How we learn

One of their biggest challenges is that teachers can’t see learning happening.

However, we can understand what happens in the brain when pupils learn.

In this blog we’ll find out how learning happens. This gives us insights we can use to make better decisions about pupil learning.

Defining learning

You might’ve heard learning defined as a “change in long-term memory” (Kirschner et al., 2006, p.75). From a neuroscientific perspective, this means memories are stored in the brain. Stored memories are our knowledge.

When we talk about ‘storing’ memories/knowledge, we mean:

Figure 1.

When first encoded, memories are fragile. This means they’re prone to forgetting or interference from other memories (Dudai, 2004).

Then, memories may go through a process of ‘consolidation’ where they are stored for the long term. This is one of our goals for pupils: long-term storage of important knowledge.

The memories we are interested in (semantic memories for facts, concepts and ideas) are stored in knowledge networks called schemas (Gilboa & Marlatte, 2017).

This means consolidation adds new memories/knowledge to schemas (Shing & Brod, 2016). Adding to what we know is learning.

Therefore: consolidation = learning.

Let’s see how consolidation happens.

Capturing fast

Imagine watching ‘Romeo and Juliet’ at the theatre. We are paying attention to what’s going on in the famous balcony scene. What’s happening in our brains?

Our hippocampus (Figure 2) quickly captures the event as episodic memories, i.e., memories with contextual details like the actors’ clothes, props etc (Yassa & Stark, 2011). Think of it like an imprecise snapshot of what’s happening.*

But there’s a problem. The hippocampus isn’t designed for long-term memory storage (Kroes & Fernández, 2012). In fact, the memories of the balcony scene are fragile (prone to forgetting/interference).

Figure 2.

Image by Levent Efe via

This fast hippocampal system needs a complementary learning system that can store memories for the long term (McClelland, 2013).

Luckily, we’ve got just this.

Learning slower

The outer layer of the brain (neocortex – see figure 2), supports networks of stored memories (schemas). They span different brain regions (Paller et al., 2020). This means that to be stored, memories must transfer from the hippocampus to the neocortex.

The hippocampus transfers memories by replaying them to the neocortex, ‘teaching’ relevant schemas what it knows. This teaching is called ‘neural replay’ (Deuker et al., 2013). It’s thought to drive consolidation.

Not all the replayed memory is stored in schemas. The contextual details of the event, e.g., the actor’s costume/the scenery, can’t enter schemas because unique details don’t fit in networks of related knowledge (Kroes & Fernández, 2012).

Let’s imagine eventually our brain extracts and stores the metaphor ‘Juliet is the sun’ and the knowledge that ‘this is said during the balcony scene’.

The metaphor isn’t stored as an episodic memory of an actor saying ‘Juliet is the sun’.

It’s stored as semantic memory for the metaphor itself, detached from most contextual details.**

Consolidating memory does two things:

Transfers the memory from hippocampus to neocortex for longer-term storage.***

Transforms the memory: episodic detail is lost, commonalities are extracted into schemas, i.e., episodic memory transforms into semantic memory.

Figure 3.


The hippocampus quickly captures episodes.

It replays them to relevant schemas (neural replay).

Parts of the replayed memory are extracted/stored in relevant schemas.

This process is called consolidation. It’s one way we learn.

How long does learning take?

The answer is…

…we’re asking the wrong question.

Asking how long learning takes treats consolidation like an event with an end point. In fact, consolidation is a process with stages (Dudai & Morris, 2000).

Even when memories seem to have consolidated, they can still change. Time can weaken our access to memories. Activating memories through retrieval (incidental or deliberate) can, under certain circumstances, update memory through a process called ‘reconsolidation’ (Nader & Hardt, 2009). Reconsolidation (updating memory) may continue indefinitely (Dudai, 2012). This means learning is never finished.

Instead of how long learning takes, we could ask, ‘how long into consolidation does memory become stabilised/less prone to forgetting?’. We can say a couple of things on this.

First, consolidation probably won’t happen in a lesson. Neural replay, thought to drive consolidation, happens during wakeful rest and sleep (Findlay et al., 2020). Pupils won’t rest/sleep until sometime after the teaching episode.

Secondly, whilst consolidation to the point of stabilisation can take a while (hours/days or longer) (Dudai et al., 2015), it’s probably less about time and more about what we already know:

more relevant prior knowledge = faster storage/learning

This means we can speed up learning (more on this in the next blog).

Insights for teachers

Knowledge builds on knowledge

Learning is a process of continual consolidation and reconsolidation: knowledge builds on prior knowledge. This means how best to grow pupils’ knowledge will depend on what we’re building upon.

Imagine starting a company. At first, you advertise, recruit a small staff and deliver a bespoke service. As your company grows, new techniques are needed to continue growing: employing more staff, relying more on word-of-mouth and using the same service for similar clients.

In other words, the best techniques depend on the stage of development.

This is just like knowledge:

When little knowledge exists, explicit instruction works well. Once they’ve grasped the basics, we should switch tact, adding greater challenge by fading support, to grow pupils’ knowledge (Clark, 2009).

