Situated Learning¶

Treat competence as something formed and recognised inside a community of practice rather than transferred through a curriculum, so a novice learns by doing real low-stakes work at the periphery and deepening toward the centre as the community recognises their growing skill.

Core Idea¶

Situated learning is the theoretical commitment, developed by Jean Lave and Etienne Wenger in Situated Learning: Legitimate Peripheral Participation (1991), that learning is not a transfer of decontextualised information from an instructor into a learner who then applies it elsewhere, but is instead constituted by participation in the authentic activities, social relationships, and material arrangements of a community of practice. What is learned is inseparable from where, with whom, and toward what purpose it is learned, because emerging competence is calibrated against the goals, tools, standards, and recognition practices of the host community rather than against an abstract curriculum.

The structural mechanism is legitimate peripheral participation: the novice enters the community at its periphery doing real, low-stakes work that genuinely contributes to the practice — ironing finished trousers before cutting cloth, answering straightforward patient questions before conducting assessments, first-chairing uncomplicated cases before managing complex ones. The periphery is not a waiting room for the real learning; it is the site where participation is legitimate because the work is real, where observation of full-membership activity is continuous, and where the novice's understanding is shaped by the same community standards that will later evaluate it as competence. Over time, participation deepens and centralises, with the trajectory and pace governed by the community's recognition of growing competence, not by a pre-specified curriculum clock. The claim is that this trajectory is not one route to competence among many; it is the structural description of how durable occupational and craft competence forms, across tailoring apprenticeships in Liberia, medical residencies, legal clerkships, and academic training alike.

Structural Signature¶

Sig role-phrases:

the community of practice — the social unit with shared goals, tools, standards, and membership signals that constitutes the site where competence is formed
the authentic activity — the practice's real work performed for its actual purposes, against which emerging competence is calibrated
the legitimate peripheral entry — the novice admitted at the periphery doing real, low-stakes work that genuinely contributes, not a holding pen before the real learning
the centrality trajectory — the deepening from periphery toward full membership, paced by the community's recognition of growing competence rather than a curriculum clock
the recognition criterion — competence judged by the community from inside the practice, not measured by a test standing outside it
the context-distance scalar — the gap between learning context and practice context that governs whether competence holds and travels, and that syllabus editing cannot close

What It Is Not¶

Not the claim that classroom instruction is worthless. Situated learning does not assert that decontextualised teaching never works; it asserts that competence is constituted at the practice and recognised by its community, so the prediction is conditional on context-distance. Where the learning context closely resembles the practice, content delivery can carry; the frame indicts the gap between learning and practice, not instruction as such.
Not an instructional method or teaching technique. It is a theory of where and by what competence is formed — participation in a community of practice — not a classroom strategy one adopts alongside others. Treating it as a technique mistakes the account of the site of learning for one more item on the pedagogy menu.
Not merely "learning in a realistic setting." Authentic surroundings are necessary but not sufficient: the novice's contribution must be legitimate (real work that genuinely counts), framed by a community whose standards recognise it, with a trajectory toward centrality. A simulation that is realistic but in which nothing the learner does matters to anyone supplies the scenery without the mechanism.
Not a probationary holding pen at the periphery. Legitimate peripheral participation is not a waiting room to be endured before the "real" learning begins; the periphery is the site where competence is formed and calibrated, because the low-stakes work is genuine and observed against the same standards that will later judge it. Reading the periphery as mere antechamber inverts the claim.
Not the substrate-free notion of learning. The substrate-general pattern — durable update of an agent from experience — is the parent prime learning, which recurs in gradient descent, conditioning, and selection. Situated learning is that update plus the human-social-practice stack, and the stack is exactly what fails to generalise; machine-learning "situated" variants borrow the word while the engine is statistical, not legitimate peripheral participation against community recognition.

Scope of Application¶

Situated learning lives across the learning-theory and sociology-of-education subfields wherever a human learner participates legitimately in a recognising community of practice; its reach is bounded to that human-social-practice substrate (the bare durable-update that recurs in machine learners or selection belongs to its learning parent, and machine-learning "situated" variants borrow only the word).

