This research concerns policy, and the application of Critical Discourse Analysis (CDA) as a fit-for-purpose approach to policy analysis in the field of education. It is set against the background of an evolving understanding of how the policy process is understood, from one in which decisions are made in a series of defined steps, described by Deborah Stone (2012) as “the rationality project”, to an understanding of the policy making process as the struggle over ideas in political communities (Ibid). A public policy such as an education policy comes about through the many struggles and contestations inherent in the activities of policy actors, both national and international who wish to influence policy in many different ways. At a certain point in the policy process a policy text is produced. The written policy text is important because it marks a point in the policy process where a coalition of ideas emanating from the ‘struggles’ mentioned above, make their way in the form of discourse into a written policy text. Discourse is a centrally important concept in this study and discourses are defined by Van Leeuwen as “social cognitions, socially specific ways of knowing social practice” (2008, p.6). An overall objective in this research is to analyse a specific education policy, that of literacy, with a view to furthering our understanding of how certain ideas are reflected in a number of key literacy related policy texts, and certain other ideas are absent. A further objective is to understand why this happens. This paper is therefore based on questions of ‘language use’ and ‘power’ - the exercise of power is necessary if certain ideas are to be accommodated and reflected in policy, while others are not. Clearly the ability both to make and to influence policy is a significant source of power in modern societies. The texts in question include curriculum and policy texts and were produced between 1971 and 2015.