Evaluating morphosemantic demotivation through experimental and distributional methods - CNRS - Centre national de la recherche scientifique Accéder directement au contenu
Article Dans Une Revue Lingvisticae investigationes : International Journal of Linguistics and Language Année : 2022

Evaluating morphosemantic demotivation through experimental and distributional methods

Résumé

Summary The lexicalization of morphologically complex words, i.e. their inclusion in the lexicon, can involve a loss of semantic compositionality. Such a phenomenon, called demotivation, has been overlooked in both morphological and lexical studies, notably regarding its gradual nature. This paper compares two measures of demotivation based on experimental and distributional semantics approaches. It builds on the evaluation of 78 pairs of French verbs and derived nouns selected to represent three levels of demotivation. The comparison of the two approaches using speakers’ judgements and word vector similarity indicates convergence on the identification of demotivation degrees within a continuum, while also highlighting specific aspects of each method. The study provides direction to further research on morphosemantic demotivation, bridging together semantic, morphological and methodological considerations.
Fichier principal
Vignette du fichier
PHC_preprint_complet_Fribourg.pdf (684 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03897593 , version 1 (14-12-2022)
hal-03897593 , version 2 (19-01-2023)

Licence

Paternité - Pas d'utilisation commerciale - Pas de modification

Identifiants

Citer

Alizée Lombard, Marine Wauquier, Cécile Fabre, Nabil Hathout, Lydia-Mai Ho-Dac, et al.. Evaluating morphosemantic demotivation through experimental and distributional methods. Lingvisticae investigationes : International Journal of Linguistics and Language, 2022, 45 (1), pp.83-115. ⟨10.1075/li.00068.wau⟩. ⟨hal-03897593v2⟩
63 Consultations
88 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More