Information-Seeking
Imagine you awake on the floor in a pitch black room and you don’t know where you are, how you got there, or anything else about the situation. It could be said that you lack “information.” You could be surrounded by a moat of crocodiles, or a buffet table. Perhaps you’re here to be murdered, or you simply passed out and a nice person brought you somewhere safe to sleep it off. What would you do once you finished sobbing?
If you leap to your feet and sprint screaming in a random direction, you might happen to pick the hallway, and a trajectory for the exit, but you might also find the moat, or knock over the food, or run face-first into a wall. A thoughtful person might first feel the ground to determine whether they’re indoors or out, and if the floor is marble or shag carpet (the preference of serial killers). Anything to get a piece of information to act on. Then she might stand and extend her hands slowly into the dark on all sides. Finding nothing, she might take a careful step and do it again, and again, until finally she contacts a wall. Then feel along the wall, maybe listening for sounds and sniffing for smells that might offer clues.
The only thing like a reward in this scenario is more information to act again. The only thing like a punishment might be a hand that fails to extend, i.e. hits a wall, providing more information at very little cost. Gradually, she builds a map of her surroundings, finds moulding, then a wooden door, a knob, and a lighted room in the home of a friend who was apparently not thoughtful enough to place her on a couch.
Reward and Punishment
We act to seek pleasure or avoid pain. These two types of feedback work mostly at the levels of physical sensation and of the emotions—signs of firstness and secondness. It’s possible but difficult to use our reasoning faculties to make ourselves do something that we know will hurt, even if it pays off later on.
A baby searches the room with its eyes, flails its arms and legs, and grasps at anything it can hold. It teases out shapes, textures, and tastes. By trial and error, it learns how to sense the world, and what movements result in a warm touch or a drink of milk, versus something unpleasant. We learn how to act through rewards and punishments.
At first, consequences are tied to single acts. Crying gets us held. But soon, we realize that no two acts are the same, nor can we expect the same consequences. There are many different contexts. Babies have different “cries,” and a mother’s response varies slightly to each, as well as time of day, place, and level of fatigue.
So we develop classes of action, tied to classes of rewards and punishments, to account for the variety of results. By this method we form habits and expectations, which take more factors into account as we grow, until we have complex societies erected on the foundation of reward and punishment. It’s a powerful system. While it may sound animalistic, I don’t think we can get rid of it. But when we treat it too simply, it directs our thoughts and actions down corridors that can cause problems. This happens when we direct our lives according to rewards and punishments in a mechanical and linear fashion.
If you put a dollar in a vending machine, you expect your chips every time. This simple input/output relationship works if you only consider the man, the dollar, and the machine in isolation. As soon as we bring it the context in which they exist, many possibilities emerge. Would you expect your food if the power was out, or if the machine was empty? Few people would be surprised to hear that machines can malfunction, but they would still be irate when the chips got stuck and no amount of shaking could knock them loose.
On the basic level, rewards work great for training children and puppies. Every time they act a certain way, they get a treat. As the habit builds, the treat comes only every third time, then every tenth. Eventually, we hope that the behavior is understood within some higher-order context, and performed for the intrinsic value, rather than a mechanical reward. For example, we stop “behaving” to earn an ice cream cone and start doing it because we understand how good behavior benefits us and those around us.
While many of us learn this lesson in general, we still miss it in certain domains. It’s good luck if the vending machine accidentally sends down two bags of chips, and unfair if it takes our money. Again, that can only be true if we dissect a handful of elements from their context, which is not a model of reality—it’s a simplistic way of understanding, useful to beginners until they can grasp the higher-order complexities.
The man and the vending machine exist within a building, in a city, on the Great Plains, in 2023, on Planet Earth, etc. What we call luck for good or ill is the confluence of context beyond what we can predict. What we call justice is our attempt to impose simple relationships on those systems. But human justice differs from the justice of nature. We want to take into account only certain pieces of information, like the suspect, the victim, the events leading to the crime, and the jurisdiction, while ignoring other things like race, gender, and what the killer had for lunch last week. This is useful for wrapping our heads around things. It also gives us a false impression of the world, and misguided expectations for the future.
The justice of nature is simply the consequence of all actions whirling together. What goes up must come down, not because the leap was evil or arrogant, but because it didn’t sustain itself against a weak but persistent force. If we pay attention to the workings of gravity, we can do incredible things like send a man to the moon, but we can’t simply jump to it. Was the initial leap rewarded? The fall, punishmed? Or could we consider the entire sequence to be something else entirely?
Small acts give us information. Sure, returning to Earth might not be what we expected, but we can learn from that, and many other acts, not to gain an immediate reward but to gain a better understanding of the higher-order contexts in which our actions are nested. Rather than reward, we seek information.
Gregory Bateson defined information as “differences that make a difference.” It hardly matters what the information is, because it informs our next action, and our next one. Over time, we have enough of it to briefly land a capsule on the moon and return it safely home.
