What Learning Actually Is – at a Concrete, Physical Level in the Brain

Learning is a positive change in long-term memory. By creating strategic connections between neurons, the brain can more easily, quickly, accurately, and reliably activate more intricate patterns of neurons. Wiring induces a "domino effect" by which entire patterns of neurons are automatically activated as a result of initially activating a much smaller number of neurons in the pattern.

Want to get notified about new posts? Join the mailing list and follow on X/Twitter.

In order to develop a good intuitive sense of how learning can be optimized, it’s crucial to understand – at a concrete, physical level in the brain – what learning actually is.

At the most fundamental level, learning is the creation of strategic electrical wiring between neurons that improves the brain’s ability to perform a task.

When the brain thinks about objects, concepts, associations, etc., it represents these things by activating different patterns of neurons with electrical impulses.

Whenever a neuron is activated with electrical impulses, the impulses naturally travel through its outward connections to reach other neurons, potentially causing those other neurons to activate as well.

By creating strategic connections between neurons, the brain can more easily, quickly, accurately, and reliably activate more intricate patterns of neurons.

As one might expect, it is extraordinarily complicated to understand what these specific brain patterns are, how they interact, and how the brain identifies strategic ways to improve its connectivity.

However, to some extent, these are just nature’s way of implementing cognition – and the overarching cognitive processes of the brain are much better understood.

At a high level, human cognition is characterized by the flow of information across three memory banks:

1. Sensory memory temporarily holds a large amount of raw data observed through the senses (sight, hearing, taste, smell, and touch), only for several seconds at most, while relevant data is transferred to short-term memory for more sophisticated processing.
2. Short-term memory, and more generally, working memory, has a much lower capacity than sensory memory, but it can store the information about ten times longer. Working memory consists of short-term memory along with capabilities for organizing, manipulating, and generally "working" with the information stored in short-term memory. The brain's working memory capacity represents the amount of effort that it can devote to activating neural patterns and persistently maintaining their simultaneous activation, a process known as rehearsal.
3. Long-term memory effortlessly holds indefinitely many facts, experiences, concepts, and procedures, for indefinitely long, in the form of strategic electrical wiring between neurons. Wiring induces a "domino effect" by which entire patterns of neurons are automatically activated as a result of initially activating a much smaller number of neurons in the pattern. The process of storing new information in long-term memory is known as consolidation. At a cognitive level, learning can be described as a positive change in long-term memory.

These memory banks work together to form the following pipeline for processing information:

1. Sensory memory receives a stimulus from the environment and passes on important details to working memory.
2. Working memory holds and manipulates those details, often augmenting or substituting them with related information that was previously stored in long-term memory.
3. Long-term memory curates important information as though it were writing a "reference book" for the working memory.

In the context of solving a math problem:

1. Sensory memory captures visual data that lets us read the problem or any intermediate work that we’ve written down, thereby allowing the written information to be loaded into working memory. It also filters out any distractions (e.g. background noise) as we solve the problem.
2. Working memory holds the relevant pieces of the problem, requests additional information from long-term memory, and applies that information to incrementally transform the pieces of the problem into the solution. The problem-solving narrative takes place within the working memory.
3. Long-term memory -- upon request from working memory -- produces definitions, facts, and procedures that we learned previously. It is like an internal "reference book" that we can use to look up additional information that would be helpful while solving the current problem.

Note, however, that there is a crucial conceptual difference between long-term memory and a reference textbook: long-term memory can be forgotten. The text in a reference book remains there forever, accessible as always, regardless of whether you read it – but the representations in long-term memory gradually, over time, become harder to retrieve if they are not used, resulting in forgetting.

It’s also worth re-emphasizing that the problem-solving narrative takes place within the working memory. Sensory and long-term memory supply working memory with information, which working memory combines, transforms, and uses to guide our behavior to solve the problem.

As researchers Roth & Courtney (2007) elaborate:

Okay… but what is working memory at a biological level?

Recall that when the brain thinks about objects, concepts, associations, etc., it represents these things by activating different patterns of neurons with electrical impulses.

Loosely speaking, the brain’s working memory capacity represents the amount of effort that it can devote to activating these neural patterns and persistently maintaining their simultaneous activation.

As summarized by D’Esposito (2007):

When the brain is initially learning something, the corresponding neural pattern has not been “wired up” yet, which means that the brain has to devote effort to activating each neuron in the pattern. In other words, because the dominos have not been set up yet, each one has to be toppled in a separate stroke of effort.

This imposes severe limitations on how much new information the brain can hold simultaneously in working memory via rehearsal.

Most people can only hold about 7 digits (or more generally 4 chunks of coherently grouped items) simultaneously and only for about 20 seconds (Miller, 1956; Cowan, 2001; Brown, 1958). And that assumes they aren’t needing to perform any mental manipulation of those items – if they do, then fewer items can be held due to competition for limited processing resources (Wright, 1981).

