win-PTCP-3SG opposition-ART-SG.F

“The opposition won the elections.”

     2. Decata (gi) bojadisaha stenite.                                                                         Clitic doubling

child.PL.N CL.ACC.3PL paint.PFV-PRF.3PL wall-PL.F-ART-PL.F

“The children painted the wall.”

In (1), the object of the sentence is placed in a sentence-initial position typically reserved for subjects and hence marked as an object with a short pronominal clitic. In (2), the sentence follows the canonical word order and yet the object is doubled with co-referring clitic. It is not quite clear which principles enforce the deviant word order and doubling in (1) or the doubling in canonical word order in (2). Previous research has shown that CD/CLLD are typically related to or dependent on referentiality (definiteness or specificity), animacy, familiarity, topicality or a combination of them. For Bulgarian, it was suggested that CD/CLLD are disambiguation markers, (secondary) definiteness markers or topic markers. So far, the claims for either hypothesis were not addressed sufficiently from an empirical perspective. Additionally, taking into account current perspectives on information structure, prominence principles and argument structure processing is valuable for a more detailed understanding of CD and CLLD in Bulgarian.

In my project, I investigate some underlying principles determining clitic doubling constructions in Bulgarian by combining behavioural methods (acceptability judgments, interpretation tasks) and electroencephalography (EEG). I focus particularly on the relation of clitic doubling constructions to topicality, givenness and word order in order to research the functions of the constructions further. In a second step, I investigate how processing predictions for argument interpretation are build up in regard of CD/CLLD based on the information structure levels just mentioned and how predictions are altered by the presence of a clitic in doubling constructions. This research ultimately contributes to our functional understanding of the syntax-discourse interface and how interactions at this interface influence predictive mechanisms in regard of argument interpretation during language processing.

Short biography

Paul Compensis (*1992) studied Chinese studies and Economics (B.A.) at the Friedrich-Alexander-University in Erlangen from 2011 to 2014. From 2014 to 2017, he studied Linguistics (M.A.) in Erlangen and at the University of Cologne. He wrote his M.A. thesis on definiteness in Molise Slavic, a South Slavic language spoken in Southern Italy. From 03/2017 to 03/2019, Paul worked as the coordinator of the Cologne Centre of Language Sciences (CCLS) and became an affiliated member of the CRC1252 “Prominence in Language”. During this time, he was co-organizer of two international conferences. Paul started his PhD project in the a.r.t.e.s regular track in April 2018 under the supervision of Prof. Dr. Petra Schumacher (Department of German Language and Literature I). In April 2019, Paul became a research fellow in the a.r.t.e.s. EUmanities programme. From April 2020 to September 2021, he will spend his mobility phase at the University of South Australia in Adelaide where he will continue his project under the supervision of Matthias Schlesewsky and Ina Bornkessel-Schlesewsky. Paul is also board member of the Cologne Centre of Language Sciences.

Contact: paul.compensis(at)uni-koeln.de

Presentations

3^rd Vielfaltslinguistik conference “Clitic doubling in Molise Slavic. A corpus-based classification.”, March 29, 2019, University of Cologne

DeMiNeS Masterclass “Syntactic accounts of clitic doubling and scrambling.”, March 19, 2019, Aristotle university of Thessaloniki

Events organized

International conference “LingCologne2019: Multimodality”, June 6-7, 2019, Cologne Center of Language Sciences, University of Cologne

International conference “Second International Conference Prominence in Language”, July 11-13, 2018, CRC1252 Prominence in Language and Cologne Center of Language Sciences

Awards

2018: Reinhold Olesch Prize of the Department of Slavic Studies for the M.A. thesis “Definiteness in Molise Slavic

a.r.t.e.s. EUmanities has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 713600.

Details:
Call: H2020-MSCA-COFUND-2015 | Proposal: 713600 – artes EUmanities
CORDIS: http://cordis.europa.eu/project/rcn/203182_de.html