The evolution of the earliest cities, states, and civilizations is an enormous topic and writing about it is made no easier by my discomfort with the term “evolution” itself. Although I criticize “neo-evolutionary” theory – that is, the attempt to create categories of human progress, which in anthropology stems from the nineteenth-century work of Edward Tylor and Lewis Henry Morgan and which was revivified in the mid-twentieth century by Leslie White and Julian Steward and others – I do not reject the term evolution or social evolution.

Economically stratified and socially differentiated societies developed all over the world from societies that were little stratified and relatively undifferentiated; large and densely populated cities developed from small habitation sites and villages; social classes developed from societies that were structured by kin-relations which functioned as frameworks for production, and so forth. These changes must be explained, and archaeologists have been doing the job with remarkable success for more than a century, with the pace of research quickening in the last decades. As I discuss throughout this book, it doesn't much matter what we call things, as long as we explain clearly what we mean, and as long as our categories further research, rather than force data into analytical blocks that are self-fulfilling prophecies.

This book is about the earliest states, particularly the constellations of power in them, and also about their evolution, that is, where varieties of power came from.

The term “agency” is much in vogue in the archaeology of the early third millennium AD (Dobres and Robb 2000), but archaeologists have not agreed on its meaning. It has been, for example, the subject of a recent essay in which an archaeologist discusses how great men achieved status and transformed traditional societies by getting access to guns or iron axes (Flannery 1999). This is held to explain how the old order succumbed to the forces of European progress. “Agents,” however, are not just powerful men, but can be anyone. They are people who belong to several kinds of social groups simultaneously and who must negotiate their economic and social status and even their identity in certain historical moments. Modern archaeologists want to study how individuals experience material conditions and how new beliefs and meanings are inscribed in individual lives, especially in times of social change (Meskell 2003). Archaeologists who infer beliefs from material culture subscribe to the motto of William Carlos Williams, “no ideas but in things,” and historians have always been able to study the lives and actions of individuals as depicted in texts. Historical archaeologists are doubly armed.

The old rules of social evolutionary theory that were used to explain the rise of the earliest states haven't worked – as I have argued in the preceding chapters. Indeed, the neo-evolutionist model developed by social anthropologists in the 1950s and 1960s and then operationalized and employed by archaeologists in subsequent decades now hinders research or, more often, is simply ignored by contemporary archaeologists. The old neo-evolutionist game was played on the central assumption that modern “traditional” societies represent stages in the development of the earliest states. These old rules were developed within American departments of anthropology as archaeologists, seeking respect from their social anthropological colleagues (as well as jobs, promotions, grants, and status), attempted to model prehistoric societies via analogies to cases described by ethnographers.

These archaeologists, who thus claimed to be genuine anthropologists, were surprised when their colleagues insisted that “traditional” societies had histories of their own and could not be inserted as “models” into a prehistoric, neo-evolutionary trajectory. In the 1980s and 1990s the whole edifice of social evolutionary theory was abandoned by non-American archaeologists, who were never part of an academic anthropological establishment. The remaining neo-evolutionists were disquieted further when many archaeologists turned to evidence provided by ancient historical sources, from Mesopotamian cuneiform tablets to New World relaciones and visitas, new decipherments of Maya glyphs and Chinese oracle bones, and art historical evidence from Egypt to Teotihuacan, evidence of particulars that confounded facile analogies.

1988 was a banner year for the archaeological study of the “collapse” of ancient states and civilizations: two books on the subject appeared in that year, The Collapse of Complex Societies by Joseph Tainter (1988) and The Collapse of Ancient States and Civilizations, edited by George Cowgill and myself (Yoffee and Cowgill 1988). These books marked exceptions in social evolutionary theory, which was mainly concerned with the origin, development, and growth of particular states. Although I have criticized such studies as failing to explain the changes over time in social roles and the relations of domination, the proliferation of new, multiple, and overlapping identities in early urban communities, and the development of new ideologies of centralization, it is equally clear that discussion of the collapse of ancient states and civilizations, with the notable exceptions of the Maya and Roman cases, was, before 1988, conspicuous by its absence from the literature on social evolution.

