Abstract: During the past decade, media researchers have intensified the study of media scandals and the role of journalism as an institution that holds social actors responsible for malfeasance and wrongdoings. On a micro level of analysis, the main attention has beendirected towards the journalistic interview and its use to promote the impression of guilt and journalistic neutrality. However, such studies have not been able to address the editing dimension of TV journalism that transforms conversation to another type of communicativepractice – that of communication through TV-flows composed of speech sequences, pictures,and sounds. This doctoral thesis develops a theoretical framework for analysis of the functions of editing inthe process of guilt attribution by journalistic TV-flows – e.g. investigative TV reporting. The purpose is also to contribute to an understanding of the relationship between the communicative competences of viewers and the contextualization of speech acts through the composition of TV-flows. The developed perspective consists of three parts: 1) A division of viewers’ reception of TV-flows in two types of interpersonal relations (to a speaker and to the composer) that involves six levels of cognitive activities. This division is based on the Habermasian notion of communicative rationality; 2) An intent-model, that lists communicative intentions expressed by the composer when speech sequences are merged and pictures are inserted; 3) A guilt-model, that encompasses guilt as a mental structure of ontologically separated elements (e.g. deed,intention, norm) and the associative relations that the viewer uses to create a meaningful whole– a fabula of guilt. The conveyed analysis of three cases of investigative reporting illustrates how the developed framework can be applied in the study of guilt attribution. The analyses also describe several compositional strategies by which the viewer is encouraged to make certain meaning, evaluate, and judge. The strategies concern the following areas: promotion of certain understanding of speech, promotion of certain evaluation of the validity claims, and promotion of certain understanding of the speaker’s intentions. Also strategies of positioning of the reporter in constructed discourses, that enhance the impression of her performances and argumentation, are explored. Furthermore, the composer’s strategies for masking intentions to interfere with the speech acts, by increasing intent ambiguity, are described. The guilt-model is used to understand the workings of the TV-flow on an overreaching level of meaning (the fabula level). Here, the analysis explains the interplay between portrayed intentions and acts, and the different ways in which condemning norms can be activated and highlighted. Furthermore, the model explores the possible employment of categorization in theprocess of guilt attribution (e.g. when properties of an individual are transferred to a group). In sum, this thesis contributes to a new way of understanding the reception of current affairs programs and TV journalism, as relation building between composer and viewer, by means of contextualization of speech acts.