Christian Number and Its Implications
Keith Hopkins
Journal of Early Christian Studies 6.2 (1998) 185-226

1. Introduction

This paper is an experiment in both method and substance. Substantively, I want to show that, in all probability, there were very few Christians in the Roman world, at least until the end of the second century. I then explore the implications of small number, both absolutely, and as a proportion of the empire's total population. 1

One tentative but radical conclusion is that Christianity was for a century after Jesus' death the intellectual property at any one time of scarcely a few dozen, perhaps rising to two hundred, literate adult males, dispersed throughout the Mediterranean basin. A complementary conclusion (of course, well known in principle, but not often explored for its implications) is that by far the greatest growth in Christian numbers took place in two distinct phases: first, during the third century, when Christians and their leaders were the victims of empire-wide and centrally organized persecutions; and then in the fourth century, after the conversion of Constantine and the alliance of the church with the Roman state under successive emperors. The tiny size of the early church, and the scale and speed of its later growth each had important implications for Christianity's character and organization.

My methods are frankly speculative and exploratory. For the moment, [End Page 185] I am interested more in competing probabilities, and in their logical implications, than in established or establishable facts. That may not be as problematic as it at first appears. Facts require interpretation. Only the naive still believe that facts or "evidence" are the only, or even the most important ingredients of history. What matters at least as much is who is writing, or reading the history, with what prejudices or questions in mind, and how those questions can best be answered. Facts and evidence provide not the framework, but the decoration to those answers. 2

One of my main objectives in this paper is to show how the same "facts," differently perceived, generate competing, but complementary understandings. For example, leading Christians were highly conscious of their sect's rapid growth, and understandably proud of their "large numbers." But many Romans, both leaders and ordinary folk, long remained ignorant of and unworried by Christians, probably because of their "objectively" small numbers and relative social insignificance. Such differential perceptions often occur, then and now. Perhaps these discrepancies were all the more pervasive in a huge and culturally complex empire, with very slow communications. So, the Roman or religious historian has the delicate job of understanding and analyzing these networks of complementary but conflicting meanings--and at the same time, the exciting task of finding, inventing, or borrowing best methods for constructing critical paths through or round our patchy knowledge of what inevitably remains an alien society.

My first task is to calculate the size and growth in the number of Christians during the first four centuries c.e. But before I do that, a word of caution. The term Christian is itself more a persuasive than an objective category. By this, I mean that ancient Christian writers may often have counted as "Christian" a number of people who would not have thought of themselves as Christian, or who would not have taken Christianity as their primary self-identifier. As I imagine it, ambiguity of religious identity was particularly pervasive in a polytheistic society, because polytheists were accustomed to seek the help of strange gods occasionally, or in a crisis, or on a wave of fashion. Or put another way, [End Page 186] it was only in a limited number of cases or contexts in ancient society that religious affinity was a critical indicator of cultural identity. But monotheistic Christians, whether out of hope, or the delusion of enthusiasm, chose gratefully to perceive Jewish or pagan interest as indicative of a commitment, which Christians idealized as exclusive. It is this exclusivism, idealized or practiced, which marks Christianity off from most other religious groups in the ancient world

So ancient Christian leaders (and modern historians) may have chosen to consider as Christian a whole range of ambiguous cases, such as occasional visitors to meetings, pious Jewish god-fearers who also attended synagogue, or ambivalent hypocrites who continued to participate in pagan sacrifices and saw nothing particularly wrong in the combination of paganism and Christianity, or rich patrons, whose help early Christian communities wanted, and whose membership they claimed. In my view then, the term Christian in the early church is a persuasive, hopeful and often porous category, used optimistically to describe volunteers in a volatile and widely dispersed, though very successful, set of small cult-groups. 3 And of course, as is now commonly agreed, there were always in the early church a fairly large number of different Christianities, gnostic, docetist, heretical; Epiphanius lists 80, Augustine 88, Philastrius of Brescia more than a hundred and fifty varieties of heretic, some of them claiming to be, and thinking of themselves as the true Christians. 4 Now that I have made this point about the porosity and fluidity of Christianity at its periphery, and the diversity of its core, in the rest of this paper, I shall, for the sake of argument, treat the category "Christian" as broadly unproblematic.

2. The Limitations of Induction

And now to number. The conventional method is heavily inductive. Scholars string together snippets of testimony from surviving sources. This has been done with exemplary skill and intelligence by Adolph von Harnack in successive editions of Die Mission und Ausbreitung des [End Page 187] Christentums. 5 The basic difficulty here is that ancient writers, whether pagan, Jewish or Christian, did not think statistically, and confused cool observation with hope, despair and polemic. As a result, to put it bluntly, most ancient observations about Christian numbers, whether by Christian or pagan authors, should be taken as sentimental opinions or metaphors, excellently expressive of attitudes, but not providing accurate information about numbers.

There would be no profit in going through all the same testimony in detail and seriatim again. But even at the risk of going over well-worn ground, let me illustrate the difficulties of interpretation, and my preferred path, by briefly running through five well-known examples. First, St. Paul (Rom 1.8), writing before 60: "your faith is proclaimed in the whole world." Secondly, the Acts of the Apostles, written towards the end of the first century, recounts a speech to Paul in Jerusalem by James the brother of Jesus: "you see, brother, how many tens of thousands of the Jews have believed" in Christ (21.20). The RSV translation perceives and gets over the difficulty of exaggeration here, by translating the Greek muriades (i.e., tens of thousands) by thousands. It is widely accepted that we should not take such statements about the extent and number of early Christians literally. 6

Next, the famous exchange of letters in 112 between the Roman emperor Trajan and a provincial governor Pliny, who consulted him about what to do with Christians in northern Asia Minor (Pontus). This is the oldest surviving account by a pagan writer about the practices of early Christians and an official Roman reaction to them. 7 It is, outside the New Testament, the most frequently cited authentication of early Christian success and persecution in their struggle with pagans. The Roman governor, then just in the second year of his governorship, asked the emperor whether all Christians were to be executed, irrespective of [End Page 188] age, except of course for the Roman citizens, who [like St. Paul] were sent for trial to Rome. If those discovered to be Christian foreswore their faith, should they be pardoned? Pliny himself had devised successive tests for those who claimed not to be, or to be no longer, Christian. They were required to pray to the gods, to burn incense, pour a libation of wine and supplicate a statue of the emperor, specially brought by Pliny into court, along with other statues of gods, and to curse Christ.

Pliny clearly indicated that merely being a Christian was in itself sufficient grounds for execution, though the obstinacy with which some Christians clung to their perverse superstition (superstitionem pravam, immodicam) afforded additional justification. 8 But reports by some repentant apostates and confessions wrung by torture from two slave-women revealed no criminal activities (such as infanticide or incest), only regular prayer meeetings and simple meals eaten together.

According to Pliny, the publicity surrounding the cases which he had already tried stimulated further accusations, and in particular, an anonymous accuser's list of alleged Christians. Pliny was uneasy about the implications of further action; so he wrote his letter to the emperor, finishing with a polite suggestion of a way out. Actually, since these are highly edited letters, Pliny may have changed his ending in the light of Trajan's reply. Pliny wrote:

"many of all ages and ranks, and of both sexes, have been or will be summoned on a capital charge. The infection of this superstition has spread not only to the towns but also to the villages and countryside. But it does seem possible to stop it and put matters right. At any rate it is absolutely certain that temples previously deserted have begun to be frequented again. Sacred rites long neglected are being revived, and fodder for victims is once again being sold. Previously buyers were very scarce. So I conclude that a multitude of men could be reformed, if opportunity were given them for repentance." (Letters 10.96) [End Page 189]
The emperor replied briefly that he would not make a general rule about procedure; Christians should not be sought out, anonymous accusations should not be admitted, those who said and proved that they were not Christian by worshipping the gods were to be set free, and those who admitted that they were Christians should be executed. Trajan may have been thinking that anonymous denunciations were what marred the reign of his tyrannical predecessor, Domitian. Trajan's reign was to be more civil. So Rome's central political concerns influenced how even peripheral Christians were treated. But later Christian writers waxed indignant that merely being a Christian was sufficient grounds for execution, whereas real criminals were punished only after they had been proved guilty of crimes committed. 9 They had a good point in equity, but the emperor was being practical.

I read Trajan's letter as recommending an almost benign neglect: don't get too worked up, don't look for trouble, ignore it if you can; confront it if you have to; it's not a serious problem. A Christian apologist would probably interpret Pliny's letter quite differently. Here we have a high-level pagan administrator, disinterestedly reporting, that even in this insignificant corner of northern Asia Minor, Christianity had already succeeded on such a scale that it had been emptying pagan temples, and was widespread in towns, villages, countryside. It was already well-launched on its voyage to eventual success.

This interpretation is possible, but I think suspect. The sequence--many Christians, everywhere, can be cured, I've taken effective action, once deserted temples now filled, long neglected rites now restored--seems disproportionate to the care with which Pliny claimed to have proceeded at the initial trials (more care, less throughput), and the subsequent single anonymous set of accusations described in the first part of Pliny's letter; pagan rites neglected seems more a literary cliche than precise reporting; Paul, according to the notoriously unreliable Acts (19.23ff.), had exactly the same impact in the large city of Ephesus in the mid-fifties. If the temples were deserted (and in a polytheistic culture, temples have, and claim fluctuating fortunes), it was probably not because of Christianity, nor were they recently frequented just because Pliny's show trials had made new Christians lose their faith. In short, I suspect (but it is a matter of judgment) that Pliny's Christians were numbered in dozens rather than in hundreds. And even if his account is more accurate than I think, the situation was not typical. Pagan temples [End Page 190] elsewhere in the Roman empire flourished, or fluctuated in their popularity, for the next two centuries. In my view, Pliny's account is either inaccurate and/or describing something atypical.

Finally, three brief quotations from somewhat later Christian writers, Justin, Tertullian and Origen--I cite them to illustrate an important point of method. Since some writers lie consciously, others unconsciously mislead, some are factually correct and others are misinformed, the criteria of usefulness, acceptance or rejection cannot be the source itself, but must be the nature of the problem at issue, and the critical intelligence and relevant knowledge, in the light of which modern historians understand and interpret the sources. 10 History should not be, pace the practice or presenting-style of many colleagues, an amalgam of sources. Or perhaps rather, it depends what you want, a pre-packed meal from a factory (Listenwissenschaft), or a crafted confection from a chef. The ingredients are partly the same, the results significantly different.