The concern with rise, to the near exclusion of collapse, in social evolutionary studies, had important theoretical ramifications: social change was perceived as a process of mutually supportive interactions that produced an irreversible succession of levels of holistic sociocultural integration (Adams 1988). Studying collapse requires that levels be broken down into institutional groupings of partly overlapping and partly opposing fields of action that lend the possibility of instability, as well as stability, to overarching societal institutions.

Just as the analysis of Livy's religious material required some awareness of his broader aims and methods, so too with Tacitus it is essential to acknowledge the way in which he undertook his project and what that was. Of course society had changed profoundly with the transition to empire and we should not expect that the eminently political act of writing history had not also adapted. Nonetheless we shall see that if circumstances and methods had changed, many similarities remain. Like his predecessor, Tacitus constructed a representation of Roman state religion from the events of the past: he ‘made sense’ of what had gone before and produced an account that reflect his idealised religious system while organising his coverage of events to argue a case. His was not the only possible version of events, even if we find it plausible historically: the historian/emperor Claudius, for instance, might have left us a very different version, given the chance. We shall therefore not only explore the way that religious institutions are represented, but also explore the agenda that helped to ‘inform’ the facts at Tacitus' disposal.

What emerges is a coherent programme, shaped by selectivity, powerful timing and presentation, with typically Tacitean vigour. He knew his own mind on religion, though this has not generally been the accepted viewpoint. Though many of the religious notices are apparently neutral, once we appreciate Tacitus' techniques of juxtaposing contradictory information, it will be clear that there is no such thing as ‘mere’ inclusion: virtually all ‘religious’ notices are pertinent and combine to create a picture of what is usually best described as incompetence.

Introduction

Over 250 years passed between the works of Tacitus and Ammianus Marcellinus. How, exactly, we are to think of the religious aspects of the period in which the latter wrote is a complex question that could easily merit a book in its own right. We can point to the survival of many pagan cults and practices for centuries after Constantine. Or, with our long hindsight, we could point to symbolic moments such as the death of Julian, the last pagan emperor, in 363, as the end of a thousand-year-long era. Perhaps the ‘real test’ of the slippery process of Christianisation was the position of the aristocracy, which gives us a slightly later and appropriately more nebulous date. We might see the end of state paganism in the refusal of Gratian to be Pontifex Maximus, probably in 376, or the failure of the so-called ‘pagan revolt’ of 394 and the death of the ‘pro-pagan’ usurper Eugenius in the west. Alternatively, we might adopt a different approach: older models of an all-out contest between paganism and Christianity have been supplanted by models of assimilation, where most aristocrats gradually, and (on the whole) peaceably, became part of the new order. Whatever our criteria, clearly something drastic happened during the fourth and early fifth centuries and any text or other evidence is potentially a part of, and witness to, these processes. Even a single inscription can be a doorway into the politics and religion of the empire on the grandest scale and across generations.

Religion and its reception

This book is an exploration of the form(ul)ation of knowledge in a given context – a process which might well be called education. More specifically, it is about the presentation of religious knowledge in an ancient historiographical context, though, like its subjects, it is occasionally given to ‘digressions’ which either enhance or detract from the text, depending on the reader's expectations and criteria for relevance. Like its subjects, it can be plundered for individual items of information but will only make proper sense when taken as a whole, where each item is defined to a large extent by its context.

It assumes that the historians in question were highly intelligent and deliberate men who went a long way towards achieving a fundamental cohesion in their works. It also works on the premise that they built up an image of religious systems as a whole, not by describing a system ex nihilo for outsiders. They represented their model of religion to their world by offering refinements of the understanding that they assumed would be brought to their text by their readers. They presumed to know roughly what this understanding was, though their frequent and deliberate refinements indicate that they also acknowledged that the details would be more or less negotiable and could be debated. What they did not cater for was a fundamentally different matrix of understanding, such as the modern reader brings to bear when reassembling the worlds they created. Without a compatible matrix of knowledge as a context, any statement as an intentional communication is doomed.

