What ancient religion was not

What is attested

Greek and Latin writers had a large vocabulary for the things a modern reader files under religion, and no word for the file itself. Greek uses ta hiera (the sacred things — meaning both the rites and the objects and precincts), ta theia (matters concerning gods), ta nomizomena (the customary observances), ta patria (ancestral usages), eusebeia (the correct conduct owed to gods, parents and the dead), and hosia (what divine sanction permits). Threskeia, cultic observance, becomes commoner in Hellenistic and imperial Greek and is the term Greek-speaking Jews and Christians later put to heavy work. Latin has religio, but the definitions the Romans themselves offered point at scruple and obligation rather than at a system of doctrine: Cicero derived it from relegere, to go carefully back over what belongs to the worship of the gods, and three and a half centuries later the Christian apologist Lactantius derived it instead from religare, to be bound, and said explicitly that he was correcting Cicero. Both etymologies are questionable philology, and that is not the point. The point is that Romans argued about what religio named because it named a disposition and a set of scrupulous practices, not a creed one subscribed to.

The institutional evidence shows the same absence from a different angle. Inscribed sacrificial calendars — the fourth-century calendar of the Attic deme of Erchia and the calendar of Thorikos, both substantial stelai, and the Athenian civic calendar re-inscribed during the law revision of the years around 410 to 400 BC, which survives only in fragments whose reconstruction and attribution to that revision are disputed — are essentially schedules and budgets: month, day, deity, epithet, victim, where it is to be offered, what it costs, who may eat. They contain no statements of belief because that is not the kind of document they are. Priesthoods were offices rather than vocations: at Athens many were held by lot or by descent within particular families, at Eleusis the two senior priestly lines were the Eumolpidae and the Kerykes, at Rome the pontificate and augurate were held by senators alongside political careers, and in a number of Hellenistic cities priesthoods were sold at auction with the terms inscribed. There is no seminary, no ordination into a separate estate, no congregation of members, no baptism or equivalent entry rite for civic cult, and no scripture. Homer and Hesiod were authoritative poetry, quoted, taught and also openly attacked — Xenophanes on the gods' immorality, Plato on the poets' theology in the Republic. Herodotus could say that Homer and Hesiod were the ones who gave the Greeks their account of the gods, and that they lived not more than four hundred years before his own day; that is a statement about poets, not about prophets, and nobody treated their texts as legally binding revelation.

The practice was also not separable from public life, which is why the modern category fits so badly. Athenian Assembly meetings opened with purificatory and sacrificial acts; the Roman Senate met in an inaugurated space and public business could be halted by adverse auspices; treaties were sworn by named gods and inscribed with the oath; armies sacrificed before battle and generals travelled with diviners; magistracies at Rome carried sacral duties, and the calendar of days on which business could be transacted was itself a sacred calendar. Nothing in the surviving record marks off a religious sphere from a political one, because the boundary a modern reader wants to draw was not drawn.

This is where the negative claim has to stop, and it is worth stating positively what the evidence does show about belief. Votives are among the most abundant surviving classes of religious object, and at many excavated sanctuaries the most abundant of all: an offering inscribed, in Greek, with the formula that the dedicant vowed it, and in Latin with the standard abbreviation for paying a vow willingly and deservedly. What the formula attests is that the dedicant recorded the discharge of a vow; that the dedicant took a specific god to have acted is an inference from the transaction, though a strong one. Curse tablets, folded lead sheets pushed into graves and wells and springs across the Mediterranean, presuppose powers who receive and act on written instructions. Oaths presuppose divine punishment for perjury; the healing inscriptions at Epidaurus narrate in the third person what the god is said to have done to a named patient. Plato's Laws legislates against three distinct positions — that gods do not exist, that they exist but do not care about human affairs, and that they can be bought off by offerings — which at least shows that a fourth-century Athenian could formulate these as distinct positions worth legislating against. Athenians prosecuted men for not acknowledging the gods the city acknowledges. To say that the ancients had no bounded category called religion is defensible. To say they had rites without convictions is not; the convictions are visible in the documents, they simply are not organized into a creed and were not enforced as one.

Two further qualifications matter for anyone writing about this. First, the practice was not uniform: alongside the cults of a city there were elective associations (orgeones, thiasoi, the collegia of Roman Italy), initiatory cults with restricted membership, household observance, professional divination and binding magic, and the theological argument of philosophers, some of which reached conclusions incompatible with civic practice while their authors continued to take part in it. The influential model that treats the polis as the frame that gave all Greek religious activity its meaning has been criticised precisely for how much of this it leaves at the margins. Second, the boundary was not invisible to the ancients themselves. Roman law distinguished res sacrae, res religiosae and res sanctae; Varro wrote a large systematic treatise on res divinae, divine matters, organized under headings; Cicero wrote a whole dialogue on the nature of the gods. They could and did treat the sacred as a subject with an internal order. What they did not have is the modern assumption that this subject is fundamentally about believing propositions, and that a person has one religion at a time.

How the rite worked

The routine unit of practice was the sacrifice, scheduled by calendar and funded by identifiable money. The inscribed calendars specify the animal, sometimes its sex, colour and price ceiling, and whether the meat was to be eaten on the spot or carried away. Funding came from public treasuries, from the rents of sacred estates, from the sale of priesthoods where that applied, from the sale of the hides of sacrificial victims, and at Athens from liturgies — wealthy citizens assigned to underwrite a chorus or a delegation. The record of these arrangements is an accounting record. It tells us who paid, how much, on what day and to which god, and it tells us nothing about what anyone thought while it happened.

Personnel were part-time and civic. A priest or priestess typically served one deity at one sanctuary, often for a fixed term, sometimes for life, with the duties and perquisites — which cuts of meat, which fees, which seat at the theatre — spelled out in the appointment document. Expertise existed in specific technical bodies rather than in a clergy: the Athenian exegetai who advised on sacred law, the Roman pontifical college which controlled the calendar and the archives of formulae, the augurs who read the signs, the quindecimviri who consulted the Sibylline books when the Senate directed. None of these bodies taught doctrine or examined anyone's belief; they answered procedural questions about what had to be done and whether it had been done correctly.

Outside the civic frame, practice was conducted in the house and in voluntary groups. Domestic cult is visible archaeologically — hearths, courtyard altars, the painted household shrines of Pompeii and Herculaneum with their bronze figurines — and almost invisible textually, because nobody wrote down what was said at a family sacrifice. Voluntary associations left inscribed rules about dues, meetings, funerary provision and discipline, which is an important reminder that where membership did exist it was regulated in the way that a club or a burial society is regulated, not by doctrinal examination. Where an initiation was involved, the material record stops at the door: the cult buildings survive, the content does not.