This thesis is about television news. I conducted a qualitative study of the decoding of television news on an opportunistic sample of 38 participants with whom I watched news programmes and then conducted individual or paired focused discussions about their thoughts and feelings as they watched. While problems of representativity and scale preclude our making demographic statements as to the prevalence of decoding practices, this database enables me to perform a critical interrogation of two seperate strands of scholarship relating to TV news. I am concerned to interpose a series of caveats as to the complexity and subtlety of interpretive practice which intervenes between the news and any ideological effect it might exert. Secondly I wish to indicate some problems in that genre of empirical studies concerned with 'learning from news' and 'information gain', which do not exhaustively capture the decoding process. I look at how we might study reasoning and inference in relation to the news, and what happens when people confess themselves unable to remember or understand, since these are areas which are not fully probed by information gain studies. I focus on the resources of meaning and reasoning strategies employed in understanding news. I also extend attention to some areas not normally considered in news audience studies, namely the expression of emotions in relation to news, particularly news about South Africa; and ludic or playful decoding. Memory is the crucial structuring construct of most mainstream research on the television news audience. I begin to problematise the nature of memory by indicating how memories are mutually produced, rather than originating entirely in internal psychic storage. I am also trying to develop ideas of social cognition and how they might be applicable to the business of decoding and the meanings which are developed between people rather than the conventional focus on decodings which are produced by individual viewers as finished products, I also try to develop a reflexive conception of how the conduct of the discussions might construct the thinking and behaviour of the particpants, particularly by reference to their apologies and the expectations they seemed to entertain about the research situation.