Jump to: navigation, search

Jan Rybicki

The Visibility of Translator Gender?


In an analysis of 18th century English writers of both genders, performed by means of multivariate analyses of word tokens (Burrows’s Delta and Zeta methods, 2007), it has been shown that there exist words characteristic of each gender; that these words are stereotypically male and female (quite an expected phenomenon in 18th- century fiction); and that a comparison of their frequencies allows to detect a strong gender signal in a mixed corpus (Rybicki 2012). The present study goes a step further: its aim is to see if gender is preserved on the lexical level in translations, and if the gender signal, once detected, is that of the original author or of the translator. Obviously, in this way, four “translational genders” can be said to exist: male translated by female, male translated by male, female translated by male, female translated by female.

This was tested for a parallel corpus of English novels from Swift to Dickens and their French translations. For the English originals, it was quite easy to detect the gender signal, and again the stereotypically “male” and “female” words allowed to divide the corpus according to gender. The French translations presented a much more complex material. It was possible to identify a purely male-by-male group that included most of the male writers except Dickens; translations of Dickens usually clustered together irrespective of the translators’ gender; texts by other writers also tended to cluster together, but there was no regular group of translations of novels by women. When translations of the author of Little Dorrit were removed from the corpus, the results were much clearer, and the translations divided into two discrete groups by original author gender.

These results do not come as a total surprise. First of all, Dickens’s “transgender” lexical behaviour has already been observed stylometrically. Indeed, this might be attributable to the broad social perspective of his novels, and a multitude of female characters, as compared to the other texts in the corpus. Secondly, several studies have already shown that the authorial lexical signal remains strong in translations (Rybicki 2012a), and that the signal of the translator is usually only discernible when translations of texts by the same author are compared (Rybicki and Heydel, 2013). Other, possibly non-lexical methods are needed to highlight the signal of the translator and of his/her gender.

Works Cited:

  • Burrows, J.F. 2007. “All the Way Through: Testing for Authorship in Different Frequency Strata,” Literary and Linguistic Computing, 22 (1): 27-47.
  • Rybicki, J. 2012. “Visualizing the femininity of the Chawton House corpus,” Transcultural, Transnational, Trans-disciplinary Perspectives on Women’s Literary History. International conference presenting the 3rd COST-WWIH Milestone, 26-28.11.2012, Adam Mickiewicz University, Pozna?.
  • Rybicki, J. 2012a. “The great mystery of the (almost) invisible translator. Stylometry in translation.” In: Oakes, Michael P. and Meng Ji (eds.), Quantitative Methods in Corpus-Based Translation Studies: A practical guide to descriptive translation research, 231–248.
  • Rybicki, J. and Heydel, M. 2013. “The Stylistics and Stylometry of Collaborative Translation: Woolfs 'Night and Day' in Polish,” Literary and Linguistic Computing (in print).

SvD, 29 May 2013

Personal tools