II. Above the Ladder Essay 8
Learning to Learn to Change
The levels of learning, and the one that changes the chooser
Ask a student how they revise for an exam and most of them will say they reread. Notes, textbook, notes again. It works on the first reading, sort of. The material feels familiar, the pages feel known, and the feeling of familiarity gets mistaken for the feeling of knowing. Then the exam arrives and the familiar material produces nothing, and the student concludes that they did not revise enough, and rereads more.
Nothing about that student is stupid, and nothing about that method is hopeless. But rereading does one specific thing. It makes the material feel known. It does not make the material usable. The gap between feeling familiar and being able to produce an answer under pressure is where most revision quietly fails, and it fails for everybody, because the method works well enough that nobody questions it.
What the student needs is not a better study technique. It is a change to the habit that chose the study technique in the first place. And that is a rarer and more valuable change than it sounds, because habits of learning are usually invisible, and because the same structure appears at every scale: a response that changes, a set of options that changes, and a way of choosing between options that changes.
Look at the rereading student at four levels, and ask what changes at each one.
Level zero, the answer. The student sits the exam and does not use what they read. Nothing about the student changes. The response is zero. The next exam, the same thing happens, with a slightly different question.
Level one, the response changes. The student reads the question and gets it wrong, and then adjusts. This is ordinary corrective learning: you attempt, you see that you were wrong, you try something different next time. Rereading sits somewhere around here, because rereading is an attempt of a sort. It produces a change in the response, although not reliably the right change, and it stays inside a fixed set of options. The set of possible study methods available to this student is still the set they had before. They are choosing differently from within it, which is genuine learning, and it is also the limit of it.
Level two, the options change. The student learns that rereading does not work for them, and that remembering requires producing rather than receiving. The set of available methods is now larger. Recall practice, self-testing, explaining the idea out loud, spacing. This is what the learning science calls learning how to learn, and it is a real and important level. The student has changed what they can do.
Level three, the frame changes. The student stops treating the exam as a test of whether they have met the material before, and starts treating it as a question about what the question is asking. They notice that the questions they fail are the ones where the wording is unfamiliar, not the ones where the content is hard. They start looking at what a question is for before trying to answer it. Nothing about the study method has changed at all, and yet the study method now includes examining the question, which it did not before. The way of seeing what a test is has changed, and everything else followed.
Where these levels come from
The four levels described here are a simplified version of a scheme developed by Gregory Bateson, an anthropologist whose work ran from animal behaviour to cybernetics to family therapy. His scheme is about how a system responds to the possibility of being wrong: whether it changes its response, changes the set of responses available to it, or changes the frame within which it selects responses.
This summary is offered as a rough and simplified version, and it should be treated as such. The original is more careful, uses different terminology, and makes distinctions this version smooths over. What is kept here is the shape, because the shape is useful: it gives a vocabulary for describing how much has changed, and a way of asking for more when a correction is not sticking.
How this relates to the ladder
The ladder and the levels of learning are not the same thing, and treating them as one is a mistake worth making deliberately once so that it is not made by accident.
The ladder is about the subject matter. It says that thinking can be about one particular thing, about what several such things share, about a rule that covers them, about the structure that produces them, or about the loop that keeps them going. The levels of learning are about the learner. They say how much of the learner's way of responding has changed.
The two are close enough to be tempting and different enough to be dangerous. They are not the same, they do not have the same number of levels, and one does not map onto the other. The mapping is rough, as the figure says, and the roughness is not a defect to be smoothed away. It is the reminder that a description of a subject and a description of a learner are two different descriptions, and that confusing them makes both useless.
One thing they do share is a shape. Both are ladders that people mistake for ladders, meaning that people treat the top as a place of permanent residence rather than a position to move through and return from. Both reward the mistake, because both make it easy to describe somebody as thinking at a level, as though level were a property they had rather than a choice they were making.
That shared shape is also the reason the two are worth keeping near each other. When a teacher asks a student to reconsider how they revise, that is a move up the learning levels. When the same teacher asks what the question is really asking, that is a move up the ladder. And the two moves feed each other: the student who has changed what they look at will choose better methods, and the student who has changed their methods will notice more about what they are looking at. Separately they are two different ideas. Together they are one account of how a person can get better at something they are already, on the evidence, quite good at.
Why the top level is rare
Level three is rare, and there are ordinary reasons for that.
First, level one is usually enough. If you get the answer right more often after a correction, you have learned. Most of the time that is all anyone needs, and it is a good deal cheaper than changing how you frame a question. The upper levels are not virtues. They are expensive.
Second, and more seriously, level three is invisible from the inside. A frame feels like the way things are. You cannot tell that you are looking through a particular frame, because the frame is what looking feels like. So the person whose frame has just changed usually describes it in terms of something else entirely: I started reading the question differently, I started asking students to explain the idea before they write it, I started looking at the marks instead of the work. These sound like technique changes. They are frame changes wearing technique as a disguise, and the disguise is what makes them repeatable.