This severe limitation of the working memory when processing novel information is known as the narrow limits of change principle (Sweller, Ayres, & Kalyuga, 2011).

An intuitive analogy by which to understand the limits of working memory is to think about how your hands place a constraint on your ability to hold and manipulate physical objects.

You can probably hold your phone, wallet, keys, pencil, notebook, and water bottle all at the same time – but you can’t hold much more than that, and if you want to perform any activities like sending a text, writing in your notebook, or uncapping your water bottle, you probably need to put down several items.

In the same way, your working memory only has about 7 slots for new information, and once those slots are filled, if you want to hold more information or manipulate the information that you are already holding, you have to clear out some slots to make room.

Long-term memory solves this problem by providing a means by which the brain can store lots of information for a long time without requiring much effort.

By creating strategic connections between neurons, the brain can more easily, quickly, accurately, and reliably activate more intricate patterns of neurons. Wiring induces a “domino effect” by which entire patterns of neurons are automatically activated as a result of initially activating a much smaller number of neurons in the pattern.

As information becomes more ingrained in long-term memory, it becomes easier to activate.

When the information becomes ingrained to the fullest extent, it can be activated automatically without conscious effort.

This is known as “automaticity”, the ability to execute low-level skills without having to devote conscious effort towards them.

Automaticity is important because it frees up limited working memory to execute multiple lower-level skills in parallel and perform higher-level reasoning about the lower-level skills.

As a familiar example, think about all the skills that a basketball player has to execute in parallel: they have to run around, dribble the basketball, and think about strategic plays, all at the same time. If they had to consciously think about the mechanics of running and dribbling, they would not be able to do both at the same time, and they would not have enough brainspace to think about strategy.

This extends to academics as well. As described by Hattie & Yates (2013, pp.53-58):

You can even see the effect of automaticity in brain scans.

At a physical level in the brain, automaticity involves developing strategic neural connections that reduce the amount of effort that the brain has to expend to activate patterns of neurons.

Researchers have observed this in functional magnetic resonance imaging (fMRI) brain scans of participants performing tasks with and without automaticity (Shamloo & Helie, 2016).

When a participant is at wakeful rest, not focusing on a task that demands their attention, there is a baseline level of activity in a network of connected regions known as the default mode network (DMN).

The DMN represents background thinking processes, and people who have developed automaticity can perform tasks without disrupting those processes:

When an external task requires lots of focus, it inhibits the DMN: brain activity in the DMN is reduced because the brain has to redirect lots of effort towards supporting activity in task-specific regions.

But when the brain develops automaticity on the task, it increases connectivity between the DMN and task-specific regions, and performing the task does not inhibit the DMN as much:

In other words, automaticity is achieved by the formation of neural connections that promote more efficient neural processing, and the end result is that those connections reduce the amount of effort that the brain has to expend to do the task, thereby freeing up the brain to simultaneously allocate more effort to background thinking processes.

Now, it’s important to realize that automaticity goes beyond simple familiarity.

If you truly “know” something, then you should be able to access and leverage that information both quickly and accurately.

If you can’t, then you’re just “familiar” with it.

And when learning hierarchical bodies of knowledge – whether it be math, chess, a sport, or an instrument – it’s important to truly know things, not just be familiar with them.

Why?

Because you can’t build on familiarity. That’s what the term “shaky foundations” refers to. You can only build on a solid foundation of knowledge.

You need to learn things so well that you effectively turn your long-term memory into an extension of your working memory.

That’s how you break free from the narrow limitations of working memory.

It’s kind of like how in software, you can make a little processing power go a long way if you get the caching right.

he natural follow-up question is “how do you develop your long-term memory to the point of automaticity,” and the answer is retrieval practice (spaced, interleaved retrieval practice to be precise).

Further Reading: Retrieval Practice is F*cking Obvious

References

Brown, J. (1958). Some tests of the decay theory of immediate memory. Quarterly journal of experimental psychology, 10(1), 12-21.

Cowan, N. (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity. Behavioral and brain sciences, 24(1), 87-114.

D’Esposito, M. (2007). From cognitive to neural models of working memory. Philosophical Transactions of the Royal Society B: Biological Sciences, 362(1481), 761-772.

Hattie, J., & Yates, G. C. (2013). Visible learning and the science of how we learn. Routledge.

Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological review, 63(2), 81.

Roth, J. K., & Courtney, S. M. (2007). Neural system for updating object working memory from different sources: sensory stimuli or long-term memory. Neuroimage, 38(3), 617-630.

Shamloo, F., & Helie, S. (2016). Changes in default mode network as automaticity develops in a categorization task. Behavioural Brain Research, 313, 324-333.

Sweller, J., Ayres, P. L., Kalyuga, S., & Chandler, P. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23-31.

Wright, R. E. (1981). Aging, divided attention, and processing capacity. Journal of Gerontology, 36(5), 605-614.

Want to get notified about new posts? Join the mailing list and follow on X/Twitter.