In many ways Roman religion emerges from the standard modern accounts of it as a dry and highly ritualized religion with few if any concessions either to religious self-expression or imagination about the role of gods in the life of men. Historians have tended to associate this character with the success of the Romans in practical aspects of life – warfare, engineering, town-planning – and to suggest that they organized their religious lives with the same kind of brutal efficiency, striking crude bargains with the narrow-minded gods and goddesses that they had themselves created. It will be clear later on that there are some reasons for taking this view, but there are also reasons to think it only a part of the truth, and profoundly misleading if it is mistaken for the whole truth.

The fundamental principle on which the religion of the Romans used to be interpreted was the idea that the Romans were an unusually conservative society. In some respects, this is perfectly true: we can for instance show that some of the rituals that were still being practised regularly in the first century BC, and even later than that, were already in place in the sixth century BC; again we know from later Roman orators such as Cicero that they placed a great emphasis on the ancestral customs and ways of the Roman people (the mos maiorum) and it is a reasonable guess that this was not a new idea in Cicero’s day but a long-cherished attitude; again, it is quite clear that the Romans placed a great deal of emphasis on getting their rituals precisely right in every detail, so that the slightest error invalidated the whole ceremony of which it formed part. If so, and if they did this year after year, there should have been no change at all. How could conservatism be taken further?

The Republic is the name we use to describe the characteristic political system of the Romans from the fall of the monarchy until the establishment of the new, though disguised, monarchy of Augustus and his successors. It is a system that we know quite well from its last century or so, because the writings of Cicero, Caesar, and Sallust as well as those of many others survive from that period. From the earlier years of the Republic we have only the historical tradition preserved by Livy and Dionysius of Halicarnassus, both of whom wrote in the period of Augustus; their tradition is in fact dependent on earlier writers, but even these only take us back to the middle of the second century BC at the earliest. There is therefore a great gap between the creation of the system in the late sixth century BC and the recording of it at least three centuries later. For the most part what follows describes the situation in the second and first centuries; how far the picture can safely be projected backwards into earlier centuries is, as we have seen, very much an open question. Given the Romans’ reluctance to make basic changes in their institutions, it would not be surprising if the evolution from the monarchic system of archaic Rome to the developed Republican system we know took some generations to achieve.

Roman deities are not very easy to classify. They were never organized into a pantheon in ancient Rome, so that we cannot say there was a fixed number of them or that they fulfilled particular functions divided between them. Some of them are clearly part of the Roman tradition from the earliest periods of the city; some are later additions, introduced at specific dates with specific ceremonies, which the Romans carefully recorded. Some have very complicated histories, having multiple associations with different areas of life and often different additional names, defining a particular aspect of the deity. Others seems to exist only because of a particular function or particular moment. Some have special priests, or in the single case of Vesta, priestesses; many, however, do not, but fall within the ambit of the major college of the pontifices.

Nobody can doubt that rituals of various kinds were a crucial part of the interactions between men and women and gods and goddesses in the religion of Rome. Rituals marked all public events and celebrations: some of these we would classify as religious occasions – annual festivals, the taking and fulfilling of vows, the anniversaries of temple foundations; others as secular – elections, the assembly of legions for warfare, the census of Roman citizens; still others challenge our own criteria – the games, or dramatic performances, which certainly had ritual elements in the programme, even though we are inclined to regard entertainment as their primary purpose. It is important to see that these are our problems of interpretation: we have no reason to believe that the Romans made such distinctions or found the existence of different types of ritual in itself problematic.

The Sibylline Books played a very central but elusive role in the religious history of the Roman people. Originally, according to the tradition, they consisted of a set of Greek oracles kept by the Romans as one of their holiest texts. The story ran that an old woman offered nine books to King Tarquin (the fifth king, conventionally 616–579 BC) and asked for a price. He refused to buy them at her price; she reacted by destroying three of them and offering the remaining six at the same price; he refused again and she reacted in the same way again, offering the last three for the same price as before. At this point, he was impressed, consulted the priests, and at last realized his mistake, thus agreeing to buy the last three books at her original price.

It is easy to think that pagan religion was essentially tolerant of other forms of religious activity. The evidence we have just considered seems at first sight to support just such an interpretation. If the Romans could accept a steady flow of new cults into their city over such a long period of time, that must surely imply that their attitudes were basically tolerant. There are, however, qualifications to be made before reaching such a conclusion. It is clear that, if there was tolerance, it was not tolerance born of principle. So far as we know, there was no fixed belief that a state or an individual ought to tolerate different forms of religion; that is the idea of far later periods of history. The truth seems to be that the Romans tolerated what seemed to them harmless and drew the line whenever there seemed to be a threat of possible harm; only, they saw no great harm in many cults of their contemporary world, where many individuals and cities worshipped gods and goddesses much like their own.

As the argument of the last chapter suggested, one level at which religious change had its most profound and lasting effects was in the character of the language used in religious discourse. This has a double impact on our understanding: first, it is one of the key areas in which we can see religious change taking place; but, secondly, understanding the different uses of language is a precondition of interpreting the different religions themselves. The Christians spoke a far more direct language about the character of the deity, the character of priesthood, about the beliefs that an individual ought to hold or not hold and about the character of the interaction between divine and earthly beings. To some extent, the pagans were forced into making explicit statements by way of competition; but, far more clearly, we should regard them as disadvantaged in the competition with Christianity by the nature of their own traditions of religious discourse. It is easy but dangerous to jump from this observation about pagan habits of speech to making simplifying assumptions about their religious experience. Just because they are inexplicit and guarded in what they say, we have no right to assume that they were either hypocritical or uninterested in regard to the gods and goddesses they worshipped or that the rituals they carefully maintained and practised carried no meanings for them.