Apprenticeship traditions — Lave and Wenger's source cases: tailoring, midwifery, blacksmithing, fisheries, where the novice does real work on real artefacts around real practitioners and competence is recognised by the community, not by external testing.
Professional and clinical training — medical residency, legal clerkships, teacher-training practicums, engineering co-ops; classroom knowledge is treated as preparation while the load-bearing competence forms through case rotations, supervised practice, and team-based work.
Vocational and trade education — the German and Swiss dual-system apprenticeships, culinary schools integrated with restaurants, and software boot-camps pairing instruction with portfolio projects.
Workplace learning theory — induction, onboarding, and mentorship, framing the observation that most workplace competence is acquired on the job rather than from training courses.
Higher-education reform — problem-based learning, capstone projects, and undergraduate research, all efforts to embed academic content in tasks that resemble its disciplinary use.
Sociology of education and identity — the recognition layer where community membership, not a test score, is the criterion of competence, and learning is bound up with becoming a recognised member of the practice.

Clarity¶

Naming situated learning makes legible an asymmetry that decontextualised, curriculum-centred accounts of education cannot see: competence is recognised by a community of practice, not measured by a test standing outside it. A surgeon is judged a surgeon by other surgeons, with the operating theatre, the patient, and the team forming an irreducible part of what is being assessed; once the criterion is community recognition rather than a transferable score, the periphery stops looking like a holding pen before the real learning and becomes the site where competence is actually formed and calibrated. That reframing dissolves a standing puzzle — why so much durable occupational skill is acquired "on the job" and so little in the classroom — by showing it is not a failure of instruction to stick but the expected consequence of where the competence is constituted.

The frame also sharpens the reading of transfer failure. When a learner can recite content but cannot deploy it in the actual practice, a curriculum view diagnoses insufficient teaching of the content; situated learning instead asks how far the learning context stood from the practice context, treating the gap itself — not the learner's grasp of the material — as the variable to act on. This converts the vague worry "will this education work?" into a set of answerable design questions: is the activity authentic to the target practice, is the novice's contribution legitimate however peripheral, is there a community whose standards frame the work, and is there a trajectory from periphery toward centrality. Distinguishing the site of learning (the practice) from its content is what keeps the analyst from trying to fix a context mismatch by editing the syllabus.

Manages Complexity¶

The design and evaluation of training spans an unruly variety of cases — tailoring apprenticeships, medical residencies, legal clerkships, vocational dual-systems, classroom-to-practicum bridges, workplace onboarding — each with its own content, its own institutions, and its own apparent peculiarities, and each carrying a separate-looking puzzle: why durable occupational skill accrues "on the job" rather than in the classroom, why a learner who can recite the material cannot deploy it, why one assessment predicts real competence and another does not. Situated learning compresses that heterogeneity by supplying a single locus — the community of practice as the site where competence is constituted — and reducing any learning arrangement to a small, fixed parameter set the analyst can score directly: is the activity authentic to the target practice; is the novice's contribution legitimate, however peripheral; is there a community whose standards frame and recognise the work; and is there a trajectory from periphery toward centrality. A design audit becomes scoring these four; a competence-acquisition prediction becomes reading their conjunction. The recurring puzzles then resolve to one further scalar — the distance between the learning context and the practice context — and stop being separate mysteries: the on-the-job advantage is what large classroom-to-practice distance predicts; transfer failure is read off that distance rather than re-diagnosed as a defect in the learner's grasp of content; assessment validity becomes the degree to which the assessment context resembles the practice context, since competence is recognised by the community, not measured by a score standing outside it. Instead of modelling each training programme's full content-pedagogy-institution map and treating its anomalies one at a time, the analyst tracks four design parameters and one context-distance, and reads off whether durable, transferable competence will form. The branch structure is correspondingly clean: all four conditions met and the gap small predicts competence that holds and travels; a failing parameter localises the defect to a specific stage and tells the designer that fixes aimed elsewhere — most often, editing the syllabus to repair what is in fact a context mismatch — cannot move it.

