Although the importance of documentary and archaeological evidence for the study of the empire has long been widely recognized, in certain important respects it is only within the last 30 years or so that researchers have begun to exploit its full potential.
The main reason for this development lies in the increasing influence of the social sciences. The new fields of inquiry and the new methods of analysis that have been developed in the disciplines of geography, economics, sociology, and cultural anthropology have inspired scholars of the Roman Empire to treat documentary and archaeological evidence not merely as a supplement to the literary sources, but as the basis for completely new avenues of investigation. Like the disciplines noted above, the social sciences originated in the nineteenth century; their impact on Roman history, however, has been much more gradual. Indeed, Rostovtzeff’s Social and Economic History was perhaps the first major work to show their influence, for all that his use of sociological and economic terminology was vague and problematic. For it was not merely Rostovtzeff’s interest in the new archaeological research of his day that made his work so seminal, but even more his willingness to formulate hypotheses about historical development that neither depended on the evidence of the literary sources nor were defined by the issues on which they focused.
But despite Rostovtzeff, the influence of the social sciences was initially much stronger in other areas of history. This was particularly true in France, where Marc Bloch and Lucien Febvre, in the journal Annales: economies, societes, civilisations, which they founded in 1929, championed a historiography that stressed underlying environmental, economic, and social factors rather than the details of political and military developments. This Annales school of historiography, as it is called, emphasized issues of climate, geography, technology, agricultural practice, manufacturing, and trade, and made greater use of statistical analysis and quantification than had hitherto been the norm. It was also concerned with mentalites, the pervasive structures of religious and social belief that characterize different societies. In all these respects, Annales historiography applied the concerns and techniques of the social sciences to the study of past eras. The most famous work associated with this school was Fernand Braudel’s La Me'diterrane'e et le monde mediterrane'en a l’epoque de Philippe II, first published in 1949 (Braudel 1972-3). Braudel insisted on the primacy of what he called the longue durce, the long term effects of climate, geography, and technology on human life and social organization. Since the Second World War, and particularly in the last 30 years, the approach to historiography advocated by the Annales school has been increasingly applied to the study of the Roman Empire.
It is difficult to summarize or even survey the impact of the social sciences on the study of the Roman Empire (for a stimulating discussion of some particular aspects, see Phillips 1986: 2681-97). I shall instead make some selective observations on three general issues: environmental studies, the use of social scientific methodologies, and cultural/ideological analysis. Several of the studies that I shall mention, particularly in the first two areas, are not limited to the specific period of the Roman Empire. This fact is in itself significant, since one effect of much interdisciplinary work has been to question the convention of defining significant historical periods by political events (e. g., the battle of Actium in 31 bce and the death of Constantine in 337 ce).
One of the most striking results of Annaliste influence has been a drive to expand historical analysis into the study of human interaction with the natural environment, to investigate both how the environment shapes human activities and how these in turn alter the environment. In this regard there has been a very fruitful interaction between Annaliste historiography and archaeology (see Bintliff 1991). Several recent archaeological projects have aimed to map out human interactions with the environment in fairly limited areas over long stretches of time. These projects typically rely heavily on the technique of survey archaeology, often supplemented with some limited excavation, and combined with scientific disciplines such as geomorphology, paleobotany, paleozoology, and paleoclimatology. For example, one of the most ambitious of such projects studied the Biferno valley in central Italy from Paleolithic times to the modern era. Its goals, in the words of the project director, were to examine how ‘‘the topography, climate, resources, and natural communications of the different parts of the valley offered different constraints and opportunities for settlement and land use,’’ and to map out ‘‘the complex ways in which different kinds of societies have reacted to these from early prehistory to the present day’’ (Barker et al. 1995: 308). Intensive local studies of this sort over the last few decades have allowed for a more detailed exploration of the historical interaction of human society with its physical environment than was ever before possible.
