As successive recalls accumulate, knowledge stored in memory becomes increasingly solid and intertwines with other knowledge, until forming coherent clusters of memories.

📌 This phenomenon of progressive aggregation of concepts is called chunking (Gobet et al., 2001). It is the hallmark of stabilized learning and easily accessible knowledge.

💡 Until chunking has occurred, the learner processes each element separately, which saturates their working memory. Once chunking is in place, they handle meaningful blocks: faster, more fluid, more efficient.

Chunking translates into a performance gain that can be spectacular, as illustrated by the example of chess grandmasters (De Groot, 1946, 1978).

📌 Example: memorizing a chess board. Grandmasters are capable of memorizing entire game configurations, whereas novices can only retain the position of a few pieces.

Why? Because configurations implicitly make sense to experts: they recognize a Sicilian opening, an Indian defense, and so on. Novices, on the other hand, process each piece individually and explicitly.

⚠️ Revealing detail: this superiority only holds for "plausible" configurations. When pieces are arranged randomly, experts do not memorize better than novices (Chase & Simon, 1973). Chunking only works on meaningful structures.

Learning to drive perfectly illustrates the transition from conscious effort to automaticity.

📌 At the start of training, the learner driver devotes all their energy to coordinating their movements: clutch, gear lever, mirrors, brake pedal. Focused on so many elements, they struggle to pay attention to road signals.

📌 After a few years, all these procedures are automated. The cognitive resources previously dedicated to supervising each action are freed up and can be used for other things: anticipating other drivers' behavior, following a conversation, planning the route.

💡 That is the power of chunking: less effort for what has been mastered, more attention available for what remains to be learned.

Explicitly presenting connections

Rather than leaving your learners to spot the links between concepts on their own, highlight them directly:

💡 Offering analogies allows your learners to progress much faster than if they had to notice these connections on their own.

Beware of your own expertise

⚠️ A classic trap for trainers: your own knowledge is highly automated. What seems obvious to you is in reality the result of years of chunking. For your beginner learners, nothing has yet been aggregated.

📌 Best practice: take the time to break down your content finely and spend considerable time on certain concepts. What is quick and obvious to you is not obvious to them.

Keywords: pedagogical tips, memory, chunking, automaticity, expertise, spaced repetition, cognitive load.