Justin, in the middle of the second century wrote that "more Christians were ex-pagans than ex-Jews" (1 apol. 53), and I think (for reasons discussed below) that during his life-time this had probably come to be true, though he cannot have had enough information to know so accurately. Tertullian in the beginning of the third century wrote of Christians: "In spite of our huge numbers, almost a majority in every city, we conduct our lives in silence and modesty" (ad Scapulam 2). I doubt if either claim can have been true; and I doubt if anyone ever accused Tertullian of modesty. Origen, in the middle of the third century, wrote: "It is obvious that in the beginning Christians were small in number" (Cels. 3.10). But even a hundred passages of this quality do not allow us to trace the pattern of Christianity's growth with any confidence.

Harnack made the best possible use of such impressionistic sources. He was very reluctant to plumb for a single overall estimate of the number of Christians in the Roman empire as a whole. He thought that at the beginning of the fourth century, on the eve of the Constantinian revolution, the density of Christianity varied so much between different provinces, as to make an overall estimate useless. In Asia Minor, Harnack reckoned that almost half the population was Christian, while the proportion of Christians, for example, in France or Germany, was insubstantial or negligible. But then in a footnote, he surrendered and [End Page 191] declared that between 250 and 312, the Christian population probably increased from 7-10% of the empire's total population. 11 But any such estimate, however well informed, can inevitably be only that, on a guess.

3. Seduction by Probability

Other scholars have not been so cautious as Harnack, but have generally more or less followed his lead. Their general opinions seem to hover around a gross estimate that in 300 about 10% of the total population of the Roman empire was Christian. 12 With Harnack's qualification about variation in mind, let's tentatively and without any commitment as to its truth, take this overall estimate (that in 300, 10% of the population of the Roman empire, i.e., roughly 6 million people were Christian) as a benchmark, and see where it leads us. We can call it arguing by parametric probability, that is, by setting an arbitrary boundary against which to test other conclusions. 13 It is as though we set about estimating the weight of an elephant, by first imagining it to be a solid cube.

We have an end point. Now we need a beginning. It is obvious that Christianity began small. And Origen says so (Cels. 3.10)! Let us make an arbitrary estimate that in 40 about 1000 people were Christians 14 --though of course at this stage of Christian evolution, it is probable that they would have envisaged themselves as Jews, who also believed in the divinity of Jesus. Actually, not a lot hangs on the exact numbers either at the beginning or the end, as will become clear when we consider Figure 1. Our primary purpose overall in this article is to think through the implications of Christian growth, not to measure it precisely (that is impossible), nor even to explain it. 15

Figure 1 sets out a constant growth line implied by simple intrapolation between our starting number, 1000 Christians in 40 and our end number, [End Page 192] six million Christians in 300. I have plotted the growth in Christian numbers on a semi-log scale, because that allows us to envisage huge growth from 1000 to 6 million at a glance. 16 But to avoid misunderstanding, let me stress that my initial acceptance of these estimates is only a heuristic device. Initial acceptance implies no final commitment to the estimates' truth. To help matters along, I have also set out the implications of this consistent growth-line, by reading across the graph to specify the Christian numbers implied, at successive intervals between 50 and 350. 17 [End Page 193]

Of course, in reality, Christian membership probably fluctuated. It probably grew faster in some periods, while in others, for example, during persecutions, it even lost numbers. 18 In reality, growth was probably not consistent. We can easily imagine three competing probabilities:

a) perhaps in the beginning growth was faster, and then slower later, (i.e., above the first part of the line in Figure 1), or
b) perhaps it was slower at the beginning and even faster later (below the first part of the line in Figure 1); or

c) perhaps growth fluctuated at different periods (above and below the line in Figure 1). Drawing a single path of consistent growth is merely an intellectual economy in the face of competing probabilities, and in the absence of reliable data.

My general procedure here is obviously experimental. Instead of being inductive, moving from the evidence to a conclusion, I start with a parametric pattern, which is like a limiting case, against which the fragments of evidence can be tested, or around which they can be fitted. I then wonder what the implications of this parametric pattern are for understanding early Christianity. I hope you will be persuaded that this experimental and unashamedly speculative method is a useful supplement to, though of course not a replacement for, common inductive practices. And it will not have escaped you, that I am behaving rather like an early Christian in pagan society, trying to upset fellow scholars, by non-conformity.

But what is the use of so speculative a line, so arbitrarily drawn? What is its epistemological status? These questions are completely reasonable. My answer is that the straight line in Figure 1 is like a set of goal posts in a game of football; arbitrarily placed, but good to measure the game against. So let's play. Five gambits deserve attention: 1) absolute numbers and proportions over time, 2) community numbers and size, 3) distribution by sex and age, 4) literacy, and 5) comparison with Jews. Let us deal with each in turn. [End Page 194]

4. Absolute Numbers, Proportions and Persecutions

According to Figure 1, in 100, there were only about 7000 or so Christians, equal to barely 0.01% of the empire's population (roughly say 60 million). And in 200, there were only just over 200,000 Christians, barely 0.35% of the total population. 19 Let me stress once again, that these are not truth statements; they are crude probabilities attached to very rough orders of magnitude. They are numerical metaphors, good for thinking about Christians with.

Such estimates imply that, practically speaking, for the whole of this period, Christians were statistically insignificant. Of course, an objector might say, numbers by themselves do not necessarily equate with importance. Perhaps not, but the number of members in a religious movement is one measure of its importance; or rather it is one factor in the discrepancy between self-importance and importance as perceived by others. Even if we accommodate all Christians in 200 in the urban population of the central and eastern Mediterranean (a very strong and probably incorrect assumption), they still constituted only about 1/30 of the probable urban and metropolitan population. 20

The statistical insignificance of Christians, in relation to the rest of the empire's population, allows us to complement and correct the perspective of surviving Christian writers. Christians themselves could properly see that their religion was expanding successfully and very fast. And they sometimes, as we have seen, made exaggerated and self-inflating claims [End Page 195] to that effect. 21 But their absolute numbers long remained small. The same facts, differently perceived, generated variant accounts. From an official, upper-class Roman point of view, Christians did not matter, except as occasional individual or local nuisances, or as scapegoats, sacrificed to placate unruly crowds. 22 For example, Herodian's political history of the Roman empire, written in the early third century and covering the period from 180 to 238, does not mention Christians at all. From a Roman government point of view, it was not worthwhile persecuting Christians systematically. And from a Jewish perspective, as we shall see in a moment, Christians were only a minor annoyance.

But what of Christian stories about being persecuted, repeatedly and from the earliest days, by Romans, Jews and pagans, everywhere? 23 As I see it, the image of persistent persecution which Christians manufactured for themselves was more a mode of self-representation, or a tactic of self-unification than an objective description of reality. I am not saying that persecutions did not happen. Sure they did, occasionally and sporadically. And the fear of persecution probably sat like a huge cloud over Christian prayer meetings. It may even have kept many Christians from openly professing their faith. But persecutions were also useful. Fear of them pulled Christians together, sorted the sheep from the goats, decreased the risk of insincere hangers-on, and helped enthuse the survivors that being a Christian was really worthwhile. Being persecuted was collective proof of Christian radicality, and an instrument of togetherness. Besides, martyrdom was a special, Christian type of [End Page 196] heroism. Mostly, you didn't actually have to die for your faith, though you could parade your willingness--if the need arose. But you had to admire those who, like Christ, were willing to, or had died, for their faith. 24

So the traditional question: "Why were the Christians persecuted?" with all its implications of unjust repression and eventual triumph, should be re-phrased: "Why were the Christians persecuted so little and so late?" Our answer should recognize that for most of the first three centuries c.e., Christians were protected from persistent persecution, both by the Roman government's failure to perceive that Christianity mattered, and by its punctilious legalism, which prohibited anonymous denunciation through the courts. At a formal level, Roman legalism protected Christianity against large-scale persecution, for well over a century. Informally, in unofficial assaults and mass disturbances, Christians were persecuted, but, as I have said, only occasionally and sporadically. So too were Jews. 25

In these unofficial attacks, it was, I suspect, pagan perception of Christians' behavior as idiosyncratic (their refusal to attend traditional public festivals, their private meetings, their rigid morality and secret gestures) more than their beliefs, which provoked repression. 26 In a publicly committed, polytheistic society, Christians seemed to those who noticed them, a new-fangled and odd-ball group of monotheists. Besides, Christianity could expand so fast, only by winning adherents from old-established practices/gods, and by drawing attention to how very different Christians were from everybody else. 27 Small wonder if this [End Page 197] combination of ostentatious difference and successful proselytism provoked occasional outbursts of hostility.

In the first two centuries after Jesus' death, Christians needed Roman persecutors, or at least stories about Roman persecutors, rather more than Romans saw the need to persecute Christians. Christianity survived and prospered, partly because of its intrinsic virtues, but partly also because Roman persecutions allowed Christians to nurture a sense of danger and victimization, without there ever having been a real danger of collective exstirpation. Christianity was also often protected by Roman officials' insistence on a legalism which effectively shielded Christians against arbitrary prosecutions. And that protectivism itself persisted, because the Roman government long failed to realize that it needed to protect itself against religious subversion as much as, or more than, against barbarian invasions. The religious frontier was largely undefended, because well-organized attacks along it were unexpected.

But it is only when we play this game of numbers and proportions, that we see most clearly that the third century was the critical period of Christian growth. According to the figures tentatively projected in Figure 1, Christian numbers grew in the third century from about 200,000 to over six million. Or put another way, it was only in the third century that Christianity gained the prominence that made it worthwhile persecuting on an empire-wide scale. But by the time the Roman government finally began to realize that Christianity posed a significant threat, and started systematic persecution of Christian leaders and their property (in 250-51 under Decius, in 257-60 under Valerian, after 303 under Diocletian), Christianity was too embedded to be stamped out easily. And it was particularly in this period of persecutions, in spite of temporary losses, that Christianity grew fastest in absolute terms. In other words, in terms of number, persecution was good for Christianity.

5. Communities: Number, Size and Dispersion

First a word of caution, "community," like the term "Christian," is a persuasive and porous category. In modern histories of the early church, community is often used as a category of expansion and idealism. For example, when we have a text, it is understandably tempting to assume [End Page 198] that the author and his immediate audience constituted a "community." Hence the commonly touted concept of Pauline communities, Johannine communities, Gnostic communities; each text is assumed to have had a matching set of the faithful, who formed solidary communities, and these communities putatively used particular texts as their foundation or charter myths.

In fact, we have very little information about how early Christian followers organized themselves, or how these so-called communities used early Christian writings. We can argue quite plausibly that successive changes in reporting Jesus stories in the gospel texts (e.g., from Mark to Matthew/Luke to John) reflected the new and varying needs/interests of successive communities. But plausibility does not equal truth. All we have are the texts. The invention of communities is a defensible, but abusable, tactic of inflating the text into social history.

