Tim Whitmarsh
This paper addresses what is arguably the single most important question for contemporary studies of ancient religion: can we imagine a critical language for analysing ancient polytheisms without retrojecting onto them modern theological categories?
That has been central to the field since the late nineteenth century, when classics’ anthropological turn shifted the emphasis away from appropriative readings based in idealism and Christian theology and towards stressing the radical otherness of antiquity. It has thus become standard to present Greek religion as wholly unlike Christianity, particularly in its protestant guise. To take but one example, Robert Parker’s recent On Greek Religion includes chapters called ‘Why believe without revelation?’, and ‘Religion without a church’ (Parker 2011). Yet while such tactics accurately diagnose the problem, they also risk exacerbating it by constructing ancient polytheism as merely the photographic negative of Christian monotheism.
Probably the most important area of contestation has been in the area of religious cognition. Price 1984 declared ‘belief’ an irredeemably Christian category, and therefore inadmissible in studies of Greco-Roman polytheism. Yet as Versnel 2011 has recently reminded us, the dogmatic assertion that ancient ‘orthopraxies’ could never accommodate notions of belief and disbelief flies in the face of ancient evidence – even if that evidence is not of the kind that exactly fits modern categories (esp. pp. 548-51; also Kindt 2012: 36-54; Harrison 2015). In the face of such discussion, the denial of ‘belief’ begins to look more like modern-western exceptionalism. More recent studies have thus begun to explore the idea that Greeks and Romans ‘believed’ (Morgan 2015) or even ‘disbelieved’ (Whitmarsh 2015) in their gods. This phenomenon is part of a wider shift away from the structural, anthropologically influenced models that dominated 20th-century scholarship, and towards actors’ categories – towards an understanding of how and why the ancients themselves understood their world.
This paper confronts an even more fundamental question. In 2013, Brent Nongbri argued that we entirely misconceive the apparatus of non-western, non-modern divinity when we use the very language of ‘religion’ (Nongbri 2013). Beginning with the Greeks, Nongbri demonstrates that thrēskeia had a very different, and much more limited, meaning than ‘religion’. If he is right, then even the study of ‘Greek religion’ as a discrete area already commits us to misreading ancient culture. My contention, however, is that he is wrong: the Greeks did indeed have an appropriate actors’ category. The concept of to theion or (more commonly) ta theia gained visibility in the fifth century. Semantically, it covered broadly the same field as modern ‘religion’ (i.e. human institutions, conventions and collective beliefs relative to the worship of the gods) – but it also covered more besides, since it was also used to designate aspects of the nature of divinity itself.
In this paper I argue that the issue is not that the Greeks lacked a proper concept of religion; quite the opposite, to theion was an expansive and powerful model for understanding the divinity of the world. The challenge for us, however, is to understand actors’ categories for what they are, i.e. not simply as proxies for ‘social structure’. Building upon Latour 2005 (the classic account of actor-network theory), I argue that the Greeks understood better than we often do that ‘religion’ is not a ‘thing’ that is witnessed passively or epiphenomenally by agents but an open-ended, plastic field within which actors creatively and stochastically improvise their own relationship to the gods. Greek religion certainly had a strong theoretical, even ‘theological’ (Versnel 2011) dimension to it; but to grasp that, we need to move decisively away from systems and structures and towards the more agile models offered by actor-network theory.