The Intelligent Design Series || Part 01
Making correct predictions in pursuit of a goal is a pretty good definition of “intelligence”.
Steven Pinker
Mistakes are, after all, the foundations of truth, and if a man does not know what a thing is, it is at least an increase in knowledge if he knows what it is not.
Carl Yung
Give me a fruitful error any time, full of seeds, bursting with its own corrections. You can keep your sterile truth for yourself.
Vilfredo Pareto
I have a radical opinion about instructional design: I think it should start from a coherent theory of how people learn. This didn’t seem radical to me when I first migrated from my original field of artificial intelligence into training and education, but I quickly discovered that when I nattered on about how learning works in instructional design sessions, people tended to look at me as though I had gone off on a tangent about metaphysics. The topic of how people learn just doesn’t come up much in mainstream instructional design, and when does, it is usually in the form of rules of thumb like “people learn better when you help them to relate your lesson to their experience.”
The main reason, I think, is that instructional design is often seen as a kind of trade, almost like, say, plumbing. No one expects plumbers to be whizzes at hydrodynamics. Similarly, a lot of instructional designers seemingly prefer to be left alone to get on with their work without having to be responsible for messy theories of how the brain works.
Following standard practices in the construction industry more or less gets you a building that works; following standard practices in instructional design often gets you a course that doesn’t.
The problem with this is that learning is not a solved problem in the sense that building construction is. Following standard practices in the construction industry more or less gets you a building that works; following standard practices in instructional design often gets you a course that doesn’t. To do better, I believe that instructional designers need to develop a better understanding of how learning works and use this as a guide to better design.
So in this column I want to do my part to move towards a brighter future by giving my answer to the question, how do people learn?
I want to begin by considering an example, and because I think discussions like this often go wrong from the start due to a modern tendency to see learning as something that happens in a classroom, I’m going to set my example in a time before formal education. So, for the sake of argument, imagine that you are a homo erectus living in the Indus river valley in, oh, say, about 500,000 BC. Let’s say that you are walking home from the fishing hole in the twilight when suddenly, in a blur of motion, a tiger leaps from the brush directly into your path.
There’s of course a good chance you won’t survive this encounter. In this case the removal of your possibly maladapted genetic strain from the gene pool may help your species adapt through natural selection. It’s cold comfort to you, but over time, as tigers take their toll, the surviving population will come to feature traits that are better adapted to a tiger-rich environment—being able to run fast, say, or being smart enough to get home before the sun sets.
But let’s be optimistic, and assume you escape. In this case, there is better news: rather than sacrificing yourself to help your species adapt, you will instead get the chance to adapt as an individual. Of course you can’t suddenly grow claws or develop the ability to run 40 miles per hour (65 kph), or anything like that. But you will change your behavior. For example, it is more than likely that you won’t dawdle at the fishing hole in the evening anymore. You may choose a different path home as well. And so on. This is the phenomenon of learning.
Two Keys to How People Learn: Surprise and Emotion
The key to learning lies in your ability to remember your experience. You will remember the tiger episode for the rest of your life and recall it at times when it may help you to make better decisions, like the next time the sun starts going down when you are away from the village. If your brain is healthy, it records every episode that you experience, but only a fraction of those episodes seem to be remembered for the long term. A complete theory of what causes a particular episode to be retained in memory doesn’t exist, but two factors are known to have a big impact: surprise and emotion. Clearly, the tiger episode has plenty of both, which is why we can be certain that you will remember it.
Why Emotion Matters
I’ll return to the issue of why surprise matters in a bit, but for now let’s think about why emotion would matter. The best answer seems to be that the things that evoke emotions tend to be things that affect, or threaten to affect, our goals, either positively or negatively. Encountering a tiger generates a lot of emotion because it threatens the goal of self-preservation. It makes sense to remember times when one of your goals was threatened, because those memories can help you identify situations you want to avoid in the future. Likewise, it makes sense to remember episodes in which a goal was furthered—say, for example, a time when you tried a new fishing spot and caught an unusual number of fish—because those memories help you identify situations that you want to recreate.
Now let’s think in more detail about what happens when you remember a past episode. First of all, it is important to note that due to a remarkable property of your memory, you typically recall past situations that are very similar to the one in which you presently find yourself. This is useful because you can expect similar things to happen in similar situations. If you go fishing, it makes sense to expect that it will be like the last time you went fishing; if you are headed down a particular trail, it makes sense to expect it to be a lot like the last time you walked down that trail; if you are talking to a particular person, it makes sense to expect it to be a lot like the last time you talked to that person; and so on. The notion that you are constantly recalling previous times you did something similar to what you are doing now seems obvious, but when you think about it, it is quite amazing that you are able to do this.
