While my own conflicting readings of literacy advice shaped this analysis, I do not assume that others who read this study will share in all or any of my interpretations. This was not the aim of this research, nor is it the outcome. This analysis did not account for the complex and diverse ways in which mothers, fathers and families may take up literacy advice. A genealogical approach to critical discourse analysis does not lend sufficient insight into the mediation of texts in everyday life to permit this. This analysis was primarily concerned with how mothering discourses reflect the literacy ideals of institutions rather than the everyday literacy practices of mothers, fathers, or children. However, this analysis may nevertheless provide an understanding of the discursive web in which domestic literacy work (and also family literacy research) is caught.

As discussed in Chapter Two, the analytic tool of multi-vocality was used to interpret texts not as integrated and unified, but as contradictory, and equally caught in the web of often conflicting voices and discourses. As Flint (1993) confirmed in her study of the Woman Reader, this multi-vocal strategy picks up the contradictions that often exist between the content of its advice, and its intent. Indeed, if reading to their children came as naturally to mothers as advice claimed, it would not be necessary to provide so much advice on its benefits and how to best carry it out. But while a multi-vocal analysis contributed to exposing the internal contradictions in advice, it also suggests the need for further investigation into the ways in which literacy advice is taken up by mothers and fathers in their everyday lives, and how domestic literacy work is negotiated and carried out in the context of complex and changing gender relationships.

A Foucauldian approach to discourse analysis privileges discontinuity. Yet so powerful was the continuity in advice discourses that more and more texts were included in the analysis to provide evidence of breaks or new discursive formations in literacy advice. This made it necessary to expand the breadth of historical time included in the study, as well as the themes and categories that were generated. Indeed, as Phillips and Jorgenson have stated, the point of discourse analysis is not to exhaust categories but rather generate them (2002, p. 74).