After being asked how people came together to build cooperative societies beyond kinship, Haidt asserts that "morality" was the key:
"A big part of Haidt's moral narrative is faith. He lays out the case that religion is an evolutionary adaptation for binding people into groups and enabling those units to better compete against other groups. Through faith, humans developed the "psychology of sacredness," the notion that "some people, objects, days, words, values, and ideas are special, set apart, untouchable, and pure." If people revere the same sacred objects, he writes, they can trust one another and cooperate toward larger goals. But morality also blinds them to arguments from beyond their group."
If we take ethnohistoric hunter-gatherers for our model of how people formed larger and more cohesive groups in the ancient past, Haidt's "morality" answer is patently wrong. These groups were held together by kinship ties first and by extended or fictive kinship second. Their "religions" (i.e., shamanisms) weren't grounded in morals and weren't much concerned with morals. While such groups had moral norms and ethical rules, these weren't twined with supernaturalism and had an independent, non-spiritual basis.
Large communities held together by religion-faith-morals are a recent development in human history, no more than a few thousand years old. The kind of community that Haidt describes is a post-Neolithic formation that has its origins in the Axial Age. So does the idea that religion is a matter of "faith." These are not ancient or evolutionary ideas. Moralizing gods and religions are relative newcomers to the supernatural world.
Haidt's mistake here is a common one: observe modern or relatively recent cultural formations and then uncritically project them back into the ancestral or evolutionary past. This mistake has other consequences, which are evident in what Haidt calls "innate" or evolutionary moral foundations: "care/harm, fairness/cheating, liberty/oppression, loyalty/betrayal, authority/subversion, and sanctity/degradation." These "innate" concerns sound suspiciously modern; I suspect at least a few are products of post-Neolithic and Western societies. I hate this notion that "religion" is an ancient evolutionary construct that arose to bind people together. First of all, "religion" is a modern notion; this category is not recognized by most societies we know about, including China and Japan as recently as AD 1850. Second, as Campbell points out, the religions of hunter-gatherers have very little moral content. (What was the common morality of the Mongol empire?) Third, religion, despite what secular professors like to say about it, is not a social phenomenon, but something that resides primarily within the minds of believers. People believe things because they make them feel better, not because they want to be part of a big group. (Much more on this here.)
I think all of this is modeled on the survival of the Jews and the extraordinary expansion of Islam, and those are very interesting historical events. But they are not more important than, say, the survival of the Kurds or the Han Chinese, or the extraordinary expansion of the British empire.