Abstract Reasoning¶

Situated learning licenses a set of inferential moves in learning theory, all routed through one relocation: competence is constituted at the community of practice and recognised by it, not measured by a test standing outside it — so the analyst reasons about learning arrangements in terms of their relation to the target practice rather than their curriculum content.

The predictive move scores a learning arrangement on four design parameters and reads competence-formation off their conjunction: is the activity authentic to the target practice; is the novice's contribution legitimate, however peripheral; is there a community whose standards frame and recognise the work; and is there a trajectory from periphery toward centrality. All four met predicts competence that holds and travels; a failing parameter predicts a specific shortfall. A fifth scalar — the distance between the learning context and the practice context — sharpens the forecast: large distance predicts that durable skill will accrue "on the job" rather than in the classroom, and that competence will degrade across the gap. The same move yields a prediction about assessment: an assessment's validity for predicting practice competence is forecast to rise with how closely the assessment context resembles the practice context, because the criterion is community recognition, not a transferable score — so a test standing far from the operating theatre is predicted to misjudge a surgeon.

The diagnostic move runs from observed failures back to context distance rather than to the learner's grasp of content. When a learner can recite material but cannot deploy it in the actual practice, the curriculum view diagnoses insufficient teaching; situated learning instead infers a gap between the learning context and the practice context and treats that gap as the variable to act on. The frame also discriminates two transfer failures that look alike from outside: inert knowledge (recitable content that fails in practice because it was encoded without practice-relevant cues) versus skill without comprehension (task performance bound to the host context that the learner cannot explain or transfer because situatedness was over-applied with no decontextualisation pass). The diagnosis names the direction of the imbalance and therefore the corrective stage.

The interventionist move follows from locating the site of learning in the practice: to build durable competence the designer makes the learning environment be, or closely resemble, the target practice and admits the learner as a legitimate peripheral participant whose contribution gradually centralises — real low-stakes work at the periphery, continuous observation of full-membership activity, and a recognition-governed trajectory toward the centre. Each design choice is a prediction tied to the parameters: closing the context distance should raise transfer, and granting legitimate peripheral contribution should produce competence that classroom instruction at the same content level would not. The sharp warning the concept issues is the mismatch — a context-mismatch failure will not move under syllabus editing, so fixes aimed at the content stage cannot repair a defect that lives in the gap between learning and practice.

The boundary-drawing move keeps these inferences inside a substrate stack of human learners capable of legitimate-participation trajectories, communities of practice with recognised norms and membership signals, and authentic activities with real stakes. Remove the community, the practice, or the legitimacy of the contribution and the parameters have nothing to score. The boundary also separates situated learning from neighbours by scope and locus: it is broader than making expert reasoning visible through modelling and fading (one instructional strategy, on the cognitive dimension), it adds to constructivism the claim that the site of construction is the practice rather than the individual mind, and it specifies one condition under which the general durable-update of learning occurs in humans. Pushed onto machine learning's "situated" variants (curriculum learning, in-context learning, embodied learning), the mechanism is representational and statistical, not legitimate peripheral participation, so the analogy is rhetorical and the inference does not carry.

Knowledge Transfer¶

Within the human-learner / community-of-practice substrate the frame transfers as mechanism, not as analogy, and its reach there is wide. Across the apprenticeship crafts that were Lave and Wenger's source cases — tailoring, midwifery, smithing, fisheries — across the professional-training schools (medical residency, legal clerkship, teacher practicum, engineering co-op), across vocational dual-systems, workplace onboarding and mentorship, and the problem-based and capstone reforms of higher education, the same machinery does the same work: the four design parameters (authentic activity, legitimate contribution, framing community, periphery-to-centrality trajectory) and the one context-distance scalar are scored the same way, and predict competence-formation, transfer failure, and assessment validity the same way. The intervention vocabulary carries intact — legitimate peripheral participation, communities of practice, boundary-crossing, scaffolding and fading, recognition by the community — because every one of these settings shares the load-bearing stack: human learners capable of legitimate-participation trajectories, communities with recognised norms and membership signals, and real work with real stakes. The Liberian apprentice tailor, the surgical resident, and the legal clerk are not analogies of one another; they are the same diagram scored on different practices, which is exactly why the on-the-job competence advantage, the inert-knowledge failure, and the operating-theatre-vs-test assessment asymmetry recur across all of them.