Two major works of scholarship, both very recent although completely different in character, perhaps indicate that study of the physical environment will gradually become more widespread among ancient historians. The first is Peregrine Horden and Nicholas Purcell’s massive study of the Mediterranean, The Corrupting Sea (2000). Horden and Purcell draw on a wide range of disciplines, especially environmental studies, archaeology, and anthropology, to build up a new global interpretation of Mediterranean civilization. Their overriding emphasis is on the distinctiveness of Mediterranean topography, which combines extreme fragmentation into a myriad of distinct geological and climatic zones, what the authors call microecologies, with the possibility of intensive small-scale movement by land and especially by sea, what they call connectivity. Within this framework, they argue, it is possible to achieve a better understanding not only of issues such as food production and trade, but also of religion and social structure. The appearance of such a grand synthesis is bound to stimulate further attention to and debate about the role of the environment in the study of the Roman world (see B. D. Shaw 2001).
The second work is the Barrington Atlas of the Greek and Roman World (Talbert 2000). Whereas most historical maps used in the study of the Roman Empire have tended to emphasize political over physical geography, the maps in the Barrington Atlas are based on the Operational Navigation Chart and the Tactical Pilotage chart, aeronautical series derived from satellite-generated data, which have been corrected so far as possible to show the physical conditions of antiquity. Their generous scale allows users easily to locate cities, roads, and other constructed features in relation to their physical surroundings. Naturally, the information provided on the maps is generalized and often highly selective; as a result it may even be misleading in certain respects (see, e. g., Alcock et al. 2001). But by combining up-do-date scientific mapping techniques with the results of recent geomorphological, archaeological, and epigraphic research and by presenting the whole in a readily accessible format, the Atlas makes it much easier than ever before for even the most casual student to have a greater appreciation for the physical aspect of the ancient Roman world.
As noted above, the adoption of social scientific methodologies has transformed the use of documentary and archaeological evidence. We may consider in particular three interrelated methodologies: quantitative analysis, the use of comparative data, and the formulation of interpretive models. An interest in quantification is in itself nothing new: even Gibbon was keen to compute the number of men under arms in the imperial forces (Gibbon 1994: 1: 47). But the development of mathematical statistics in the twentieth century led to a new use of quantitative data in sociology and economics, allowing researchers to measure, for example, the degree of causal connection between two variables. The application of these techniques to the study of the Roman Empire, however, runs up against a formidable obstacle: the available quantitative data are almost always woefully inadequate. Hence the importance of the second methodology, the use of comparative data, which can with due caution be used to fill at least some of the gaps. For example, specific information about crop yields in Roman antiquity is not available; such information does exist for other times and places, however, and can provide a point of departure. The same is true for comparative data about population trends, transportation, trade, and technology. Clearly, the use of such evidence involves potential dangers, and one must keep in mind possible variations in a whole range of conditions. What can in large measure provide an overarching framework for both these techniques is the formulation of interpretive models. An interpretive model may be defined as an abstraction, a hypothetical formulation that allows the researcher to step back from specific empirical data and to focus instead on the significant relationships that may underlie them. Such models provide a framework for both analysis and the use of comparative data, although they are themselves not without their pitfalls (see, e. g., the cautionary remarks of W. V. Harris 1999: 63-4). I will consider the role of these methodologies in one particular area of Roman history: demography.
Demography is a discipline that depends in large part on the availability of quantitative data, but one that can also make extremely effective use of comparative evidence and statistical models since, as one important scholar has pointed out, ‘‘in demographic history the number of probabilities is strictly finite’’ (Hopkins 1983a: xii). Although there do exist some quantitative data, they are generally hard to come by, especially for more complex demographic issues such as questions of age composition, marriage patterns, fertility, and mortality (see in general Parkin 1992, and the concise survey of Frier 1999). The most significant source of such data is the body of census returns preserved on papyri from Egypt; the returns typically include information about the sex, age, and relationships of individuals in households and survive for a period of about 250 years. Roger Bagnall and Bruce Frier, by analyzing this data in light of comparative evidence and model life tables, have been able to provide fairly reliable information about such matters as household structure, female life expectancy, patterns of male and female first marriage, and female fertility rates (Bagnall and Frier 1994: 170-3).
