ParaDis and Demonette From Theory to Resources for Derivational Paradigms

摘要

This article traces the genesis of the French derivational database Demonette,_(v2) and shows how current architecture and content of derivational morphology resources result from theoretical developments in derivational morphology and from the users' need. The development of this large-scale resource began a year ago and is part of the Demonext project (ANR-17-CE23-0005). Its conception is adapted from theoretical approaches of derivational morphology where lexemes, units of analysis, are grouped into families that are organized into paradigms. More precisely, Demonette_(v2) is basically an implementation of ParaDis, a paradigmatic model for representing morphologically complex lexical units, formed by regular processes or presenting discrepancies between form and meaning. The article focuses on the principles of morphological, structural and semantic encoding that reflect the methodological choices that have been made in Demonette_(v2). Our proposal will be illustrated with various examples of non-canonical word formations.