But there is more to it than that--early Christian communities are often imaged in modern pious thought, and in much scholarly literature, as models for modern believers. In the beginning, the myth seems to go, early Christians faithfully followed the prescriptions of Jesus and the apostles; the earliest Christian communities were close-knit, pious, mutually supportive and devoted; in short, the earliest Christians were "true Christians." And, of course, early Christian writers themselves idealized the community/ies (koinonia, ekklesia) of Christians. The concept community plays a crucial role in the self-representations of early Christian collectivities.

Needless to say, practice diverged from the ideal, even if ideals of community played a significant role in influencing practice. Paul's letters to the Corinthians, for example, amply indicate the internal tensions which affected, and divided groups of early Christians. 28 Inevitably, some early communities were riven by internal differences, social and doctrinal, and partly so, exactly because they contained fervent idealists. Some individuals thought that they had already been saved, so that they were free from ethical strictures. Others differed in their practice, commitment, and teachings. Some teachers even were greedy and exploitative. 29 [End Page 199] In sum, the concept community is used to disguise these internal divisions and shifting boundaries, and to project the legitimacy and effectiveness of Christianity's exclusive claims over its members, as though all early Christians must have been full members of a community of Christians.

But the concept still has its uses. Let us proceed by trying to estimate how many Christian communities there were. The normal procedure is of course inductive. Harnack listed as the location of a Christian community any place mentioned in early Christian texts as having had Christians. This procedure yields estimates of about fifty Christian communities in 100, and about one hundred Christian communities in 180. But this inductive procedure is suspect. Such listings are liable to be seriously incomplete, as Harnack himself fully realized. 30 Surviving sources are only a small fraction of what was once written.

Once again we can play with probabilities in a scissor-argument. As a heuristic device, without commitment to its truth, let us assume that these fifty Christian communities wrote/received on average two letters per year during the period 50-150. That is surely a low level of inter-community correspondence; less and there was little hope of securing inter-community coherence; more, then my argument holds a fortiori. But if the average inter-community correspondence was only two letters per year, then in this period, ten thousand letters were written, of which barely fifty survive. I do this calculation, exempli causa, merely to illustrate how hazardous conventional inductive procedures are, when scholars so carefully reconstruct church history only from surviving sources. Or put another way, those who think, as I do, that the earliest Christian communities, corresponded about their religion quite frequently, i.e., more than twice a year on average, must also recognize the appalling unrepresentativeness of their sources, and the limitations of induction.

My own guess is that in 100 and 180 respectively, there were significantly more than the 50/100 Christian communities listed by Harnack. I have two principal reasons for increasing his numbers. First, I see no reason in principle why Christian success was limited to those towns mentioned in the scarce surviving sources. Secondly, early Christian groups (through lack of resources and fear of persecution) typically [End Page 200] met in private houses. 31 So in larger towns, there were probably several distinct Christian gatherings, by which I mean groups of Christians who regularly worshipped together, but who may or may not have thought of themselves as linked with all other local or regional Christian groups.

I prefer to think of these early Christian nodules as "house cult-groups," rather than as communities. The term captures the image of enthusiasm, radicality and fear of persecution which perhaps characterized some early Christian gatherings. Ideally, of course, these house cult-groups may have been loosely coordinated, by cooperation, or hierarchically under a priest or bishop, into a community. However, I suspect that in the conditions of early Christianity, close coordination of dispersed house cult-groups would have been difficult to achieve. The different house cult-groups within each town were more likely to reflect Christian diversity than homogeneity. Some Jewish evidence, though not strictly comparable, illustrates the dispersion of the faithful among groups inside towns. In Sepphoris and Caesarea, each of them middle-size Palestinian towns, there were seventeen and nineteen synagogues respectively. 32 A principle is easily deducible: the larger the number of Christians within any town and the larger the town, the greater the probable number of house cult-groups.

How big were these communities or house cult-groups? We do not know. So, once again, I think the most sensible procedure is to play probabilities with a scissor argument. Three preliminary considerations seem important. First, we should take into account the diversity of primitive Christianity, its incapacity to control fragmentation, and the probability that there were several separate house cult-groups in larger towns. Secondly, the larger the community in each town, the more separate house cult-groups there probably were, since at least up to the end of the second century, Christians usually met in private houses and not in dedicated, stand-apart religious buildings. Thirdly, above a certain size (perhaps a few dozen), the larger the house cult-group, the less possible it was for all members to meet together regularly in a private house. Larger size involved a diminution of attendance or commitment. [End Page 201]

If we follow Harnack, then in 100, there were about fifty Christian communities; each Christian community therefore (according to the numbers set out in Figure 1), had a membership on average of 140 people (7000/50 = 140). 33 But if we follow the arguments outlined above, there were significantly more than fifty communities and/or house cult-groups. I suspect that even by 100, there were probably more than one hundred Christian house cult-groups dispersed over the eastern Mediterranean basin, with an average size of less than seventy people. This reconstruction surely fits better with the idea of early Christian radical commitment, and the probable size of houses used by a non-elite sect (see below).

Let us move ahead in time. By 180, according to Harnack, there were a hundred or so Christian communities recorded in surviving sources. 34 As before, it seems reasonable to think, because of the accidents of loss and survival in the sources, that this is an underestimate; and if only because of intermittent persecutions, meetings were still held in houses or house-churches, so that there were many more house cult-groups than communities. And of course, by this time there was more heaping in the density of Christian membership. In the huge cities of Rome and Alexandria, and in Antioch and Carthage, each with a population of above 100,000, Christian communities were probably substantial. Each metropolitan church (considered as a single collective or community) probably had several (e.g., 5-10) thousand members, enough to support a hierarchy of professional and dependent clergy, and a visible program of support for the poor. 35 But in many other towns, Christian communities, and their associated house cult-groups, must have remained still quite small. The house cult-group, even towards the end of the second century, was still the norm.

We could, as before, simply and arbitrarily, double Harnack's estimate, and say that there were Christian communities (and many more house cult-groups) in say 200 towns, with an average membership of 500 people ( Figure 1: 100,000/200 = 500). But according to this [End Page 202] reconstruction, the vast majority of the two thousand odd cities of the Roman empire, 1800 out of 2000, had no Christian community at all. 36 If the historical reconstructor has to choose between, on the one hand, relative concentration and larger average community size, and, on the other hand, dispersed smallness, with a handful of exceptionally large metropolitan communities, I myself favor the second choice. As I see it, Christianity towards the end of the second century was more pervasive; i.e., it had more small cells in more towns, say 200-400 of the two thousand towns in the Roman empire. This dispersion was a significant factor in the character of early Christianity, both because it considerably increased the difficulties of controlling diversity, but at the same time stimulated attempts among Christian leaders to control it.

Christianity was still probably concentrated in towns in the central and eastern Mediterranean basin, although there were some Christian communities in southern Gaul. And by this period, Christianity had begun to attract some, though very limited numbers from among influential provincial supporters and contributors, including knights and town-councillors. It now had some well-educated members and sponsors (but see section 7). Its liability to sporadic persecution, its general shortage of funds and the recurrent need to keep discreetly quiet about its activities, kept its normal cell size still within the bounds of house meetings. It seems no accident therefore that the earliest dated church building to survive comes from the mid-third century, and that very few ostensibly Christian, burial inscriptions date from the third century or earlier. 37 Christianity in the early third century still had the aroma of a once secret society. In the third and fourth centuries, as Christianity expanded, Christians came more out into the open, built large churches, but inevitably many of them became actually, though not ideally, more like other Romans. 38 [End Page 203]

6. Age, Sex and the Role of Women

According to modern historical demographers, ancient populations were usually made up, roughly speaking, of 30% adult males, 30% adult females, and 40% children of both sexes under age seventeen. 39 Mortality was particularly high among infants and children under five, but by modern standards continued to be very high in adult populations. For example, roughly speaking, half of those surviving to the age of fifteen, died by the age of fifty. Sickness and death, and presumably the fear of death, were pervasive. Hence, crudely speaking, the significance and appeal of immortality.

These basic figures are fundamental for understanding the structure and growth of early Christian communities and house cult-groups. So, for example, if by 100, there were one hundred Christian communities, then the average community consisted of seventy people ( Figure 1: 7000/100 = 70) with perhaps twenty adult males, twenty adult females (or twenty families), and thirty children. Of course, early Christian house cult-groups were probably more numerous, and correspondingly smaller (perhaps averaging a dozen or so families?), depending as they did on the sizes of houses, owned by Christians and available for meetings.

But some ancient critics of Christianity and modern scholars have argued that women were particularly prone to conversion to Christianity; and it is clear from the earliest Christian writings, that women played an important role in primitive Christian house cult-groups. 40 Of course, it is arguable that women, marginalized in a male-dominated Roman [End Page 204] society, were more likely to join a marginal religion, such as Christianity, as a covert form of rebellion. But to my eyes, the homology (marginal women, marginal religion) seems more rhetorical than descriptive. And ancient pagan criticisms that Christianity was particularly attractive to women and slaves were a literary cliche, expressing a depreciatory attitude towards women and Christianity more than cool observation.

Modern evidence on conversion to religious cults also suggests that young adults (sometimes of both sexes, sometimes of females primarily, with males as secondary converts through the female converts) are prime customers for conversion, through personal social contacts. It seems likely that the pattern of religious recruitment to Christianity in the Roman empire was similar, if only because young adults could and sometimes did feel they wanted to break away from what they perceived as repressive familial norms. So in a rapidly growing cult, there may be a tendency to overrecruit young adults (and arguably more women than men). 41

But a religion growing as fast as Christianity is supposed to have done (according to Figure 1, 3.4% compound increase per year), needed both men and women. Demographically, the new religion can be understood as being like a colony, which receives lots of young immigrants. It benefits from the fresh converts' higher (age-specific) fertility, compared with the general population, and providing that the converts' children themselves continue as Christians, this age imbalance among Christians may account for some (though it cannot account for all) the growth in Christian numbers. 42 But the greater the degree that the religion depends [End Page 205] on children of Christians as recruits (and how else could a cult grow so rapidly?), the smaller the probability of persistent sexual imbalance. Or put another way, the larger the number of Christians, the more likely that their demographic and social composition reflects that of the larger population.

Once we take all the considerations which we have discussed together (sex and age composition, dispersion, variety of belief and practice, fission, the fear of persecution, the need for secrecy, the prevalence of house cult-groups, and the availability of houses for meetings), we can plot a plausible path of Christian evolution. In 100, there were perhaps about one hundred Christian communities, dispersed in towns, mostly in the eastern and central Mediterranean basin; and many of these communities were further split into house cult-groups. On average, each community had seventy members, and many of these were children. House cult-groups were, by definition, even smaller, with an average size of a dozen or so families. By 200, Christian numbers had grown to over 200,000, spread in several hundred (say 200-400) towns out of the two thousand-odd towns in the Roman empire. So the average size of each community was in the range of 500-1000. But some metropolitan communities were very large (several thousand strong) and hierarchically organized. Even there and elsewhere, house cult-groups were still the dominant norm.

