Cognitive apprenticeship
A way of teaching that borrows from craft apprenticeship — watching an expert work, doing the work under supervision, getting feedback, taking on more responsibility — and applies it to mental skills like reading, writing, and mathematics. The defining problem it tackles is that the real work of a thinker happens out of sight, inside the head; its teaching moves exist to pull that hidden process into the open where a novice can see and copy it.
An instructional framework, articulated by Allan Collins, John Seely Brown and Susan Newman in the 1989 Glaser Festschrift paper, that adapts the methods of traditional craft apprenticeship — watching the master work, doing the work under supervision, receiving feedback, taking on increasing responsibility — to teaching skills that are cognitive rather than physical, and so are ordinarily invisible. The framework specifies six teaching moves that make the invisible visible: modelling, coaching, scaffolding, articulation, reflection, and exploration. The argument's premise is that the cognitive work of reading, writing, mathematics, and research is real craft work, susceptible to the same apprenticeship logic that makes carpentry teachable; what makes it harder to teach is only that the work cannot be watched directly.
Etymology§
The phrase cognitive apprenticeship is the 1989 paper's coinage. The pedagogical claim it makes is older — Lev Vygotsky's zone of proximal development (developed in the 1920s and 1930s, available in English from 1978) supplies the developmental-psychology underpinning, and Jean Lave's situated cognition programme (her 1988 Cognition in Practice with Etienne Wenger's later Communities of Practice work) supplies the anthropological case that learning happens through participation in real practice. What the 1989 paper added was the instructional-design specification: six concrete moves a teacher can make to run an apprenticeship-style classroom where the practice is cognitive.
The framework's anchor is the contrast it draws with traditional classroom instruction. In a carpenter's shop, the apprentice can see the master plane a board: the work is in the world, the mistakes are visible, the corrections are demonstrable. In a mathematics classroom, the work the expert is doing — choosing which method, monitoring whether the current approach is working, backing out of dead ends, recognising patterns — is happening inside the expert's head and the student sees only the finished solution on the board. The 1989 paper argues that the reason mathematics is hard to teach is partly that the real work remains invisible; the cognitive-apprenticeship moves are the work of making it visible.
The six moves are the operational definition. Modelling: the expert performs the task and narrates the cognitive process out loud, so the novice can hear the false starts, the dead-end checks, the choices that get discarded — the part of the work that finished solutions hide. Coaching: the novice attempts the work while the expert observes and intervenes on specific weaknesses. Scaffolding: the expert provides temporary support (a prompt, a partial solution, a checklist) that is removed as the novice gains capability. Articulation: the novice is required to put the reasoning into words, which both consolidates the schema and makes weaknesses inspectable. Reflection: the novice compares their own performance to the expert's, often via a video recording or marked-up artefact. Exploration: the novice transfers the technique to new domains under decreasing supervision.
The framework's empirical case in the 1989 paper drew on three worked applications. Alan Schoenfeld's problem-solving instruction in mathematics put modelling and articulation at the centre — Schoenfeld solved problems on the board while narrating the metacognitive moves (am I sure this is the right approach? what would tell me it wasn't?), then required students to do the same in pairs. Marlene Scardamalia and Carl Bereiter's procedural facilitation programme provided written prompts that scaffolded the writing process (what do I want my reader to learn from this?) until students internalised the prompts. Annemarie Sullivan Palincsar and Ann Brown's reciprocal teaching method handed off the role of the expert reader to students in turn, with the rest of the group asking comprehension questions — making the cognitive work of skilled reading externally visible.
Cognitive apprenticeship sits in conversation with deliberate practice as a parallel account of skill acquisition. Where Ericsson's framework foregrounds the practitioner's effortful, feedback-rich work at the edge of present ability, the cognitive-apprenticeship framework foregrounds the visible expert process the novice has access to and the scaffolded transition from observation to independent practice. The two are not contradictory — both require feedback, both require working at the edge of capability — but they emphasise different parts of the same loop.
Cognitive apprenticeship is in the vocabulary of most teacher- education programmes; the structural implementation in K-12 classrooms is rarer, constrained by the same time and student- to-teacher ratios that limit mastery learning. Where the framework operates intact, it is most often in graduate education (lab apprenticeship in the sciences, supervisor-and- student in the humanities), in vocational training, and in contemporary tutoring and coaching practice.