Beyond that substrate the transfer should be reported in two registers. The honest shared-mechanism observation is that the genuinely substrate-general pattern is not situated learning but its parent, learning — the durable update of an agent's internal state from experience — which really does recur in machine learners, in associative conditioning, and in selection-driven biological adaptation. But what travels to those substrates is the bare update, not situated learning's own named machinery: legitimate peripheral participation, community recognition, and the practice-as-site-of-construction claim all presuppose a community of practice with norms and membership, and there is no such community in a gradient-descent training loop, an associative reflex, or a fitness landscape. So the cross-substrate lesson should be carried by the parent prime, not by "situated learning" — situated learning is precisely learning plus the human-social-practice stack, and the stack is what fails to generalise.

In the second register the cross-domain use is plain analogy (A): machine learning's "situated" variants — curriculum learning, in-context learning, embodied learning — borrow the word and the shape (graded exposure, context-dependence) while the mechanism is representational and statistical, not legitimate peripheral participation against community recognition. Calling in-context learning "situated" renames the components (community → training distribution, recognition → loss, peripheral contribution → easy examples) and keeps the silhouette while dropping the social-epistemic engine that gives the original its predictive content, so it should be marked as rhetorical. The clean boundary is the substrate stack itself: situated learning transfers literally wherever a human learner participates legitimately in a recognising community of practice, and everywhere that triad is absent the reach belongs to learning (and, for the visible-reasoning strategy, cognitive_apprenticeship), not to the named concept. (See Structural Core vs. Domain Accent.)

Examples¶

Canonical¶

Lave and Wenger's source case is the apprenticeship of Vai and Gola tailors in Liberia, drawn from Jean Lave's ethnographic fieldwork. A boy entering a master tailor's shop does not begin by learning to cut cloth — the highest-skill, highest-stakes operation. He begins at the finishing end of the production sequence: ironing completed garments, sewing on buttons, attaching pockets. This work is genuinely useful to the shop, low-risk if imperfect, and performed in continuous view of the master executing the whole craft. Over roughly the span of the apprenticeship the novice moves "backward" through the production process toward cutting, taking on more consequential steps as the master recognizes readiness. Competence is judged by the master and the shop, not by any external certificate, and the trajectory's pace is set by that recognition rather than a fixed syllabus.

Mapped back: The tailor's shop is the community of practice; ironing and button-sewing are the authentic activity at the legitimate peripheral entry — real contributions, not busywork. Movement from finishing toward cutting is the centrality trajectory, paced by the recognition criterion (the master's judgment). Because the learning site is the practice, the context-distance scalar is near zero, which the theory predicts yields durable, transferable craft competence.

Applied / In Practice¶

Surgical residency is the frame's paradigmatic modern deployment. A first-year resident does not begin by leading a complex operation; under graduated supervision they first close incisions, place sutures, and manage straightforward post-operative care — real clinical work that genuinely contributes to patient care while carrying limited risk. As attending surgeons observe growing skill, the resident is entrusted with progressively more of the operation, eventually first-assisting and then leading cases, in the classic "see one, do one, teach one" progression. Competence is certified through the judgment of supervising surgeons and board examination tied to demonstrated operative experience, not by classroom testing alone. The years of case rotations, not the medical-school lectures, are where the load-bearing operative competence forms.

Mapped back: The surgical team and department are the community of practice; suturing and post-op management are the authentic activity of the legitimate peripheral entry. Graduated entrustment toward leading cases is the centrality trajectory governed by the recognition criterion (attending judgment). Learning inside the operating theatre keeps the context-distance scalar small, which the frame ties to competence that holds in real practice.

