What Even Is Learning?
By: Stephen M. Ryan
The Perfect Textbook
When I was a young teacher, tasked with teaching English Conversation to classes of 30-40 students, I found the perfect textbook. It checked all the boxes for me. Suitable for the students’ level? Check. Interesting and engaging layout/illustrations? Check. Needed little explanation from me? Check. Practice activities that went on long enough that I could get round the whole class to check students were doing as asked? Check. It worked great. I used it in all my classes. The publisher loved me for ordering in volume. Only one thing I forgot to check: were the students actually learning anything?
It was a few years later that I heard Dick Allwright speak about the two kinds of teachers (he may have said “British teachers,” I’m not sure): the ones who move on to the next activity by saying “OK” (“OK. Let’s do the second exercise now.”) and the ones who do the same thing by saying “Right.” “Just what is ‘OK?’” he asked. “What has been done ‘right’?” He went on to suggest that these little words were evidence of a conspiracy between teachers and students. A conspiracy to agree that learning has occurred.
No Longer into Teaching
I’m older now. Older and much less interested in teaching than I was. In fact, the older I get, the more focussed I seem to be on students’ learning. Why? For selfish reasons, of course. Because that’s where the action is, the buzz, the excitement of lights going on behind eyes, the satisfaction of seeing that someone is more able to do something today than they were yesterday and knowing I had a hand in that process. Ease of implementation, attractiveness of activities, even students’ engagement are no longer my measures of success. Student learning is.
But what even is learning? This is going to take a bit of explanation but bear with me, as the pay-offs that come from a deep understanding of learning are real, tangible, and literally where the action is.
Inchoate World
We are born into an inchoate world. It buzzes and flashes and smells and never seems to stop. None of it makes sense. None of it is self-explanatory. Worse, our bandwidth for accessing everything that’s out there is severely limited: in information processing terms, we can deal with less than 0.0004% of the bits and bytes of sensory information in our environment (Zimmerman, 1985)—but of course “information” is the wrong name for it, as it doesn’t make sense.
Even worse, we are born with needs (food, comfort, hygiene) that can be taken care of only by engaging with the buzzing, stinking, flashing stuff of the world and also have to figure out which parts of this sensory morass will help with this, and which will endanger our very survival.
Fortunately, we are also born with brain tools for making sense of all this. We have an innate drive to seek patterns in sensory intake (pattern-recognition). We also have mechanisms for storing the neural patterns evoked by the sensory intake for future reference (memory). And, most helpfully, we have ways of using the stored patterns to make extrapolations about future sensory patterns (prediction). (Otherwise, what would be the point in recognising and storing patterns?)
These simple but powerful tools allow us to do something very important: to learn, to gain enough information about the world to survive and thrive in it.
Let’s look at the process in detail:
Pattern Detection. First, we pay attention to patterns: repeated elements of the intake, elements that seem to co-occur, elements that occur in sequence. Oval shapes with two dark holes in them keep peering into my crib. Sometimes there is a cooing sound and I feel comforted. Something is put in my mouth and quite soon the emptiness in my tummy goes away.
Memory. Instinct prompts us to remember these patterns. We lay down memories of anything that seems to make sense, especially if it occurs frequently or is associated with strong emotion. In fact, emotion itself is simply a way of tagging a pattern of experience as being “potentially useful in the future” i.e., worth remembering. We are incapable of learning without emotion (pace Mr. Spock). Such memories are the key to our survival, so we can well imagine that ancestors who did not have this ability did not live long enough to actually become our ancestors.
Reinforcement. A certain combination of neurons fires in the brain each time we encounter the same pattern in our environment. Each firing makes that combination more likely to fire together in the future. (Famously, what fires together, wires together.) The cooing sound often goes with a smile. The more often the two co-occur, the stronger the connection between the neuronal firing pattern caused by hearing the coo and the pattern caused by seeing the smile.
