| Abstract [eng] |
This master’s thesis analyses Todd Field’s film Tár (2022) and its Lithuanian subtitles, examining how power is constructed multimodally and rendered in translation. The study is grounded in Allen’s (1998) theory of power modalities (Power Over, Power To, Power With), Butler’s (1999) concept of gender performativity, cinematic gaze theories by Mulvey (1975), Kaplan (2010), and McNealy (2021), and a multimodal framework combining Gottlieb (1998) and Kress and van Leeuwen (2001). Six scenes are analysed, each aligned with a power modality, revealing that power in Tár is fluid, performative, and shaped by verbal, visual, and auditory semiotic resources. Lydia Tár’s authority emerges not only through speech, but also through posture, gaze, silence, spatial dominance, and collaborative or introspective interactions. The research compares a professional subtitle version with an AI-generated one produced through ChatGPT using structured prompt. Applying Gottlieb’s (1992) subtitling strategies, the study finds that both versions largely preserve verbal expressions of power through transfer and paraphrase. However, professional subtitles better reflect rhetorical tone and multimodal nuance, while AI-generated subtitles tend to follow source syntax more rigidly and sometimes flatten meaning. |
