A close relationship between image and text is a key prerequisite for the unproblematic perception of television programmes. The addition of respoken subtitles for hearing-impaired recipients intensifies this relationship in several respects. In this paper, we address two main issues of image-text relations by referring to examples of respoken news items from “10 vor 10”, a Swiss Television newscast.
1. The temporal delay of subtitles constitutes a serious problem for reception, especially when information is conveyed primarily by means of an image (e.g., a diagram or chart that needs to be ‘read’ by the recipient). At the same time, the macrostructural sequenceability of a newscast into thematic sub-units is disrupted for the viewer if the subtitles lag behind the shift from one image block to another.
2. The alteration between on-screen and off-screen speakers presents a significant problem for respoken news items. Indicating the sources of spoken comments is a prerequisite for establishing coherence with regards to content.
In this paper, we highlight and differentiate the problem of not only the quantity but also the consequent density, in addition to the specific quality of information created by text-image relations. Responses to the complex informational challenges of respeaking thus must be sought elsewhere than in simplistic guidelines such as limiting the number of subtitle lines to three (De Linde & Kay 1999: 25).