Two of the earliest martyr narratives, both of which appear to be substantially authentic, place their martyrdoms in Smyrna at a time described as “Great Sabbath”:

Mart. Polycarpi:

1. 8.1: ὄντος σαββάτου μεγάλου

2. 21: σαββάτῳ μεγάλῳ (dated 23 February)

Mart. Pionii:

1. 2: ἐνισταμένου σαββάτου μεγάλου (dated 23 February)

2. 3.6: ἐγεμίσθη πᾰσα ἡ ἀγορὰ καὶ αἱ ὑπερῷαι στοαὶ Ἑλλήνων τε καὶ Ἰουδαίων καὶ γυναικѽν̇ ἐσχόλαζον γὰρ διὰ τὸ εἶναι μέγα σάββατον.

The interpretation of this festal date has long been a subject of controversy, as can be readily discerned from two recent and important discussions: W. Rordorf, “Zum Problem des Großen Sabbats im Polykarp- und Pioniosmartyrium,” in Pietas: Festschrift für Bernhard Kötting, Jahrb. f. Ant. u. Christ. Ergänzungsbd. 8 (Münster, 1980), pp. 245–9; and P. Devos, “Μέγα Σáβατον chez Saint Epiphane,” Anal. Bolland. 108 (1990), 293–306.

Stephen is traditionally designated the first martyr, πρωτομáρτυρες. But in the earliest centuries of the Church this was by no means the case. In the account of the martyrs of Lyon in AD 177, as preserved by Eusebius in what is arguably an authentic document (see Appendix iv below), the first to advance to their deaths in the persecution are simply and plausibly called πρωτομáρτυρες (HE 5.1.11). There is no suggestion that the singular of the noun might describe the first of all martyrs in the history of Christianity.

The word does not appear in the New Testament. In the detailed narrative of Acts 7 on the stoning of Stephen those who witness the event are called οἱ (μάρτυρες (7.58).

The early martyrdoms in the period down to Constantine are a conspicuously urban affair. They do not occur in the mountainous regions of Greece, or in the remote parts of central Anatolia, or in the near eastern steppe, or on the fringes of the Sahara in North Africa. For the most part, they take place in the greatest cities of the Roman world, predominantly in the eastern part of it. Apart from Justin at Rome and the group of martyrs at Lyon in France, the early martyrdoms provide a check-list of the most prosperous and important cities of the eastern Roman empire: Pergamum, Smyrna, Caesarea by the Sea, Carthage, Alexandria. In Greece, it is Thessalonica that has its martyrs, not Athens, and this is a proper reflection of the relative importance of the two places at that time. It was only in the period just before Constantine that there was a conspicuous deviation from this pattern of urban martyrdoms.

The spread of Christianity into the rank and file of the Roman army (and perhaps also the increasing importance of soldier-emperors in the third century) led to the first group of soldier-martyrs, as reflected in the martyrdom at Durostorum on the Danube.

On four luminous days in May of 1993 I had the honour and the joy of delivering the Wiles Lectures at the Queen's University in Belfast. Before an audience of broad interests and deep intelligence I touched upon the historical context of a phenomenon that had, even as I spoke, a powerful resonance in the political life of Northern Ireland. Martyrdom, I argued, first came into being in the Roman empire and was inextricably rooted in a society and culture peculiar to that world. The later transformation of the concept and the practice of martyrdom lay outside my theme (and my competence), but I know that it was never far from the minds of my listeners.

It is a tradition of the Wiles Lectures that the lecturer meet each evening with university colleagues and invited guests for discussion of the afternoon's lecture. I am immensely indebted to all who contributed to the discussions for their insight, criticism, and benevolence. The late Professor Lewis Warren led each meeting with wisdom and skill. For this and for hospitality of many kinds I remain permanently in his debt. I thank as well those distinguished friends and colleagues who came to Belfast to hear the lectures and participate in the nightly colloquies: David Braund, Averil Cameron, Werner Eck, Edmond Frézouls, Keith Hopkins, Christopher Jones, François Paschoud, David Potter, and Lellia Cracco Ruggini.

Nearly two decades ago I queried the extraordinary prescript of the great letter, preserved by Eusebius, concerning the martyrdom at Lyon in the year ad 177. The letter, clearly full of many authentic details concerning the martyrdom, begins with a prescript containing a striking formulation of the home territory of the recipients: οἱ ἐν Βιέννη καὶ Λουγδούνῳ τῆς Γαλλίς παροικοῠντες δοῠλοι Χριστοῠ τοῐς κατὰ τὴν Ἀσίαν καὶ Φρυγίαν τὴν αὐτὴν τῆς ἀπολυτρώσεως ἡμῐν πίστιν καὶ ἐλπίδα ἔχουσιν ἀδελϕοῐς. The conjunction of Asia and Phrygia seemed odd, inasmuch as Phrygia was an integral part of Asia in normal second-century usage, although a few texts of apparently later date displayed such phrasing as “Asia and Phrygia,” “Asia and Lydia,” or “Asia and Caria.”