What are the implications of the small average size of early Christian house cult-groups and communities? First, in small groups, it is easier to enforce discipline, to foster internal collusion about the benefits of belief, to give mutual reassurance, and to diminish the role of free-riders, i.e., those who undermine collective commitment, by seeking the benefits without paying the costs of membership. In other words, small groups can more easily maintain a collusive sense of the superiority of their own vision, and of the benefits of their own beliefs and lifestyle. Secondly, the relative importance of women in the workings of the primitive church, albeit disputed, may have been a function of the small numbers in each cult-group, as well as of differential recruitment.

But, per contra, it is extremely difficult for dispersed and prohibited house cult-groups and communities to maintain and enforce common beliefs and common liturgical practices, across space and time in pre-industrial [End Page 206] conditions of communications. 43 The frequent claims that scattered Christian communities constituted a single church was not a description of reality in the first two centuries c.e., but a blatant yet forceful denial of reality. What was amazing was the persistence and power of the ideal in the face of its unachievability, even in the fourth century. On a local level, it is also unlikely that twenty households in a typical community, let alone a dozen households in a house cult-group, could maintain even one full-time, non-earning priest. Perhaps a group of forty households could, especially if they had a wealthy patron. But for most Christian communities of this size, a hierarchy of bishop and lesser clergy seems completely inappropriate.

7. Literacy and Stratification

The concepts literacy/illiteracy cover a broad range of techniques (from inability to read or write, barely reading, or writing slowly and with difficulty, artisanal/instrumental reading or writing of a limited range of words, reading and writing fluently, to reading/writing poetry or theology), and correspondingly different levels of competence and understanding. William Harris, in his ground-breaking and synoptic survey of ancient literacy, cautiously estimated that ancient literacy rates after about 100 b.c.e. in the Roman world were on average no more than 10-20% among males (much less for females). The general literacy rate in the Roman empire as a whole was kept down by the gap between various native languages (Egyptian, Aramaic, Punic, etc.) and the administrative and upper-culture language of the Roman conquerors, Greek and Latin. Urban literacy rates were in all probability significantly higher than rural rates; and there was considerable regional variation (the eastern Mediterranean was more literate than the western Mediterranean). Most literacy was at the basic, slow and functional end of the literacy range. 44 Fluent, sophisticated literacy was concentrated in, but was not the exclusive privilege of, the ruling strata.

A brief analysis of Roman stratification might be helpful here. The Roman empire was a preponderantly agricultural society, with 80% or [End Page 207] so of the population engaged in farming, and 15% of the population living in towns. 45 The stratification pyramid was very steeply pitched, i.e., there was a huge gap between a small, powerful and rich elite and the mass of rural and urban poor. For example, a middling senator at the end of the first century c.e. had an income sufficient to support two thousand families at subsistence level. 46 In between the elite and the mass, there was a sub-elite (inevitably a shadowy, but still a useful concept) of unknown size, which comprised middling land-owners, merchants, professionals, such as lawyers, doctors, architects, professors of rhetoric and philosophy, middling and lesser administrators, army officers, scribes, school-teachers, and eventually Christian ideologues. These sub-elites were probably particularly concentrated in the metropolitan centers (Rome, Alexandria, Antioch, Carthage), in the larger cities (such as Ephesus, Corinth or Milan) and in merchant ports (Puteoli, Ostia, Cadiz) and the university town of Athens.

The steepness of the stratification pyramid, and the relatively small size of the Roman middle class, meant that people in intermediate positions could both be despised by their superiors, and appear privileged to those beneath them. It is also worth stressing that sophisticated literacy correlated significantly with wealth and high social status, but high status, literacy and wealth did not completely coincide. There were some slaves and ex-slaves, for example, who were low in status, but who were literary sophisticates, just as there were rich land-owners, who were, or were thought to be, cultural boors. It is sometimes argued that Christianity particularly appealed to people with high status inconsistency; it may be correct, and particularly important for the first phase of Christian expansion, but cannot account for the rate of expansion in the empire as a whole. 47

Now for proportions and numbers. As usual in Roman history, little is known for sure. But the ruling elite of senators, knights and town-councillors (decuriones) can be estimated at just over 1% of the adult [End Page 208] population, comprising some 210,000 adult males. 48 There is no particular advantage in estimating the size of the sub-elite, since its bottom boundaries are necessarily fuzzy. But I speculate that it constituted say another 2% of the total population, of whom at most half (another 200,000 adult males and far fewer females) possessed a sophisticated and fluent literacy. This relatively low percentage of literary sophisticates, compared with the modern industrial world, reflects the level of Roman social evolution (the percentage of literates at any level in the Mediterranean basin as a whole had been near zero a thousand years earlier), and the relative absence from Roman society of a middle class. 49 That said, the proportion of sophisticated literates may seem low, at <2% of adult males, but it is also, I think, a generous estimate, if they constituted between a fifth and a tenth of all literates at whatever level (and if literates constituted 10-20% of the male population). By this tentative reckoning, there were about 400,000 literary sophisticates (of different levels) in the Roman empire. 50

Let's now apply these general, albeit hypothetical, literacy rates to Christians. The basic problem is that we know very little about the social standing of early Christians. But we can follow several clues. It seems generally agreed that Christianity did not initially attract converts from among the ruling strata of senators, knights and town-councillors, or not in significant numbers, at least until the third century. Complementarily, the self-presenting profile of primitive Christianity is repeatedly anti-rich (Luke 6.24: "Woe to you that are rich"), anti-ruling powers (e.g., Revelation 17, in which Rome is portrayed as "Babylon, the great mother of harlots and of earth's abominations"), and artisinal. 51 Jesus himself is represented as the son of a carpenter, a simple man at home in [End Page 209] the villages of Galilee, Paul is proud of earning his living as a tent-maker, the apostles are drawn from a set of illiterate fishermen and tax-collectors. Pagan critics of Christianity accused them of avoiding the educated (a charge which the third-century Origen denied strenuously), and of recruiting particularly among tradesmen, illiterates, women and children. Or put briefly, in this view, primitive Christianity was aimed at the poor, and was led by the underprivileged. 52 It was, and was seen as, a religion of opposition.

These arguments have both strengths and weaknesses. To be sure, as Christianity grew, it had to recruit from among the poor; and Christian writers themselves acknowledged that the bulk of the faithful were illiterate. 53 How could it have been otherwise, if the sect was to grow so fast? But two counter-arguments also seem compelling. First, the texts of the New Testament itself, the New Testament apocrypha and early Apostolic Fathers must have been written by members of that small stratum, within the top 2% of Roman society, who could write Greek fluently. The New Testament writings are of course not part of high classical culture; they do not match the careful court writings of essayists like Seneca, historians like Tacitus, or rhetoricians such as Dio Chrysostom. The gospels are written in ostensibly, one might even say ostentatiously, simpler, instrumental prose; but Matthew and John, at least, are consciously artful, while Paul is idiosyncratically inventive.

Complementarily, the rhetoric of simplicity and the appeal to the foolish and poor was just that, a rhetorical play. It made the best of Jesus' humble background in the urban world of Hellenized culture, in which the gospel message was sold. But why was the message so successful, how could it remain virtually unchanged in its primary focus, as Christianity went socially up-market? 54 I wonder if the answer lies partly in the steepness of the social pyramid, and in the tiny size of its middle class. Roman society demanded an uncomfortable mixture of [End Page 210] pervasive deference to superiors and openly aggressive brutishness to inferiors, not just slaves. It was a world of deference and condescension, of curt commands and pervasive threats. It was in this world that nearly everyone, even a middling senator with an income which could support thousands, could imagine himself to be poor. Poverty is best seen as a subjective, not an objective category.

Two subsidiary points need to be dealt with briefly. It might be argued that early Christians were disproportionately literate, partly because sacred texts were central to Christian religious practice, and partly because they inherited a commitment to education and reading from Jewish tradition and practice. But the centrality of sacred texts to liturgical practice is no proof of widespread or disproportionate literacy; believers can participate by listening, as well as by reading. The development of the post of lector (i.e., reader) in the early Christian church indeed suggests that most believers could not read, and had the text read to them. Besides, the Greek translation of the Hebrew Bible, which was the exemplary Christian sacred text well into the second century (or even of selected passages from it) would have been too expensive for most people to afford. And as to the New Testament, it is doubtful if many/most Christian communities had a full set even of the core texts (Gospels, Acts, Letters) before the second half of the second century. 55

As to the Jewish tradition of widespread literacy, I suspect assessment of it is often apologetic and idealistic. Prescription was conflated with practice. Ideally, of course, Jewish fathers had a duty to teach their sons to read. Rabbinical sources emphasize how many schools there had been in the old days, in first century Palestine: for example, 480 schools in Jerusalem and 500 in the undistinguished town of Beitar, each with 500 students. Even if we grant that Jews in the first century were exceptionally educated, compared with pagans, and that this tradition had some initial effect on primitive Christianity (after all, it too became a religion [End Page 211] of the book), the characteristic was not central to Christian self-identity. Early Christians did not establish their own specifically Christian elementary or secondary schools. 56 Therefore, it seems reasonable to conclude that Christians were, roughly speaking, no more literate, or only marginally more literate than the sub-populations from which they were recruited.

In sum, let us suppose, generously, that 20% of Christian adult males were literate at some level or other, and that 2% of Christian adult males were sophisticated, fluent literates. Female literacy was, I assume, very significantly lower, even from a statistical point of view, negligible. The estimate for sophisticated literacy is especially generous, if our argument is granted that almost no Christians, in the first two centuries c.e., were recruited from the ruling elite of senators, knights and town-councillors (though obviously some came from the sub-elites). The consequences of these proportions can be analyzed for Christians as a whole and for typical communities and house cult-groups, at different periods.

The implications of these literacy rates are quite startling. In 100, there were, according to the numbers estimated in Figure 1, about 7000 Christians, of whom about 30% = 2100 were adult males. Of these, say 20% = 420 could read and write at varying levels of literacy. But only 2%, that is 42 adult Christian males were fluent and skilled literates. Of course, the reasoning is too speculative to be trusted in detail; the number 42 is here a symbol for a small number of unknown size. But even if we double or treble it, in order to flatter the social composition, or literary skills of primitive Christianity, and add in some female skilled literates, we can still see that intellectual Christianity, that is, the part of Christianity which is preserved and transmitted in the sacred texts, was composed, explained and developed, by a tiny group of specialists, very thinly spread across the eastern and central Mediterranean basin. 57

If we split these 7000 Christians of 100 among 100 or so communities (and more house cult-groups), each on average with seventy members, [End Page 212] the implications are striking. Each community had, on average, twenty adult male members, of whom two were literate at some level. But many or most Christian communities (and a fortiori even more house cult-groups) simply did not have among them a single sophisticated reader or writer. After all, sophisticated literate Christians were likely to concentrate in the bigger towns.

