Computational Discourse Analysis
Audrey Alejandro, Computational Discourse Analysis: A Methodological Framework. Preprint. 2026.
Abstract:
Computational Discourse Analysis is an emerging methodological approach that bridges discourse analysis with natural language processing and quantitative text analysis. While computational discourse analysis is gaining traction across a growing number of case studies, this momentum has not yet been matched by the development of dedicated methodological resources. This article addresses this gap by developing a methodological framework that makes sense of the diverse and scattered works falling under the scope of computational discourse analysis to legitimise this innovative approach and make it accessible to a broad audience. I begin by tracing the methodological traditions that underpin computational discourse analysis: discourse analysis and quantitative text analysis/ natural language processing. I then identify two primary modes through which it is currently practised: “mixed-method” computational discourse analysis and “text-as-discourse” computational discourse analysis. Building on this foundation, I address terminological and practical confusions and clarify what computational discourse analysis is and is not. Finally, I outline the core rationale and key benefits of adopting a computational discourse analysis approach: enhanced rigour, greater efficiency, and increased analytical value. In doing so, the article provides conceptual grounding for a broader methodological conversation at the intersection of qualitative and computational text analysis.
Key words: Computational Discourse Analysis; Quantitative text analysis; Qualitative text analysis; Natural language processing; Mixed methods research; Computational Qualitative Inquiry; Methodological pluralism.
Abstract:
Computational Discourse Analysis is an emerging methodological approach that bridges discourse analysis with natural language processing and quantitative text analysis. While computational discourse analysis is gaining traction across a growing number of case studies, this momentum has not yet been matched by the development of dedicated methodological resources. This article addresses this gap by developing a methodological framework that makes sense of the diverse and scattered works falling under the scope of computational discourse analysis to legitimise this innovative approach and make it accessible to a broad audience. I begin by tracing the methodological traditions that underpin computational discourse analysis: discourse analysis and quantitative text analysis/ natural language processing. I then identify two primary modes through which it is currently practised: “mixed-method” computational discourse analysis and “text-as-discourse” computational discourse analysis. Building on this foundation, I address terminological and practical confusions and clarify what computational discourse analysis is and is not. Finally, I outline the core rationale and key benefits of adopting a computational discourse analysis approach: enhanced rigour, greater efficiency, and increased analytical value. In doing so, the article provides conceptual grounding for a broader methodological conversation at the intersection of qualitative and computational text analysis.
Key words: Computational Discourse Analysis; Quantitative text analysis; Qualitative text analysis; Natural language processing; Mixed methods research; Computational Qualitative Inquiry; Methodological pluralism.