Perhaps the most remarkable thing about Ignatius of Antioch, as we know him from his letters, is that his zeal for death at the hands of the Roman authorities is wholly untouched by the language of martyrdom. As Charles Munier recently observed, “Ignace ne connait pas encore les acceptions techniques des termes μáρτυς, μαρτύριον, pour désigner le témoignage sanglant; bien mieux, il ne connaît aucun terme réservé à cette signification.” Ignatius's career is assigned by Eusebius to the early second century (HE 3.36.2–4), and his sanguinary demise at Rome is put, not perhaps with complete accuracy, in AD 107 (Chron.). If μáρτυς had meant martyr at that time, Ignatius would undoubtedly have availed himself of the word.

Towards the end of the reign of the Roman emperor Commodus, in the last years of the decade of the 180s AD, a Roman governor in the province of Asia was conducting his normal judicial activities when a throng of excited people pushed forward to stand before his tribunal. Without provocation or prior accusation they all voluntarily declared themselves to be Christians, and by this declaration they presumably showed themselves unwilling to sacrifice to the Roman emperor — a test to which governors regularly put professing Christians. The pious mob encouraged the governor to do his duty and consign them all promptly to death. He obligingly had a few of them led away to execution; but, as the remainder clamored ever more loudly to be granted the same reward, he cried out to the petitioners in exasperation, “You wretches, if you want to die, you have cliffs to leap from and ropes to hang by.” The Roman official, who was a well-known member of a famous senatorial family at Rome, would hardly have confronted Christians for the first time on this occasion. He must have known their enthusiasm for death at the hands of the Roman administration. The philosophic emperor Marcus Aurelius had, not long before, wondered to himself in his Meditations why it was that the Christians were so unreasonable and disorderly.

In the first poem of his anthology of lyrics on the crowns of martyrdom (Peristephanon Liber), the Christian Latin poet Prudentius finds that his zeal for recounting the excruciating tortures suffered by the martyrs is severely limited by the loss of records about them. “Alas,” he exclaimed, “for the all too common forgetfulness of the voiceless past [o vetustatis silentis obsoleta oblivio]!” When we are denied the details, the story itself may be extinguished. Long ago, according to Prudentius, a blasphemous soldier took away the records of martyrs so that subsequent generations, trained in the preservation of memory by written account, should not disseminate to posterity, in sweet language, “the order, time, and manner that was handed down about the martyrdom.”

The fundamental written texts, what Prudentius called the tenaces libelli, obviously contained the raw material for those inspiring stories of martyrdoms that circulated in the Roman empire from the middle of the second century onwards and perhaps even earlier. These were texts that were adapted, expanded, altered, and imitated throughout the Roman imperial and Byzantine periods. Since the formative period of martyrdom was over by the early fourth century, when the empire became Christian, there could be no more documents of the struggles of the early Church against an intolerant and polytheist bureaucracy.

From Jean Bodin to Lewis Morgan to contemporary scholars, Rome has provided the paradigm of patriarchy in western thought. The paterfamilias, with his unlimited legal powers over members of his familia, has been interpreted as the extreme case in which “ paternal authority passed beyond the bounds of reason into an excess of domination.” Family relationships are often conceptualized as falling somewhere along a spectrum from afrectionless power at one end to loving concern at the other, and the movement along the spectrum is then historicized as social development. The Roman father, who in legend would execute his disobedient son without flinching, is taken to represent afrectionless power – the starting point from which the affectionate family gradually evolved.

My study has suggested the inadequacy of such a simple evolutionary view of family history. The Roman family was unquestionably patriarchal, in the sense that it was defined with reference to the father, who was endowed with a special authority in the household. But, I have argued, the characterization of an “excess of domination” has been the result of both a misinterpretation of the legends and, above all, an overly legalistic approach to the family. The law endowed the pater with a striking potestas encompassing extensive coercive and proprietary rights, yet a purely legal understanding of the Roman family is as incomplete and misleading as would be a solely legal understanding of the twentieth-century family. In certain circumstances, to be sure, legal rights and powers were important in determining the nature of family interactions, but in many other contexts the Romans appear to have been as oblivious to formal legal definitions of power in the family as we are today.

The relation of authority, obligation, and coercion within the Roman household constitutes the subject of this chapter and the next. Over the centuries the Roman paterfamilias has served as a paradigm of patriarchal authority and social order; patria potestas has been seen as the embodiment of arbitrary, even tyrannical, power. By “ arbitrary,” I mean here the sort of power a master exercises over his slave, a power more or less unanswerable to higher principles or formal procedures. On this view, Roman family relationships were almost wholly asymmetrical, with power in the hands of the father, and the obligation of obedience imposed on the rest of the household and underwritten by the core family value of pietas. In this chapter I will first analyze how the Romans construed pietas: was it a value designed in such a way as to encourage obedience in asymmetrical relationships between the paterfamilias and other members of the household? Next, the formal elements of patria potestas will be enumerated. Finally, consideration will be given to how those powers were exercised within the field of family obligations, the norms of the wider society, and the social and economic dynamics of the household.

The image of the Roman father endowed with nearly unlimited power over his household goes back to antiquity, especially Greek commentators. Dionysius of Halicarnassus (2.26.4) enumerated the powers that Romulus granted to fathers for life over their children: the power to imprison, to beat, to hold in the country, even to kill their sons.