By 178, according to the numbers posited in Figure 1, there were about 100,000 Christians, of whom 30,000 were adult males, split among say 200 or more town communities, and significantly more house cult-groups. By this time, the total number of sophisticated literate adult males who were Christian had burgeoned to six hundred. And by the end of the second century, it was, by these calculations, well over one thousand. Indeed, we can see in the surviving literature that Christian writers were now trying to assimilate their writings to classical upper-class pagan culture. 58 And there were enough Christian sophisticated literates overall, even with bunching in larger towns, for us to imagine that each community had one sophisticated literate leader. I think the imaging of Christian growth proposed here has some implications for the evolution of the episcopacy. Only towards the end of the second century, was it possible to find an educated leader for each Christian community.

8. Christians and Jews

Before we desert number, let us take a brief look at the Jews. Modern scholars, from hopelessly inadequate data, customarily guess that Jews in the early first century c.e. constituted about 7-8% of the population of the Roman empire. According to these guesses, which may be inflated, there were about 4.2-4.8 million Jews in the Roman empire in the mid-first century c.e. 59 The great majority of these Jews lived outside Palestine, because the carrying capacity of Palestine in ancient economic conditions (and by no means all the inhabitants there were Jews) was about a million people. 60 I myself would be happier with a much lower [End Page 213] estimate of 3 million Jews than with the higher estimate, but with either there is a high probable margin of error. However, numerical precision is not important here; for the moment, we are concerned only with very rough orders of comparative magnitude.

Jews outside Palestine but in the Roman empire, like Christians initially, were concentrated in towns in the eastern and central Mediterranean basin. For all the differences between Jews and Christians, Jews constituted the most obvious target customers for evangelical Christians, particularly after the destruction of the temple, and three disastrously unsuccessful rebellions against Rome (66-74, 117-18, 132-35). By then, many Jews must have been disenchanted, disaffected and despondent, ready to receive alternative messages, or even to desert their Judaism. 61 Some Jews must have been tempted, as the original followers of Jesus were, to join a radical renewal movement. After all, Jews knew half the Christian story, some expected or hoped for a messiah, and believed in an interventionist God; they largely shared Christian ethics, and thought that religious piety involved religious control over private life.

If I had been a hungry, wandering Christian beggar-missionary in search of success and food in the first hundred years after Jesus' death, I like Paul, perhaps even at the risk of a beating, would have made for a synagogue or house of a pious god-fearer, in preference to the market-square or the temple of Jupiter. 62 In sum, it seems reasonable to suppose that Jewish-Christians, who awkwardly straddled both Judaism and [End Page 214] Christianity, to the eventual indignation of both, probably for a significant period constituted the central, numerical core of Christians.

Three arguments of unequal weight support this claim. First, modern studies of cult-group conversions in North America show that conversion flows principally along lines of social networks. 63 Relatives and friends are primary targets as converts. Few conversions are made cold, for example, on doorsteps or by telephone. Of course, the social conditions of ancient and modern cultures are different. The New Testament and early Christian writings dwell on mass, or exemplary conversions, after miracles and healings. 64 But for me these sound like stories told to bolster the faith of the faithful, not descriptions of reality. Since the first Christians were Jews, then ethnic Jews and their associates at synagogues, the god-fearers, were the most probable clients for early Christian missionaries in towns throughout the eastern Mediterranean and beyond.

Secondly, Justin (1 apol. 53) wrote in the mid-second century that by then the number of ex-gentile Christians outnumbered ex-Jewish Christians. I am not concerned here with the statistical element in Justin's formulation, but with the historical process to which he alludes. In all probability, he did not know the proportion of ex-Jews and ex-gentiles in dispersed Christian groups, but did think that ex-Jews formed a substantial portion of Christians, even in the mid-second century. And to me, that seems highly probable. It makes sense.

Finally, I cite, exempli causa, the preoccupations of the gospels with things Jewish, the great body of pseudepigraphic writings of Jewish origin preserved by Christians, and the early Christian ethical writings like the Didache and the Epistle of Barnabas, which illustrate the continuity and overlap between Judaism and early Christianity. In this view, Christianity was to a large extent ethical Judaism, without circumcision and detailed rules of observance, plus a belief in Jesus as Messiah. It was a religion which in its early form was more likely to appeal to Jews than pagans. Indeed, Christian preoccupation with the wickedness of the Jews, from Pharisees to High Priest, and with [End Page 215] establishing their moral inferiority illustrates the urgency of Christian leaders' needs to differentiate themselves from their prime rivals.

Back, once again, to parametric (im)probabilities. My general argument here is that Christianity should have appealed particularly to Jews rather than to pagans. But according to the figures which I have cited, if only 3.3%, or one in thirty Jews (i.e., according to my low estimate, 3,000,000/30 = 100,000) embraced Christianity before 175 (and some could have done that as Judaeo-Christians without necessarily thinking that they were deserting Judaism), then ex-Jews and their descendants constituted all Christians existing in 175 (see Figure 1). Let me stress immediately that I am not claiming as a fact that 3% of Jews did convert to Christianity, or that all Christians in 175 were ex-Jews or their descendants. To me, each of these two estimates appears far too high.

What I am arguing instead is that these exemplary speculations can be useful as exploratory devices. They illustrate the boundaries of probability, and the reasonable derivation of radically divergent interpretations. For example, whatever the associative affinities between Judaism and Christianity, and whatever the sympathetic appeal of monotheism to both, and however distressed Jews were after 70 at the apparent failure of God's special relationship with Israel, we can now easily say that most Jews stayed Jewish, or at least they did not embrace Christianity. So Jews and their leaders could have sensibly considered Christians as only a minor irritation. To be sure, in some Jewish prayers, heretics (including by implication Christians) were routinely cursed every day; but it is by no means clear (although mentioned by Justin) that such curses were universally practiced by all Jewish groups in the second century. 65 Much more striking is the absence of explicit mention of Christians in the mass of Rabbinical writings. Or put another way, most Jews did not become Christians, and most Jews before 300 did not obviously care about Christianity. But complementarily, in the early period, I suspect until about 150, most Christians were ex-Jews or their descendants, and that is one reason why Christians fixated on the Judaeo-Christian boundary as a major problem. Or put crisply, Jews mattered much more for Christians until the fourth century, than Christians did for Jews. [End Page 216]

9. On The Social Production of Religious Ideology

At this point I want to change tack, and investigate the social production of religious ideology. My argument is that the number of Christians and the number and size of Christian cult-groups or communities, materially influenced the style of Christian ideology. By ideology, I mean here a system of ideas which seeks to justify the power and authority of a set of ethical prescriptions and metaphysical explanations, and also, of course, to justify the power and authority of a particular set of intepreters of these ideas. Let me proceed by crudely contrasting Judaism and Christianity.

Christianity was different from all religions of the Roman world. Like Judaism, it was (or claimed to be) monotheistic. Like Judaism, it was exclusivist, in the sense that its leaders claimed that believers in the one true god could not, or should not, pay homage to any other god. Unlike Judaism after the destruction of the Temple, Christianity was dogmatic and hierarchical; dogmatic, in the sense that Christian leaders from early on claimed that their own interpretation of Christian faith was the only true interpretation of the faith, and hierarchical in that leaders claimed legitimacy for the authority of their interpretation from their office as priests or bishops. "Obey your bishop," Ignatius of Antioch ordered (allegedly in the early second century), "so that God may heed you." 66

Admittedly, individual Jewish leaders claimed that their own individual interpretation of the law was right, and that other interpretations were wrong. But systemically, at some unknown date, Jewish rabbis seem to have come to the conclusion, however reluctantly, that they were bound to disagree, and that disagreement was endemic. Truth for them (as for Roman jurists) came to lie, or was represented as, a balance of competing opinions. Now, of course, this systemic property of sceptical balance is a characteristic of the system. Each individual rabbi in his own group could be, and probably was, as dogmatic as he dared be. But each lacked power over a large, pluralistic, dispersed and ethnically embedded set of followers. The balanced incapacity to enforce a single interpretative view, and a broad acceptance of that incapacity became a characteristic of Judaism, considered as a system, not a characteristic necessarily of each member of the system. 67 [End Page 217]

Christianity, by contrast, never accepted tolerance of diverse belief as an ideal, though of course Christians too as individuals were often inclined or forced to accept variety in practice. And it was this very intolerance as a defining characteristic of Christianity which eventually made it such a useful, if expensive, tool of state control. Christian ideologues, from Paul onwards, repeatedly attempted to lay down the law. Each claimed that his own interpretation of Christian belief was right, and that any one who disagreed was wrong, and should be excommunicated. 68

In the beginning, leaders of the primitive church had little (or insufficient) power to enforce their views. But the very idea that correct belief identified the true Christian and that incorrect belief pushed the believer who wanted to be a Christian beyond the pale became entrenched as a core defining characteristic of early Christianity. By the end of the second century, leaders tried to enhance their authority by claiming that the catholic church had held constant and unified beliefs since apostolic times. There was a direct line of legitimacy stretching from God to Jesus to the apostles and from them to bishops of the "united" orthodox church. Christians invented, or gave unprecedented force to the idea of orthodoxy and heresy. And as soon as the Church gained extra power from its alliance with the state in the fourth century, Christian leaders persecuted those Christians whom they considered deviant (and the boundaries shifted unpredictably) more assiduously than pagan Romans had usually persecuted Christians.

