Weighing words across languages: An empirical Usage-based Construction Grammar investigation of weight effects on constituent order in English and Slovak

• Applicant: Dr. Jakob Horsch
• Subject Area: Individual Linguistics, Historical Linguistics ; General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 546954102

How does cognitive processing efficiency interact with morphological typology to influence constituent order across languages? This project addresses this question with an empirical investigation of Heavy NP Shift (HNPS) in English and Slovak, which are at opposite ends of the analytic-synthetic cline. HNPS is a change in constituent order from [V NP PP], which in English is "basic" (Hawkins 1994: 20), to [V PP NP] when the NP is heavier. This happens due to processing efficiency/weight effects; speakers tend to place heavy (long) before light (short) constituents. While HNPS is thoroughly investigated in English, an analytic language with rigid word order, there is a gap regarding Slavic languages, which are more synthetic/more flexible in word order. The study focuses on Slovak, where only information structure has been discussed as influencing word order. However, Goldberg’s Tenet #5 (2003: 19) predicts general cognitive restraints, e.g. weight effects, will result in cross-linguistic generalizations, making it unlikely that Slovak is not susceptible to HNPS. Also, Hawkins’ Performance-Grammar Correspondence Hypothesis (PGCH; 2004: 5) predicts that fixed orders in analytic languages (the basic English [V NP PP] order), due to performance preference, will be mirrored in synthetic languages. Thus, it can be expected that there is a preferred, if not basic [V NP PP] order in Slovak. Nevertheless, it stands to reason that these phenomena are less pronounced in Slovak due to ist syntheticity. Assuming a Usage-based Construction Grammar approach where syntactic patterns such as [V NP PP] are Argument-Structure constructions (ASCs) (Goldberg 2002, 2006), this is the first contrastive constructionist study of HNPS, using a corroborative approach (Hoffmann 2007, 2011) drawing on ‘big data’ from corpora and grammaticality judgments. A new corpus family will be built and published to this end in cooperation with the Ludovít Stur Institute of Linguistics at the Slovak Academy of Sciences, using texts from Radio Slovakia International. These texts, which are composed for oral delivery, allow investigating weight effects on speech planning, not just parsing. Further collaboration will be with the Institute of Czech Language and Theory of Communication of Charles University, Prague. The acceptability judgments will be obtained at two partner universities in the US and Slovakia using the MET method (Bard et al. 1996; Cowart 1997: 73-84; Hoffmann 2013). I predict that Slovak will turn out to be susceptible to weight effects/HNPS much like English (Goldberg’s Tenet #5), and have a preferred [V NP PP] order mirroring that of English (PGCH), but less significantly due to ist typology (synthetic, freer word order). The study addresses questions about cognitive and typological effects on word order, contributing to our knowledge of English and Slovak, and to our understanding of how language works in general.

• DFG Programme: WBP Fellowship
• International Connection: Czech Republic, Slovakia
• Hosts: Dr. Jana Levická; Professor Dr. Radek Simik