Does gender affect translation? : analysis of English talks translated to Arabic

Abstract

When a text in a foreign language is translated into English, many of the features of the original language disappear. The tools described in this paper can give people who work with translators and translations an insight into dimensions of a culture that may escape the notice of someone not familiar with the source language or culture. A set of computer programs are described that analyze both English and Arabic texts using each language's function word or closed-class words categories. First, the LIWC (Pennebaker, Booth, & Francis, 2007) text analysis program was translated into Arabic. Then, the grammatical dimensions of Arabic function words was determined that served as a basis for the Arabic LIWC designed for Arabic texts. These same Arabic dimensions were used to fit English words into the same categories. A large corpus of Modern Standard Arabic and English text files that have been translated in both directions were used to establish the equivalence of the translated word lists. Then, the uses and applications of the dictionaries for computer-based text analysis within and across cultures are described in the study of influence of gender on translation of TED talks between English and Arabic. Differences were identified in language style between men and women in their English language TED talks, and these features were examined whether they were faithfully maintained in translations to Arabic. The rates of function word use was employed to measure language style. Function words (e.g., pronouns, prepositions, conjunctions) appear at high rates in both English and in Arabic, and they have been shown to provide social, demographic, and psychological information about authors and speakers in English and a variety of other languages. The sample included 328 (196 male and 132 female) TED talks delivered in English from 2004 to 2010 and their translations to Arabic. Rates of function word use in the original and translated texts were examined using the English version of the word counting software. The function word use compared between male and female speakers, male and female translators, and their interaction. The results confirmed gender differences in language style for English texts found in previous studies in English. For example, women used more pronouns, more negatives, and fewer numbers than did men. It was further found that several of the distinguishing language style features between men and women in English disappeared in Arabic translations. Importantly, there was a significant gender difference in the language style of male and female translators: first person singular pronouns, second person pronouns, conjunctions, and prepositions were used more by female translators, and quantity words were used more by male translators, regardless of the gender of the original speaker. This study presents one application of computerized text analysis to examine differences in language style that may be lost or gained in translations. Future research and applications within personality, forensic, and literary psychology, linguistics, and foreign language studies are discussed.