This is why the change so often gets described in the wrong words, and why the description matters practically. If somebody says they changed their technique, they will go looking for better techniques next time. If they can say they changed what they were looking at, they will know what to look at next time. The words a person uses for their own change determine what they will try the next time something goes wrong, which is the moment it matters most.
Third, changing a frame means giving up things. A frame is comfortable partly because it has already been applied successfully a hundred times. A new frame has to earn its keep from scratch, and until it does it feels slower and more effortful, which people read as a sign that it is worse. This is the most reliable barrier to learning in any organisation, and it is why the level three move is usually made by somebody who has been forced to, by a failure big enough to make the old frame untenable.
There is also a cost in the other direction, which gets less attention. A frame that has been changed is a frame you are worse at. For a while after the change you see less clearly, not more, and the work is slower and the results are worse, and the temptation to go back to the old frame is strongest exactly when the new one would soon start paying. Almost everybody who has made this change has gone back to the old one at least once first. That is not failure. That is the cost being paid.
Level three in ordinary work
The same structure turns up outside education, and it is worth seeing it there because it is easier to watch.
A small workshop has a rule: every piece goes out with a photograph. The rule exists because a customer once complained about a chipped rim and the workshop had no way to show what it looked like when it left. Level one, the rule changes the response.
Level two, when the rule is found not to settle the argument, the options change: they start keeping a sample, so the next time they can put the actual cup in front of the customer. Level three is the point where the workshop stops asking how to prove the piece was fine and starts asking why the customer could not believe it in the first place. That might mean changing how the pieces are packed, what is written on the invoice, or what the workshop is actually selling. Nothing about the photograph changes, and everything about the purpose behind it does.
In a school, the same move looks like a teacher who stops marking essays and starts looking at what the essays are for. Not a new marking scheme. A different question.
What if the rereading student is given a new tool, a revision app that spaces practice and tests recall? That is a level two intervention and it works. The student uses it, the app shows them what they have forgotten, they practise accordingly, and their exam results improve. Nothing about how they think about exams has changed, and nothing needs to.
Now change the condition: what if the same student is asked to look at their own exam scripts and mark the questions they got wrong by writing down what the question was asking, before writing down what they had answered? That is a level three intervention, and it is a much smaller change than it sounds. The result is not better revision. The result is that the student starts noticing that a particular kind of question is where they fail, which is a fact about the exam they could not have learned by revising better, because revising better does not change what you look at.
"Learning how to learn is the top level." It is not, and treating it as the top makes it hard to see the level above. Learning how to learn is a level two move: you have changed the set of tools available to you. You still look at problems in the same way, you still choose your approaches in the same way, and you still get stuck in the same places. The student who has learned to learn is a better student within the frame they already had.
The other misconception: "you can will yourself to a higher level." Nobody sits down and decides to reframe how they see a question. The move happens when something makes the old frame stop working, usually after a failure that is bigger than the frame can absorb, and it happens over months. What you can do is make the failures survivable, keep the questions visible, and notice the moment when the familiar way stops producing what you need.
Limits
The levels of learning are a scheme for describing change, not a ladder of virtue. Nobody is "at level three" in general. A person can be at level three about how they run a meeting and at level zero about how they argue with a friend, and the difference has nothing to do with their character.
Second limit: the scheme cannot distinguish a change in understanding from a change in vocabulary. A student who has learned to use the word "entropy" has not thereby learned to think about energy dissipation. Levels of language and levels of understanding are different ladders, and conflating them is the most common way people convince themselves they have grown.
Third limit: the levels are described from the outside, and from the outside everything looks like a change in the learner. It may be a change in the situation instead. The student who suddenly studies differently may have had a teacher who changed what was expected of them, and the learning moved with the environment. This is a real problem for any account of learning, and it is a reminder that the thing being described is a system, with the learner inside it rather than above it.
Stuck
You are stuck when you keep trying harder inside the frame that has already failed. The tell is that the same methods keep getting tried with more effort, more hours, more colour coding, more highlighting, and the results do not move. You are correcting at level one, inside a set of options that does not contain the thing that would help.
The way out is to change the question rather than the effort. Instead of asking "how do I study this better", ask "what is this subject actually asking me to do, and what have I been assuming it wants". Write the answer down in a single sentence, the way you would explain it to somebody who has never seen the course. If the sentence is easy to write, you have probably already learned it. If it is hard, the hard part is the learning, and the rereading was never going to get you there.
Sources
- The four levels of learning are a simplified version of Gregory Bateson's scheme of learning, or double learning. The summary here is deliberately loose and should be read as a rough version rather than a faithful account. The original terminology and the distinctions between the levels have been flattened, and anyone relying on this should read Bateson directly.
- The claim that levels of learning and levels of thinking are not the same, and do not map onto each other, is the author's own.
- All classroom and workshop examples are hypothetical or ordinary operational situations.