Investigation into such issues in other parts of the empire is inevitably hampered by the absence of adequate data sets. The evidence that does exist, however, can be used to examine more limited questions. For example, Keith Hopkins and Graham Burton applied statistical sampling techniques to the available information about Roman consuls and their family relationships in order to determine ‘‘how far Roman consuls were chosen from among the descendents of consuls or themselves had consular descendents’’ (Hopkins and Burton 1983: 127); this in turn provides a useful indication of the extent to which the imperial Senate constituted a hereditary aristocracy. Their conclusion that ‘‘during the first three centuries ce membership of the Roman Senate was to a large extent not hereditary’’ (Hopkins and Burton 1983: 194) runs counter to the impression created by the literary sources, and thus allows us to see the latter as expressive more of their authors’ ideals than of reality. Another example: Brent Shaw has applied similar techniques to the corpus of Christian inscriptions from Rome and Italy in order to chart the seasonal variations in rates of mortality and marriage (Shaw 1996, 1997). He has shown that most marriages occur in the winter months and that this pattern can be correlated with the agrarian cycle. His conclusion overturns the long-standing belief that June was the most popular month for marriages, a belief based on certain passages of Ovid’s Fasti. In this case as well, the use of sociological methodologies has not only corrected a misunderstanding, but has allowed us to see more clearly the ideological agenda of a literary source.
The last topic that I shall consider in connection with the impact of the social sciences is what may broadly be described as cultural and ideological analysis. Simply put, many current interdisciplinary approaches to the Roman Empire proceed from two of the fundamental insights of cultural anthropology and critical theory. Firstly, every element in a culture is in some way significant: the way people dress and eat, the way they structure their family, household, and social relations, the way they organize their days, their houses, and their communities, the way they conceive ofand relate to the divine. All these things are meaningful and help us to characterize the underlying nature of that culture. Secondly, these meanings were not merely passive or neutral, but in many cases had the active function of producing and reinforcing certain social structures, especially relationships of class, status, gender, and ethnicity. As a result, most, if not all, aspects of a culture are susceptible of an ideological analysis. The impact of these two insights on the study of the Roman Empire, especially in the last 20 years, has been dramatic. On the one hand, scholars have begun to examine the ideological role of various aspects of high culture, such as art and literature, that were traditionally studied on a purely formal, aesthetic basis. On the other hand, they have started to analyze elements of Roman culture that were previously the province of antiquarianism, such as clothing, housing, and dining practices, or were dismissed as meaningless superstition, such as magic, astrology, and dream interpretation. I will discuss a few examples from each of these areas.
The study of Roman art has long been somewhat peripheral to the historical study of the empire, for all that the art itself was an enormously important part of Roman cultural life. There are again disciplinary reasons for this, since the study of Roman art typically falls under the discipline of‘‘art history’’ rather than ‘‘Roman history,’’ and the concerns of art history have traditionally centered on describing and ordering stylistic and formal developments within and between particular historical periods. Inevitably, then, there was in the past relatively little overlap between this field of study and traditional Roman history, with its focus on political and military developments. In recent decades, however, the influence of the social sciences and critical theory has transformed art history just as it has Roman history, so that there is an increasing interest in analyzing Roman art as a medium for communicating and constructing significant social meanings. R. R. R. Smith provides an excellent statement of this approach in his study of honorific portrait statues in the eastern Roman Empire. In his view, these statues ‘‘display a received and recognizable statue and portrait language to make and project plausible-looking statements about selected social, cultural, and political aspirations.... The statues, their inscribed bases, their portrait heads, and their architectural settings were all parts of quite complex statements, with a vocabulary and grammar to be read. The language of these monuments was understood unconsciously by an ancient viewer brought up amongst them. For the modern viewer some reconstructive interpretation is required’’ (Smith 1998: 92). Hence, we can analyze these statues not only in terms of their style and iconography, but also as elements in a broader discourse about cultural, political, and social status.