As a pre-transition society, Rome undoubtedly had far higher birth and death rates than those of the industrialized world today: more than 35 births and deaths per 1,000 per year, as compared with less than 10 births and deaths per 1,000 today. The aim of this chapter is to review and evaluate the problematic evidence and arguments for mortality and fertility rates and ages at marriage in the Roman population, and thus to justify the choices for the values of the parameters used in the microsimulation.

Mortality

The mortality experience of a population can be expressed in various ways. Most directly, the mortality rate can be summarized as the number of deaths per thousand per year – a statistic of interest to the historian studying broad trends in the population as a whole. For the historian concerned with the individual and family experience, it is more useful to think in terms of mortality rates at given ages, from which average life expectancy (e) may be derived. Average life expectancy is the average number of years those of a particular age will live and is calculated by adding together the additional years of life of those of a given age and then dividing by the number of individuals in that age group. It is important to keep in mind that average life expectancy is not the typical, or modal, experience. In a population with an average life expectancy at birth of twenty-five (eo = 25), the most common age at death will be under one year and relatively few will die at age twenty-five.

Dowry was an institution of legal, economic and social dimensions that Roman fathers and mothers used to transmit property and status to their daughters. Although legitimate marriage in Roman law did not require validation by a dowry, the provision of one was a paternal duty (officium patemum) enforced by social expectation and law among the propertied classes. Like a creditor who automatically expected repayment of a debt, a husband expected payment of a dowry, “ since he would not take a wife without one (indotatam uxorem)” (Dig. 42.8.25.1, Venuleius). Since lack of a dowry could effectively prevent a woman of the propertied classes from entering marriage, Augustus legislated against fathers who tried to stop a daughter's marriage by refusing to finance a dowry. My aim in this chapter is not to repeat the recent systematic treatments of the legal rules and other evidence for dowries. Rather, I wish to examine dotal exchange within the Roman legal and demographic context as one element related to others in the devolution of familial property. Although no systematic evidence for dowries is available, and the scattered testimony points in various directions, it may be possible to generalize beyond the assertion that “almost everything about Roman dowry is ambivalent.”

As one element in the devolution of property before the death of the parents, dowries in historical societies have varied in function and size – so much so that it could be argued that “dowry” as a catch-all category is misleading.4 Since dos is a specific category in Roman law, even though dowries could differ widely in size and purpose, it does not seem sensible to drop it from the analysis.

The Romans considered the bonds of family and kinship to be biologically based but not biologically determined. It is the biological basis that opens the possibility of simulating the kinship universe with the aid of a computer. Roman authors saw the beginnings of kinship bonds and of the wider society in the biological reproduction of the married man and woman. Roman law, to be sure, offered citizens a flexibility in restructuring their kinship bonds that was remarkable by later European standards: not only were divorce and remarriage easy in the classical period, but adoption permitted change of filiation. Nevertheless, adoption was apparently not so common as to vitiate a model of the kinship universe based on biological reproduction.

Anthropologists have stressed that, within the universe of those linked by reproduction or marriage, not only do particular cultures systematically stress certain bonds over others, but individuals make choices about which relationships to maintain and cultivate out of the kinship system as culturally defined from the biological kinship universe. Generally, choice becomes more evident among more distant kin. The significance of cultural and individual choices in defining kinship roles is beyond question, but to understand the choices and the definition of kinship roles it is important to have a sense of the kinship universe that flowed from marriage at certain ages and from the biological events of birth and death. For example, Roman law suggests that paternal uncles had a special culturally defined role; the social significance of that role, however, depended in part on whether Romans at certain stages of life had a living paternal uncle to fulfill the role.

In his powerful interpretation of slavery in the American South, Eugene Genovese discusses the conception of the plantation as a “ family, white and black.” Like the domus in Rome, the planter's family encompassed his wife and children and his slaves, “and therein lay dangerous implications.” In suffering merciless beatings, the slaves “did not always fare much worse than the master's wife and children. From ancient to modern times we hear this theme. According to Roman legend, Manlius Torquatus beheaded his son, who had just returned victorious from combat, for breaking ranks.” Genovese quotes planters on the virtues of corporal punishment of children and slaves alike, concluding: “The slaveholders' vision of themselves as authoritarian fathers who presided over an extended and subservient family, white and black, grew up naturally in the process of founding plantations.” In the cultural matrix of the slave society, then, the categories of the master's slaves and of his children were assimilated, both subject to patriarchal authority enforced by violent coercion. As Genovese's reference to Manlius Torquatus shows, the stereotype of the Roman paterfamilias invites projection of such an assimilation back to Roman society, apparently confirmed by the similarities in the legal position of the slave and the filiusfamilias.

Despite the law, the Romans did not assimilate children and slaves in their reflections on the nature of authority. Cicero, following Greek philosophers, wrote that “different kinds of domination and subjection (et imperandi et serviendi) must be distinguished.” A father governs his children who follow out of readiness to obey (propter oboediendi facilitatem), but a master must “coerce and break (coercet et frangit) his slave” (Rep. 3.37).