Unveiling Hidden Cognitive and Ideological Architectures in Language Teaching

Abstract

Modern foreign language (MFL) teaching is informed by cognitive schemas, ideological discourse and technology-mediated instructional approaches, but these dimensions are often addressed separately. This study set out to explore how cognitive, ideological, and technical architectures affect experiences of language learning in the current digital environment and to advance an integrated model that repurposes language teaching as a multi-layered semiotic system. A mixed-methods design was employed using a tri-varied research framework that combined Cognitive Linguistics Analysis, Critical Discourse Analysis, and Technology-Enhanced Learning. Data were obtained from 120 hours of classroom discourse, 86 teaching materials, and 214 student responses collected between 2024 and 2026. Results reveal that classroom language is influenced by implicit mental schemas, discursive power relations, and algorithmic mediation. Quantitatively, the density of conceptual metaphors in classroom discourse positively correlated with learner understanding (r = .62, p < .05). Teacher explanation, measured by time spent on didactic instruction, decreased student engagement (β = .01), while dialogic discourse, measured by the frequency of teacher-student exchanges, increased engagement (β = .36, p < .05). Conversely, prescribed use of modality showed a negative association with engagement (β = -.29, p < .05). At the qualitative level, modality, transitivity and evaluative language indexed authority and transmitted vertical epistemic relations of knowledge. The AI-mediated feedback also exhibited tendencies toward standardisation; for instance, 83% of non-standard language forms were marked as incorrect. The authors argue that cognitive, ideological and technological dimensions are co-constitutive rather than separate. It has to do with more critical, responsive and equitable language pedagogy.