Translation choices and their variability: The role of network structure and entrenchment
(2025) 58th Societas Linguistica Europaea (SLE) Conference — Location: Université Bordeaux Montaigne, Bordeaux (26.August.2025)
Files
No attached file found for this publication.
Details
- Authors
- Abstract
- Halverson’s (2003, 2017) Gravitational Pull hypothesis integrates insights from cognitive linguistics and bilingualism (Langacker 2008; Diessel 2019; De Groot 2011) to provide a usage-based theoretical model for translational behavior. Given the notion that translators’ bilingual knowledge is organized in networks whose structure is determined by the number of nodes and the salience of intra- and cross-linguistic connections, Halverson posits that the more entrenched connections (which are more readily activated) may attract translators to particular linguistic choices. Subsequent work has considered aspects of salience besides entrenchment, including recency and cross-linguistic similarity (Halverson 2024; Lefer & De Sutter 2022; Heilmann et al. 2022). My presentation discusses how this model helps to explain translator choices in French translations of English [N+N] sequences, e.g. research firm → société de recherche, entreprise dédiée à la recherche, agence de notation (among other solutions). These items pose particular difficulties (Lefer & De Clerck 2021) due to cross-linguistic formal differences, potentially ambiguous semantic relationships, and the availability of near-synonyms in French. These difficulties can lead translators to different translation solutions. Following Malmkjær’s suggestion that “investigation might prove particularly fruitful” where translator’s “varied realisations” differ most (Malmkjaer 1998: 539), I focus on translation variability as a phenomenon of interest in its own right, taking advantage of the recent Multilingual Student Translation corpus (MUST; Granger & Lefer 2020), which contains multiple translations for each source text. From 10 source texts (specialized in sustainable finance) and 168 translations, I have identified 117 noun sequences with 4,444 French translations. The main research question concerns how translation choices and their variability are influenced by network structure, i.e. by the number and relative salience of different connections. I combine multiple operationalizations to examine the role of entrenchment. Monolingual and bilingual reference corpora provide frequency data for general and specialized language. As simple token frequencies do not capture entrenchment well, my analyses also consider dispersion, association strength, type frequencies (morphological family size), and surprisal and entropy (Gries 2022). These corpus data are complemented with smaller-scale, preliminary data drawn from elicitation and judgment experiments to test for individual entrenchment of [N+N] combinations, using Psytoolkit (Stoet 2017) to collect responses and reaction times. To date, I have collected such data for 4 participants (students who also contributed to the MUST corpus). This combination of corpus and experimental methods reflects a growing trend (Gilquin & Gries 2009; Dąbrowska 2016; De Sutter & Lefer 2020; Serbina & Neumann 2021) and obtains both collective and individual evidence of usage (cf. Schmid’s (2020) Entrenchment and Conventionalization model). My talk presents qualitative and preliminary quantitative results. Descriptions of multiple translations illustrate how semasiological and onomasiological patterns combine to produce the observed variability. The quantitative analyses are expected to confirm that the number and distribution of translation solutions will be congruent with the number and relative salience of connections within and between the source and target languages.
- Affiliations
Citations
Prinzie, T. (2025, August 26). Translation choices and their variability: The role of network structure and entrenchment. 58th Societas Linguistica Europaea (SLE) Conference, Université Bordeaux Montaigne, Bordeaux. https://hdl.handle.net/2078.5/280365