Discourse (Gee's Capital-D)

In the AI Story

Gee distinguished Discourse (capital D) from discourse (lowercase) in his earliest major work to capture something the ordinary word could not. Ordinary discourse refers to spoken or written language. Capital-D Discourse refers to the entire socially recognized configuration of practices, values, and identity-markers that, together, constitute membership in a community. You cannot master the Discourse of medicine by reading medical textbooks. You must practice medicine — make diagnoses, confront failures, participate in the communal rituals of rounds and case conferences, perform the identity of "physician" in ways the medical community recognizes as legitimate.

The distinction between acquisition and learning is foundational to how Discourses are mastered. Acquisition, in Gee's framework, is the immersive, practice-based process through which first-language speakers develop their native Discourse — fluid, automatic, deeply integrated with identity. Learning is the explicit, effortful, rule-based process through which second-language speakers develop a Discourse in a classroom — conscious, effortful, prone to breakdown under pressure. Both processes produce competence, but the competence differs in kind. Acquired Discourse membership is what senior practitioners possess. Learned Discourse membership is what AI-augmented practitioners risk producing in its place.

AI enables practitioners to produce the surface features of a Discourse without undergoing the acquisition that genuinely constitutes membership. The code looks like code written by a competent engineer. The brief looks like a brief written by a competent lawyer. The essay looks like work from a student who has engaged deeply with the material. The vocabulary is deployed correctly. The structure conforms to convention. The register is right. But the embodied, tacit, identity-constituting mastery that these surface features traditionally signified may be absent — because the practices through which mastery is acquired were bypassed in the production of the output.

The organizational consequence is a specific form of fragility. Under normal conditions, the distinction between acquired and learned Discourse membership is invisible. Both produce adequate outputs. The distinction becomes visible under stress — when the situation is novel, when the AI fails in ways that require the practitioner to fall back on her own depth, when a problem arises that the learned rules do not cover. At those moments, the acquired practitioner draws on situated intuition. The learned practitioner consults documentation, asks Claude, and may arrive at a correct answer through slower, more dependent, more fragile processes.

Origin

Gee introduced Discourse with a capital D in Social Linguistics and Literacies (1990), where the distinction between Discourses and lowercase discourse became the foundational conceptual move of his sociolinguistic framework. The capital-D usage was controversial within linguistics — some colleagues considered the typographic distinction arch — but the concept proved durable because it named something other vocabulary could not capture: the identity dimension of communicative competence.

Key Ideas

Discourses are identity kits. Membership includes not just language but values, practices, and recognized ways of being.

Acquisition differs from learning. Immersive practice produces fluid, automatic mastery; explicit instruction produces conscious, effortful competence.

Surface features can be simulated. AI enables the production of Discourse-conforming output without the acquisition that constitutes genuine membership.

Community validation matters. Discourse membership is recognized by other members; the recognition is part of the identity.

New Discourses emerge. AI-augmented practice is itself an emerging Discourse with its own practices, values, and forms of recognition.

Debates & Critiques

Whether AI-augmented practice constitutes a genuine new Discourse — one that can produce deep identity formation and sustained community recognition — or whether it remains a learned overlay on older Discourses that still require acquisition through traditional practice is a question the current generation of practitioners is living through. The answer is unlikely to be uniform. In some domains, the AI-augmented Discourse will cohere into something genuinely new. In others, it will function as surface competence layered over hollowed-out depth. The difference will depend on whether the affinity spaces that support Discourse formation are deliberately built for the new practices.