The centrality of correct dogma, as a defining characteristic of Christian praxis, was a religious innovation. It arose, I think, from the circumstances in which Christianity evolved. Two factors seem important: first, number and dispersion, and second, the continuously rapid rate of growth. Let us deal with each factor in turn. Members of the Christian church were spread in small groups all over the Mediterranean basin. Numbers, as we have seen, are necessarily speculative; but it seems reasonable to imagine that in the first century or more after its birth, [End Page 218] Christianity was typified by having more than one hundred smallish house cult-groups or cells, each with less than a handful of fluent literates. Indeed, on the figures crudely proposed in the first part of my paper, it is possible to think that in 100, Christian ideology was the intellectual possession of barely fifty fluent literates. It was the tiny size of this creative body, and the small cult-groups within and between which they worked, which together account for the exclusivist and dogmatic character of their self-representation. 69

In its early stages, say during the first century and a half of its existence, Christianity was a set of small and vulnerably fissile cult-groups. Internally, each group may have been held together by a demanding ethic, communal worship, and an encouraging message of hope. And all the groups, as a set, may have been held together by shared oral traditions and a thin stream of beggar-missionaries. But if Christianity was to survive over time as a recognizable entity, some mechanism had to be found to unify these small, scattered and volatile communities. Writing and belief, or rather writing about belief, became the prime instrument of unification. And the dogmatic style of exclusivism (only my version of the truth is acceptable) was, I argue, partly a function of the small average size of each cell and the rarity-value of literate leaders within each. In these circumstances, single teachers might feel encouraged to be dogmatic.

Of course, the drive towards unification did not succeed completely, ever. The house cult-groups and communities were too diverse and too diffused over different regions with their own cultural traditions, and individual Christian believers were too passionate and inventive for unity ever to be achieved in reality. But the ideal and illusion of unity as a church and as a grand (apostolic) tradition persisted, and had a powerful effect on Christian organization and self-representation. Christian church leaders repeatedly tried, at least from the middle of the third century onwards, to achieve unity of belief and practice.

The continuous, rapid rate of growth of Christianity, envisaged in Figure 1 (3.4% cumulative per year), implies that at any one time about [End Page 219] two fifths of all adult Christians had become converts, and so new members of house cult-groups or communities, during the previous ten years. 70 This rate of continuous growth put a tremendous strain on the absorptive and instructional capacity of older members. And it helps us understand the idea, which so differentiated Christians from pagans and Jews, that Christians were made, not born. At any one time in the first three centuries of Christianity, if the numbers in Figure 1 are anywhere near right, a significantly large proportion of the adult members of the Christian church were new members, pupils, volunteers.

But volunteers could both join and leave, or be ejected. So Christianity shared with devotees of a polytheistic cult (but not with Judaism) the possibility of temporary attendance. But membership of pagan cult was by and large a function of locality and performance, not belief. Pagans performed local cult, as a matter of course, by living in a city or village, by growing up as polytheists. They could voluntarily opt into extra religious performances, as the desire or need took them. In saying this, I do not want to collude with a Christianizing distinction between belief and behavior. Inevitably, religious behavior consciously or unconsciously, involves mental attitudes. Jewish thinkers and pagan ideologues expected religious performances to be accompanied by appropriate thoughts, such as pious reverence or purity of heart (or at least the absence of hate). 71

Christian leaders too expected this internal piety from their followers. But, in addition, they expected and exacted formal commitment to specific beliefs about Jesus' godhead and redeemership, and their own hopes of salvific redemption and immortality. This demand constituted a radical break from both Judaism and polytheism. Why? Two explanations seem important, one genetic (in the Genesis sense), the other functional.

Genetically, Christian leaders' fixation on their common beliefs arose from their extraordinary nature: Jesus was both human and divine, he suffered death to save humanity; by believing in him as the son of God, [End Page 220] we will be saved. By both Jewish and pagan standards, this message was extraordinary. No wonder it played a crucial role in Christians' self-definition. Functionally, concentration on formal statements of belief made it much easier to join communities spread around the Mediterranean. A simple test (do you believe x and y?) could be administered, and their justifications could be elaborated, by letter. The dispersed and vagrant leadership of the primitive church could maintain the illusion of homogeneity through writing about their beliefs. Of course, it took some time to decide exactly what the identifying beliefs were (the creed was not formalized until the fourth century), and what the test of belonging should be, or how it should be administered. But the innovative principle that their religion was founded on a shared belief (rather than, or as well as on, a shared practice) remained constant for centuries.

Concentration on belief rather than on practice originated in part as a device for differentiating Christians from Jews, just as Sunday and Sabbath differentiated them (though both sects remained similarly distinctive in the Roman world, by having a weekly holy day). But functionally speaking, belief statements were an economy travel-package, so much more easily transportable between widely dispersed, fast growing and freshly established cult-groups, than detailed rules of legal observance which needed a solid body of long-time practitioners to socialize new adherents. Simple capsules of Christian belief statements could be so much more easily absorbed by a constant flood of new recruits than complex rules of daily life or even of liturgical practice. Or put another way, it was much simpler to learn how to be a Christian, than to learn how to be a Jew. And we must remember that according to the crude numbers outlined in Figure 1, 40% of the adult members of any Christian community had become new members in the previous decade. Christianity was a religion which, because of the rapidity of its expansion, always had to be questioning its members about the nature and degree of their adherence.

This strategy of privileging belief over practice carried high risks. The high risks arose from the need to maintain coherence by expelling (or threatening to expel) deviants. Expulsion or the threat of expulsion seriously increased the risk and incidence of heresy and schism. I don't think that anyone in the middle of the second century could have reasonably predicted that the policy of dogmatic exclusivism would end up with a triumphant monopoly. The success of the strategy was discovered only over time; it was not purposively invented as a marketing device. Yet, in balance, the costs of maintaining orthodoxy were mitigated, especially in small groups, by the advantages of inculcating [End Page 221] a heightened sense among the survivors, that they were the sole inheritors, and the correct propagators of the one true Christian message. And in the medium term, concentration on belief allowed a constant elaboration and sophistication of what these beliefs implied, and how they fitted in with each other. This elaboration of belief, which we call theology, allowed a gradual rapprochement between Christian leaders and pagan philosophy. And this gradual rapprochement gave the Christian message a socially acceptable veneer. Sophisticated elaboration of Christian ideology also allowed or even encouraged an internal differentiation among the Christian faithful, so that ideological specialists could gain symbolic capital, material rewards and ecclesiastical power from their intellectual proficiencies. For all its pretentions at universality, Christianity particularly rewarded its elite; indeed, such differential rewards were a necessary part in Christianity's political success and influence.

10. The Implications of Mass Conversion

Rapid growth in the absolute numbers of Christians occurred only in the third and fourth centuries. According to the estimated numbers in Figure 1, there was an increase of about one million Christians in the first half of the third century, and of five million new Christians in the second half of the third century, thirty million plus in the fourth century. From such figures, it becomes easier to judge the scale of what church historians often claim for the growth of Christianity. The problems of internal adjustment, and the cumulative impact of paganism on Christian practice, must have been tremendous. But for the moment, let us put scepticism aside. Let us assume that the straight growth line is roughly right. 72

Now we can see that it was only in the third century, when the number of Christians grew from say 220,000 to over six million, that the church gained the resources and numbers to justify building churches devoted to meetings. And in most towns, it was only then that internal differentiation evolved to a degree which could maintain a hierarchy of bishops and priests, working exclusively as priests, and supported by the contributions of the faithful. [End Page 222]

But the increase in numbers brought its own troubles. Visibility and bulk provoked the first serious attempts by the central Roman government to destroy Christianity in 250-51, 257-60 and after 303. And amazingly, if these growth figures are anywhere near right, then the persecutions or contemporary conditions (civil wars, barbarian invasions, rampant inflation, repeated plagues, urban decline), or their combination, encouraged an unprecedented growth in the numbers of Christians. Success in achieving growth also prompted a battle royal among Christians themselves, between the traditional rigorists who wanted to maintain the old ways of the devoted small community, and the laxists, who wanted growth in numbers, even if that meant sacrificing moral standards.

The conversion of the emperor Constantine, the continued Christianity of his successors and the alliance of the church triumphant with the Roman state brought about a still more dramatic increase, of say thirty million people, in the membership of the church during the fourth century. But by then, for most people being a Christian must have meant something quite different from what it had meant in the first three centuries c.e., and the nature of some conversions may have been, must have been, superficial.

The utility of the Christian church to the state was, I imagine, discovered only over time, by Constantine and his successors. It was not necessarily foreseen by any one of them. But it is worth noting that successive Sassanian kings towards the end of the third century moved the Iranian empire towards religious (Mazdaean) exclusivism and the systematic persecution of religious "deviants." 73 It seems that the two rival and hostile empires, Iran and Rome, were moving along a similar path at roughly the same time. Was it because both empires needed a greater degree of symbolic unity in order to squeeze greater resources from their subjects?

In the long-run, Christianity gave to the Roman state a degree of symbolic unity and exploitable loyalty, which it had previously lacked. Christianity had more combinatory power and more power to demand self-sacrifice than the previous combination of localized polytheisms, [End Page 223] vague henotheism, and emperor cult. Christian rulers and their henchmen now had the legitimacy and authority of a powerful and interventionist God to support their authority and the enforcement of state regulations. The Roman state endorsed and then borrowed Christian intolerance.

In the medium-term, a unified religion helped the Roman state to secure the self-sacrifice required of both soldiers and tax-payers. Sacrifice, so their leaders said, was demanded by God. The Christian religion became, in other words, a supplementary weapon of political and social control, used alongside law, violence and taxation. Christianity also helped provide a cohesiveness of religious discourse among enthusiasts, which, rather like political discourse in modern developed states, bound competitors together via their minor differences. One advantage of religious over political discourse was that, at least overtly, enthusiasts were not discussing the redistribution of resources (always a tricky issue in a pre-industrial society with a limited disposable surplus), but the irresolvable issues of the nature of God or of life after death.

One significant short-term advantage of switching to Christianity as the state religion lay in the possibility it opened up to pillage the stored reserves of pagan temples. Only one source, the fifth-century church historian Sozomen, states that Constantine took money from pagan temples. 74 But it is difficult to envisage how Constantine managed to find the huge sums needed to found the new capital of Constantinople without using the funds of the temples, which increasingly over the next sixty years were destroyed. Christianity also allowed successive emperors to switch patronage from an over-privileged traditional aristocracy to fresh swathes of often newly Christianized supporters.

The disadvantages of Christianity, from a political point of view, lay mainly in transition costs, in the alienation of the conservative and pagan upper-classes, and in the difficulty of effecting a thorough-going mass conversion. The superficiality of Christianization in broad areas of the Roman world was revealed only three centuries later, when Islam swept triumphantly through exactly those regions where it is alleged Christianity first took deepest root. The church was also prone to zealous schism and dissent. The combination of alienation, superficiality and division meant that the church could not always deliver to the state the political loyalty of its putative believers. It has also been argued that the Christian church was itself expensive to the state in tax-immunities, diversion of [End Page 224] ability, and non-productive life-styles, particularly at a time when the empire was desperately trying to defend itself against barbarian incursions. There may be something in these arguments, but the survival of the eastern empire, for close on a thousand years after the conversion of Constantine, indicates that the internal costs of religion were not excessive.

