In 1991, two researchers named Jean Lave and Etienne Wenger published a book that should have changed corporate learning forever. The book was called Situated Learning: Legitimate Peripheral Participation, and its central argument was both simple and radical: you cannot separate learning from the context in which it will be used. The two are not merely connected — they are inseparable. To teach something in isolation from its real-world application is not to teach it at all. It is to create the appearance of knowledge without its substance.
Three decades later, most corporate eLearning continues to ignore this entirely. Learners sit in front of slides that describe processes they will perform in environments that look nothing like the slides. They answer questions about scenarios they have never experienced in contexts that do not resemble their jobs. And then we are surprised when the knowledge does not transfer.
The apprentice and the master — learning as participation
Lave and Wenger's original research was not about corporate training at all. They studied traditional apprenticeships — midwives in Yucatan, tailors in Liberia, meat cutters in American supermarkets. What they found, in every case, was the same thing: learning did not happen when the apprentice sat down and was taught. It happened when the apprentice began to participate — first at the edges of the community of practice, then closer to its centre.
The midwifery student did not learn by reading about childbirth. She learned by being present at births, first in a minor role, then progressively more central ones. The tailoring apprentice did not learn by studying fabric theory. He learned by performing small, real tasks — pressing, finishing, observing — and gradually taking on greater complexity as his competence grew. The context was not a setting for the learning. The context was the learning.
"A fish does not learn to swim by being told about water. It learns by moving through it — feeling the resistance, adjusting, developing an intuition that no lecture could create."
Why abstraction is the enemy of transfer
The reason context matters so much comes down to how memory works. We do not store information in a neutral, context-free archive from which we can retrieve it in any situation. We store it in association with the context in which we encountered it. This is called context-dependent memory, and its implications for learning design are profound.
If you learn a process in a classroom, your memory of that process is associated with the classroom — its sights, sounds, the questions that were asked, the instructor's voice. When you later encounter a situation that requires that process in a noisy factory or a hospital corridor or a customer's home, the retrieval cues are different. The memory does not surface easily, because the context has changed. The knowledge is there, technically. But it is not accessible in the moment it is needed.
This is why situated learning is not merely a philosophical preference — it is a practical necessity. When learning occurs in a context that resembles the context of performance, the memory traces are encoded with richer, more transferable cues. The learner who practised a troubleshooting conversation in a realistic simulation will retrieve that knowledge more readily when they are standing in front of a real piece of equipment than the learner who read about it in a module.
What situated learning looks like in practice
The good news is that situated learning does not require a physical environment. It requires a designed environment — one that is rich enough in authentic detail to activate the same mental models the learner will use in the real world. This is the promise of well-designed simulation, scenario-based eLearning, and business game mechanics.
Consider the difference between these two approaches to teaching a clinical research associate about patient data verification. The first approach presents a slide that explains the SDV process and lists the steps in order. The second places the learner inside a virtual hospital room, asks them to review a patient record, and presents a discrepancy between the source data and the trial database — then asks: what do you do?
Both approaches convey the same information. Only one of them activates the same cognitive and emotional processes the learner will use in the field. Only one of them creates a memory that will surface when they are standing in front of a real patient record, under time pressure, with a clinical trial's integrity on the line.
The community of practice — learning is social, not solitary
There is a second dimension of situated learning that corporate training almost universally ignores: the social dimension. Lave and Wenger's concept of the community of practice describes how knowledge in any professional domain is held not only in individuals but in the relationships, conversations, and shared practices of the community. A new employee does not just learn from a course — they learn from watching how experienced colleagues handle difficult situations, from the war stories told in break rooms, from the informal advice offered before a challenging client meeting.
Designing for situated learning means designing for this social dimension too. Mentoring programmes that pair new hires with practitioners. Communities of practice where learners can share challenges and solutions. Structured reflection activities that ask learners to connect training content to their own professional experiences. These are not add-ons to a learning programme — they are, in the situated learning framework, the programme itself.
Implementing situated learning without a full simulation budget
The most common objection I hear to situated learning principles is a practical one: simulations are expensive. Full business simulations with branched scenarios and authentic environments cost time and money that not every L&D budget can accommodate. This is true. But it misses the point.
Situated learning is a design principle, not a technology requirement. A well-written case study that places learners in a recognisable professional dilemma is situated learning. A structured peer discussion that asks learners to connect training content to a recent real-world experience is situated learning. An OJT checklist that asks a supervisor to observe specific behaviours in the actual work environment is situated learning. The investment required scales with the fidelity of the simulation, but the principle applies at every budget level.
Context is not the setting for learning. Context is the learning.
When you design a learning experience, the most important question is not "what content do I need to cover?" It is "what environment do I need to create?" The closer that environment is to the context in which learners will perform, the more likely it is that what they learn will transfer when it matters most. Lave and Wenger gave us the theory thirty years ago. The only remaining question is whether we are willing to use it.