Measuring free word order: some empirical and modeling perspectives

predavanje na Inštitutu za bolgarski jezik, Sofija, Bolgarija, 28. 1. 2021

Abstract

Languages manifesting flexibility of word order (within the sentence's compositional meaning) have always presented a challenge for modern theories of syntax requiring any deviation from the canonical word order to be grammatically motivated. Parasyntactic motivations such as information structural or stylistic requirements may account for some portion of this flexibility, but not all of it. In addition, native speakers do not necessarily accept canonical and non-canonical word orders to an equal extent. In fact, the latter typically receive lower acceptability scores than the former, albeit above the subjective threshold for what would count as "ungrammatical". Some of the combinatorially possible word orders are not acceptable at all. In this experimental study we scrutinize different word order sequences in a free word order language (Serbo-Croatian) and attempt to isolate independent displacement factors responsible for various elements of the sentence appearing away from their canonical structural positions. We explore differential and cumulative effects of these independent factors to predict speakers' acceptability scores.

Keywords

Free word order;experimental syntax;Serbo-Croatian;sentence acceptability task;

Data

Language:	English
Year of publishing:	2021
Typology:	3.14 - Invited Lecture at Foreign University
Organization:	UNG - University of Nova Gorica
UDC:	81
COBISS:	51120131
Views:	1817
Downloads:	0
Average score:	0 (0 votes)
Metadata: