With Meier’s reconceptualization of the Roman Republic’s tendency to endure our understanding of politics and the rules by which it functioned was substantially deepened. But Meier had concentrated on the political dimension, where the senators played a special role. Although Meier had indeed thoroughly discussed the equites (‘‘knights,’’ essentially the vast majority of the wealthy who were not senators) and the plebs, he had done so to demonstrate that any fundamental reconstruction or reform of the Republic could not have originated with these classes. Indeed, according to Meier the Republic fell into the ‘‘crisis without alternative’’ precisely because the potentially powerful group of the equites could safeguard their vital interests without changing the system and the ordinary people could not really attain power despite - or because of - their partial integration into the state by means of the popular assemblies. It was partly only a natural reaction that after long years of the dominance of the political aspect as well as of research into the upper class, interest grew in the 1960s in social history, and particularly in the lower classes. But this was also favored by the general political climate in the West, where the reduction of social inequality had moved higher on the agenda.
Several works now elucidated the harsh living conditions of the Roman plebs and described the sometimes violent ways in which they responded;33 others emphasized the deep fracturing of Roman society owing to social conflicts.34 That the broad mass of the rural citizen population, which had been largely deprived of their rights, played a decisive role as soldiers in dissolving the aristocratic Republic was seen as an ironic consequence of the relative indifference of the upper class toward the interests of the poor.35 But in order to understand better the situation of the lower classes it was necessary to investigate issues such as life expectancy, family size, the division between city and hinterland, migration, the burdens of military service and taxes, and the threat of plagues and failed harvests. Karl Julius Beloch’s early interest in demographic questions had, however, initially not been taken up by others,36 and so it was an epochal innovation when Keith Hopkins in the 1960s introduced the methods of historical demography into ancient history.37 Scholarship at this time became generally somewhat more open to the theoretical stimulus of the social sciences, a change that I cannot pursue here in detail.38 However, a particular appeal of Hopkins’s approach was that quantitative methods of social analysis, which had been considered inapplicable to antiquity because of the very limited and unrepresentative nature of the sources, were now applied to Roman history.
The central problem, however, was a methodological one from the start. As everyone was aware, the usual documentary basis for demographic statistics did not exist for Rome, and even today debate continues as to whether the observations in our sources - for example, concerning population decline - reveal some aspect of actual developments or only about the perceptual patterns and obsessions of the educated classes from whom these statements originate. Statistics based on inscriptions are largely a dead end. Analyses of bones from individual graveyards do not permit as exact a determination of age as one would like, and do not yield precise dates of the time of burial; furthermore, there is always the question of whether or not they are representative. So there was some basis for Hopkins’s radical skepticism in excluding all data that were not entirely reliable, attributing no significance to consistency with data from other sources, and essentially relying on comparison with better known pre-modern demographic developments as represented in the Model Life Tables of life expectancy, which are extrapolated by computer modeling from censuses and other quantitative data from pre-industrial societies of the recent past. By this method it is possible to generate different types of demographic development and to see very clearly the consequences of slight changes in some basic parameters like fertility rates or marriage ages. However, it is not easy to prove that Roman demography should be modeled on one type of development rather than another, and the variations are not irrelevant. In the meantime less pessimistic approaches have been advocated that do attribute some validity to the ancient evidence, at least to the extent that clues may be gleaned from it as to which pre-modern type of demographic development the Roman Empire seems to resemble most closely. Now there seems to be some preference for the employment of Model West Level 3.39 Since in this approach the papyrological evidence from Roman Egypt takes a particularly important place, these simulations and models are oriented to the Imperial period and their details are therefore not central for the purposes of this volume.4
While Hopkins’s use of the Model Life Tables to formulate hypotheses about Roman demography was focused on the Imperial period, his approach always had significance for the Republic since there is no reason to suppose that the relevant parameters of demographic development had fundamentally changed. This was accepted by Peter Brunt in his monumental study Italian Manpower, published already in 1971, in which he had gathered and carefully interpreted all of the data relevant to demographic development from 225 BC to ad 14.41 In this work Brunt largely wanted to update Beloch’s work, but he was also able to make use of Hopkins’s first articles. Brunt’s book long dominated this area of research; it was the standard work to consult for information on matters such as the scale of mobilization for military service, the nature of population shifts and migrations during the Republic, what was known about the age at which Romans customarily married, and the like. However, in recent years the basis of Beloch and Brunt’s analyses, namely their calculation of the citizen population, has been thrown into question above all by Elio Lo Cascio, who roughly triples their figures for the citizen population of the Augustan period. The basis of his reconstruction is the assumption that the Augustan census-figures give the number of male citizens, as was quite traditional.4 Neville Morley has recently worked out the repercussions that such a population increase since the second century, if accepted, would have on our ideas of the developments of the middle and late Republic.43 Second-century Italy would then hardly have been marked by a drop in the rural free population but on the contrary by an increase; and this, according to Morley, makes the hunger for land and the Gracchan program of agrarian distribution much more understandable than Beloch’s and Brunt’s model, according to which sufficient land should actually have been available. But as interesting and promising as these consequences of the ‘‘high count’’ of citizen numbers seem to be, Walter Scheidel has now convincingly demonstrated that the implications of such a densely populated Roman Italy do not fit our knowledge of demographic development and, moreover, contradict some of the other evidence we have.44 So the better solution seems to be to accept that Augustus changed the meaning of census figures by including not only adult males as before, but also women and children, in accordance with the principle we know to have been followed in the provincial censuses he established for the first time.
Stimulated by this demographic research, and also by the increasingly refined findings of landscape archeology as well as by the search for a better understanding of the conflicts of the Gracchan age and their effects down to the fall of the Republic, scholars have turned increasing attention in recent years to the distribution of property and to the modes of agricultural production and thus embarked upon a closer investigation of the concrete facts of lower-class existence. Much is in flux, and I cannot trace here the wealth of suggestions, hypotheses, and rebuttals. I might single out one new approach: according to Willem Jongman, the great estates that have traditionally bulked so large in accounts of social and economic change in the Middle and Late Republic were not the dominant feature of the Italian countryside, a massive slave population was perhaps more an urban than a rural phenomenon, and the displacement of grain cultivation by the vine and olive may instead have been a marginal development.45 For land tenure an unusual body of sources is available in the writings of the Roman land surveyors,46 which had already prompted Max Weber to undertake a seminal investigation. Important studies have now been published of the forms of land division and their symbolic and social significance,47 and the rituals that attended the foundation of a colony have been made the subject of a stimulating investigation.48 Nathan Rosenstein shows in his newly published book how the disposition of farmland, family structure, and demographic development interact, and how our reconstructions of specific agricultural forms directly determine our picture of the potential for social and political conflict. Building upon the conclusion that the average age of marriage for men was quite late, he demonstrates that for average Romans the demands of peasant small-farming were more consistent with frequent and long-term military service than had previously been thought.49 He notes that the high military death rate also brought relief in the competition for ever-scarce farmland, and points out that the survival of soldiers also increased the risk of poverty for their families if they did not succeed in acquiring additional land.50
These highly controversial investigations into the size and development of the population have established an important branch of research into the history of the Republic. There is potential here to make a very considerable contribution to social history. For this purpose the most important sources are the Model Life Tables for pre-industrial societies, which alone make a quantitative approach possible; and demographic assertions without a quantitative basis remain impressionistic and of limited validity. That the application of the Model Life Tables has been accepted in general despite some criticism is also connected with a clear adjustment of goals found already in Hopkins’s work. That is, the goal is not to find one uniquely valid model with which to portray exactly the structure of the Roman population. Rather, it is to assign the Roman world to a group of such model statistics in a well-reasoned manner, not in order to calculate the Roman numbers precisely but rather to produce a probable range within which Roman circumstances fell. Above all, in this way one can prove that various ancient opinions or modern reconstructions and models are unrealistic - and that is no small thing.
World History