Introduction

The discussion thus far has focussed on re-examining Livy's presentation and has argued for a more nuanced approach in reading his religious material. We can now begin analysis of his religious material in earnest. Much of it will lend itself with ease to the kinds of issues raised in the introduction. As a working principle, coherence and intelligibility is maximised, not least as a corrective to previous discussions. This is not to imply that incoherence is absent, or should be explained away; all cultural systems are notoriously contradictory in manifold respects. But the very observation of incoherence relies upon the assumption of a meaningful structure of concepts that have a relationship to one another: much of this system may not be explicit at any given moment. The following account therefore represents what is fundamentally an experimental analysis of the religious system deployed by Livy. No other authors are taken into account – which can be treated as either a strength or a weakness; no essay could attempt to incorporate all sources on these topics in more than a few authors at best, and those that did would run the risk of failing to appreciate the various contexts of the material. This analysis is therefore limited, and thereby all the more precise. It is structured along lines that are familiar to the scholar of Roman literature, though it might be said that the end result is a subversion of that structure. In that it is intended to be a reasonably full answer to many of the criticisms of typically problematic areas, it is more usefully considered a starting point than a final result in itself.

The reception of Livy

Religious material in Livy

Our reception of Livy has tended to be determined by our requirements. As a source of historical information, he has been considered disappointing when judged by modern standards and methods. Though there are undoubtedly problems recovering historical facts from our author, recent, more ‘literary’ studies have brought more favourable results that show Roman identity to have been a key factor in shaping the Ab Vrbe Condita (AVC). The following discussion belongs firmly in this latter category.

Livy is of course familiar to students of religion: without him our knowledge of Republican religion would be infinitely poorer. The historian is generally treated as a store of material that can be taken, by and large, as it comes: little discussion of Livy's specifically religious methodology is thought to be necessary. But there is also a tradition of scholarship examining Livy's ‘belief’ as an object of study in itself. For this school, religion has been a puzzling and contradictory phenomenon, and no clear consensus has been reached: the most recent – and probably the fullest – attempt to examine the material is the work of Levene in his Religion in Livy which therefore merits some attention. Levene makes central to his argument the issue that had previously confounded most attempts to understand the presentation of the religious material – namely that the author appears to contradict himself at various points on religious matters: Levene endeavours to retain this tension rather than favour one side to the detriment of the other.

Religion is a powerful theme in the three authors scrutinised here: indeed it might be said to be the backbone both of the historical record and of the transaction of Roman identity. Virtually every major battle and a great many other events are explained with reference to the gods. Historiography, in its role of explanation and characterisation, appropriated a particular role to itself in its synthesis of the competing religious knowledges: these self-appointed spokesmen for Rome's religious tradition were not directly ‘informing the public’ of a central canon, thrashed out at some mysterious policy-making thinktank. These accounts are each individual yet deliberately placed at the centre of religious authority: this was, self-evidently to the audience, a construction. The historical accounts created and negotiated a normative position and aimed at procuring enough general assent to facilitate any excursions into expert or controversial areas. Ironically then, the supposedly sceptical historians might be the closest thing we have to the voice of the ‘state religion’, at least in terms of text; not a specific formulation of a particular detail or set of circumstances (e.g. as Feeney argues for Horace's Carmen Saeculare), but a general framework of practice and interpretation.

Religion is central to the texts under scrutiny here: proper methods, pitfalls, the enactment of the gods' will – all represented over and over again for the reader's edification. Noting prodigies, their expiation, propitiating the gods, not offending the gods in one's actions (such as stripping a temple) or manner (such as boasting of one's felicitas), finetuning one's interpretation – all this is Roman religion in practice.