Here is a model to help explain how a fairness-based social organization can evolve.

Suppose there is a group of 10,000 humans, newly captured strangers who must now live together. They are faced with a common problem. Maybe they are all enslaved by some other kind of people who have the capability of taking any of them away for punishment or execution.

If one of these prisoners betrays another prisoner in order to gain some advantage from the guards, all the other prisoners will soon know of it. If one prisoner helps another prisoner evade punishment by a guard, that makes the helper have some value to the one who was helped.

The prisoner who is especially good at defeating the guards may befriend and help several fellow prisoners. The whole group of them may coordinate their actions under the guidance of the most capable among them.They will come to be bound together by bonds of both trust and responsibility. Failure to act responsibly may merely cause ill feeling or it may cause the rest of that group to ostracize the irresponsible one.

Among a large population of prisoners, many such co-responsibility groups may form. If their leaders are wise, they will bind themselves into similar co-responsibility groups, and out of that may come a leader at a higher level who will be able to coordinate tasks alone the two or more lower-level groups.

If everyone keeps in mind that the main problem in their lives is the suppression and harm coming from their enslavers, then a single leader may emerge to direct the activities of them all. That is not to say that this leader will tell them individually what to do, but that s/he will tell leaders on the second level of some objective that needs to be achieved, and they will work out the practical details on their own.