Digital Critical Discourse Analysis — Analyzing Power and Ideology in Online Communication

Standard CDA was developed for print and spoken media and focuses on language as social practice. Digital CDA retains all of CDA's theoretical commitments — power, ideology, critique — but extends the analytic toolkit to address digital-specific features: platform affordances, algorithmic curation, multimodality, hyperlinks, user interactivity, and the blurring of producer and consumer roles. The theoretical foundations overlap heavily; Digital CDA is best understood as CDA applied and adapted to digital communicative environments.

Yes, but it typically requires a mixed-method design. Researchers often use computational methods — corpus linguistics, keyword-in-context analysis, topic modeling, or social network analysis — to identify patterns across large datasets, then apply close CDA reading to a purposive sub-sample of theoretically significant texts. Purely manual close reading is only feasible for smaller corpora.

No, and it does not claim to be. Like all CDA traditions, Digital CDA is explicitly value-committed: it takes the side of those disadvantaged by power and aims to expose and challenge ideological domination. Researchers are expected to declare their normative stance and practice reflexivity. This is a feature of the approach, not a flaw — but it does mean Digital CDA findings should be read as critical interpretations, not neutral descriptions.

Key ethical concerns include: informed consent (public posts may still involve privacy expectations, especially in sensitive communities); platform terms of service compliance; data anonymization to protect participants; and power asymmetries between researchers and communities under study. Researchers should consult their institution's ethics board and follow platform-specific data access guidelines.

The choice depends on your research question. Fairclough's three-dimensional model (text, discursive practice, social practice) is well-suited to institutional and media discourse. Van Dijk's socio-cognitive approach is particularly useful for studying the mental models and ideological representations in political or populist discourse. Wodak's Discourse-Historical Approach is strong for tracing how discourses evolve across time and genres. All three have been applied productively in digital contexts.