Project 3

Form and meaning as factors in the identification and learning of constructional slots – English phrasal verbs and verb-preposition combinations

While substantial evidence supports the idea that linguistic knowledge is present in the mind in the form of constructions, as defined in Construction Grammar (CxG) and usage-based approaches, there is limited research on how specific constructions are stored and processed in the brain. Cappelle, Shtyrov, and Pulvermüller (2010) argue, based on mismatch negativity, that phrasal verbs (e.g., “look up”) are processed in the brain similarly to single lexemes, cautioning against completely discarding the lexicon-syntax distinction (Pulvermüller, Cappelle & Shtyrow 2013: 415). This project aims to investigate whether phrasal verbs, as idiomatic combinations, exhibit unique processing characteristics compared to other constructions containing lexical elements, such as argument structure constructions with prepositions (e.g., “agree on”) or clauses (e.g., “think that…”, “aim to do”). Additionally, the study will explore how these effects compare with constructions lacking such lexical elements, like the ditransitive construction (e.g., “X VERB Y Z”).

The research, utilizing fMRI, MEG, and EEG data, intends to contribute valuable insights into delimiting specific constructions, slot-fillers within constructions, and the nature of chunks in relation to both lexical associations and more schematic slots. Behavioral experiments, including eye-tracking and neurolinguistic measurements, will be conducted. The project’s initial phase involves designing a narrative text and isolated test sentences featuring the constructions of interest. Test items will be identified through cluster and frequency analyses of present-day English corpora. Subsequently, eye-tracking and brain activity measurements will be performed on approximately 50 native English speakers to ascertain reactions to phrasal verbs and assess any differences compared to other construction types.

This research goes beyond previous studies by incorporating running text, allowing for a more natural examination of subjects’ performance in comparison to isolated sentences. The methodology, developed at FAU, involves multi-modal neuroimaging measurements during continuous speech perception and employs multivariate cluster analysis of resulting spatiotemporal neural activation patterns (Schilling et al. 2021; Kriegeskorte, Mur, & Bandettini 2008; Krauss et al. 2018; Schilling et al. 2021).

This project is done by Hassane Kissane and is supervised by Prof. Dr. Thomas Herbst and Dr. Patrick Krauß.