Allomorphs of French de in coordination: a reproducible study

Abstract It is known that French de ‘of’ can take wide scope in coordination – that is, the coordination can optionally be reduced by omitting the second de : de X et/ou (de) Y , meaning roughly ‘of X and/or (of) Y’. De has an allomorph d’ that is used when the following word begins with a vowel. Th...

Full description

Bibliographic Details
Published in:Linguistics Vanguard
Main Author: Zuraw, Kie
Format: Article in Journal/Newspaper
Language:English
Published: Walter de Gruyter GmbH 2015
Subjects:
Online Access:http://dx.doi.org/10.1515/lingvan-2014-1017
https://www.degruyter.com/document/doi/10.1515/lingvan-2014-1017/xml
https://www.degruyter.com/document/doi/10.1515/lingvan-2014-1017/pdf
Description
Summary:Abstract It is known that French de ‘of’ can take wide scope in coordination – that is, the coordination can optionally be reduced by omitting the second de : de X et/ou (de) Y , meaning roughly ‘of X and/or (of) Y’. De has an allomorph d’ that is used when the following word begins with a vowel. This paper shows, using a large written corpus, that the two allomorphs, de and d’ , do not behave the same when it comes to reduction/wide scope. Two main factors seem to be at play: resistance of the d’ allomorph to taking wide scope, and hiatus avoidance between et/ou (which are both vowel-final) and a following vowel-initial word. The existence of phonological factors that affect reduction rate implies that the grammar and/or processing architecture must retrieve some phonological information about X and Y before the final “decision” about reduction is made– or that the phonology is powerful enough to delete the second de on its own. This paper also aims to make a methodological contribution to reproducibility. The web materials accompanying the paper (scripts, documentation, and intermediate-stage data files, available at TROLLing, the Tromsø Repository of Language and Linguistics, opendata.uit.no/dvn/dv/trolling) allow the reader to reproduce all the steps of the data processing analysis, starting from a publicly available corpus.