Cooperation First Principles

In the previous post (10JAN12) we reviewed a few kinds of cooperation: educational, a military unit, sports, and corporations. Here I want to back away from the specifics and look at cooperation from first principles, mostly from communication — the transfer of information and the construction of new information.

In a group of cooperating humans communication is essential. No communication, no cooperation. In a cooperating group of humans chunks of information are sent from one person to one or more other people in the group. Information is also sent from any one person to him or herself when that person is quietly thinking by him or herself. Chunks of information received cause a reaction of one kind or another in the person receiving the information. The information goes into the brain of the receiving individual. How far in does it go? What does it do in there? Sometimes information heard “makes no sense”. If the receiver does not know the language of the sender then the information can’t go in too far, it can’t be connected with any other information already there, except for the receiver to note that he can’t understand the information. Or maybe the receiver has been exposed to the language casually in the past so that the receiver can recognize that the language for example is Chinese or German from the sounds alone. But the specific information in the spoken words cannot get into the receiver’s mind. Similarly if the sounds are not loud enough or they are distorted by background noise or the speaker, in which cases the receiving person asks the sender to repeat the message, to resend the chunk of information. So the chunk of information received now makes sense to the receiver. This means that the receiver connects, associates this incoming information to some information already there in his or her mind, to some category that the new information is about. Next the receiver makes further associations and connections to and from this new information, and after some amount of processing — thinking — decides to accept or reject — says yes or no to — the new information. The acceptance or rejection may depend on whether the new information has been sufficiently connected to the truth category (all the information in the person’s mind that he or she considers true). Or the information may be accepted on the basis of the respect or trust the receiver has in the sender. Or there can be many other reasons, many other methods, other processes, the receiver uses in the decision to say yes or no. If the decision is yes, then the information received is connected more strongly to other information the person uses regularly in running his or her life — it becomes operational, it is used or useable in thinking and other activities.

In summary new information is accepted or rejected depending on how well it fits in with information already in the receiver’s mind.

If the new information doesn’t fit in well, the receiver may respond to the sender expressing his disagreement partially or totally, asking for clarification, or suggesting a reformulation. In this way, this back and forth, can continue for some time between two people communicating with each other, cooperating toward some purpose, some goal or goals. A beautiful example is Amos Tversky and Daniel Kahneman who collaborated for years. See Daniel Kahneman’s book “Thinking, Fast and Slow”.

In this communication the chunks of information going back and forth are changing as the conversation continues. And as the conversation continues the information content of the two minds is changing. Each person to the conversation is fitting into his or her mind information received from the other. They each will often have developed a more or less similar understanding of the subject matter of their conversation. Or they may have developed a plan that they each more or less agree upon to reach their individual and joint goals. Or they may be building something as they are communicating like building a house. Or they may be doing something together like hunting.

The above discussed two people communicating cooperatively. It can easily be extended to several people. But not to too many people. One thousand people couldn’t have a casual conversation where each person could have his say at some point after many other people have spoken. Nor could each person even hear each other person. There are just too many people. Twenty people are probably too many. Fifty surely are. To have a cooperative group of more than about 20 people some structure for the group is needed. Structure means rules. Rules as to who can speak when, and for how long, and on what subjects, etc. There are formal and informal rules. Formal rules are written down somewhere. Informal rules are not. Informal rules are what people do by habit without thinking about it. Even in a two person conversation there are informal rules such as: ask for clarification if you don’t understand; and generally begin speaking only after the other person has stopped speaking or has paused; and a good conversationalist pays attention to the emotional responses of his listener in order to avoid having a conversation only with himself, etc.

There are many possible different structures for a cooperating, communicating group. A group could have a facilitator whose job is to keep the group following its rules. It could have a leader who directs or dominates the discussion. A leader (in contrast to a facilitator) leads the group discussion in directions he chooses rather than letting the directions evolve from the group discussions. This can produce groupthink where the other participants avoid expressing ideas they know the leader does not want to hear, and so the group may not benefit from the knowledge of all the members. See for example the formal rules for general assemblies in Occupy Wall Street and the other Occupies.

Beyond this kind of structure of rules and different roles for some of the individuals in the group, a large group may be partitioned into subgroups, with each subgroup having a sub goal. For example a design team for a complex system such as an automobile may have sub-teams for various subsystem designs such as for example: the electrical system, the power system (engine, drive train, etc.), interior, exterior, brakes, etc. Of course the sub teams must communicate cooperatively with each other and the overall design team.

Almost all complex human activities requiring the cooperating work of large numbers of people will have their communication structures divided into substructures paralleling the system/subsystem structures of the things they are building, making, producing — automobiles, satellites, houses, buildings, movies, food production and distribution, education, research and development, provision of health care, sports, etc. This is because information must precede actions.

This is not to say that any one person, or even a few people designed, or could design any of these systems or subsystems. Present day designs build on previous designs. There is evolution of both systems and their designs (the information and knowledge used to make, build, produce, etc). Compare a Model T Ford with a Prius. No way could there be a Prius if there had been no other automobiles before it. So when we speak of the design team for a Prius we must include a huge number of people no longer living — huge numbers of people who designed and built automobiles in the past. This is not just about automobiles. It is about all complex human activities. All are dependent on the similar activities that came before. Any large cooperating group of people is the result of an evolution, a building on top of whatever came before it. And it would not be the way it is if what came before had been different.

