Identifying gender ideology in web content: Debates about feminism

This dissertation aims to contribute to the development of a critical language awareness for the web. Drawing on the fields of Computer-mediated communication (CMC), Gender studies, and Feminist media studies, it examines ideological discourse on the web by analyzing web sites on controversial gender issues. The primary research question addressed is: What are the textual discourse features that indicate particular ideological positions in web sites addressing two controversial issues: abortion and domestic violence? Data were drawn from 15 web sites about abortion and 15 web sites about domestic violence. Three searches were done for each topic using the Google search engine, and data were collected from the top five sites in the results from each search. The first 200 words from each of three pages (Home, About Us, FAQ) were subjected to three textual analyses: (1) word frequency, (2) overall lexical field, and (3) evaluative lexicon. The hypothesis was that each web site would have an identifiable lexical structure that could be associated with a specific ideology. The textual features of the selected sites were analyzed utilizing methods of computer-mediated discourse analysis (CMDA) (Herring, 2001, 2004) grounded in the research tradition of critical discourse analysis (CDA). Critical discourse analysts make the case that ideology is evident through verbal discourse representations. The present research extends CMDA by applying it to the identification of ideological elements in web pages designed for audiences with varying political perspectives. The application of CMDA to a sample of web pages about controversial topics revealed on-going struggle over key concepts. Results demonstrate that similar discursive strategies -- such as the selective deployment of terms of abuse or endearment -- are used by people and groups holding radically different perspectives. However, analysis of word frequency and lexical structure exposed striking differences in how the central issues in the debates were defined, and these differences tended to correspond to identifiable ideological positions. This study argues that an information economy requires a new level of critical language awareness and tools that aid in the identification of underlying ideology in everyday discourse. CMDA appears to be readily adaptable to the study of everyday ideologies on the web.