While language style is considered to be automatic and relatively stable, its plasticity has not yet been studied in translations that require the translator to “step into the shoes of another person.” In the present study, we propose a psychological model of language adaptation in translations. Focusing on an established interindividual difference marker of language style, that is, gender, we examined whether translators assimilate to the original gendered style or implicitly project their own gendered language style. In a preregistered study, we investigated gender differences in language use in TED Talks ( N = 1,647) and their translations ( N = 544) in same- versus opposite-gender speaker/translator dyads. The results showed that translators assimilated to gendered language styles even when in mismatch to their own gender. This challenges predominating views on language style as fixed and fosters a more dynamic view of language style as also being shaped by social context.
more » « less- PAR ID:
- 10545161
- Publisher / Repository:
- SAGE Publications
- Date Published:
- Journal Name:
- Social Psychological and Personality Science
- Volume:
- 12
- Issue:
- 1
- ISSN:
- 1948-5506
- Format(s):
- Medium: X Size: p. 131-142
- Size(s):
- p. 131-142
- Sponsoring Org:
- National Science Foundation
More Like this
-
Translations help people understand content written in another language. However, even correct literal translations do not fulfill that goal when people lack the necessary background to understand them. Professional translators incorporate explicitations to explain the missing context by considering cultural differences between source and target audiences. Despite its potential to help users, NLP research on explicitation is limited because of the dearth of adequate evaluation methods. This work introduces techniques for automatically generating explicitations, motivated by WikiExpl: a dataset that we collect from Wikipedia and annotate with human translators. The resulting explicitations are useful as they help answer questions more accurately in a multilingual question answering framework.more » « less
-
Canonical morphological segmentation is the process of analyzing words into the standard (aka underlying) forms of their constituent morphemes.This is a core task in endangered language documentation, and NLP systems have the potential to dramatically speed up this process. In typical language documentation settings, training data for canonical morpheme segmentation is scarce, making it difficult to train high quality models. However, translation data is often much more abundant, and, in this work, we present a method that attempts to leverage translation data in the canonical segmentation task. We propose a character-level sequence-to-sequence model that incorporates representations of translations obtained from pretrained high-resource monolingual language models as an additional signal. Our model outperforms the baseline in a super-low resource setting but yields mixed results on training splits with more data. Additionally, we find that we can achieve strong performance even without needing difficult-to-obtain word level alignments. While further work is needed to make translations useful in higher-resource settings, our model shows promise in severely resource-constrained settings.more » « less
-
null (Ed.)The study of gender in engineering continues to be highly relevant due to the persistence of the field’s domination by men and masculinity. Mainstream discourse on gender in STEM, however, has been kept in a “black box” for decades according to Allison Phipps [1]. She states that the reliance on a simplistic gender binary unaccompanied by racial, cultural, or sexual identity nuances undermines engineering’s own political aims of gender equity. One large gap in our existing body of gender research and discourse is how the highly gendered landscape of engineering education is experienced transgender or gender nonconforming (TGNC) people. We are opening the “black box” on gender in our research project “Invisibilized Gendered Experiences: Transgender and Gender-nonconforming Experiences in Engineering Education.” The research project contains three key objectives: 1) To infuse queer studies and feminist research methodologies into engineering education research practice 2). To record, examine, and share the wide range of experiences from TGNC engineering students to our research community, and 3). To collaborate with the student community to inform the research products for engineering educators and researchers. This presentation will first introduce the audience to conceptualizations of gender informed by contemporary queer theory, which defines gender as a fluid and dynamic social system beyond biological binaries. Next, we will use our own research project as an example of how we can transform our approach to the study of gender through feminist research methodologies that place the subject community as the experts on their lived experiences. We end by sharing prominent themes from the 300 national participants in our initial community outreach questionnaire. This final portion provides ideas offered by the undergraduate TGNC population on how to build a more supportive and inclusive environment for their success. Altogether, this presentation provides a comprehensive look at our project on gender resiliency by TGNC students in engineering education.more » « less
-
null (Ed.)Neural Machine Translation (NMT) performs training of a neural network employing an encoder-decoder architecture. However, the quality of the neural-based translations predominantly depends on the availability of a large amount of bilingual training dataset. In this paper, we explore the performance of translations predicted by attention-based NMT systems for Spanish to Persian low-resource language pairs. We analyze the errors of NMT systems that occur in the Persian language and provide an in-depth comparison of the performance of the system based on variations in sentence length and size of the training dataset. We evaluate our translation results using BLEU and human evaluation measures based on the adequacy, fluency, and overall rating.more » « less
-
The goal of the current study was to better understand the development of gender typicality in young adulthood by applying the dual-identity approach to gender typicality, previously developed with children, to a university sample. Participants ( n = 215, Mage= 20.20 years; 62% female) were asked to rate their perceived similarity to both own- and other-gender peers. They also completed questionnaires assessing sexist attitudes, internalized sexualization (females), adherence to male-typed behaviors in the context of interpersonal relationships (males; adherence to physical toughness and restrictive emotional expressivity), gender-based relationship efficacy, friendships, self-esteem, social self-efficacy, and social anxiety. Results indicated that self-perceived gender typicality involves comparisons to both gender groups, and that meaningful typologies can be created based on similarity to own- and other-gender groups. As with children, results indicated that identifying with one’s own gender was advantageous in terms of low social anxiety and relationships with own-gender peers. For adults who identified with both own- and other-gender peers, we identified additional social benefits (i.e., efficacy and friendships with other-gender peers). Further, we identified a downside to own-gender typicality: individuals who identified only with their own gender had more sexist attitudes than those who identified with the other gender. Findings support the viability of the dual-identity approach in young adults, and have implications for researchers assessing gender typicality across development.