So cooperation itself is evolving. So any explanation of cooperation which does not take account of this evolution is incomplete, deficient.

Is System 1 educable?

Maybe I am “prone to overconfidence” — most of us are — but there are many reasons that the extensive fact based scientific knowledge about how our thinking actually works, and sometimes makes mistakes, as described by Daniel Kahneman in his book “Thinking, Fast and Slow”, can be used by us to improve our thinking, our choices, and our decision making.

Why do we acquire new knowledge? To use it. How will we use this new knowledge about how we humans actually think? First we must use it in any theories that depend upon how we think. Any such theories must be modified to take account of this new knowledge — particularly economic theories, sociobiological theories, but really all social theories.

In practice, education will be changed to reflect these new facts, interpersonal communication will be changed, mass communication will be changed, art and science will be changed. This assumes that these new scientifically established facts will diffuse, will be spread far and wide, through most of our cultures, to almost all people. This will take time, but it will happen because this knowledge is useful. People who acquire this knowledge will think better. Their choices and decisions will better correspond to reality. They will get more of what they want. They will want more of what is good for them because they will better know what is good for them. They will be happier, healthier, and live longer than people who continue to think and communicate and choose and decide crudely and poorly.

Groups, organizations, societies, and cultures that acquire and use this new knowledge will be more effective, more efficient, more likely to attain their goals.

Daniel Kahneman is a proper scientist. He and other psychologists conjecture, test, validate, and methodically record and report the results psychological experiments. This is their job, this is how they see their jobs as scientists. They do not project to the future. But we can project to the future on the basis of sound principles of cultural, societal evolution. One such principle is: Knowledge — useful information — spreads through human communication. It will spread on its own through diffusion, person to person. And it will spread faster if those who have it deliberately spread it to more people.

So we can’t expect Kahneman, in concluding his book, in the quotes from yesterday (repeated below), to be as confident as I am here. Indeed there is an overconfidence mistake we often make. And most of us, on many occasions, are “prone to be overconfident”. But many of us are sometimes depressed, even overdepressed. This is a serious mistake too since it leads to inaction and sometimes death.

“What can be done about biases? How can we improve judgments and decisions, both our own and those of the institutions that we serve and that serve us? The short answer is that little can be achieved without a considerable investment of effort. As I know from experience, System 1 is not readily educable. Except for some effects that I attribute mostly to my age, my intuitive thinking is just as prone to overconfidence, extreme predictions, and the planning fallacy as it was before I made a study of these issues. I have improved only in my ability to recognize situations in which errors are likely: “This number will be an anchor …,” “The decision could change if the problem is reframed …” And I have made much more progress in recognizing the errors of others than my own.”  — Kahneman p. 417.

“The way to block errors that originate in System 1 is simple in principle: recognize the signs that you are in a cognitive minefield, slow down, and ask for reinforcement from System 2. … We would all like to have a warning bell that rings loudly whenever we are about to make a serious error, but no such bell is available, and cognitive illusions are generally more difficult  to recognize than perceptual illusions.” — Kahneman p. 417.

Kahneman says above “System 1 is not educable”. System 1 itself may not be educable but at least some of what it works on may be. Part of what System 1 works on is associative memory. This is information we acquire through learning and experience. Part of it consists of heuristics — rules of thumb, little rules, little associations, little connections we have learned and which we use automatically via System 1. Since this is all information we have learned one way or another, it may very well be able to be improved. Indeed Kahneman gives examples of vast improvements in the associations, the heuristics of System 1 in his chapter 22 “Expert Intuition: When can we trust it?” So experts can improve their associations and heuristics that System 1 operates on, and although not everyone can become an expert in everything, most people do become more or less expert in some few areas. So Kahneman’s own chapter 22 demonstrates that the quality of at least some of the information that System 1 works on can be improved. So System 1 can give better answers for us if we improve the information it works on.

When can we trust an expert’s intuition? Kahneman worked with another scholar, Gary Klein, on this question:

“At the end of our journey, Gary Klein and I agreed on a general answer to our initial question: When can you trust an experienced professional who claims to have an intuition? Our conclusion was that for the most part it is possible to distinguish intuitions that are likely to be valid from those that are likely to be bogus. As in the judgment of whether a work of art is genuine or a fake, you will usually do better by focusing on its provenance than by looking at the piece itself. If the environment is sufficiently regular and if the judge has had a chance to learn its regularities, the associative machinery will recognize situations and generate quick and accurate predictions and decisions. You can trust someone’s intuitions if these conditions are met.” — Kahneman p. 242.

The environment must be sufficiently regular and the expert must have learned its regularities.

The environment must be sufficiently regular and the expert must have learned its regularities.

“… [Some] experts may not know the limits of their expertise. … [they] … do have intuitive skills in some of their tasks, but they have not learned to identify the situations and  the tasks in which intuition will betray them. The unrecognized limits of professional skill help explain why experts are often overconfident.” — Kahneman p. 242.

So we need to learn to evaluate our own intuitions by asking ourselves: Is the environment, the subject matter (stock prices, psychological evaluations, politics, chess, medical diagnosis, etc.) sufficiently regular and if it is, have I really learned its regularities.

Can people learn these things? Can we learn how and when and where to be skeptical about System 1’s answers? Of course we can.