Search

Technological advances continue to enable the creation of more and more audiovisual (AV) products, i.e., films, TV shows, podcasts, etc., which are widely disseminated online and accessible to millions of diverse users across the globe. The translation of these AV products remains a significant challenge, resulting in increasing numbers of individuals and groups becoming online volunteers to translate foreign audiovisual products into their domestic markets, i.e., fansubbing – a contraction of ‘fan’ and ‘subtitling’. However, when translating culturally-sensitive information, particularly swearwords, there is reported convergence and divergence between fan-produced subtitles and professional/official subtitles produced by government-owned companies. This dissertation aims to explore the nature and impact of these similarities and differences. To do so, initially a corpus-based translation approach was employed to identify translation patterns of swearwords by these two groups. A corpus of 549,349 words in original English subtitle scripts was collated and aligned with 528,889 professionally and 543,522 non-professionally translated Chinese words from a diverse sample of 57 recent English films. The results showed mostly convergent practices, but fansubbing appeared to adopt a more vulgarising approach when rendering swearwords (55%) than prosubbing (46%). Informed by this corpus information, the study then employed an eye-tracking approach to investigate how audiences receive and cognitively process translated swearwords in films. An eye-tracking experiment collected data from 150 participants who were allocated into one of the five subtitling groups to watch four representative film clips: four different groups for translation strategies of swearwords ranging from low to high profanity, and one control group which saw only the original English same-language subtitles. Established measures of Total Fixation Count, Total Fixation Duration, Mean Fixation Duration and Time to First Fixation were analysed using a series of one-way ANOVAs. In general, these eye-tracking results showed no significant differences between the swearword translation strategies in terms of processing and reception. Pre- and post-task questionnaires were also employed to collect demographic information and qualitative feedback, in addition to the receptive measurement of immersion, satisfaction, enjoyment, comprehension and offensiveness. Generally, there were no significant differences among these measurements across all clips. Further, participants reported functional awareness of swearwords, which helped them to understand the observed characters’ emotions and feelings. As a result, they reported an expectation for a more authentic translation of swearwords that closely reflected the real-life usage of the target audience. The dissertation concludes with a presentation of the empirical and methodological contributions of the research, followed by a critical reflection on its limitations and the identification of future avenues of study.