Reflections On the Populousness of the Roman World
David Vincent Kimel
I. The History and Significance of the Questions at Hand
Prior to the popularization of the work of T. R. Malthus (1766-1834), it was widely believed among seventeenth and eighteenth century thinkers that a sprawling population was evidence of a prosperous society governed by just institutions. Although he did not agree with this idea in its entirety, David Hume wrote in 1777 that “if every thing (sic) else be equal, it seems natural to expect, that, wherever there are most happiness and virtue, and the wisest institutions, there will also be most people.”1 This notion linking a state’s metaphorical and literal vitality inevitably informed early scholarly opinions on the size of the Roman Empire’s population. The period of the Antonine emperors in particular (96 CE-180 CE) was associated by such historians as Edward Gibbon with notions of hyperbolic grander thanks to its supposedly enlightened political leadership. Citing an opinion undoubtedly antithetical to contemporary stirrings in the American colonies, he declared in the third chapter of the first volume of The History of the Decline and Fall of the Roman Empire in 1776 that “If a man were called to fix the period in the history of the world, during which the condition of the human race was most happy and prosperous, he would, without hesitation, name that which elapsed from the death of Domitian to the accession of Commodus.”2 With his imagination fired by the holistic grandeur of antiquity, Gibbon wrote in chapter two that the Roman Empire constituted “the most numerous society that has ever been united under the same system of government.” This opinion was echoed with just as little grounds by earlier scholars like Isaak Vossius, who estimated that the city of Rome at its height housed some 14 million people “with an area twenty times greater than that of Paris and London combined.”3 No less an authority than Montesquieu wrote in 1721 that Europe was depopulated compared to the days of the Caesars, with the eighteenth century population likely representing one fiftieth of the ancient total.4
It would be comic understatement to suggest that common assumptions about the size of Rome and its empire have somewhat altered over the past three centuries. In the wake of work such as David Hume’s groundbreaking study on the populousness of antiquity and, most importantly, Julius Beloch’s Die Bevoelkerung der griechisch-roemischen Welt published in 1886, altogether smaller numbers began to be imagined for antiquity, with the entire population of Italy during the reign of Augustus (27 BCE-14 CE) amounting to perhaps no more than 6 or 7 million people, with three quarters of a million to 1 million people crowded into Rome itself.5 Hume worked with a wide body of knowledge about ancient literary sources, pointing out how the often extravagant totals mentioned by ancient authors were unrealistic and contradictory. For his part, Beloch revolutionized approaches to the issue by turning to quantitative analysis; specifically, he sought to measure changes in the size of the Roman population by studying fluctuations in the size of the city’s public grain doles, proceeding to estimate what percentage of the population those receiving the grain represented. According to modern authorities like Walter Scheidel, Beloch’s conclusion that the Roman Empire at its height contained some 35 to 80 million people definitively set the parameters for all future discussion on the subject.6 While acknowledging the existence of important contributions to the question of Rome’s populousness since 1886, scholars like John C. Caldwell believe that “much of classical demography, originally deduced from literary sources and burial inscriptions, remains essentially unchanged.”7 Whatever the validity of this claim, it is unlikely that mainstream academic opinion will ever favor the assertions of Vossius and Montesquieu again. After all, they appear at odds with well-known trends in comparative demographic history, which, for better or worse, only admit to limited gains in world population until the advent of the industrial era.
Of course, the question of whether the Roman Empire contained 35 or 80 million people seems to leave a great deal of room for meaningful debate. Unfortunately, personal bias often appears to motivate authors toward defending lower or higher estimates. For example, in an attempt to highlight the productivity and populousness of pre-Roman Gaul, C. Jullian averred that before Caesar’s invasion, the population probably stood at some 20 million people, which proceeded to double over the course of the next century thanks to “the long famous fertility of Celtic women.” Likewise, E. Lo Cascio’s rejection of Beloch’s totals and his insistence on a population of 7-14 million for Augustan Italy have been branded patriotic hogwash by the late Keith Hopkins.8 Nationalism certainly becomes a particularly thorny issue when it comes to the scale of the Roman Empire’s population compared to that of Han China. The notion that one empire was significantly more populous than the other invariably reveals bias in favor of the “progressiveness” of Roman or Chinese culture, since evidence to suggest any fundamental differences in size simply does not exist. (Admittedly, recent efforts seem aimed at building bridges and accentuating the similarities between the two imperial systems, though the thematic emphases of this approach might arguably conceal its own kind of bias shaped by the fear of stepping on professional toes.9) At other times, pride in one’s academic discipline, such as Medieval Studies, might tempt some to underrate the Roman period’s supposed luster relative to subsequent history. There is surely something deliberately revisionist in the air when it comes to Angus Maddison’s 2001 attempt to suggest that far from representing a height in Europe’s population, the number of people on the continent might have stayed the same or even slightly increased over the course of the period formally known as the Dark Ages.10 Irene Barbiera and Gianpiero Dalla-Zuanna wrote disparagingly of the very notion of the “Dark Ages,” claiming that the bleakness of the era was largely a rhetorical trope developed by Christian authors longing for the imperial order of the past.11 However, other “authorities have posited the population of Europe halving during the first six centuries of the modern era,” maintaining that the decline of Roman civilization was indeed accompanied by a fall in population.12
Fundamental questions concerning the quality of life in ancient Rome, the scale of the empire’s economy, and the ways in which urbanization transformed the provinces are all bound to debates over population size; for example, a lower population might reveal an unexpected source of economic strength, with more benefits for everyone to go around and less competition for jobs.13 Unfortunately, in many ways our state of knowledge remains woefully speculative. In the words of Scheidel: “Our ignorance of the size of ancient populations is one of the biggest obstacles to our understanding of Roman history. After generations of prolific scholarship, we still do not know how many people inhabited Roman Italy and the Mediterranean at any given time.”14 Attempts to glean the likely populations of major cities from vague references in ancient literary sources can be compared to similar attempts to understand the scale of Pre-Columbian American society by assembling a constellation of random quotations and trustingly treating them as scientific evidence. In an effort to draw attention to the ludicrousness of such methods, David Henege jokingly attempted to calculate the population of elves and orcs in Middle Earth by analyzing references from J. R. R. Tolkien’s Lord of the Rings.15
In this study, I present important sources of evidence about the size of Rome’s population over time, discussing various broad indicators of growth and then examining approaches to the question of populousness grounded in a diversity of different sources, from the analysis of bones to studies of comparative DNA profiles. The fact that debate persists to this day with an intensity belying the poverty of the available evidence is telling, though few scholars disagree with the broad parameters established by the work of Beloch with regard to the grain dole and with Harkness (1896) and McDonnel (1913) vis a vis funerary inscriptions.16 Ultimately, I will show that while there exists a general consensus that the Roman world was far removed from modern population dynamics, the methods of evaluating the data are all extremely problematic, and one’s conclusions about the size of the empire’s population often reveals more about the nature of the researcher and his or her academic interests than historical truth.
II. Harnessing Evidence on the Populousness of the Romans
During the reign of the emperor Trajan (98 CE-117) , the Roman Empire stretched 3000 miles from the Atlantic Ocean to the banks of the Euphrates and encompassed some 1,750,000 square miles, approximately half the territory of the contemporary United States. Roman civilization facilitated the spread of Hellenistic civilization around the Mediterranean basin and its hinterlands, creating a great cultural melting pot solidified by centuries of general peace.17 Even after a millennium and a half of neglect, the ruins of cities like Leptis Magna and Pompeii are impressive enough to awe millions of tourists a year; the urban landscapes of the empire at its height, before centuries of theft and collapse took their toll, must have been magnificent indeed. Surveying the rural landscape of Turkey and noting the many hulks of abandoned Roman cities, Gibbon took the ruins as evidence of the populousness and vigor of antiquity compared to the state of affairs under the Ottomans:
(The Asian provinces) of the east present the contrast of Roman magnificence with Turkish barbarism. The ruins of antiquity scattered over uncultivated fields, and ascribed, by ignorance, to the power of magic, scarcely afford a shelter to the oppressed peasant or wandering Arab. Under the reign of the Caesars, the proper Asia alone contained five hundred populous cities, enriched with all the gifts of nature, and adorned with all the refinements of art. Eleven cities of Asia had once disputed the honour of dedicating a temple to Tiberius, and their respective merits were examined by the senate. Four of them were immediately rejected as unequal to the burden; and among these was Laodicea, whose splendour is still displayed in its ruins.18
The notion that the Pax Romana was an era of unprecedented prosperity has definite implications with regard to opinions concerning the size of the population that enjoyed its fruits, even if the effects are difficult to quantify. In the second century BC, the archeological record shows great ranches dotting the Italian countryside where once there had been barren fields, suggesting demographic change. These are the so-called latifundia, cash-crop plantations manned by hundreds of thousands of imported slaves.19 The suicides of Antony and Cleopatra in 30 BCE represented the final act in a century of tragedies and brought an end to the Civil Wars. Until the early third century CE, battles were subsequently altogether more infrequent and smaller in scale. Given the destructiveness of warfare in pre-modern societies, the introduction of peace associated with the rise of Rome might conceivably have facilitated a growth in population.20 The idea that the Celts, Berbers, Teutons, and Illyrians who had once inhabited Rome’s provinces lived like noble savages in blissful harmony with nature is flatly contradicted by the fact that 90-95% of all world societies, statistically speaking, were involved in periodic warfare to some extent or another and that violence was part and parcel of everyday life for the people of most pre-modern civilizations. Indeed, the historical record attests to unending combat among indigenous peoples before the Roman occupation.21 At the same time, Roman culture celebrated fertility and encouraged early marriage among women, with the mothers of three children granted special economic privileges by the emperor Augustus.22 According to Frier, surviving Egyptian census data suggested that the vast majority of women married during the Roman period. Specifically speaking, he estimated that some 80% were wedded by the age of 20.23
Added to these trends was the introduction of innovations like “iron tools, iron knives, screw presses, rotary mills, even water mills…silver and bronze coins, money taxes, chattel slavery, writing, schools, written contracts, commercial loans, technical handbooks, large sailing ships, shared risk investment, (and) absentee landlordship,” all speaking to possibilities for enhanced productivity and the accommodation of a large population.24 Grain imports and handouts, bathhouses, aqueducts, gymnasia, sewage systems, written laws, and paved roads facilitating travel and migration might easily be added to the list. Keith Hopkins explained that by raising taxes and spending money on the defense of distant frontiers, the empire facilitated long distance trade and enhanced possibilities for social mobility.25 Authors such as Horace, Martial, Juvenal, and many others all affirmed that these economic opportunities drew significant numbers of migrants into Italy, with the city of Rome ballooning to ever larger heights, a trend confirmed by rising numbers of insula type high-rise apartments discovered in the suburb of Ostia dating to the first two centuries CE.26 In certain parts of the empire, sources of evidence even seem to suggest that Roman rule was associated with long life-spans (and presumably a large population). For example, a graveyard from the North African site of Castellum Celtianum was found to contain 1,258 individuals with an average lifespan of 60.2 years in a time period where the average life expectancy of most world societies was in the high teens or early 20s.27 Although the site is unique, graveyard inscriptions from the salubrious provinces of North Africa in general suggest life expectancies closer to 40 than 20.
While all of this seems compelling enough, how can one go about attempting to actually quantify the Roman population? Comparisons to other historical epochs, informed conjectures, and old-fashioned common sense in the face of extremely limited evidence are the rules of the day. In 2 CE a census in Han China counted 12,233,062 families, which has been used to suggest that some 60 million people lived under imperial rule. An Augustan census of 14 CE included 4,937,000 citizens. This has been interpreted to suggest rough parity with China, for “assuming that full-fledged citizens of Rome constituted less than 10 percent of the empire’s total population at that time, it is (thus) reasonable to conclude that the Roman Empire contained 50 to 60 million people in the early decades of the first century CE.”28 Adding to these numbers were streams of slaves from beyond Rome’s borders, an insidious source of population growth. Sir William Smith’s epic nineteenth century dictionary mentioned that the Roman Empire saw the system of slavery augmented “to a prodigious extent.” Quoting Book VI of Athenaeus, the author reflected upon the idea that “very many Romans possessed 10,000 and 20,000 slaves and even more.”29 Even if this total seems exaggerated, the number of slaves owned by certain aristocratic Romans was likely to be very high indeed. Pliny the Elder recorded in Book XXXIII.10 of his Natural History that 4,116 slaves were left to the heirs of a single Augustan freedman who (paradoxically) had seen his estates greatly diminished during the Civil Wars; Dio Cassius reported in Book V.1.27 of his History that Augustus allowed a man to take 40 slaves or freedmen with him into exile out of sympathy for his impending loneliness. In his paper “Believing the Ancients: Quantitative and Qualitative Dimensions of Slavery and the Slave Trade in Later Eurasia,” Timothy Taylor took the words of Athenaeus at face value when he declared that Scheidel’s estimate of slaves at 10% of the classical population was likely too low; when it came to classical Athens, after all, Athenaeus described a populace of 21,000 citizens, 10,000 resident metics, and 400,000 slaves, implying that 93% of that city’s population was enslaved.30 Even if these numbers are off, they imply a world in which it was possible to imagine sprawling numbers of unfree people toiling alongside a small core “in-group” of citizens, such as was famously the case in fifth century Sparta.
The Romans themselves took great interest in numbering their subjects for the purposes of taxation and (during the Republic at least) conscription, though most of the information drawn from these censuses and, indeed, how the surveys were even conducted in the first place is tragically lost. Nevertheless, from the first century BCE to the fourth century CE, literary sources mention the numbers reached by selected censuses. (Please refer to Table A for a representative sampling.31) Twenty-five different censuses are recorded from the third century BCE to the end of the second century BCE, ranging between some 137,000 to 395,000 people.32 These numbers clearly do not approximate the entire population of the empire; most scholars assume that they represent the number of adult male citizens, though even this is not uncontroversial. Whatever the case, the numbers rise dramatically to 910,000 in 70/69 BCE and a whopping 4,063,000 in 28 BCE when the method of taking the census itself evidently changed. 14 CE saw 4,937,000 people counted. Claudius’ census of 47 CE totaled 5,984,072, further evidence of long-term growth.
But can these numbers be trusted? Basing his work on a seminal 1971 study by P. A. Brunt, P. M. G. Harris insisted that the general trends to which the data spoke made perfect sense in light of Roman history. For example, from 465 BCE to 493 BCE the population of Rome seems to have increased by two-thirds if the census was accurate, a trend associated with an extension of citizenship rights to allied states and an increase in the cultivation of the ager publicus, or land for public use. By contrast, 218 BCE to 203 BCE saw steep losses in the wake of the Second Punic War so grave that the author compared them to demographic trends in Aztec Mexico after the arrival of the conquistadors. The Lex Julia of 90 BCE enfranchised several allied states up to the banks of the Po River, adding nearly a million people to the empire. Finally, by the time of the 47 CE census, Claudius had begun to extend Roman citizenship to the people of southern Gaul, further driving up the numbers.33 Estimates about the size of Rome’s population often toy around with this data, asserting that various census totals represent different hypothetical percentages of the total populace.
The 1886 work of Beloch on the size of the Roman grain dole served to contextualize these numbers. (Please refer to Table B for data related to the public distribution of goods in Rome.34) In 123 BCE, Gaius Gracchus instituted the practice of doling out grain to the urban masses, and in 58 BCE Cicero’s enemy Publius Clodius Pulcher made the practice permanently free of charge. Beloch put the information that 150,000 to 320,000 men were eligible to receive the various doles to good use by attempting to guess how many dependents (wives, children, slaves, etc.) these men might have had, and how many foreigners likely lived in the city alongside them. He ultimately concluded that 800,000 inhabitants for the city of Rome seemed consistent with the levels of grain imported during the age of Augustus. This number likely increased over time. Dionysius of Halicarnassæus, for example, attested in Book IV.3 that the ancient walls of Rome had nearly the same circumference as those of Athens, but that by his time, Rome’s suburbs were so extensive that it was impossible to tell where the city ended or the countryside started. Also consistent with a narrative of increasing population is Gerda de Kleijn’s work on the water supply of imperial Rome.35 The completion of the Aqua Claudia and the so-called New Anio aqueducts begun by Caligula in 38 CE and completed by Claudius in 52 CE suggest an increasing demand for fresh water, just as Claudius’s construction of a second harbor at Portus to supplement the one at Ostia suggests a growing urban market for grain and other goods. Altogether, the population of the city likely peaked at 1.2 million people, making it the largest urban center in Europe (and according to some sources, the largest in the world) until the early nineteenth century.
When it comes to the first two centuries CE, the archeological record is unanimous throughout the lands bordering the Mediterranean that the volume of goods traded dramatically increased, which might be consistent with a rising population prospering during peacetime. Sites like the colony of Cosa grew greatly in size and, like Rome, gradually acquired high-rises and suburbs.36 We know there were 430 so-called urban centers in Italy during the age of Augustus. Using this information, Elio Lo Cascio took issue with Scheidel’s statement that Italy as a whole probably contained 6 million people total (the so-called “low count”). Pointing out that if the estimates were raised for 25 major towns from 15,000 to 20,000 inhabitants and the average of the 405 minor cities from 2000 to 3000, a total of 3 million would be reached, which would mean that half of Italy’s pre-industrial population was urbanized—a number which comparative demography suggests is altogether too high.37 Thus, he estimated that the total population must have been somewhat greater than 6 million if even 35-45% of Italians lived in cities, since ancient agriculture was rudimentary and a great deal of food would have to be produced by many hands to feed the sprawling populace. One is struck by how even a slight shift in assumptions can radically affect an interpretation of the extant archeological evidence. Nevertheless, taken as a whole, a degree of agreement has been reached concerning certain issues in Roman demography. For example, despite statistically outlying sites like Castellum Celtianum, Harkness’ 1896 work with funerary inscriptions suggested a stationary population for the empire with an average life expectancy of 18 years, and McDonnell’s widely cited study in 1913 harnessed a still more extensive corpus of inscriptions to raise life expectancy in Rome “to 22 years for males and 21 for females, in the Iberian Peninsula 39 and 34 years respectively and in Africa (not including Egypt) 48 and 46.”38 In 1966, Keith Hopkins used United Nations model life tables to reach an empire-wide life expectancy of 20-30 years.39
Archeological fieldwork in Egypt has proved to be especially informative thanks to the discovery of papyrological records containing information of interest to demographers. The literary record is unfortunately erratic when it comes to contextualizing this data, which is certainly disheartening considering the centrality of this kind of evidence to conjectured numbers for Roman Italy. For example, Josephus in his Jewish War II.385 suggested that 7.5 million people lived in Egypt outside of Alexandria; Diodorus of Sicily, however, said in his Library of History I.31.6 that the number for the entire country was a paltry 3 million. Roger Bagnall and Bruce W. Frier took Diodorus’ number seriously and used it as the basis for all of their work; however, one is struck by the fact that they might have just as easily based their findings on Josephus’ number.40 The surviving archeological evidence cannot provide definitive answers and is often more tantalizing than edifying. For example, while scattered examples of birth registrations have been discovered in Egypt, they are few in number and the practice in general seems to have been optional. At the same time, we know that local administrators took detailed tax records with a large number surviving in clumps dating to the reign of Claudius (41- 54 CE), but almost all of the archives have been lost.41 Scheidel put the number of Egypt’s people during the Roman period at 4.75 million people, with 35% of the people inhabiting urban areas, though the categories of “urban” and “metropolitan” often bleed into each other.42 Nevertheless, as in his account of the population of Italy, Scheidel’s estimate might have been too low. Joseph Manning, for example, explained that during the Roman period as a whole, growth in population was reflected in gradually increasing agricultural and craft production.43 And according to some ancient sources, the city of Alexandria came to rival that of Rome in size and splendor.
Recent years have seen further refinements in the debate over the size of the Roman population. Comparative genetic analyses of individuals hailing from former imperial provinces represents a particularly exciting, nascent field. Eric Faure in 2008 turned to Roman history to explain the distribution of chemokine receptors related to the CCR5- Delta 32 allele. Homozygosity for the CCR5-Δ32 allele results in resistance to R5- tropic HIV-1. The frequency of this allele is lowest in areas corresponding to the lands of the Roman Empire. 10% of Europeans on average have the gene, but only 4% of Greeks, and almost no one in North Africa. To explain the data, Faure suggested that feline zoonoses might have spread among provincial populations as the Romans brought increasing numbers of cats to new areas with them to serve as pets and to control pests.44 Although he suggested that gene flow between colonizers and the colonized was “low and indirect,” this data suggests that the scale of Roman occupation was extensive enough to leave fundamental and permanent marks on Europe’s genetic landscape.
III. The Limitations of Existing Demographic Models of the Roman Population
While, as we have seen, eighteenth century models of the Roman population were informed by the underlying assumption that the empire represented a period of unprecedented prosperity, current estimates of the civilization’s population are caught up in a sort of self-fulfilling prophecy that it could only have been a large natural fertility regime with rampant disease similar to other pre-industrial societies. The problem is that this approach, while grounded in reasonable assumptions, denies the possibility that the Roman Empire was somehow uniquely ahead of its time, which was once assumed as a given—after all, Europe did not see such extensive urbanization and such large metropolitan centers until over a millennium and a half later. Tim G. Parkin declares that “no one today would seriously regard as accurate” Pliny’s estimate of 600,0000 total people for Seleucia, but this is an arbitrary assumption grounded in nothing but the belief that the Empire contained few such extensive cities.45 The dismissive tone of the author is particularly telling. After all, had his assumptions about the inherent trustworthiness of Pliny been different, he might have forgiven the author under the grounds that he could have been referring not only to the city of Seleucia itself, but also to its extensive surrounding hinterlands.
Nevertheless, multiple sources of evidence discussed in the previous section of this paper can safely be labeled problematic. For one thing, even if the Roman Empire saw extensive urbanization, large-scale internal migration, and the formation of suburbs, this redistribution of people geographically is not necessarily synonymous with population growth. Moses Finley seems to have been correct in his critique of Keith Hopkin’s model of Roman commerce, declaring that opportunities for exploitation could increase without a corresponding growth in productivity.46 Studies such as those by Barbiera and Dalla- Zuanna attempting to understand population size with reference to burial sites are often hampered by the paucity of the existing sources of evidence; for example, they use eleven cemeteries to represent the entirety of the period between the first century BCE and the fifth century CE, but have ten data points for the sixth to seventh centuries alone. They also systematically ignore the fact that the bulk of the Western Roman population practiced cremation during the first two centuries CE, while early Christians (who practiced burial but lived on the fringes of society and likely did not have the best diets) were probably over-represented.47 Evidence for enhanced nutrition might be grounds for believing that Italy was becoming a more salubrious and populous place. It might alternately, however, be evidence for economic collapse as lands formally devoted to cash crops were turned over to the production of fruits and cereals and formally massive urban populations broke up into smaller groups whose nutrition did not rely on grain doles. There is simply no way to know the truth, though the fact that the paper was written by medievalists rather than classicists perhaps informed the ultimate thesis. Whatever the case, a high population for Rome and its empire might be interpreted as a mixed blessing vis a vis long term growth. At I.12, Herodian described how “because of its very high population, and because it took in immigrants from all over,” mortality was highest in Rome during times of plague. And contrary to the naïve impulse that a large population implies a prosperous nation, Bruce W. Frier aptly explained that with higher population can paradoxically come lower living standards and less opportunity for economic advancement.48 Paradoxically, the lower the population of Roman Italy is estimated to be, the more urbanized and prosperous its people might seem to appear.
It is impossible to know to what degree rising birth rates and immigration were responsible for the dramatically increasing numbers seen in Table A. An increasing percentage of a stable pre-existing population might simply have been granted citizenship over time with no corresponding growth in population size. Even the epigraphic data, deemed over-analyzed by Scheidel, can prove to be deceptive. For example, epigraphic patterns differed depending on location in the empire. Old ages seem to have systematically not been recorded in Noricum, but were a popular typos in Africa.49 The number of surviving grave sites, just as in the case of the number of surviving documents from Egypt, is not enough to make accurate estimates for the state of the empire as a whole. Even the little evidence that survives is in some degree inherently unrepresentative. For example, young males between the ages of ten and fourteen were underrepresented in Egyptian papyri dealing with the census, likely because the census was taken regularly and once it became known that a boy reached the age of fourteen, he became liable for taxation; females were under-registered as a rule.50 It is not even clear what category of people were eligible for being counted in censuses of Roman citizens. Saskia Hin, for example, pointed to the possibility that Augustan counts might have included widows, children and grandchildren liberated from patria potestas, and freed slaves.51 At the same time, while there is some evidence for the growth of the Roman population in urban contexts, there is also evidence for a drop in fertility among certain subsections of the populace. A full three-quarters of the senatorial families of the early Roman Empire disappeared after a single generation. This extraordinary fact has been variously interpreted, but it was likely, at least in part, the result of deliberate birth control, delayed marriage, and even infanticide.52 While Augustus passed laws punishing bachelors and rewarding fertility among Roman wives, two children were exposed in his own family: the child of his granddaughter Julilla, and an infant whom the future emperor Claudius suspected was illegitimate.
Compounding these problems is a lack of sensitivity to just how untrustworthy the numbers mentioned by ancient literary sources can be. As we have seen, scholars such as Bagnall and Frier somberly employed numbers mentioned by Diodorus of Sicily in their estimates of the size of the Egyptian population. However, this ignores the fact that Diodorus has been called one of “the most accomplished liars of antiquity” and was condemned to hell in one of Lucian’s satires for his poor scholarly standards.53 In fact, some have even assumed that Diodorus (or a scribe) simply made a typo and meant to write 7 million rather than 3 million, since he used just that number at an earlier point in his history.54 Ultimately, even the most respected and meticulous authors of antiquity were liable to make mistakes. In Book I.2 of his History of the Peloponnesian War, Thucydides mentioned that 200 triremes were dispatched by Athens to Egypt in 460 BCE, which would mean that 40,000 men were sent to the Nile in the midst of the First Peloponnesian War when there were incredible strains on manpower. A Persian source, Ctesias, says that the number was actually 40 triremes, and he may (or may not) have been closer to the mark.55 Duncan-Jones readily accepted Cicero’s assertion that the state income of Ptolemaic Egypt was 300 million sestertii. However, we know that the entire Roman Empire’s annual income ranged between 650 and 900 million sestertii, and the notion that 35-40% came from Egypt alone seems improbable.56 To make matters worse, ancient sources in general cannot be fully trusted when it comes to any number. Scribal errors and the use of letters to represent digits resulted in maddening variations across the manuscript traditions of many ancient authors, which sometimes only survive in their present form from a single copy that may or may not have been accurately transcribed.57
Gibbon’s notion that the Roman Empire represented the most populous state in the history of the planet seems hyperbolic, yet his use of evidence was actually quite reasonable and rather similar to contemporary approaches, albeit informed by different assumptions about the possibilities for the veracity of the source material and the proportion of people who were enslaved and/or otherwise unrepresented in the data set. He wrote, for example:
We are informed that when the emperor Claudius exercised the office of censor, he took an account of six millions nine hundred and forty-five thousand Roman citizens, who, with the proportion of women and children, must have amounted to about twenty millions of souls. The multitude of subjects of an inferior rank was uncertain and fluctuating. But, after weighing with attention every circumstance which could influence the balance, it seems probable that there existed, in the time of Claudius, about twice as many provincials as there were citizens, of either sex, and of every age; and that the slaves were at least equal in number to the free inhabitants of the Roman world. The total amount of this imperfect calculation would rise to about one hundred and twenty millions of persons; a degree of population which possibly exceeds that of modern Europe.58
It is singularly discouraging that Gibbon’s use of the same source material as Walter Scheidel could result in an estimate twice the size of contemporary guesses. Be that as it may, imaginative applications of common sense to ancient data can sometimes generate compelling arguments indeed, shedding light on obscure demographic forces. For example, when considering whether or not the deliberate breeding of slaves raised the population of Roman Italy, Hume ingeniously concluded that the effect was likely minimal:
At the capital, near all great cities, in all populous, rich, industrious provinces, few cattle are bred. Provisions, lodging, attendance, labour are there dear; and men find their account better in buying the cattle, after they come to a certain age,from the remoter and cheaper countries. These are consequently the only breeding countries for cattle; and by a parity of reason, for men too, when the latter are put on the same footing with the former. To rear a child in London, till he could be serviceable, would cost much dearer, than to buy one of the same age from Scotland or Ireland; where he had been bred in a cottage, covered with rags, and fed on oatmeal or potatoes. Those who had slaves, therefore, in all the richer and more populous countries, would discourage the pregnancy of the females, and either prevent or destroy the birth.59
Hume backed up his argument by noting that individuals bred into slavery, so-called vernae, enjoyed special legal rights compared to other kinds of slaves, so there was likely not many of them. (owners preferring to maximize the possibilities for exploitation). Moreover, Cato, Varro, and Columella mention nothing about the profitability and desirability of breeding slaves. Xenophon’s Oeconomicus actually suggests that male and female slaves sleep separately. While it is impossible to tell whether or not Hume was correct about the impact of slave-breeding on the size of the population, the silence of these iconic texts on the issue is difficult to ignore. To Hume, as to most modern authorities on the subject, the eighteenth century image of an empire of 120 million people fuelled by the breeding of massive numbers of slaves seemed somehow intuitively unreasonable.
IV. Final Thoughts: “Losing the Trees for the Forest?”
Keith Hopkins warned that economists and demographers exploring the Roman past often make simplifying assumptions in an effort to see where they lead without paying attention to the complexities and ambiguities of the real world situation: “It is as though, in order to guess the weight of an elephant, you first imagine it to be a solid cube.”60 As we have seen, assumptions about the overall size of the Roman Empire have shifted dramatically over time, with the same problematic evidence interpreted through various thematic prisms forged by one’s academic interests and general beliefs about the nature of pre-industrial civilization as a whole. Be that as it may, Beloch’s work in 1886 broke real ground by harnessing reported census returns to make reasonable guesses about the size of the population, and since then, the general parameters of the debate have been set; the Roman Empire likely contained c. 60 million people, with the city of Rome boasting somewhere around a million inhabitants at its height. The size of individual provinces— even the relatively extensively documented Egypt—remains controversial. However, if the Roman Empire was not a natural fertility regime and consistently showed life expectancy above the mid twenties, it would be unique in the history of the pre-industrial world.
There is perhaps an unfortunate tendency in modern scholarship to ignore the possibilities for such uniqueness, or to make generalizing claims about the size of the ancient population without paying attention to the ways in which cultural practices might shape demographic realities. The fundamental problem is that the evidence for institutions such as birth control, infanticide, etc., is limited to passing references in literary sources, complicating the possibilities for evaluating the “modernity” or lack thereof of ancient practices, to say nothing of their demographic effects.
However, in the tradition of Hume, I believe that there still exists the possibility to make use of common sense and a strong imagination to add new information to our knowledge of otherwise extremely inaccessible states of affairs. For example, what is one to make of the ancient Roman saying “sexagenarios de ponte deicere”—“(to) hurl sixty year old men from the bridge”? Modern scholars who treat the subject almost unanimously assume that the phrase refers to a remote period in Roman history when an over-eager youthful populace attempted to monopolize voting rights, casting elderly men off the planks that Roman citizens would cross to reach voting places.61 However, at least one ancient author, Festus, admitted that the saying might have referred to ancient practices of senicide. The answer to the puzzle seems lost to time, but there are some clues to guide us. First, we know that bridges in general were invested with religious significance in early Rome—to this day, the pope, like the Roman emperors before him, is known as the Pontifex Maximus. Moreover, on the 15th of every March, a series of ancient purification ceremonies began whose origins were purportedly obscure; the Vestal Virgins would throw life-size dolls (argei) from the Sublician Bridge into the Tiber. Some have theorized that this was a proxy for former human sacrifices, though as far as I know, few have specifically connected the practice with the ancient saying.62 Human sacrifice in general was banned by Crassus and Lentulus as late as 97 BCE, and Romans in the time of Cicero and Augustus seem to have preferred to ignore the issue. However, I personally suspect that it was no coincidence that Julius Caesar was “sacrificed” on the Ides of March, the very day when the ancient ceremonies associated with senicide and the freedom of an independent youthful electorate took place.63
This is a situation in which an ancient practice which may or may not have existed was likely limited to a small circle of old men in extremely early Roman history. There certainly exists no way to measure its overall demographic effects on the size of the population (which, anyway, might have been negligible to begin with.) However, a sense of imagination with regard to the surviving ancient evidence can highlight unexpected quirks which made the structure of Roman civilization unique rather than a cookie-cutter example of a pre-industrial society. While archeologists pine for new discoveries and many scholars believe that our best hope for new knowledge of Roman demography will derive from this source, an ability to creatively reconstruct past social practices in the tradition of Beloch and Hume perhaps suggests another avenue for hopefulness with regard to our understanding of the structure of ancient society at large.
1 See David Hume, Essays Moral, Political, Literary, edited and with a Foreword, Notes, and Glossary by Eugene F. Miller, with an appendix of variant readings from the 1889 edition by T.H. Green and T.H. Grose, revised edition (Indianapolis: Liberty Fund 1987). Chapter: ESSAY XI: OF THE POPULOUSNESS OF ANCIENT NATIONS, accessed at http://oll.libertyfund.org/title/704/137546.
2 Edward Gibbon, The History of the Decline and Fall of the Roman Empire, 6 vols. (London: Printed for W. Straham; and T. Cadell, in the Strand, 1776). Chapter 3.
3 See the opening essay in Isaac Vossius, Isaaci Vossii Variarum Observationum Liber (Londini: Prostant apud Robertum Scott bibliopolam, 1685).
4 See the Baron de Montesquieu, The Persian Letters, Letter 112, 1748. Roger B. Oake points out that the 1758 edition of the letters was edited to read a tenth rather than a fiftieth. See also Roger B. Oake “Montesquieu and Hume,” Modern Language Quarterly 2 (March 1941): 25–41.]
5 In 1973, the famous classicist Moses Finley still branded Beloch’s study “The fundamental work on ancient population figures.” See M. I. Finley, The Ancient Economy, Sather Classical Lectures, (Berkeley,: University of California Press, 1973). Pp. 182.
6 Walter Scheidel and Sitta von Reden, The Ancient Economy, Edinburgh Readings on the Ancient World (Edinburgh: Edinburgh University Press, 2002). Pp. 201.
7 John C. Caldwell, “Fertility Control in the Classical World: Was There an Ancient Fertility Transition?,” Journal of Population Research 21, no. 1 (2004). Pp. 11.
8 Scheidel and Reden, The Ancient Economy. Pp. 182. See E. Lo Cascio, ‘The Size of the Roman Population,’ JRS 82 1994. Pp. 115. See also C. Jullian Histoire de la Gaule, (Paris; 1920) Vol. 5. Pp. 25-28.
9 For a cultural, political, and economic comparison between the two empires in a spirit of building bridges between subfields of history, see Walter Scheidel, Rome and China : Comparative Perspectives on Ancient World Empires, Oxford Studies in Early Empires (Oxford ; New York: Oxford University Press, 2009). Donald Kagan criticized the book to me in a conversation for its unwillingness to dwell on the implications of fundamental differences vis a vis conceptions of liberty between the two empires.
10 For the development of this argument, see Angus Maddison and Organisation for Economic Co-operation and Development. Development Centre., The World Economy : A Millennial Perspective, Development Centre Studies (Paris, France Development Centre of the Organisation for Economic Co-operation and Development,2001).
11 Irene Barbiera and Gianpiero Dalla-Zuanna, “Population Dynamics in Italy in the Middle Ages: New Insights from Archeological Findings,” pp. 369.
12 Caldwell, “Fertility Control in the Classical World: Was There an Ancient Fertility Transition?.” Pp. 2.
13 Quoted in L. de Ligt and Simon Northwood, People, Land, and Politics : Demographic Developments and the Transformation of Roman Italy 300 Bc-Ad 14, Mnemosyne Supplements, (Leiden ; Boston: Brill, 2008). Pp. 17.
15 David P. Henige, Numbers from Nowhere : The American Indian Contact Population Debate (Norman: University of Oklahoma Press, 1998). Pp. 287-280.
16 Caldwell drew my attention to their work. See Harkness, A. G. 1896. Age at marriage and death in the Roman Empire. Transactions of the American Philosophical Society 27. Pp. 35-72. See also McDonnell, W. R. 1913. On the Expectation of Life in Ancient Rome, and in the Provinces of Hispania, Lusitania, and Africa. Biometrika 9. Pp. 366-380.
17 M. I. Finley, The Ancient Economy, 2nd ed. (London: Hogarth Press, 1985). Pp. 29-30.
18 Gibbon, The History of the Decline and Fall of the Roman Empire. Chapter 2.
19 M. I. Finley, Ancient Slavery and Modern Ideology (London: Chatto & Windus, 1980). Pp. 84.
20 For a description of the destructiveness of periodic old wars and the old order, see Alan Macfarlane, The Savage Wars of Peace : England, Japan and the Malthusian Trap (Oxford, UK ; Cambridge, MA: Blackwell Publishers, 1997).
21 See Lawrence H. Keeley, War before Civilization (New York: Oxford University Press, 1996). pp. 28.
22 For the most comprehensive contemporary evaluation of the institution of Roman marriage, see Susan Treggiari, ACLS Humanities E-Book (Organization), and American Council of Learned Societies., “Roman Marriage Iusti Coniuges from the Time of Cicero to the Time of Ulpian.” (Oxford New York: Clarendon Press; Oxford University Press, 1991), http://hdl.handle.net/2027/heb.01470.
23 Bruce W. Frier. 2000. “Demography” in A. K. Bowman, P. Garnsey, and D. Rathbone (editors), The High Empire: AD 70-192. Cambridge Ancient History Volume 9. Cambridge University Press.
24 Peter Garnsey, Keith Hopkins, and C. R. Whittaker, Trade in the Ancient Economy (London: Chatto & Windus : Hogarth Press, 1983). Pp. 12.
25 For an analysis of this important element of Hopkins’ contribution, see John R. Love, Antiquity and Capitalism : Max Weber and the Sociological Foundations of Roman Civilization (London ; New York: Routledge, 1991). Pp. 215.
26 Fritz-Heiner Mutschler and Achim Mittag, Conceiving the Empire : China and Rome Compared (Oxford ; New York: Oxford University Press, 2008). Pp. 127.
27 Tim G. Parkin, Demography and Roman Society, Ancient Society and History (Baltimore, Md.: Johns Hopkins University Press, 1992). Pp. 8-9.
28 Alfred J. Andrea, James H. Overfield, The Human Record: Sources of Global History: To 1700, Houghton Mifflin Harcourt, 2000. Pp. 146.
29 See the entry under slavery in William Smith, A Dictionary of Greek and Roman Antiquities, 2nd ed. (Boston: Little, Brown, 1870).
30 Timothy Taylor, Believing the Ancients: Quantitative and Qualitative Dimensions of Slavery and the Slave Trade in Later Prehistoric Eurasia, World Archaeology, Vol. 33, No. 1, The Archaeology of Slavery, (Jun., 2001), pp. 27- 43.
31 See Table A taken from Tenney Frank, “Roman Census Statistics from 225 to 28 B. C,” Classical Philology 19, no. 4 (1924).
32 Ligt and Northwood, People, Land, and Politics : Demographic Developments and the Transformation of Roman Italy 300 Bc-Ad 14. Pp. 19.
33 P. M. G. Harris, The History of Human Populations (Westport, Conn.: Praeger, 2001). Pp. 168- 172.
34 See Table B taken from
34 Gerda de Kleijn, The Water Supply of Ancient Rome : City Area, Water, and Population, Dutch Monographs on Ancient History and Archaeology, (Amsterdam: Gieben, 2001).
36 Alan K. Bowman and Andrew Wilson, Quantifying the Roman Economy : Methods and Problems, Oxford Studies on the Roman Economy (Oxford: Oxford University Press, 2009). Pp. 142-143.
37 Ibid. Pp. 101.
38 For a summary of findings on inscriptions, see Caldwell, “Fertility Control in the Classical World: Was There an Ancient Fertility Transition?.” Pp 9-10.
39 Hopkins, Keith. 1966. On the Probable Age Structure of the Roman Population. Population Studies 20(2). Pp. 245 onward.
40 Roger S. Bagnall et al., “The Demography of Roman Egypt,” In Cambridge studies in population, economy, and society in past time 23. (Cambridge England ; New York, N.Y.: Cambridge University Press, 1994), http://hdl.handle.net/2027/heb.02277.
41 See chapter 3 of Walter Scheidel, Death on the Nile : Disease and the Demography of Roman Egypt, Mnemosyne Supplements, (Leiden ; Boston: Brill, 2001).
42 Walter Scheidel, Debating Roman Demography, Mnemosyne, Bibliotheca Classica Batava Supplementum, (Leiden Netherlands ; Boston: Brill, 2001). Pp. 141.
43 Joseph Gilbert Manning, Land and Power in Ptolemaic Egypt : The Structure of Land Tenure (Cambridge, UK ; New York: Cambridge University Press, 2003).
44 See Eric Faure, “Could FIV zoonosis responsible of the breakdown of the pathocenosis which has reduced the European CCR5-Delta32 allele frequencies?” Virol J. 2008; 5: 119. Published online 2008 October 16, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2575341/pdf/1743-422X-5-119.pdf
45 Parkin, Demography and Roman Society. Pp. 65.
46 Joseph Gilbert Manning and Ian Morris, The Ancient Economy : Evidence and Models, Social Science History (Stanford, Calif.: Stanford University Press, 2005). Pp. 212.
47 Irene Barbiera and Gianpiero Dalla-Zuanna, “Population Dynamics in Italy in the Middle Ages: New Insights from Archeological Findings,” pp. 379.
48 See chapter 4 of Scheidel, Debating Roman Demography.
49 Bowman and Wilson, Quantifying the Roman Economy : Methods and Problems.
50 Walter Scheidel, Measuring Sex, Age and Death in the Roman Empire : Explorations in Ancient Demography, Journal of Roman Archaeology Supplementary Series, (Ann Arbor, MI: Journal of Roman Archaeology, 1996). Pp. 56-57.
51 See Saskia Hin’s essay in Ligt and Northwood, People, Land, and Politics : Demographic Developments and the Transformation of Roman Italy 300 Bc-Ad 14. Pp. 187-285.
52 D. S. Potter and D. J. Mattingly, Life, Death, and Entertainment in the Roman Empire, New and expanded ed. (Ann Arbor: University of Michigan Press, 2010). Pp. 100.
53 Lloyd, quoted in Robinson, Eric W., “Thucydidean Sieges, Prosopitis, and the Hellenic Disaster in Egypt,” Classical Antiquity, Vol. 18, No. 1 (Apr., 1999). Pp. 135.
54 Parkin, Demography and Roman Society. Pp. 65.
55 Robinson, Eric W., “Thucydidean Sieges, Prosopitis, and the Hellenic Disaster in Egypt,” Classical Antiquity, Vol. 18, No. 1 (Apr., 1999), 132-135.
56 Scheidel and Reden, The Ancient Economy. Pp. 196.
57 Ligt and Northwood, People, Land, and Politics : Demographic Developments and the Transformation of Roman Italy 300 Bc-Ad 14. Pp. 20.
58 Gibbon, The History of the Decline and Fall of the Roman Empire. Chapter 2.
59 David Hume, Essays Moral, Political, Literary, edited and with a Foreword, Notes, and Glossary by Eugene F. Miller, with an appendix of variant readings from the 1889 edition by T.H. Green and T.H. Grose, revised edition (Indianapolis: Liberty Fund 1987). Chapter: ESSAY XI: OF THE POPULOUSNESS OF ANCIENT NATIONS, accessed at http://oll.libertyfund.org/title/704/137546.