Prediction. So, we begin to expect that if, on hearing the cooing, we open our eyes, we will see a smile on the face of the cooer. We are using previous experiences to help us prepare for the future. This is what Pavlov’s dogs were doing when they salivated before any food was present. They were preparing for an imminent future event that patterns in their previous experience had prepared them to expect.
Self-Correction. Crucially, we have a self-correcting mechanism, so that an initial rough draft of a prediction, over time and given further experiences, becomes more and more refined. What first seemed like a simple correlation between coo and smile becomes more nuanced. We discover that there is only an 89% chance of co-occurrence. The probability is higher if it’s that nice soft face I see, but if it’s the rougher one with the moustache on the upper lip there is a smaller chance of a coo and a smile.
This is the fundamental process by which we learn: notice, remember, predict, correct/refine. It is responsible for everything we know about the world, consciously or unconsciously. It is how weather forecasts work (they look for patterns in past weather to predict future weather): it is the reason my blood pressure increases before I stand up; it is why our palms sweat before an important interview; it is why we are first shocked by and then learn to accommodate within our worldview the unexpected.
A lot of learning occurs while we are asleep. When we are awake, our brain is too busy trying to keep up with data from the world around us to focus much on pattern recognition, etc. but whenever it can, it switches to these important tasks. Not only when we are asleep, but in odd moments of rest, too (day-dreams anyone?). Most of the process is unconscious. Dreams are a (barely) conscious residue left by our brain’s attempts to make sense of new sensory data in terms of previously remembered data: trying things out; looking for patterns; finding ways to modify older understandings to make way for new experiences. No wonder babies sleep so much! They have a whole world to figure out.
Growing Up
As we get older, learning changes a little. It takes more to surprise us, to alert us to the possible existence of a new pattern or a hitherto unnoticed kink in an old one. We already have a rough understanding of much of the world around us. But the mechanism is still the same, as we can experience by moving into a new and different social or physical context (we call this experience culture shock) or simply encountering something in our usual environment that hasn’t been there before. We notice, store the memory, look for patterns, make predictions, correct them in the light of ongoing experience.
We are no longer dealing with a wholly unexplored world of sensations. Our learning consists less and less of constructing new bases for predictions and more and more of modifying existing ones. We are no longer just looking for patterns in ongoing sensory intake; we are also looking for connections between what we are perceiving now and all our memories of previous perceptions. This, too, is learning, and it is the kind of learning we do for most of our lives.
At first, the process is effortful. A new element in our environment can require conscious attention. Fitting it into existing understandings can literally keep us awake at night (Why did she look at me that way? What did he mean when he said that?). It can lead to conscious Aha! Moments (cf. Archimedes). It can demand careful calculation to develop strategies to a) gather more data to confirm, modify, or reject emerging new understandings, and b) figure out how to use these understandings to enhance our surviving and thriving. Conscious processing like this uses large amounts of energy (glucose and oxygen, in brain terms). It is why Study Abroad students tire so easily in their new environment. It is why even a small perturbation in our usual routine can send us to bed early.
Once a new experience is internalised, however, made part of a slightly modified version of our previous understanding of the world, it disappears from consciousness. It is now part of our unconscious expectations about how the world works and how we can get things done in it. It is in play whenever we encounter similar situations to the one where we first encountered it, but it is in play automatically, without any conscious effort. This automaticity is essential to our daily functioning as we could not go around being constantly surprised at every new encounter.
Kahneman conceptualises the contrast between effortful learning and the automaticity that results as being analogous to having two brains in one head: a fast-acting one and a slow-thinking one. The first is automatic, relying on previously detected patterns (and minor changes to them as it goes) to allow us to react quickly and usually appropriately to any experience that life brings us. The second is slower, more deliberate and more effortful, taking time to mull things over and develop an action plan. The usual analogy is with learning to drive a car. There are lots of new patterns to learn at first and it takes a lot of effort to master them, but once learnt the process is automatic, unless the grossly unexpected comes along (big truck in the wrong lane).
