Adapted from a comment here.
Re: the necessity of overcoming “primitive tribalism” if we are to live “in a better, freer world”. Since actual, flesh and blood men do not live in the “world” but rather specific parts of it, and since there are apparently compelling genetic and evolutionary reasons for believing that ethnocentrism will be with us for the foreseeable future, the following considerations seem in order.
Let A and B be ethnic groups, the members of which, being “primitive tribalists”, are ethnocentric: they will, to borrow the language of game theory, “cooperate” in interactions with their co-ethnics and “defect” in interactions with non-co-ethnics. Whenever any two members (say a1 and b1) of these respective groups meet, non-cooperation will be the outcome; whenever two members of the same group meet (say a1 and a2, b1 and b2) the outcome will be mutual cooperation.
Now take another ethnic group, C, whose members, have overcome “primitive tribalism” and replaced it with civilised universalism. They are humanitarians: they treat each person they meet as a unique individual. Assuming that they are also a good-natured bunch willing to give everybody a fair shake, whenever any two members of C (say c1 and c2) meet, the result will be mutual cooperation. If C remains isolated from other ethnic groups the level of voluntary cooperation in its society will be no different from – and certainly not any higher than – an ethnically homogenous ethnocentric society.
But what happens if the humanitarians of C are not isolated from other ethnic groups? What happens if members of ethnocentric group D make their home in C’s land? An ethnocentric “primitive tribalist” will defect in games with non-co-ethnics and humanitarians will cooperate in games with non-co-ethnics, so when, say, c1 and d1 meet, c1 will cooperate and d1 will defect. The humanitarian will play the sucker, the ethnocentrist the free rider (the worst and best outcomes for the humanitarian and the ethnocentrist respectively). Now as c1 is a good humanitarian not some “primitive tribalist”, he will of course not draw the conclusion that D’s are an untrustworthy people. No, he knows that the only decent thing to do is to treat everyone as an individual. So when he meets another member of D, d2, he will play cooperate and the ethnocentrist d2 will, naturally, play defect. Again c1 will end up the sucker, and again as a good humanitarian he will not draw the conclusion that D’s are an untrustworthy people; he will go on treating all God’s children as individuals. And so when he meets d3 he will end up the sucker, and when he meets d4 he will end up the sucker, and when he meets d5 he will end up the sucker, and when he meets… What is true of c1 is true of his fellow humanitarians c2, c3, c4, etc, so we can say that the strategy of C is, effectively, unconditional cooperation, whereas the strategy of D is unconditional defection – the humanitarianism of C condemns it to being bled dry by (any and all) ethnocentric groups. Whilst the non-cooperation that results whenever members of different ethnocentric groups meet is far from ideal, much worse is to play the sucker for all eternity (or until your society collapses, at least). And so we reach the paradoxical conclusion that if a humanitarian society wishes to survive, its members really ought to retain a smidgen of “primitive tribalism” vis-à-vis admitting newcomers. But if this is too distasteful, they could, as Hobbes recommends, institute “a common power to keep them all in awe” with all that that entails.