If you observe very young children, you’ll notice that they apply expectations from past episodes in a very simple and literal way—they basically expect their second experience in a particular kind of situation to be just like the first. This can create an occupational hazard for parents. I sometimes do a thought experiment in my talks where I ask people to imagine they took their kid to a fast food restaurant and got a balloon, and then took them to the same place on another day only to find there were no longer any balloons. Whenever I say this, every parent in the room instantly gets a pained expression, because we all know what the upshot will be: an unhappy surprise, followed, quite likely, by tears.
As people get older, however, they get more sophisticated about teasing apart the elements of a situation and hypothesizing the individual causes of those elements, and this complicates the picture somewhat. For example, if the last time you went fishing it was cold and there was ice on the river, you may be able to reason out that because it is warm this time, you should not expect ice. How exactly people do this sort of reasoning is a big open question—it’s the sort of thing I worked on when I was in artificial intelligence–but this shouldn’t obscure the essential fact that you remember past episodes and extract from them expectations about the present.
Expectations are useful in large part because anticipating something that might happen allows you to take action to ensure that that thing either does or does not actually happen. If you remember that the last time you walked down a particular trail you slipped on a mossy rock, for example, that allows you to step more carefully this time, while if you remember that the last time you went picking strawberries you found some juicy ones in a particular location, you can be sure to go straight to that place the next time.
This takes us back to where we started, with encounters with tigers being one of the more obvious sorts of things you might want to avoid. So we now have a simple model of how learning works that we can summarize thusly: We have an experience, we recall it when conditions are similar, we expect a set of possibilities based on what happened in the recalled experience, and we act to avoid or further those possibilities as our goals dictate.
Introducing Schemas
Although this model seems to make intuitive sense, it is wrong—or at least, incomplete—in a very important way. To see that, think about what happens when you have a lot of experiences in the same essential situation. For example, you may have been fishing hundreds of times. If so, you will have experienced many different episodes in which you learned important things. So what happens when you go fishing again? Do you remember all of these episodes at once? If so, how do you manage to mentally process all those memories rapidly enough to extract and use the knowledge they contain? This issue has been referred to as the “paradox of the expert”—the notion that an experienced person might have so many relevant memories that she would be paralyzed by the effort to make sense of them all. The label is meant to be tongue-in-cheek, because of course real experts are not paralyzed in this way. The question is, why not?
The answer lies in something that psychologists call a schema. A schema is an organizing structure in memory that collects together the set of things you should expect in a particular kind of situation. So for example, if you have been fishing a lot, you will have a fishing schema in memory that amalgamates all of the things you have learned from all of the fishing experiences you have had. You may remember which spots are best to fish from, which kinds of fish you are most likely to catch at different times of year, what kind of bait works best for each kind of fish, and on, along with miscellaneous details such as remembering that you should stay out of the water in the winter months, and that if you leave your bait unguarded a raccoon may steal it. By collecting this knowledge in one place, the schema helps you avoid the paradox of the expert.
So how do you create a schema? Clearly, it must happen incrementally, with your existing schema changing each time an episode teaches you something new. Presumably, the starting point is the first episode of a particular kind that you experience. But what causes you to update your schema, after that point, in response to a new experience? The deceptively simple answer is, when what is happening does not conform to what your schema told you to expect. Expectation failure—the failure of your schema to accurately predict what is going to happen—is the trigger for learning. This is not exactly obvious, but if you think about it you can see that there really is no other way it could work. After all, when your schema correctly tells you exactly what to expect in a situation, what is there to learn? It’s analogous to a scientific theory, which, if confirmed by experiment, remains unchanged, but must be modified when a violation occurs.
The Importance of Surprise in How People Learn
This, by the way, is where the notion of “surprise” comes into the picture. As I noted above, surprise is one of two features–the other being emotion–that tend to make an episode stick in memory. Its interesting to note that surprise triggers the production of the neurotransmitter dopamine in your brain, and dopamine has the effect of strengthening the active connections between your neurons. This is thought to have the impact of making it more likely that you will remember in the future whatever it is that you are thinking about now. Emotions—especially pleasure and pain–also trigger Dopamine production. So there is at least a plausible neurological theory of how surprise and emotion determine what you remember, and are thus important elements of how people learn.
When something happens that wasn’t anticipated in your schema, the obvious thing to do is simply add a new expectation to that schema. For example, if there is ice on the river one day, and you walk out on it to fish and fall through, an expectation about that will certainly be added to the fishing schema.
What about when something happens that contradicts what your schema told you to expect, though? For example, let’s say that past experience has led you to expect that the spot under the old pine tree is the best for catching fish, but today your companion on the turtle-shaped rock is catching more than you. Clearly it would be rash to simply cancel your old expectation and replace it with the expectation that the turtle rock is the best spot. Instead, it makes sense to look for an explanation. It could be, for example, that the kingfisher sitting on a nearby tree limb is hurting your cause. Or it could be something about the temperature, or the season, or the river currents. You may or may not be able to figure it out, but the point is, when an expectation fails, you try to explain it. If you come up with a plausible explanation, this allows you to replace your expectation with something more sophisticated. Instead of “the spot under the pine is the best for catching fish” you might expect in the future that “the spot under the pine is best for catching fish except when the water level is low, in which case the spot on the rock is better.” Or, you might form a new expectation that says, “any spot with a kingfisher near it will be temporarily bad for fishing.” Or, alternatively, “Kingfishers tend to hang out where the fish are, so fish there!” Not being a fishing person myself, I have no idea which is a more plausible theory.