Structural Tensions¶

T1: Community recognition versus external measurement (who certifies competence). The frame's central relocation is that competence is recognised by the community from inside the practice, not measured by a test standing outside it — a surgeon is judged a surgeon by other surgeons, with the theatre, patient, and team part of what is assessed. This dissolves the on-the-job-versus-classroom puzzle, but it also makes competence hard to certify portably and honestly: the recognition criterion can encode the community's gatekeeping, in-group bias, and conservatism as readily as genuine skill, and there is no view from outside to check it against. The very move that locates competence correctly also removes the external anchor that would keep recognition from drifting into mere membership. Diagnostic: Is competence being judged by the practice's own standards from inside — and does that criterion track skill, or does it track who the community is willing to recognise?

T2: Legitimate periphery versus real stakes (low-stakes yet genuinely counting). The periphery must be legitimate — real work that genuinely contributes to the practice — and simultaneously low-stakes enough for a novice. These pull against each other: work consequential enough to be legitimate carries real risk if botched (a resident's suture, an apprentice's finishing), while stripping the risk down to a pure simulation where nothing the learner does matters supplies the scenery without the mechanism. The design must thread genuine contribution through limited risk, and the two constraints tighten against each other as the stakes of the domain rise. Diagnostic: Does the novice's peripheral work genuinely count to the community, or has the risk been removed to the point where it is busywork or simulation with the mechanism gone?

T3: Context-distance closure versus decontextualisation (situatedness over-applied). Closing the gap between the learning context and the practice context raises transfer — the frame's core intervention. But over-applying situatedness produces the mirror failure: skill without comprehension, performance so bound to the host context that the learner cannot explain or transfer it because no decontextualising pass was ever made. Inert knowledge (recitable content that fails in practice) and skill-without-comprehension (context-bound performance the learner cannot abstract) are opposite failures of the same context-distance dial, and pushing it fully to either end fails. Diagnostic: Is the failure recitable-but-inert content (distance too large) or context-bound-uncomprehended skill (situatedness over-applied with no decontextualisation)?

T4: Recognition-paced trajectory versus curriculum-clock predictability (no fixed schedule). The trajectory from periphery to centrality is paced by the community's recognition of growing competence, not a curriculum clock — which is faithful to how durable craft competence actually forms. But that fidelity sacrifices the schedulability, comparability, and auditability a curriculum offers: pace varies by learner and by the community's willingness to entrust, so the process resists planning, funding against a timeline, and defending to outside stakeholders who want a completion date. The mechanism that makes competence real makes its acquisition hard to administer. Diagnostic: Is progression governed by recognised readiness, or forced onto a fixed clock that the recognition criterion does not actually honour?

T5: Context-mismatch failure versus the syllabus reflex (fixing the wrong stage). When a learner can recite the material but cannot deploy it, the curriculum view diagnoses insufficient teaching and reaches for the syllabus. But a context-mismatch failure will not move under syllabus editing — it lives in the gap between the learning context and the practice context, not in the learner's grasp of content. The intuitive corrective therefore operates on the stage that is already fine, leaving the real defect untouched, and the more reasonable the syllabus fix looks the more effort it wastes. The frame's sharpest warning is precisely this mismatch between where the failure is felt and where it lives. Diagnostic: Does the failure track the learner's grasp of content (edit instruction), or the distance between the learning and practice contexts (close the gap instead)?

T6: Autonomy versus reduction (a named theory or learning plus the social-practice stack). Situated learning is a canonical theory with its own four design parameters, its legitimate-peripheral-participation mechanism, and its community-recognition criterion. But the substrate-general pattern that recurs in machine learners, associative conditioning, and selection-driven adaptation is the parent learning — the bare durable update of an agent from experience — while situated learning's own machinery presupposes a community of practice with norms and membership that a gradient-descent loop, a reflex, or a fitness landscape simply lacks. Machine-learning "situated" variants borrow the word while the engine is statistical, not legitimate peripheral participation against community recognition. Diagnostic: Resolve toward learning (and cognitive_apprenticeship for the visible-reasoning strategy) when carrying the lesson beyond human communities of practice; toward situated learning when a human learner participates legitimately in a recognising community.