I will note here two examples of recent work along these lines, studies that focus respectively on the beginning and end of the Roman imperial period as traditionally defined. First, Paul Zanker’s study of art in the age of Augustus brilliantly shows how the traditional concerns of art history and Roman history can in fact overlap in important ways. That Augustus transformed the Roman world politically and socially has been recognized since his own day; that there was in the same period a major stylistic change has also long been acknowledged. What Zanker did was to demonstrate a relationship between these two phenomena; he argues that Augustus’ program ‘‘required a new visual language,’’ so that there was in fact a ‘‘complex interrelationship’’ between ‘‘the establishment of monarchy, the transformation of society, and the creation of a whole new method of visual communication’’ (Zanker 1988: 3). In other words, the stylistic developments of the Augustan period were not merely coincidental with the political and social transformation, nor yet simply reflective of it, but were also instrumental in bringing it about. Second, Jas Elsner has revisited the old problem ofthe formal shift that took place in Roman art from the naturalistic style of the first two centuries ce to the abstract, hieratic style of the fifth and sixth centuries ce. His approach, however, is to focus not so much on the formal qualities of the two styles as on the process of viewing; he argues that the stylistic shift corresponds to an underlying cultural shift that provided ‘‘profoundly different frames for the ways viewers formulated their responses to images’’: ‘‘the naturalism or abstraction (that is, the style) of objects is dependent on a great many conceptual, sociological and essentially historical factors rooted in the way art is viewed at particular times’’ (Elsner 1995: 9, 13). By emphasizing the process of viewing, Elsner reminds us that what people saw in a given work of art depended in large part on the way they understood the world more generally, which was in turn necessarily bound up with a variety of social and cultural factors.
The study of literature provides an interesting parallel to the study of Roman art. Like the study of art, it has traditionally focused on issues of stylistic development and the analysis of individual works or genres, although for the reasons outlined at the beginning of the chapter the study of literature has had much closer ties to Roman history. And just as in the study of art, there is now a much greater concern with the ideological role of language and literature. Of course, there has long been interest in particular questions about particular works, often formulated in rather simplistic terms; for example, whether or not Vergil wrote the Aeneid as propaganda for Augustus. But in the last 20 or 30 years there has been increased sensitivity to the fact that all literature was produced in very precise social contexts and so inevitably reflects the circumstances of its production. Some scholars would now go further, asserting that it not only reflects social relationships but also helped construct them. Thomas Habinek has recently proposed that we regard Latin literature ‘‘as a medium through which competing sectors of Roman society sought to advance their interests over and against other sources of social and political authority;’’ in particular, he suggests that ‘‘many of the characteristics of Latin literature can be attributed to its production by and for an elite that sought to maintain and expand its dominance over other sectors of the population through reference to an authorizing past’’ (Habinek 1998: 3). This social and ideological function of literature extends even to the level of correct Latinity, i. e., the formulation of prescriptions about what constitutes ‘‘proper’’ Latin vocabulary, syntax, and style. Martin Bloomer has argued that ‘‘in approaching the construction of walls of decorum within and between literary texts, we must ask who is excluding whom, and why and by what means.’’ Rules about acceptable usage in fact provided ‘‘a vehicle for anxieties about ethnicity, social order, social status, and gender’’ (Bloomer 1997: 6). In short, Latin literature, and Greek literature as well, was not simply a matter of aesthetics and style, but was part of the very social fabric; if we want to understand that society, we must understand the role of literature within it.
