CECL seminar: Matt Riemland of Dublin City University

Seminar

The CECL organizes a seminar on October 27th (12.55-13.55, Room c.228). The seminar will be presented by Matt Riemland of Dublin City University. The title of the seminar is Language power dynamics and linguistic patterns in translation: A multilingual corpus-based investigation. Here's an abstract:

This seminar provides an overview of my doctoral thesis, which I will complete in May 2024. Formative works in descriptive translation studies assert that language power dynamics – discrepancies in the “status” or “prestige” of source languages (SLs) and target languages (TLs) – broadly determine translations’ linguistic features (Baker 1996, 183; Toury 2012, 314). Still, these claims have not been tested with a systematic, empirical investigation involving a variety of languages and linguistic features. The central research question addressed by my doctoral thesis is whether translations from comparatively higher-status SLs tend to exhibit higher levels of SL influence, conceptualized as interference and foreignization.

The project applies comparable corpus methodology. It constructs a corpus of literary prose from the late 19th and early 20th century, where texts are both translated between and originally composed in English, French, German, Italian, Swedish, Croatian, and Irish. The relative status for each selected language is expressed synchronically as an ordinal variable using a novel method of assessing language status that adopts and develops the EGIDS model (Lewis and Simons 2010). The thesis subsequently conducts studies measuring the potential association between SL status and SL influence on the lexical, syntactic, and paratextual features of translations. Lexical interference is operationalized as the relative frequency (RF) of loanwords originating in the SL and attributable to the translator. Syntactic interference is operationalized using a novel metric called the syntactic interference/normalization coefficient (SINC), which measures the extent to which a translation’s RF distribution of part-of-speech (POS) n-grams resembles those of comparable SL and TL texts. Paratextual foreignization is operationalized as the RF of translator-attributed footnotes and endnotes. The studies test for the hypothesized positive association between SL status and each of the response variables using the Kendall rank correlation coefficient. Finally, the results of the three studies are synthesized to determine whether there is a positive association between SL status and SL influence on translations’ linguistic properties.