Knowledge is prone to interference and forgetting

Before memories consolidate, they are particularly prone to interference/forgetting (even when they have consolidated, forgetting sets in). This means that what pupils experience within/between lessons may alter their knowledge. Learning isn’t linear. Teachers continually need to check the state of pupils’ knowledge.

Learning never ends

Our memories want to be the best predictors of what will happen in the environment, so they adapt. Reconsolidation can continue indefinitely meaning pupils can always build upon what they know.

Learning doesn’t happen immediately

We have a slower learning system to store memories. This tells teachers that what looks like pupil learning (of new information) in a lesson is unlikely to be. Memory consolidation usually takes time, requiring sleep or rest.

To know if long-term learning has occurred, teachers need to check over the long term


There are ways we can speed up the learning process. Find our how you can use schemas to speed up learning.

*This isn’t a passive process. The hippocampus is connected to the medial prefrontal cortex which activates schemas related to the events we experience to help us make sense of them.

**Although, if over time we keep thinking of the metaphor and thinking of the scene, we will continue to recall a hazy episodic memory of the actor saying it too.

***Memory that has contextual details of time and place (like the actor on the stage saying the line ‘Juliet is the sun’) is thought to still be supported by the hippocampus.


Clark, R. E. (2009). How much and what type of guidance is optimal for learning from instruction?. In Tobias, S., & Duffy, T. M. (Eds.) Constructivist instruction (pp. 170-195). Routledge.

Deuker, L., Olligs, J., Fell, J., Kranz, T. A., Mormann, F., Montag, C., Reuter, M., Elger, C. E. & Axmacher, N. (2013). Memory consolidation by replay of stimulus-specific neural activity. Journal of Neuroscience, 33(49), 19373-19383.

Dudai, Y. (2004). The neurobiology of consolidations, or, how stable is the engram?. Annual Review of Psycholology, 55, 51-86.

Dudai, Y. (2012). The restless engram: consolidations never end. Annual review of neuroscience, 35, 227-247.

Dudai, Y., Karni, A., & Born, J. (2015). The consolidation and transformation of memory. Neuron, 88(1), 20-32.

Dudai, Y., & Morris, R. G. (2000). To consolidate or not to consolidate: what are the questions?. Brain, perception, memory: Advances in cognitive neuroscience, 149-162.

Findlay, G., Tononi, G., & Cirelli, C. (2020). The evolving view of replay and its functions in wake and sleep. Sleep Advances, 1(1), zpab002.

Gilboa, A., & Marlatte, H. (2017). Neurobiology of schemas and schema-mediated memory. Trends in cognitive sciences, 21(8), 618-631.

Kirschner, P., Sweller, J., & Clark, R. (2006). Why Minimal Guidance During Instruction Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem-Based, Experiential, and Inquiry-Based Teaching. Educational Psychologist, 41(2), 75-86.

Kroes, M. C., & Fernández, G. (2012). Dynamic neural systems enable adaptive, flexible memories. Neuroscience & Biobehavioral Reviews, 36(7), 1646-1666.

McClelland, J. L. (2013). Incorporating rapid neocortical learning of new schema-consistent information into complementary learning systems theory. Journal of Experimental Psychology: General, 142(4), 1190.

Nader, K., & Hardt, O. (2009). A single standard for memory: the case for reconsolidation. Nature Reviews Neuroscience, 10(3), 224-234.

Paller, K. A., Mayes, A. R., Antony, J. W., & Norman, K. A. (2020). Replay-based consolidation governs enduring memory storage. The cognitive neurosciences, 265-276.

Shing, Y. L., & Brod, G. (2016). Effects of prior knowledge on memory: Implications for education. Mind, Brain, and Education, 10(3), 153-161.

Yassa, M. A., & Stark, C. E. (2011). Pattern separation in the hippocampus. Trends in neurosciences, 34(10), 515-525.


7 responses to “How we learn”

  1. […] now we know ideal memories for pupils are generalised and sufficiently detailed. In the next blog we find out how our brains store memories. This provides teachers with fundamental insights about the learning […]


  2. […] is usually encoded by the hippocampus, which temporarily stores episodes we experience. In a slower process, the hippocampus replays what it knows to relevant schemas in the neocortex (outer layer of the […]


  3. […] process for storing memories in the brain is called consolidation (Dudai, 2012). Fragile memory traces are transformed into stable knowledge in networks called […]


  4. I reallyn like your blogs, but strange that you give this blog a title that’s also tephe title of the book I wrote with Carl Hendrick but don’t even mention that.


    1. Thanks for reading the blogs Paul, that means an awful lot as I’ve gained a enormous amount from your work and hearing you speak at ResearchEd! I apologise, it was not intentional to use the name of yours and Carl’s book. I can see how it would look that way as there are many ways I could have said the same, so I will change the title.


  5. […] The memory trace is usually fragile at this point. Memories can then be stabilised through a process called consolidation: the hippocampus replays some memories to the neocortex (outer layer of the brain) where they are […]


  6. […] therefore makes sense to build our schools around what we know about how children learn. Willingham’s simplified model of memory is a good starting […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: