Publication type: Conference other
Type of review: Peer review (abstract)
Title: Leveraging MT Literacy (panel)
Authors: Ehrensberger-Dow, Maureen
O'Brien, Sharon
et. al: No
Proceedings: International conference on translation, interpreting and cognition : book of abstracts
Page(s): 15
Conference details: 3rd ICTIC, Forlì, Italy, 2 - 5 November 2021
Issue Date: 3-Nov-2021
Publisher / Ed. Institution: Università di Bologna
Language: English
Subject (DDC): 410.285: Computational linguistics
418.02: Translating and interpreting
Abstract: Thanks to the ready availability of so-called free online services, neural machine translation (NMT) has recently emerged as one of the most important language resources for lay and professional users in various domains (e.g., Nurminen, 2019). The quality of the non-edited output has reached impressive levels for some language combinations, although there are still problems with sociolinguistic-, cultural-, domain and register-inappropriateness. Users with reasonable levels of proficiency in both the source and target languages can recognize such problems and intervene to fix them (i.e., post-edit) or reject the output and translate from scratch. However, the misleadingly fluent quality of the output can deceive less proficient users into assuming that problematic output is actually acceptable (cf. Martindale & Carpuat, 2018). Informed use of MT through MT literacy training has been suggested as an effective way to help prevent the risks associated with the naïve deployment of this technology (e.g., Bowker and Buitrago Ciro, 2019; Cadwell et al., 2019; Nitzke et al., 2019; O'Brien & Ehrensberger-Dow, 2020). A competent level of MT literacy can inform judgements about the suitability of using MT for certain genres, about quality expectations, about risks, and about when intervention by professional translators is required. Those judgements are based on the rich intercultural awareness that translators bring to their work, since they have been trained to recognize and deal with cultural differences, potential ambiguity, terminological inconsistencies as well as conceptual and lexical gaps (e.g., Federici & Declercq, 2019), but are likely to be missing or inadequate for users who do not have such training. In this panel, we would like to go beyond the basics of MT literacy to explore in more depth its cognitive dimensions, including, for example, questions such as: What cognitive processes are at play when lay users evaluate the output of MT? Do the concepts of adequacy and fluency, for example, intuitively form part of their evaluation? What factors are used to judge whether a text needs to be post-edited? What role does trust play? How does MT-mediated communication affect the communicative process overall? Is creativity enhanced or inhibited? What risk assessment factors come into play when MT is used for mediation? What emotional responses does the use of MT for lay users elicit? Are there implications for other AI-based technology? The panel invites papers that respond to these questions and others relating to how MT literacy can be leveraged in various translation environments.
Fulltext version: Published version
License (according to publishing contract): Not specified
Departement: Applied Linguistics
Organisational Unit: Institute of Translation and Interpreting (IUED)
Published as part of the ZHAW project: Digital Literacy im Hochschulkontext
Appears in collections:Publikationen Angewandte Linguistik

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.