60 Scheidel and Reden, The Ancient Economy. Pp. 193-194.
61 See, for example, Lily Ross Taylor, Roman Voting Assemblies from the Hannibalic War to the Dictatorship of Caesar, Jerome Lectures, 8th Ser (Ann Arbor,: University of Michigan Press, 1966). Pp. 92.
62 For a sampling of contemporary thoughts on the issue, see Tim G. Parkin, Old Age in the Roman World : A Cultural and Social History, Ancient Society and History (Baltimore: Johns Hopkins University Press, 2003). Pp. 265-268.
63 Incidentally, I believe I am the first to make this claim. It at least adds credence to the idea that Caesar might actually have been told to “Beware the Ides of March”—it was no ordinary day.
Bagnall, Roger S., Bruce W. Frier, ACLS Humanities E-Book (Organization), and American Council of Learned Societies. “The Demography of Roman Egypt.” In Cambridge studies in population, economy, and society in past time 23. Cambridge England ; New York, N.Y.: Cambridge University Press, 1994.
Bowman, Alan K., and Andrew Wilson. Quantifying the Roman Economy : Methods and Problems, Oxford Studies on the Roman Economy. Oxford: Oxford University Press, 2009.
Caldwell, John C. “Fertility Control in the Classical World: Was There an Ancient Fertility Transition?” Journal of Population Research 21, no. 1 (2004): 1-17.
Finley, M. I. The Ancient Economy, Sather Classical Lectures,. Berkeley,: University of California Press, 1973.
Finley, M. I. The Ancient Economy. 2nd ed. London: Hogarth Press, 1985.
Finley, M. I. Ancient Slavery and Modern Ideology. London: Chatto & Windus, 1980.
Frank, Tenney. “Roman Census Statistics from 225 to 28 B. C.” Classical Philology 19, 4 (1924): 329-41.
Garnsey, Peter, Keith Hopkins, and C. R. Whittaker. Trade in the Ancient Economy. London: Chatto & Windus : Hogarth Press, 1983.
Gibbon, Edward. The History of the Decline and Fall of the Roman Empire. 6 vols.London: Printed for W. Straham; and T. Cadell, in the Strand, 1776.
Harris, P. M. G. The History of Human Populations. Westport, Conn.: Praeger, 2001.
Henige, David P. Numbers from Nowhere : The American Indian Contact Population Debate. Norman: University of Oklahoma Press, 1998.
Hopkins, Keith. 1966. On the Probable Age Structure of the Roman Population, Population Studies 20(2).
Jullian, Camille. Histoire De La Gaule. 8 vols. Paris,: Hachette et cie, 1920.
Keeley, Lawrence H. War before Civilization. New York: Oxford University Press, 1996.
Kleijn, Gerda de. The Water Supply of Ancient Rome : City Area, Water, and Population, Dutch Monographs on Ancient History and Archaeology,. Amsterdam: Gieben, 2001.
Ligt, L. de, and Simon Northwood. People, Land, and Politics : Demographic Developments and the Transformation of Roman Italy 300 Bc-Ad 14, Mnemosyne Supplements,. Leiden ; Boston: Brill, 2008.
Love, John R. Antiquity and Capitalism : Max Weber and the Sociological Foundations of Roman Civilization. London ; New York: Routledge, 1991.
Macfarlane, Alan. The Savage Wars of Peace : England, Japan and the Malthusian Trap. Oxford, UK ; Cambridge, MA: Blackwell Publishers, 1997.
Maddison, Angus, and Organisation for Economic Co-operation and Development. Development Centre. The World Economy : A Millennial Perspective, Development Centre Studies. Paris, France Development Centre of the Organisation for Economic Co-operation and Development,2001.
Manning, Joseph Gilbert. Land and Power in Ptolemaic Egypt : The Structure of Land Tenure. Cambridge, UK ; New York: Cambridge University Press, 2003.
Manning, Joseph Gilbert, and Ian Morris. The Ancient Economy : Evidence and Models, Social Science History. Stanford, Calif.: Stanford University Press, 2005.
Montesquieu, Charles de Secondat. Persian Letters. The 5th ed. London,: M. Cooper, 1755.
Mutschler, Fritz-Heiner, and Achim Mittag. Conceiving the Empire : China and Rome Compared. Oxford ; New York: Oxford University Press, 2008.
Parkin, Tim G. Demography and Roman Society, Ancient Society and History. Baltimore, Md.: Johns Hopkins University Press, 1992.
Parkin, Tim G. Old Age in the Roman World : A Cultural and Social History, Ancient Society and History. Baltimore: Johns Hopkins University Press, 2003.
Potter, D. S., and D. J. Mattingly. Life, Death, and Entertainment in the Roman Empire. New and expanded ed. Ann Arbor: University of Michigan Press, 2010.
Scheidel, Walter. Death on the Nile : Disease and the Demography of Roman Egypt, Mnemosyne Supplements,. Leiden ; Boston: Brill, 2001.
Scheidel, Walter. Debating Roman Demography, Mnemosyne, Bibliotheca Classica Batava Supplementum,. Leiden Netherlands ; Boston: Brill, 2001.
Scheidel, Walter. Measuring Sex, Age and Death in the Roman Empire : Explorations in Ancient Demography, Journal of Roman Archaeology Supplementary Series,. Ann Arbor, MI: Journal of Roman Archaeology, 1996.
Scheidel, Walter. Rome and China : Comparative Perspectives on Ancient World Empires, Oxford Studies in Early Empires. Oxford ; New York: Oxford University Press, 2009.
Scheidel, Walter, and Sitta von Reden. The Ancient Economy, Edinburgh Readings on the Ancient World. Edinburgh: Edinburgh University Press, 2002.
Smith, William. A Dictionary of Greek and Roman Antiquities. 2nd ed. Boston: Little, Brown, 1870.
Taylor, Lily Ross. Roman Voting Assemblies from the Hannibalic War to the Dictatorship of Caesar, Jerome Lectures, 8th Ser. Ann Arbor,: University of Michigan Press, 1966.
Treggiari, Susan, ACLS Humanities E-Book (Organization), and American Council of Learned Societies. “Roman Marriage Iusti Coniuges from the Time of Cicero to the Time of Ulpian.” Oxford New York: Clarendon Press; Oxford University Press, 1991.
Vossius, Isaac. Isaaci Vossii Variarum Observationum Liber. Londini: Prostant apud Robertum Scott bibliopolam, 1685.