Horvath presents a four-stage model of the transition from first experience to automaticity: Active Learning (this is the most effortful part); Plasticity (when the neurons that fire together re-configure their connections so they are more likely to fire together again); Modularity (a higher level of networking between frequently fired-together neurons); and Automaticity (don’t need to think about the new knowledge/skill in order to use it). This view of the path towards automaticity helps us to see it as a gradual process rather than a sudden transition from novice to master.
Once a sensory pattern from one experience has been incorporated into our understanding of the world, how easily can that understanding be applied to other experiences that differ, however slightly, from the previous one? This is the issue of transferability. “How easily does something learned in one context transfer to other contexts?” is a question of direct relevance to language teachers. Many of the activities we have our students do are simulations: performances in one context (the classroom) that are intended to prepare them for another context (the “world”). This is a complex and imperfectly understood area of learning, but there is substantial evidence that transfer works best when the simulation and the real context of use are as similar as possible.
The other major development in learning for adults (or rather, non-baby) learners is the ability to modify understanding based not only on one’s own experience but also on the experience of other people. This increases learning exponentially. I don’t need to fly into space to have a rough idea of what it is like; I don’t need to have my own fMRI machine to learn about the workings of the brain; I don’t need to develop my own materials to help my students learn. I can learn from other people! Learning is essentially social. This is why considerable amounts of brain function and architecture are devoted to figuring out other people (aka, Theory of Mind), so that we can learn from their experience. We learn by talking with them, by watching them, by reading their books and papers, by watching their YouTube channels. Their experiences expand on what we could experience ourselves to give us more and more ways to make sense of the world. Here, language, the tool that allows us to share vicariously in the experiences of others, is key.
Into the Classroom
Pattern Detection. Memory. Prediction. Self-correction. Automaticity. Learning transfer. Social learning. How does all this help me to focus on my students’ learning? Well, the good news is that some of the things I was doing, even in my earliest days as a teacher, were by chance helping students to learn: repetition, pattern practice, pair-work. But other things were not: engagement is not the same as learning. Filling the whole 90-minute lesson with Student Talk Time feels great to the Conversation teacher, but is probably not the best use of time if we consider our main role to be supporting and encouraging student learning.
What we need to do in the classroom, when based on an understanding of learning mechanisms, should be fairly obvious from the foregoing: exercises to help students notice new language; opportunities to connect language with experience (personal when possible; vicarious when not); linguistic and socio-linguistic pattern-detection practice; lots of repetition, review, and retrieval (much more than offered by any textbook I have examined); low-stakes testing as a regular part of lessons; explicit teaching of memory strategies; lots of practice in using the language; authenticity of tasks; deliberately constructed chances for students to learn from each other. Oh, and sleep. Emphasising the importance of a good night’s sleep to sort out, internalise, and consolidate the day’s learning.
I’ll leave it to others in this issue (and the next) to fill in the details of some of these ideas. My point here is this: you wouldn’t take your car to be repaired by somebody without a good understanding of engines, or your computer to be fixed by somebody who has never explored the insides of one. Surely, the more teachers know about learning, the better they can help their students.
I laughed when I first heard teachers described as Learning Facilitators. Now, it is a badge of pride.
Further Reading
This is a necessarily brief summary of learning processes. I have tried not to encumber it with too many links and references. Readers are encouraged to pursue topics raised here in the following publications.
Bott, D., & Horvath, J. C. (2020). 10 things schools get wrong (and how we can get them right). John Catt Educational.
Clark, A. (2023). The experience machine: How our minds predict and shape reality. Penguin.
Immordino-Yang, M. H. (2015). Emotions, learning, and the brain: Exploring the educational implications of affective neuroscience. W. W. Norton.
Kahneman, D. (2011). Thinking, fast and slow. Penguin.
Lieberman, M. D. (2013). Social: Why our brains are wired to connect. Crown.
Stephen M. Ryan works at Sanyo Gakuen University, helping students to learn English and to learn from Study Abroad experiences. His interest in learning grows daily.