Structural–Framed Character¶

Situated learning sits at the framed-leaning position on the structural–framed spectrum — a named theoretical account of how competence forms, whose distinctive content is a human-social-practice stack that does not generalize, wrapped around a bare durable-update that does. The criteria pull framed. Evaluative_weight is modest but real: the account is descriptive at core (where competence is constituted) yet freighted with forced design and methodological prescriptions — organize instruction as authentic practice, not decontextualized drill; admit the learner as a legitimate peripheral participant — so it carries commitments about how training ought to be arranged. Human_practice_bound is high in the constitutive sense: the whole mechanism presupposes a community of practice with norms, membership signals, and a recognition criterion, plus human learners capable of legitimate-participation trajectories — remove the community, the practice, or the legitimacy of the contribution and the parameters have nothing to score. Institutional_origin is pronounced: "situated learning" is a dated, authored theory (Lave and Wenger, 1991) of the learning sciences and sociology of education, an artifact of a scholarly tradition, not a structure found in nature. Vocab_travels fails outside the human-learner substrate: legitimate peripheral participation, community recognition, centrality trajectory all lose their referents in a gradient-descent loop, and machine-learning "situated" variants borrow only the word while the engine is statistical. Correspondingly import_vs_recognize is analogy off-substrate: calling in-context learning "situated" renames the components and keeps the silhouette while dropping the social-epistemic engine.

The portable structural skeleton is thin and lies one level down: the durable update of an agent's internal state from experience — bare learning, the parent prime that genuinely recurs in machine learners, associative conditioning, and selection-driven adaptation. Situated learning is precisely that update plus the human-social-practice stack, and the stack is exactly what fails to generalize: legitimate peripheral participation, community recognition, and the practice-as-site-of-construction claim all require a community that a fitness landscape or a training loop simply lacks. So the cross-substrate reach belongs to learning (and, for the visible-reasoning strategy, cognitive_apprenticeship), while everything that makes "situated learning" the specific named theory stays home. Its character: an evaluatively-laden, practice-constituted, tradition-originated theory of competence-formation whose only substrate-spanning content is the bare learning update it specializes, with its distinctive community-of-practice machinery pinned to human social settings and unable to travel.

Structural Core vs. Domain Accent¶

This section decides why situated learning is a domain-specific abstraction and not a prime, and it carries the case for its domain-specificity — there is no separate section for that.

What is skeletal (could lift toward a cross-domain prime). Strip the community of practice and a thin relational structure survives: an agent's internal state is durably updated by experience, so that later behaviour is changed by earlier exposure. That is the bare learning skeleton — a system carrying the mark of its past interactions forward — and it is genuinely substrate-portable, recurring as mechanism in gradient descent, in associative conditioning, and in selection-driven biological adaptation. It travels because nothing in "durable update from experience" presupposes a human, a norm, or a recognising audience: a fitness landscape and a training loop both learn without any of them. This is the core situated learning shares with every learner, not what makes it the specific named theory.

What is domain-bound. Almost everything that makes the account situated learning in particular is a human-social-practice stack that does not survive extraction. Its worked machinery is a community of practice with norms and membership signals; legitimate peripheral participation — real, low-stakes work that genuinely counts; authentic activity against which competence is calibrated; a centrality trajectory paced by the community's recognition rather than a curriculum clock; the criterion of competence being recognition-from-inside rather than measurement-from-outside; and the operational apparatus the frame scores by — the four design parameters (authentic activity, legitimate contribution, framing community, periphery-to-centrality trajectory) plus the context-distance scalar. The decisive test is the entry's own: remove the community, the practice, or the legitimacy of the contribution and the parameters have nothing to score — a realistic simulation in which nothing the learner does matters to anyone supplies the scenery while the mechanism is gone. Strip the stack and what remains is no longer situated learning but the looser thing "durable update from experience."

