Over half of the world’s population currently follow one of two closely related religions, Christianity and Islam. The concept of a Big God, who actively monitors human behaviour and punishes immoral actions, is a major feature of these religions and might explain their success. It is suggested that belief in supernatural punishment for antisocial behaviour, especially supernatural punishment by Big Gods, makes people more cooperative and, historically, helped humans to build large, complex societies.

Though this theory about Big Gods is intuitively appealing, it’s difficult to test. The standard evidence presented in its favour – that Big Gods are associated with big societies – is not convincing. The problem is that cultures share many traits with one another because of common ancestry and cultural transmission, not just because one trait causes another. In many big societies today people eat pizza, drive cars, consume alcohol, are subject to an official police system, and live under a formal government. While some of these features, such as an official police system and a formal government are likely to be causally related, nobody is going to claim that pizza causes formal government systems. The challenge is teasing apart which associations are due to a causal relationship between the traits, and which associations simply reflect patterns of cultural transmission or inheritance from a common ancestor. Experimental studies suggest that religions are capable of increasing cooperation within groups, but don’t tell us whether Big Gods have actually led to big societies.

Fortunately, evolutionary biologists have developed powerful new computational phylogenetic methods that account for the common ancestry of species. These methods are capable of getting at causality by indicating which trait tends to come first, and whether a particular trait increases the chance of another arising. These methods have recently been used to address longstanding debates about the cultural evolution of political complexity, kinship systems and language structure. In a recent study we formally tested whether belief in supernatural punishment helped to build politically complex societies in indigenous Austronesian cultures. This family of cultures originated in Taiwan and spread across a region covering over half the world’s longitude, from Madagascar in the west to Easter Island in the east. Austronesian cultures provide an ideal sample to test whether political structures have coevolved with supernatural beliefs and practices because the indigenous features of these cultures – how they were before influence by Christian and Muslim cultures – was well documented, and highly varied. The political structures of Austronesian-speaking cultures ranged from the localised big-man societies of the Lak to the politically complex states of Hawaii. Beliefs in supernatural beings were no less varied. Nature gods were associated with features of the natural environments, an example being the taniwha of the Māori, who were found in deep waters and often preyed on humans, but could be outwitted, fought, or placated with offerings. There were countless cultural heroes, such as Porana, trickster-orphan-hero of the Solomon Islands, who performed a range of feats such as introducing humans to the pleasures of sexual intercourse, but was of little religious importance. Deceased ancestors were often believed to continue to play an active role in the lives of their descendants. Some of these ancestors were elevated to the place of powerful gods, while others had only limited powers. While some supernatural agents were interested in human actions, such as sacrifices and offerings, few were specifically supportive of human morality. Less than half of the cultures in our study believed in supernatural punishment for selfishness, and the concept of a moral big god was only found in a handful of cultures. Living in a modern culture, it is easy to assume that religion and morality are intrinsically linked, but many of the gods of pre-modern cultures were simply not interested in how humans treated each other. Using a language based family tree of Austronesian cultures we used computational phylogenetic methods to test whether political structures have coevolved with supernatural beliefs.

So did we find support for the Big God theory? Not as such. Though we found that big moral gods coevolved with political complexity, these gods followed, rather than drove, political complexity. Interestingly, most big moral gods were found in regions of Southeast Asia where there was a history of contact with the Islamic world. These gods were claimed to be indigenous, but the recent origin of these gods, and the history of the cultures in which they arose, suggests that the concept of a Big God may have been borrowed from Islamic cultures under the name of an indigenous god.

Can we conclude that religion serves no social function? Not at all! We found that belief in supernatural punishment by a broad range of supernatural agents, not necessarily by just a Big God, came before the rise of political complexity. In other words, moralising religious concepts such as supernatural punishment may have helped humans build large politically complex societies. However, we found that supernatural punishment did not help maintain politically complex societies once they had been built. While it is possible that supernatural punishment increased cooperation, there are also other ways that supernatural beliefs could have helped build politically complex societies. The spirits of deceased ancestors were often believed to be both a source of supernatural punishment and a way of justifying claims to political authority. Chiefs were often believed to have descended from prominent ancestral spirits or gods, and these deities may have provided divine justification for political leadership.

The co-evolution of political structures and supernatural concepts in Austronesia provides a powerful insight into the evolution of pre-modern societies. These cultures built big societies without Big Gods. While today belief in Big Gods is common throughout the world, in the vast majority of cases these beliefs are associated with Christianity and Islam. These religions arose and spread within the last few thousand years, well after humans developed large, complex societies. Our research shows the power of phylogenetic methods to address debates about the origins and functions of religion in human society.