I don't understand your categorization of "rationalist" and "irrationalist". EVERYBODY is BOTH, to varying degrees on different topics. "rationalism" is NOT a tribe, with goals opposed to other tribes.
If the best way to meet one's goals is to join a fraternity and get branded, that is rational. Expending energy to fight a cabal or to set up a competing cabal seems far less rational than joining it.
I'm not a big fan of tribal thinking, preferring economic and intellectual motivators because they're more fine-grained and easier to ensure that they're consistent. But that doesn't mean that when I notice benefits of tribal cohesion, I'm not allowed to partake.
To help answer this question, consider a community of ninety people, living, perhaps, during the Bronze Age. For simplicity, let's assume that all the members are purely selfish. The ninety people are governed by a village council of nine, elected annually. At the end of every year, every one of the ninety people gives a campaign speech, and whoever can persuade the most people to vote for them wins.
Why would 90 perfectly selfish people agree to subject themselves to the outcome of this collective choice procedure if it wouldn't be in their interest to do so? If they did, then they wouldn't be perfectly selfish.
Isn't it more reasonable to think that selfish people would makes deals on a case by case basis (only ones that benefit them on net) and would not bind themselves to obeying the outcome of a collective choice rule that could produce outcomes that harmed their interests.
I don't want to get too far off-track from the theme of your post, but I think adding the premise that the agents be perfectly selfish is in tension with how your hypothetical is supposed to play out.
Being able to make collective choices at all seems to be an obvious benefit, even given pure selfishness. To consider a simpler example, imagine a group of twelve purely selfish soldiers. Would these soldiers agree to appoint a lieutenant, who they would agree to obey the orders of? Well, if they do appoint a lieutenant, there's a chance that the lieutenant will order them to do something dangerous. But if they don't appoint a lieutenant, they won't be able to fight effectively and will all be killed by the enemy anyway. The selfish choice is clearly to appoint the lieutenant.
Even if they would all agree to choose a lieutenant and obey his orders, that doesn't mean they would agree to abide by a collective choice rule that effectively gives the majority a permanent blank check. The battle scenario involves a unanimous agreement (that is only a special case of collective choice and happens to align perfectly with individual choice). Your battle scenario seems more like what I meant by "case by case basis" than agreeing to do whatever the village council says (even if it hurts their individual interests).
Presumably, the selfish agents wouldn't agree to give the lieutenant unlimited power over them or even rule beyond the length of the battle. Even if everyone agrees to abide by the collective choice for a given action, why would they agree to be bound by collective choice for any action the majority agrees to take up?
I might agree beforehand to help pay for a pizza with a topping that is selected by the majority of my two friends and I, but I surely wouldn't agree to be bound to do anything that my two friends agree should be done.
Feel free to ignore this comment if it is distracting you from the goal of your post.
They are perfectly selfish, not perfectly rational. Even a selfish person can believe it is intrinsically good to have a lieutenant (king, council) above them.
Perhaps it is an established tradition to have a council and there are penalties for disobedience. One can't always destroy the social order, never mind how much selfish one is.
(This post is the first in a planned sequence, discussing the ways in which irrationality can sometimes win. I write this with the hope that rationalists can see why rationality sometimes loses, and adopt "irrationalist" techniques if the techniques are actually effective.)
I'd like to begin this sequence by introducing one of the most powerful groups of people in the world- more powerful than such legendary organizations as Skull and Bones and the Freemasons. The members of this organization control wealth, power, and influence almost beyond imagining. Among their current and former members are six U.S. Presidents, four Vice Presidents, three Speakers of the House, three Supreme Court justices, dozens of Cabinet members, and well over a hundred Congressmen. This group also has vast economic power, founding and leading dozens of billion-dollar companies. No fewer than 86 members have been university presidents, of universities including Berkeley, U. Chicago, Dartmouth, MIT, and Yale. There are members of this organization in virtually every Fortune 500 company. And yet, the entire group is no bigger in size than a smallish US city. This organization definitely knows how to win. What is it? Is it the Bilderbergers? The Council on Foreign Relations? The Gnomes of Zurich?
No. It's the Delta Kappa Epsilon (DKE) college fraternity, known for branding its members with red-hot coat hangers during initiation.
College fraternity brothers are, on the whole, not the most rational of people. (I say this as a current Yale student.) Rationalists are supposed to win. DKE is clearly winning; the members of DKE are collectively far more powerful than the entire atheist movement, despite having far fewer members. What is going on here?
To help answer this question, consider a community of ninety people, living, perhaps, during the Bronze Age. For simplicity, let's assume that all the members are purely selfish. The ninety people are governed by a village council of nine, elected annually. At the end of every year, every one of the ninety people gives a campaign speech, and whoever can persuade the most people to vote for them wins.
Now, suppose that, out of this tribe of ninety, a smaller cabal of ten people forms, and the members of this cabal agree to help each other. Why the cabal was formed doesn't matter. The cabal could be about branding people with hot irons. Or about a cosmic Jewish zombie. Or about worshiping a jug of dirty swamp water- it doesn't matter. At the next election, by pure random luck, one of the group members gets elected to the council.
What happens during the next election? The villagers stand up, one by one, and give their speeches, which extol their own abilities, their willingness to work hard for the tribe, putting the tribe's interests above their own, and so on. However, when the cabal member who was elected last year stands up, in addition to promoting himself, he also talks about how capable another cabal member is. Since this guy is already on the village council, he must be capable and important, and lots of people listen to him. And when the votes are counted, the cabal has elected a second member to the council. This continues on for a few years, until the majority of the village council is made up of cabal members. Eventually, one must be a member of the cabal in order to be elected, and the cabal and the village council merge to become the same group. (This is a big part of where political parties come from, and why one effectively must be the member of a party to get elected in a modern democracy.)
The members of the cabal weren't more capable than anyone else. But the cabal members succeeded out of proportion to their number anyway, simply by being members of a group, a group that could have been based on anything at all. And since most people are irrational, most groups turn out to be based on irrational things. Hence, a random rationalist can be less likely to win than a random irrationalist, just because the irrationalist is a member of a group of irrationalists and the rationalist is operating individually.
We have previously discussed the question of whether a group of rationalists can win over a group of irrationalists. However, even framing the question this way is a serious case of privileging the hypothesis, because it simply assumes a rationalist group and an irrationalist group, a Blue Team and a Green Team, of at least comparable power. But, most of the time, the world doesn't actually work this way. The most common situation is, not a team of rationalists against a team of irrationalists, but individual rationalists competing against groups of irrationalists.
This certainly doesn't mean that things are hopeless. (If things were hopeless, there'd be no point in writing about them.) There are many possible solutions to this problem. One can pretend to be an irrationalist to join a group of irrationalists. One can join a group of irrationalists that doesn't care much about conformity, and won't exclude you for being a rationalist. One can say "screw everyone", and try to win despite not being a member of a group. One can try to start their own group of rationalists (be warned that this is a Hard Problem). But, in order to formulate a solution, one must first consider the problem. Odds are, most of the people you know are less rational than you. How are you going to handle it?
EDIT: Although this post is related to Why Our Kind Can't Cooperate, rationalist ability (or lack thereof) to cooperate isn't the main point. The main point is that there are many more irrationalists than rationalists, so in any given population, groups of irrationalists will form long before groups of rationalists do, even assuming equal ability to cooperate.
EDIT 2: For a young, but growing, attempt at setting up such a group of rationalists, see the Existential Risks Career Network.