Notice that our rocket still has insufficient thrust to escape the pull of rewards and punishments. What we did was to replace immediate rewards and punishments with the satisfaction we get from seeking new information that leads us to more skillful testing, learning, and action in the future. We accept small, immediate failures in order to produce large future rewards that aren’t possible from a simple input.
As I said, we understand this in most cases. Few adults are so stunted that they can’t put off a little short-term pleasure in certain domains. But we all give in to the temptation at times, and in certain areas of life, we can’t even imagine something beyond the simple input-output of a vending machine. Since pleasure and pain work on us at a very basic level, they can hamstring us even when our reason tells us we should behave differently.
Unfair Is for Beginners
Where do you fall short? Look to the times you demand that the simple justice of man be imposed on the complexities of the world. The word “should” often reflects a personal preference rather than a moral imperative. Things are “unfair” when you fail to consider the context in which they exist. Nature—and by that I mean all movement from atoms to galaxies—always works out as it should, according to the confluence of all forces. It is fair beyond our comprehension. When we think otherwise, it signals that we are beginners in that field.
We still seek simple information, simple relationships, that work every time. I believe this is a remnant of childhood learning models, as well as an erroneous transposition of human justice models onto nature.
With practice, we can break into the second-order understanding that provides context for our actions, and develop classes of action that get general results, and maybe even classes of classes of action (i.e., virtues). This takes time and attention. It requires that we act to seek more information, and remain ready to change our behaviors if the feedback suggests it.
How to Seek Information
Seeking information requires acting at the second-order. That means doing something not to elicit an immediate result, but to see what result it gets, so you can apply the learnings to your next move. The reward comes from the intrinsic value of the act of seeking information, and the knowledge that skillful seeking leads to classes of action that yield great benefit down the road, even if the initial act could be called a “failure.”
In truth, it’s only a failure if you had a certain expectation that wasn’t met. As with the hand extended in the dark room, any information is a success, even the absence of information, which itself tells you that there is no wall within arm’s-length.
When we receive frequent, easy rewards for our actions, we risk growing spoiled. If those actions don’t always result in a reward, we pitch a fit, or become despondent and shut down. On the other hand, if we receive too much punishment—unpleasant outcomes—we can become traumatized and unwilling to try.
A good trainer of dogs and children can calibrate the outcomes so that they’re steered toward good behavior without being ruined in the process. But much of life happens “in the wild,” and some people never get a decent teacher at an early age. Relying on first-order rewards and punishments, therefore, is destined for confusion and misery.
Information-seeking is not approached with a goal in mind. It’s an act of exploration. True discovery, whether of new lands or laboratory transformations, is undertaken without a result in mind. The diligent explorer simply ventures into the unknown and notes the results, which guide his next venture. This has to be done with some care, as the unknown can harbor some very unpleasant things. He doesn’t rush headlong, but uses his wisdom to press in a little at a time.
If we die in the process, we can’t learn anything. The goal of information-seeking is not to find something in particular, but to find enough to try again, differently. Even bumping into a wall that blocks our path can be construed as a reward at the second order of magnitude, because it tells us where not to go.
Nassim Nicholas Taleb offers a way forward in his concept of antifragility—those things that gain from small shocks and non-fatal disorder. When we take small risks in a certain amount of variety, there’s a great chance that one will open up fantastic new avenues of exploration. Though many will fail, we never commit a critical effort all at once, so we always have enough left to try again. This risk-taking is built open a foundation of reliable action. Most of what we do is time-tested. We have already gained an excellent understanding of many things, and these principles guide our efforts. By grounding ourselves in what’s worked, we can extend the borders of our knowledge.
We refuse to explore when we fear that we may be harmed; that new knowledge may upset a precious worldview; or when we just think we know everything worth knowing and further action is a waste of time. By avoiding novelty, we become mired in bad habits, while failing to develop our good ones into tremendous resources.
When you feel you have succeeded or failed at something, ask yourself: what did I hope this act would bring me? Did I expect it right now? Or much later? Did I act to confirm my knowledge and to be made whole by my notion of justice, or did I act to inform my future actions and to seek a better understanding of a more natural justice—one that is consistent and complex beyond human notions, but which can be gleaned by expanding my context through exploration?
Exercise
9. Each day, find out one thing you would have previously let slide. As usual, pay attention to the physical sensations and emotions you experience before, during, and after the act.
The exercise you just did in which you asked people to clarify their meaning is one way to go about this, but there are others. You can look up a strange word. Find out the basic history of some place, or concept. Try two options to see which works better, without deciding in advance. Explore a new physical space, like a trail, or a building.
My favorite involves following up on some mundane detail that comes up in conversation. If a friend wonders aloud when a restaurant closes, be the one who looks it up. Or plug a foreign word on the menu into your favorite translator. Or ask what the specialty is, even if you don’t care.
Seeking information is itself a habit, and the more you practice, the more inclined you’ll be to do it in more ways, richer ways. Eventually, you’re bound to stumble upon something of incalculable value.