The scale at which humans cooperate expanded greatly over the last 10,000 years—from hundreds of people to hundreds of millions. One popular theory that explains this dramatic increase in the scale and complexity of human societies is known as the Big Gods hypothesis. The basic idea, as Ara Norenzayan explains in his book, is that “watched people are nice people.” In small-scale societies people are constantly watched by kin and kith, who will impose sanctions on them for antisocial behavior, such as free-riding on collective efforts to produce public goods. But who will watch people in large-scale societies in which people need to cooperate with complete strangers? The proponents of the Big God hypothesis have an answer: all-seeing and all-powerful supernatural beings will see when people do wrong and punish them, sometimes in this life (by bringing misfortune on their heads) and sometimes in the afterlife.
Anubis weighs the heart of a recently departed against the feather of Maat Source
This is a neat explanation, which provides a solution to the free-rider problem in very large groups of people and, apart from being somewhat reductionist, it makes sense to me (see my review of Ara’s book on this blog). Cross-cultural research confirms that there is a strong association between social complexity and belief in moralizing high gods (see this article and references in it). We see the same pattern in Seshat data:
But as any statistician will tell you, correlation is not causation. In particular, the empirical pattern we see is equally consistent with either the idea that Big Gods gave rise to Big Societies, or that Big Societies gave birth to Big Gods. Which direction does the causal arrow go? Prior analyses using “static” data, in which we see characteristics of a society at a particular point in time cannot easily resolve this question. The Seshat Databank is really a unique resource, because it traces how societies in different parts of the globe change over time. As a result, we can ask a very simple but very important question: which comes first, Big Gods or Big Societies?
On Wednesday the Seshat team published an article in Nature that answers this question. We analyzed 414 polities (politically independent societies ranging from independent villages to chiefdoms, states, and empires) from 30 different locations spread around the world:
The global distribution and timing of beliefs in moralizing gods shows that they appear appear in complex societies. The area of each circle is proportional to social complexity of the earliest polity with moralizing gods to occupy the region or the latest precolonial polity for regions without precolonial moralizing gods. For regions with precolonial moralizing gods, the date of earliest evidence of such beliefs is displayed in thousands of years ago (ka), colored by type of moralizing gods. Whitehouse, François, Savage, […] Turchin. (2019) Nature.
The time frame of the analysis spans more than 10,000 years, beginning with Neolithic Anatolians (today Turkey) in 9600 BCE. Our analysis confirmed that there is an association between Big Societies and Big Gods. In many cases we see these two cultural characteristics appear simultaneously (within 100 years of each other). But there are many instances where Big Gods trail the transition to large-scale complex societies. And in no case do we see Big Gods appear well before this transition. To see this on a single graph, lets focus on all the Seshat areas in which societies achieved large scale (to be specific, when the total polity population increases from hundreds of thousands to millions of people) before the colonization era. Let’s call the moment of transition to a Big Society “Time 0” and mark the time when each such area acquired Big Gods in relation to this point. We then see the following pattern:
Source: analysis of Seshat data by the author
The thin grey lines trace the evolution of social scale in each individual Seshat area (Natural Geographic Area). Individual trajectories have been shifted so that Relative Time = 0 is the moment when they exceed 5 on the Social Scale (corresponding to the transition to millions or more of population). The thick brown curve is the averaged, or “typical” trajectory. Orange bars show when each area acquired Big Gods (or, more technically, BSP: broad supernatural punishment or MHG: moralizing high gods). (This figure provides a somewhat different angle on the same pattern as in Figure 2 of the Nature article, because there individual trajectories are shifted so that Time = 0 corresponds to the appearance of Big Gods.)
However you slice it, the conclusion is that Big Gods do not precede Big Societies. At best (in about half the cases), they appear simultaneously, but in the rest of cases they can trail the transition to Big Society by hundreds, and sometimes even thousands of years.
This is not to say that the Big Gods hypothesis is entirely wrong — just one aspect of it, which predicts first Big Gods, then Big Societies. Additional (and as yet unpublished) analyses I’ve done on these data support a feedback loop relationship between Social Scale and presence of Big Gods (BG). There is a very strong causal arrow from Scale to BG, and somewhat weaker feedback from BG to Scale. However, the feedback starts operating only once Scale exceeds 5; that is, once societies become large-scale and complex. The major implication of this regression result (together with the timing of BG appearance) is that moralizing gods and supernatural punishment are just one, albeit important, of the social technologies that are needed to stabilize large-scale societies when they arise. Other such stabilizing cultural traits include equity institutions that reduce inequality, the common identity provided by world religions that stabilize multiethnic societies, bureaucracy for better administration, and others. Big Societies are highly fragile when they first appear and they need many such institutions to make them more resilient to internal and external shocks. Those Big Societies that didn’t acquire enough stabilizing institutions break apart and are replaced by more cohesive societies. As a result, after thousands of years of cultural evolution we see a near universal presence of such stabilizing cultural characteristics, and Seshat data show that Big Gods is one of the important ones.