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. Situated learning's transfer is bimodal. Within the human-learner / community-of-practice substrate it travels intact as mechanism — the Liberian apprentice tailor, the surgical resident, and the legal clerk are not analogies of one another but the same diagram scored on different practices, which is exactly why the on-the-job competence advantage, the inert-knowledge failure, and the operating-theatre-versus-test assessment asymmetry recur across all of them. Beyond that substrate it travels only by analogy: machine learning's "situated" variants — curriculum learning, in-context learning, embodied learning — borrow the word and the silhouette while renaming every component (community → training distribution, recognition → loss, peripheral contribution → easy examples) and dropping the social-epistemic engine that gives the original its predictive content. And when the bare structural lesson is needed off-substrate, it is already carried, in more general form, by the parent learning — the durable update that genuinely recurs in machine learners, conditioning, and selection — with cognitive_apprenticeship carrying the narrower visible-reasoning strategy (modelling and fading) that situated learning subsumes on the cognitive dimension. The cross-substrate reach belongs to those parents; "situated learning," as named, is learning plus the human-social-practice stack, and the stack — precisely what makes it distinctive — is exactly what fails to generalise and should stay home.

Relationships to Other Abstractions¶

Current abstraction Situated Learning Domain-specific

Parents (2) — more general patterns this builds on

Situated Learning is a kind of Situated Cognition Domain-specific

Situated Learning is the learning-focused species of Situated Cognition, specifying how competence forms and is recognized through authentic participation in a community of practice.
Situated Learning is a kind of Learning Prime

Situated Learning is a strict kind of Learning in which the durable experience-driven update is formed through legitimate participation in an authentic social practice.

Hierarchy paths (3) — routes to 3 parentless roots

Situated Learning → Situated Cognition → Context

Show alternative paths (2)

Not to Be Confused With¶

Situated cognition. The broader theoretical stance (from the same intellectual movement) that knowing and reasoning are constitutively of their situation. Situated learning is its learning-focused application — how durable competence forms through legitimate peripheral participation in a community of practice. Tell: is the claim about the nature of cognition/knowing in general (situated cognition) or specifically about how competence is acquired and recognized over an apprenticeship trajectory (situated learning)? The former is a stance about mind; the latter a theory of skill-formation.
Cognitive apprenticeship. The narrower instructional strategy of making expert reasoning visible through modelling, coaching, scaffolding, and fading — one technique on the cognitive dimension. Situated learning subsumes it but is broader: it is a theory of where and by what competence is constituted (the community of practice), not a teaching method. Tell: is it a specific technique for exposing hidden expert thinking (cognitive apprenticeship) or the whole account of legitimate peripheral participation and community recognition (situated learning)?
Communities of practice. The social unit — a group with shared goals, tools, standards, and membership — that is the site in situated learning's account, later developed as its own construct (Wenger). It names the container; situated learning names the process of moving from periphery to centrality within it. Tell: are you naming the recognizing social group itself (community of practice) or the trajectory-and-recognition mechanism of competence forming inside it (situated learning)?
Constructivism. The broader family holding that learners actively construct knowledge rather than receiving it. Situated learning adds the specific claim that the site of construction is the practice, not the individual mind — so it is a socially-and-materially located constructivism. Tell: does the account locate construction in the individual's meaning-making (generic constructivism) or in participation in a communal practice with its recognition standards (situated learning)?
Experiential learning (Kolb). The learn-by-doing cycle of concrete experience, reflective observation, abstract conceptualization, and active experimentation — centered on the individual's reflective processing of experience. Situated learning centers the social community and its recognition, not a solo reflection cycle. Tell: is the engine an individual's reflect-and-abstract cycle over their own experience (experiential learning) or legitimate contribution recognized by a community of practice (situated learning)?
The learning parent prime. The substrate-general durable update of an agent from experience — recurring in gradient descent, conditioning, and selection. Situated learning is that update plus the human-social-practice stack, and only the bare update travels off-substrate. Tell: strip the community, recognition, and peripheral-participation machinery and what remains — durable change from experience — is learning, not situated learning; machine-learning "situated" variants borrow only the word. (Treated fully in a later section.)

Neighborhood in Abstraction Space¶

Situated Learning sits in a sparse region of the domain-specific corpus (69^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Organizational Power & Coordination Drift (9 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12