Artificial Intelligence-Assisted Critical Discourse Analysis for Decolonizing Plant Breeding

A major focus of my visiting scholarship at Cornell University, the research project "Decolonizing Plant Breeding: Integrating Indigeneity, Participatory and Epistemic Justice Approaches to Enhance Agricultural Outcomes, Sustainability and Equity in Malawi" explores the profound influence of colonial legacies on plant breeding practices in Malawi. The primary objective is to help increase technology adoption levels in agricultural research and development projects by utilizing Critical Discourse Analysis (CDA) to uncover the colonial narratives embedded in plant breeding discourse and propose alternative narratives that elevate Indigenous knowledge and possibly other counter-narratives, participatory approaches, and epistemic justice.

Colonialism has historically marginalized Indigenous knowledge systems in favour of Western scientific methods, perpetuating inequities in agricultural practices and outcomes. This research is driven by the recognition that smallholder farmers in Malawi, who constitute the majority of the agricultural workforce, rely on traditional knowledge and practices. However, these farmers' voices are often excluded from formal plant breeding programs, leading to unsustainable and inequitable practices.

The study is structured around key research questions that seek to identify the dominant discourses in plant breeding research and policy in Malawi, analyze how these discourses reflect colonial legacies and power dynamics, and uncover the marginalized voices to promote epistemic justice and technology adoption levels. By integrating CDA, the research aims to reveal the power relations and ideological assumptions embedded in the language and narratives used in policy documents, research publications, and other texts shaping plant breeding practices.

The methodological approach involves a comprehensive corpus collection from diverse sources, including policy documents, research publications, interviews, and media articles. Textual analysis will focus on linguistic features such as lexical choices, metaphors, and modality, while interdiscursive analysis will explore intertextuality and discursive practices. Social analysis will examine power relations and ideological assumptions within the broader historical and political context of colonialism and neoliberalism in Malawi. We seek to draw on natural language processing (NLP) and large language models (LLMs) to facilitate this process.

Expected outcomes of this research include the identification of dominant discourses, a detailed analysis of power dynamics, and the uncovering of ideological assumptions in plant breeding discourse. The study aims to develop alternative narratives that promote indigeneity, participatory approaches, and epistemic justice, offering recommendations for incorporating these narratives into policy and practice. The dissemination of findings will be achieved through academic publications, community reports, policy briefs, workshops, and conferences, with a strong emphasis on advocacy to influence policy changes supporting equitable and decolonial research practices.

In conclusion, this research seeks to provide a comprehensive framework for analyzing the language and discourse in plant breeding research and policies in Malawi. By integrating textual, interdiscursive, social, and discourse signature analyses, the approach will uncover the deep-seated power dynamics, ideological assumptions, and epistemic injustices that shape the current landscape. The findings will inform the development of alternative discourses that promote more equitable and sustainable agricultural practices, contributing significantly to the decolonization of plant breeding in Malawi.

22 July 2024