We may consider one example in a little more detail. As Bloomer notes, anxieties about gender were among the issues with which stylistic prescriptions were concerned. In the last ten years there has been considerable work on the relationship between rhetorical practice and masculinity. For years scholars tended to dismiss oratory in the imperial period as trivial, artificial, and insignificant. Nevertheless, rhetoric played a central role in education and elite culture during the imperial period; especially in the Greek-speaking parts of the empire, but elsewhere as well, virtuoso speakers attracted great crowds and wielded great influence. To understand this phenomenon adequately, we must reassess what we consider significant and adopt new tools that will allow for new analyses. As Maud Gleason has pointed out, ‘‘one reason that these [rhetorical] performances were so riveting was that the encounter between orator and audience was in many cases the anvil upon which the selfpresentation of ambitious upper-class men was forged’’ (Gleason 1995: xx). We may put the accent here upon the word ‘‘men’’: oratory was an almost exclusively male activity, and one of the most effective charges one could bring against a rival was that he spoke like a woman. This was no accident: ‘‘because rhetorical skill was considered a definitive test of masculine excellence, issues of rhetorical style and self-presentation easily became gendered’’ (Gleason 1995: 160). Far from being trivial and unimportant, then, rhetoric in the imperial period turns out to have been a crucial forum for the construction of masculinity and the struggle for social and political power (see Richlin 1997c for a useful survey of the issues).
As I have noted, it is not only ‘‘high culture’’ that is currently providing the subjects of cultural and ideological analysis. I will end my consideration of this topic with two examples, one from the realm of everyday material culture, the other from the wide range of ancient disciplines that were until recently dismissed as mere superstition. Domestic architecture has long been an area of interest in the study of the Roman Empire, although less so than the monumental architecture of public buildings and temples. By and large, however, most work has been pragmatic and typological, concerned with identifying the functions of the various spaces within the house and with classifying the different kinds of structures. Since the mid-1980s, however, as the co-editor of one recent collection of papers has put it, ‘‘the study [has been] moving from questions of function to reveal new depths of understanding that stress the rlole of the house as an element of a society’s social matrix, charged with the cultural ideology of its inhabitants’’ (Laurence 1997: 7). This more recent work again stresses that the organization of space in domestic architecture was not merely a passive reflection of social structures. ‘‘What we see in Roman domestic space, including the artifacts, is not evidence that can simply be ‘read off,’ but evidence for how Roman society reinforced the categories of male: female, free: unfree, married: unmarried, or adult: child, as well as a series of status hierarchies from elite to dispossessed’’; in other words, ‘‘[domestic] space would appear to have played a riole in the constitution of gender, slavery, and the transitional stages in the Roman life cycle’’ (Laurence 1997: 14).
Turning from material to intellectual culture, I will take as my final example the ancient practice of physiognomics, ‘‘the discipline that seeks to detect from individuals’ exterior features their character, disposition, or destiny’’ (T. Barton 1994: 95). Traditionally, historians have paid little attention to physiognomics or to other ‘‘pseudo-sciences’’ such as astrology, alchemy, and dream interpretation; if they noted them at all, it was usually as evidence for the decline of rationality and the rise of superstition. More recent scholars have demonstrated that if we approach such material from a broadly anthropological perspective, we can derive from it considerable insights into the way that contemporaries viewed the world. Physiognomics employed as its chief principles of classification the dichotomies of male/female, Greek/barbarian, and human/animal, ideas that were deeply embedded in ancient culture; at the same time, it had the status of a technical discipline, like medicine or astrology, which it achieved through the systematic elaboration of analytical tools (see in general T. Barton 1994: 95-131). For example, Gleason has shown how the treatise on physiognomics by the celebrated second-century ce orator Polemo can serve ‘‘as a source for the ‘body language’ of his cultural milieu, particularly for the coercive way images of male deviance functioned in the semiotics of gender. Because gender categories were invoked as ordering principles for physiognomic data, the treatises of Polemo and his predecessors offer a unique source of insight into the way sex and gender categories could be used to sort human differences into readily comprehensible hierarchies and opposition’’ (Gleason 1995: xiii). Moreover, Polemo’s mastery of physiognomics gave him a claim to authority that he could use against rivals in the struggle for imperial favor and political influence, notably against the eunuch Favorinus, whom Polemo attacked for effeminacy. Regardless of whether or not we regard it as superstitious, then, the study of ancient physiognomics and its use in society can provide us with valuable insight into ancient cultural attitudes.
World History