11. Summary

In this paper, I have experimented with estimating the number of Christians at successive stages of Christian evolution. The figures are necessarily speculative, but nevertheless can usefully serve as a framework against which to test alternatives and implications. I have come to five main conclusions. First, the number of Christians expanded fast, but for a long time remained tiny relative to the total population of the empire. The disparity between number and proportion helped produce two different but complementary accounts; Christians thought of themselves as successful but persecuted, while leading Romans long remained ignorant of their activities. Stories about persecution, rather more than persecution itself, were an important factor in Christian self-representation and togetherness.

Secondly, Christian house cult-groups in the first century after Jesus' death were on average both small and dispersed. The small size of the groups helped maintain enthusiastic vigor and ethical rigor among converts. The rapid growth and expansion of the religion depended upon the creation of an easily administrable test of membership, encapsulated in brief statements of belief. This emphasis on belief combined with ethical practice was a significant religious innovation. It was, I argue, a function of dispersion, small numbers and rapid growth.

Thirdly, given general rates of literacy among the Roman population, and even allowing for somewhat higher rates among Christian converts, it seems likely that the development and maintenance of Christian religious ideology in the first century after Jesus' death was at any one time the intellectual property of only a few dozen men, scattered throughout the Mediterranean basin. The maintenance of identity between groups depended therefore upon writing, and particularly upon the writing of letters. The smallness of the group of educated devotees helped give early Christianity its intense internalized character.

Fourthly, the number of Jews was very large compared with the number of Christians, at least until the late third century. Because enthusiastic cult-groups, according to modern evidence, expand usually [End Page 225] along family and social networks, i.e., among relatives and friends, it seems likely that Jews were the main early customers for conversion to Christianity. But differences in experience between Christians and Jews (as between Christians and Romans) helped generate complementary, but contrasting accounts. In the early days, most Christians were ex-Jews or their descendants. So Christians, at least until the completion of the New Testament texts (roughly speaking, the middle of the second century), were preoccupied with their relationship to Judaism. But Jews, the vast mass of whom remained immune to Christianity, for a long time largely ignored the existence of what was for them a marginal group. This discrepancy of experience helps explain why Christians continued to use the Jewish bible as their main authenticating text. The main focus of Christian expansion moved to the gentiles only during the second century. This change of focus perhaps helps explain why the letters of Paul and the Acts of the Apostles, both of which celebrate the mission to the gentiles, were finally included in the canonical New Testament.

Finally, the greatest surge in Christian numbers (in absolute terms) occurred in two stages, in the third century and fourth centuries. Surprisingly, the first mass, centrally organized and reputedly severe persecutions coincided with considerable Christian growth. In terms of number, persecutions helped Christianity. And then the mass of Christian conversions, which followed the alliance of the Christian church with the Roman state under Constantine and his Christian successors was on a huge scale, and was sufficient significantly to change the nature of Christian practice. It is customary to consider Constantine's conversion and adoption of Christianity as a state-favored religion, in terms of his personal sincerity or his perception of Roman interests. It is extremely interesting that Iran, Rome's most powerful enemy, had gone along the same road of trying to create an exclusive monopoly of state religion, barely thirty years before Constantine's conversion.

Keith Hopkins is Professor of Ancient History at King's College, Cambridge.

*My warm thanks for encouragement, help and advice to Jaime Alvar, Mary Beard, Keith Carne, Elizabeth Clark, Simon Goldhill, Christopher Kelly, Seth Schwartz, and to this journal's anonymous reader.


1. A similar tactic is used by Rodney Stark, The Rise of Christianity (Princeton, 1996), 4ff. I found his book suggestive, helpful and provocative. My debt to his thinking pervades this article, though I differ from him in emphases and interpretation.

2. This opposition between what we could call interpretative or reflexive understanding and critical path analysis is sometimes conceptualized as being between soft history and hard sociology. But history and sociology, are each immensely diverse. Besides, I prefer to think of them as complementary, with many overlaps of concept and practice. That said, I should stress that my arguments in this article are predominantly of the "suppose if"/parametric probability kind.

3. I take it for granted that membership of voluntary associations fluctuates; how could it not? For historical illustration, see the excellent analysis of Shakers, Mormons and the Oneida community by L. Foster, Religion and Sexuality (New York, 1981).

4. Epiphanius, Panarion; Augustine, De haeresibus ad Quodvultdeum, Philastrius, Diversarum hereseon liber, to say nothing of the other heresiologists, such as Irenaeus and Hippolytus, who celebrated Christian centripetality and diversity.

5. Adolph von Harnack, Die Mission und Ausbreitung des Christentums in den ersten drei Jahrhunderten, fourth edition (Leipzig, 1924); English translation of the second edition, The Mission and Expansion of Christianity in the First Three Centuries (London, 1908). This is still an indispensable discussion of the surviving testimony.

6. In a similar vein, the British princess Margaret, returning from a holiday in the West Indies, is reported to have said that she had had a wonderful time: "Absolutely no one was there."

7. Letters 10, 96-97, dated about 112. For a glimpse into the enormous literature on this correspondence, see A. N. Sherwin-White, The Letters of Pliny (Oxford, 1966), 691ff. Tacitus' account of the persecution of Christians under Nero was written a few years later.

8. G. E. M. de Ste Croix in a justly famous article, "Why were the early Christians persecuted?" Past and Present 26 (1963): 6-38, reprinted in M. I. Finley, ed., Studies in Ancient Society (London, 1974), 210ff., argued that being called a Christian (technically the nomen christianum) was a sufficient criminal charge against early Christians. Sherwin-White argued less convincingly (ibidem, 250ff.) that it was the early Christians' obstinacy, mentioned in Pliny's letter (10.96) which ensured their persecution. In my opinion, de Ste Croix's superior advocacy (ibidem 256ff.) has unjustly obscured the nature of the problem. Both were partly right, though answering different questions. The first answer is to the question: on what formal charge were Christians prosecuted? The second is an incomplete answer to the more general social question: why were Christians prosecuted/condemned?

9. Justin, 1 apol. 3-4; Athenagoras, leg. 1-2; Tertullian, apol. 1-2.

10. See R. G. Collingwood's brilliant An Autobiography (Oxford, 1939), 79ff., for a long-unheeded but still all too relevant criticism of ancient history's "scissors-and-paste men," and the criteria for using evidence.

11. Harnack, Mission und Ausbreitung, 946ff.; Mission and Expansion, 2:324ff. The influential footnote which contains a confusing misprint is found at p. 807 and 2:248 respectively.

12. Stark, Rise of Christianity, 6, for several modern estimates.

13. On the tactics of model-construction in Roman history, see K. Hopkins, "Rome, Taxes, Rent and Trade," Kodai 5/6 (1995/6): 41ff.

14. Following Stark, Rise of Christianity, 5.

15. It is quite possible to think of implications without knowing the exact size of the Christian population. But that is why so many of my arguments here have the form "if x then y," or "the more x the more (or less) probable y is." For example, if Christians usually met in private houses, and if regular attendance was a condition of being Christian, then the more Christians there were, the more house cult-groups there were.

16. This graph is a reexpression of the illustrative figures given by Stark, Rise of Christianity, 7. One advantage of a graph is that it is easy to see the crudity of the linear assumption, and to read off interstitial numbers.

17. These are rounded up versions of the precise numbers given by Stark, ibid., 7 with a couple of additions. Note: they are guesstimates, not facts.

18. See, for example, on mass desertions from Christianity in the persecutions under Decius, Cyprian, De lapsis 7-9, and Dionysius of Alexandria in Eusebius, h.e. 6.41.11-12; on sacrifice by the bishop of Smyrna, see Acta Pionii 15. Even so, the overall number of Christians increased in the same general period.

19. Cf. Stark, Rise of Christianity, 7. It's worth emphasizing that no one knows the size of the population of the Roman empire. Estimates vary, though most scholars by convention use 50-60 million as plausible middling figures, following in the track of K. J. Beloch, Die Bevoelkerung der griechisch-roemischen Welt (Leipzig, 1886), 507, who estimated the total population in c.e. 14 as 54 million. For a modern view, see B. Frier, Cambridge Ancient History, vol. 11/2 (Cambridge, 1998). He argues plausibly that the population of the Roman empire grew in the first and second centuries c.e. If so, then all my proportional arguments hold a fortiori.

20. This calculation is based on a very rough estimate. Let us say that the population of the eastern half of the Roman empire, more urbanized than the west, was 35 million out of the 60 million total population. Let us say the urbanized population there was 15% or 5.25 million, which includes the large cities of Antioch and Alexandria. And as our present problem is the visibility of Christians in towns and cities, we should include Carthage and Rome in our calculations. So in total, we have to reckon say 220,000 Christians in 200 as a proportion of (urbanites in the eastern Mediterranean, plus Rome and Carthage) say 6.4 million = 3.4%. But according to Dionysius of Alexandria in Eusebius, h.e. 7.24.6, Christianity did spread to villages in Egypt.

21. For example, 2 Clem. 2.6 states that Christians were more numerous than Jews (but see section 8 below); Tertullian claims that Christianity has spread widely geographically, and socially upwards: "we have filled cities . . . villages, towns . . . town-councils, palace, senate, forum, leaving only the temples to you" (Apol. 2).

22. It is obviously risky to use Christian apologetics or martyr acts to portray relations between Roman provincial governors and Christian leaders, since what we have are Christian self-representations, not official accounts of trials. For the considerable difference between surviving court records of trials from Roman Egypt and martyr acts, see the convincing account by G. A. Bisbee, Pre-Decian Acts of Martyrs and Commentarii (Philadelphia, 1988), esp. 33ff. That said, Roman irritation with rather than anger against Christians comes out for example in Tertullian, ad Scapulam 4-5.

23. On persecutions, see the full but credulous account by W. H. C. Frend, Martyrdom and Persecution in the Early Church (Oxford, 1965), and with flair, A. J. Droge and J. D. Tabor, A Noble Death (San Francisco, 1992). On Jews as an alleged source of persecution, see, for example, Justin, Dialogue wth Trypho 16: "You are powerless to lay hands on us, because of our overlords [the Romans], but you have done so whenever the opportunity arose . . ." and Tertullian: "the synagogues of the Jews are the cause of our persecution" (Scorpiace 10).

24. Individual martyrs became a special Christian type of hero, with power, so some believed, to forgive sins, even in this world (much to the controlled indignation of bishop Cyprian, Letters 15-20). Contrarily, though it was not the winning position, some Christians thought that voluntary martyrdom was as futile as the suicide of an Indian fakir, and that real martyrdom was to be sought in daily life (Clem. Alex., strom. 4.4.17, 2.20.104).

