Search

Long neglected by research on specialized languages, diachronic approaches currently seem to experience their coming of age. However, studies on diachronic aspects of LSPs still show significant heterogeneity, both in objectives and methods. This may reflect their high investigative potential, but also supports the view that, while it is no longer necessary to make a plea for diachronic approaches, the ways in which they can be applied to LSPs warrant further investigation. We propose to adopt the angle of textual terminology to evaluate their applications in LSPs, as, for several years now, this approach has sought to build a systematic framework for tool-assisted methods aimed at analysing variations in specialized corpora, and has endeavored to establish their theoretical tenets, by factoring in actual contexts of use. In particular, textual terminology has questioned the validity of several "pillars" of diachronic studies, discussing such notions as "corpora", "tools", "experts", and "interpretation".

This article presents a first description and a proposal for the classification of the evolution phenomena involved in short-term diachrony in the field of space. It is based on the principles of Textual Terminology and relies on a tool-based analysis of two diachronic corpora. The linguistic methodology is briefly described but the emphasis is on the list of evolution phenomena revealed through our analysis. These results present an original description of knowledge evolution: 17 types of evolution are listed, revealing the heterogeneity and richness of terminology dynamics and offering a descriptive basis to start new subsequent research that would complete this typology

This doctoral dissertation explores the question of the diachronic dimension in terminology using a tool-based approach applied to a diachronic corpus. This study focuses on a request from the French National Space Agency (Cnes) where both knowledge and technology necessarily evolve over the course of “long-term” space projects (approx. 20 yrs.). This phenomenon can be tricky and a source of difficulties, such as poor communication with new coming engineers who arrive during the project or unconscious modifications of the meaning or forms of terms, etc. Learning how to identify this evolution is therefore of prime importance. Our research aims at defining a linguistic method to track down knowledge evolution within texts. This aim allows us to structure a triple point of view: a theoretical one, a methodological one and an applied one. To do so, this doctoral dissertation presents a description of four linguistic cues which can be semi-automatically identified in corpora to observe evolution: frequency, contexts of evolution, variants and syntactic dependencies. The cues are studied in two diachronic corpora from the Cnes: one corpus made up of three editions of a course on optics and optoelectronics (1994-2002) and one technical corpus composed of specification reports from the DORIS project (1989-2000). This context makes it possible to offer theoretical and methodological paths in textual terminology to track down, study and characterize the evolution that is likely to appear in specialized fields, especially within an environment that has been very little explored, i.e. short-term diachrony. The approach developed in this study also sheds light on the interpretation of corpus data, firstly, by combining linguistic cues in order to build up a reliable interpretation (or even a diagnosis) of the evolution through linguistic cues; secondly, by reflecting on the place and role of the experts in analyses and by analyst-expert dialoguing as a driving force in the co-construction of interpretations., Dans cette thèse nous abordons la question de la diachronie dans les langues de spécialité à travers la définition d'une méthode linguistique pour repérer l'évolution des connaissances en corpus. Cette recherche s'ancre dans une demande appliquée émanant du Centre National d'Études Spatiales (Cnes), où la question de l'évolution prend une dimension particulière dans le cadre de projets spatiaux dits « de longue durée » (~20 ans), au long desquels les connaissances impliquées évoluent nécessairement. Ce phénomène, inévitable mais parfois insidieux, est susceptible d'entraîner un certain nombre de difficultés telles qu'une mauvaise communication avec les ingénieurs qui arrivent en cours de projet, une modification non consciente du sens/de la forme des termes, etc. L'objectif est donc de proposer une méthode pour repérer cette évolution, objectif qui permet d'articuler un triple regard sur la diachronie : un regard théorique, un regard descriptif et un regard appliqué. Pour ce faire, cette thèse propose la description de quatre indices linguistiques repérables semi-automatiquement en corpus pour observer l'évolution : la fréquence, les contextes d'évolution, les variantes et les dépendances syntaxiques. Ces indices sont explorés à partir de deux corpus diachroniques : trois éditions d'un cours d'optique et optoélectronique (1994-2002) et un corpus de rapports de spécification du projet spatial DORIS (1989-2000). Ce contexte permet de proposer des pistes théoriques et méthodologiques en terminologie textuelle pour repérer, étudier et caractériser l'évolution susceptible de se manifester dans les domaines de spécialité, en particulier sur de très courts intervalles temporels. La démarche mise en place dans cette recherche permet également d'alimenter la question de l'interprétation des données en corpus : tout d'abord à travers la combinaison d'indices comme moyen de construire une interprétation (voire un diagnostic) fiable de l'évolution à partir d'indices linguistiques ; ensuite à travers la question du rôle et de la place des experts dans l'analyse et du dialogue analyste/expert du domaine comme moteurs d'une co-construction de l'interprétation.