By repeatedly explaining your expectation failures, you create progressively more sophisticated theories. Think, for example, about how you learn to get along with a particular person over time. At first, your schema for interacting with them may be more or less a copy of your generic schema for interacting with anyone. But over time various expectations in that schema will inevitably fail. To take a simple example, you may have a supply of jokes that make most people laugh, but you may discover your new acquaintance doesn’t find those jokes funny. From other interactions you may decide that they are quick-tempered, sentimental, iconoclastic, shy, devious, or any of a hundred other things. As you repeatedly explain observed behaviors that violate your general expectations, you can go from the simple observation of a failure—say, that person X doesn’t like your joke about the three Neanderthals who walked into a bar—with more useful general expectations, for example, that person X doesn’t like anything having to do with negative stereotypes. These kinds of generalizations accumulated over time will help you make much more accurate predictions about a particular person’s behavior that may differ substantially from what you would expect from an average person in the same circumstance. This accumulation of generalized expectations is essentially what it means to become familiar with a person, or with anything else for that matter—places, things, activities, ideas, and so on.
As I noted above, the theory of how people learn through this kind of reasoning works is far from completely worked out. The key point to hang onto here, though, is that explanatory reasoning is triggered when an expectation fails. I want to make one other quick point about this. You might ask why, if a person knows enough to explain an expectation failure, they wouldn’t have anticipated that failure in the first place. The answer is that expectation failures tell you what needs explaining. In other words, they are an attention focusing device, which is necessary because you don’t have the mental bandwidth to work out detailed causal explanations for every single thing that happens. I’ll give a quick, slightly silly example from a more modern, though still somewhat ancient era: my graduate school career. During this time, I sprained my ankle playing intramural football and ended up on crutches for the first time in my life. As you might imagine, there was a lot to learn from this new experience. One surprisingly important thing I learned was about getting sodas. It was my habit, when I got my daily diet coke out from the soda machine in the building where my office was housed, to pop the top immediately and take a quick drink. The first time I did that on crutches, I quickly realized that now that the soda was open, I was not going to be able to get it back to my desk without spilling it all over as I crutched myself down the hall. While I could in principle have figured that out in advance, it was pretty unlikely that I would ever have happened to think of it. But, once the failure focused my attention, it was simple to figure out what went wrong. And having had my attention called to it, I adjusted my expectations and did not make the same mistake again… at least, not more than once or twice. (There is still a challenge in changing a habit pattern even when you’ve realized what’s wrong with it.) That’s why expectation failure in general plays such a vital role in learning: expectation failures show you where to focus your thinking in order to improve your understanding.
The last point I want to make about learning is that in general, people seem to learn much more from doing things themselves than from watching others do them.
As it stands, my model doesn’t do much to explain that difference. After all, you can form expectations and revise them in response to failures just as easily when you aren’t involved in an activity as when you are. I think there are three reasons why you learn by doing better than by watching. The first is that, when you do something and the expected result doesn’t occur, that almost by definition means that one of your goals is thwarted. That means that you are more likely to remember the lesson. The second is that when you have to make a decision, you attend to a lot of things you might not attend to when you aren’t involved. For example, you can watch people fish all day without really paying attention to when they cast their lines, how far the cast them, how they manipulate their equipment to make the cast happen, and so on. When you are forced to make decisions, you are forced to notice a lot of things that you wouldn’t have noticed otherwise, and thus you have a much richer memory of the experience. The third thing may seem a bit subtle, but it may actually be the most important of the three. To make a decision, you have to settle on a theory of what you expect to happen, even if it is only a wild guess. When you are just observing, for example, you don’t have to have any theory at all of how hard you should pull on the line once you get a strike. But when you are fishing, you are forced to guess, and your guess will be either supported or refuted by what happens next. Your guess becomes an initial expectation, which, even if it is off base, can start you on the cycle of expectation failure and explanation that will lead you to a better theory.
So we can summarize our theory of how people learn something like this: Learning is based on the ability to remember our experiences and recall them in similar circumstances. When an experience is repeated, our mind creates a schema to represent what we can expect in the typical instance of that experience. We use the expectations from this schema to guide our subsequent behavior. When an expectation fails, we seek an explanation, and based on that explanation we revise our schema. Through the action of this process over time, our schema will become more and more accurate and detailed, and we will become better able to function in the environment that it describes.
If we want to design effective learning experiences, this is the place to start. These are the mechanisms of the human mind that lead to learning; a good design is one that effectively causes those mechanism to spring into action.