25. By modern scholarly convention, Jews rebelled, but Christians were persecuted. Statistically, Jews for a long time had more to complain about. For oppression in Antioch, see Josephus, bj 7.46ff. and 103; in Alexandria, Philo, in Flaccum 55ff. On the long anti-Jewish prejudice in Alexandria, see H. Musurillo, ed., Acts of the Pagan Martyrs (Oxford, 1954).

26. For example, repeatedly making the sign of the cross on the forehead, and not wearing or decorating door-posts with wreaths during festivals (in so far as Christians actually behaved openly as their leaders told them) must have set them apart (Tertullian, De corona 3 and 10).

27. Christian apologists in the second and third centuries (from Justin to Minucius Felix and Origen) preserve Christian versions of the (powerful) attacks, which pagans made against them. It would be foolish to assume that these rationalized arguments were the only criticisms popularly made against the Christians. For all their overt appearance as documents addressed to emperors and educated pagans, it would take a very patient pagan to read them. They are aimed at Christians, and celebrate Christian difference.

28. On the internal divisions within Paul's group at Corinth (I Cor 11.17ff.), see, for example, G. Theissen, The Social Setting of Pauline Christianity (Edinburgh, 1982), 145ff.

29. For warnings against false teachers, who want to stay in a house cult-group for more than three days without working, and who ask for money as well as food, see Did. 11-12. The notion of false prophets haunts the dispersed early Christian groups. How can they tell?

30. Harnack, Mission und Ausbreitung, 618ff.; Mission and Expansion, 2:89-96. See also Y. Aharoni and M. Avi-Yonah, The Macmillan Bible Atlas (New York, 1977), 166-67.

31. L. M. White, Building God's House in the Roman World (Baltimore, 1990), especially 105; he notes that there were in Paul's time six houses in Corinth used for meetings by Christians.

32. Palestinian Talmud, Kilayim 9.4.32b and Shabbat 6.8a. These passages may reflect fourth-century, not second-century conditions; but for my current purposes that does not matter. The principle I want to establish is that if attendance was a condition of membership in a religion, then the larger the town, the more meeting places you needed, even for a licit religion, a fortiori for an illicit one.

33. Harnack, Mission und Ausbreitung, 618ff.; Mission and Expansion, 2: 89-96. I may seem to be being a bit unfair to Harnack, since his agenda was to establish what can be known/proved about Christian expansion. The trouble is that positivist followers translate known into all.

34. Ibid.

35. According to Eusebius (h.e. 6.43.11), the church at Rome in 251 supported 46 priests, over a hundred lesser clergy and employees, plus 1500 widows and beggars. Harnack (ibid., 806; 2:247) guessed that there needed to be at least 30,000 Christians at Rome to support that number of clergy and dependents.

36. This is and can be only a very rough order of magnitude. There are two problems, our ignorance and the arbitrariness of the boundary which divides a town from a village, not notionally (town-council, baths, acknowledged status), but in fact. All that said, I think 2000 is about right. For the testimony on which this is based, see A. H. M. Jones, The Later Roman Empire (Oxford, 1964), 712ff.

37. On the relatively late coming out of Christians (at the end of the second, early third century), see, for example, P. Lampe, Die stadtroemische Christen in den ersten beiden Jahrhunderten (Tubingen, 1989), 13ff. On the earliest, archaeologically known church, an unobtrusive, mud-brick refurbished house, no longer used as a residence, built in about 230, converted to church use about ten years before its destruction in 256 (its assembly hall held 65-75 people), see C. H. Kraeling, Excavations at Dura-Europus, 8.2 (New Haven, 1932), 3, 19 and 37.

38. On the building of large churches, see Porphyry (died c. 305) cited by Macarius Magnes, 4.21.

39. See A. J. Coale and P. Demeny, Regional Model Life Tables (Princeton, 1966), table West, level 3, stationary population. To be unnecessarily precise, children aged 0-17 constituted 31.9% of the total population, adult males 28.6%, and adult females 31.9%. Mortality of adult females was lower than of males. The sex ratio from Roman Egyptian census returns is in the region of 108:100 m:f; see R. Bagnall and B. Frier, The Demography of Roman Egypt (Cambridge, 1994), 95.

40. On the appeal of Christianity to women and children, see Celsus in Origen, Cels. 3.55; on the preponderance of women among Christians, see Council of Elvira, canon 15, and for a collection of testimony, Harnack, Mission and Expansion, 2:64ff. Stark (Rise of Christianity, 95ff.) in his chapter on Christian women indicates that in modern proselytizing religious movements, women are primary converts. But in treating ancient evidence, I think he too readily equates prescription with performance, and single instances with general patterns of behavior. See also J. Bremmer, "Why Did Early Christianity Attract Upper Class-Women?" Instrumenta Patristica 19 (1989): 36ff.; the answer must surely be that before the fourth century, it did not attract many (proportionately), and the stress on their membership which we find in Christian sources arises precisely from women's social visibility and rarity.

41. See L. R. Iannaccone, "Religious Practice," Journal for the Scientific Study of Religions 29 (1990): 301-2; Stark, Rise of Christianity, 15ff. In my view, preferential female recruitment in Roman conditions was probably more a rhetorical figment than a statement of general fact. And its impact was less in an ancient society, suffering high mortality, than it would be in a modern society, enjoying lower mortality. In the Roman world, if, exempli causa, all recruitment was among young adults, of whom 75% were female and prior to the birth of their first surviving child, and 25% male at a similar age, then, if all the children of Christians became Christian, the sex ratio in the total body of Christians, growing at a constant 3.4% per year, would be 40% male, 60% female (Stark, p. 101, miscalculated by omitting children). But among adults, because of the high constant rate of recruitment, the ratio would be 33% male, 67% female. In a typical community, therefore, in 100 there would be only 13 adult males and 27 adult females, plus 30 children. This would have caused difficulties. I conclude therefore that differential recruitment was not as great as 25:75 m:f. Do other people argue like this?

42. Natural increase (excluding migration) in a pre-industrial population before the demographic revolution is unlikely to persist for a long period at above 1% per annum. The growth of Christians posited at 3.4% per annum compounded by the numbers in Figure 1 may therefore be made up by ca. 1% p.a. natural increase (allowing for the extra fertility of young adult recruits), plus ca. 2.4% p.a. increase via conversions.

43. Augustine tells the story of how Manichees at Rome, a prohibited sect in the late fourth century, were reluctant to enforce discipline against miscreant electi, because of the fear that any disgruntled member might report them to the Roman authorities (De moribus manich. 69). Similar forces must have been at work in the early Christian church.

44. William V. Harris, Ancient Literacy (Cambridge, Mass., 1989), 175ff., 272 and 323ff.

45. There was some crossover between urban and rural populations, in the sense that some people living in towns worked in fields outside towns, and some (a significant minority) of those living in villages were artisans, either full- or part-time, or engaged in other non-agricultural occupations (e.g., priests, scribes, tax-collectors, traders).

46. The income of Pliny the Younger is estimated at 1.1 million HS by R. P. Duncan-Jones, The Economy of the Roman Empire (Cambridge, 1982), 21. Subsistence is crudely reckoned at about 250 kg wheat equivalent per person/year, so that an average family of four persons would need 1 metric ton wheat equivalent, roughly 150 modii at 3 HS per modius.

47. W. A. Meeks, The First Urban Christians (New Haven, 1983), 72ff.

48. Any such calculation must be vague, since there was/is no single valid definition of Roman ruling strata. But if we combine senators, knights, and town councillors (100 for each of 2000 towns), we get a total of say 210,000 adult males (i.e., 1.2%) out of 17 million adult males in the empire. I use adult males as a unit of calculation for convenience. In fact, some towns did not have as many as 100 town councillors, and their wealth differed dramatically according to the size and wealth of the town.

49. In the third millennium b.c.e. in Egypt, less than 1% of adult males had been literate, according to C. G. Eyre and J. Baines, "Four Notes on Literacy," Goettinger Miszellen 61 (1983): 65ff. In the rest of the Mediterranean basin, presumably, literacy rates were as low or lower until very much later.

50. Since not all town councillors (or even bishops) could write, let alone rank as sophisticated literates, this estimate seems overgenerous.

51. Cf. James 2.5: "Listen, my beloved brethren. Has not God chosen those who are poor in the world to be rich in faith and heirs of the kingdom which he has promised to those who love him? . . . Is it not the rich who oppress you, is it not they who drag you to court?"

52. Origen, Cels. 3.44 and 55. This traditional view has been criticized, for example, by E. A. Judge, The Social Pattern of Christian Groups in the First Century (London, 1960), 52ff., and Meeks, Urban Christians, 51ff.

53. Origen, Cels. 1.27: "among the great number of people" converted to Christianity, "because there are many more vulgar illiterates than educated rational thinkers, it is inevitable that the uneducated should outnumber the more intelligent." Needless to say I cite this extract, not as proof, but as illustration. So too Tertullian's dictum that most Christians were simple and uneducated (Against Praxeas 3).

54. Of course, Justin, Tertullian, Clement, and Origen, are dressing Christian arguments in increasingly well-educated clothes, but the appeal to simplicity, poverty and charity, remains and persists as a rhetorical figure and as a spur to action. See P. Brown, Power and Persuasion in Late Antiquity (Madison, 1992).

55. The (later to be) canonical gospels long circulated separately, not as a set of four, and since they were not canonical, nor regularly cited by Christian writers until the middle of the second century, there was no particular reason why Christian communities should get copies, indeed if they knew of their collective existence. A case in point are the Marcionite prefaces to the Pauline epistles, written in the mid-second century, but found in a major set of Latin vulgate manuscripts; their adoption indicates that western Christian communities did not receive a text of the Pauline letters until after the mid-second century, and/or were indifferent to their heretical introductions. See D. de Bruyne, "Prologues bibliques d'origine marcionite," Rev. bened. 24 (1907): 1ff. and esp. 11ff.

56. On parental obligations, see Josephus, Against Apion 2.204; Antiquities 4.211. On ubiquitous schools, see Palestinian Talmud, Megillah 3.73d; Ketuboth 13.35c; Taanith 4.69a. See the good discussion on these points by Harry Y. Gamble, Books and Readers in the Early Church (New Haven, 1995), 6ff.

57. Now that we know the implications of the reasoning, some critical or precommitted readers may want to question it again. But as I see it, the room for maneuver is constricted. Either primitive Christians were recruited more than anyone has suspected from the sophisticated literate sub-elites, or these sub-elites constituted a much larger proportion of the Roman urban population than anyone has suspected, or there were many more Christians than the 7000 estimated for 100.

58. With Tertullian, Clement and Origen, Christian writing showed overt ambition and some success in clothing itself like pagan classical culture. See, for example, C. Bigg, The Christian Platonists of Alexandria (Oxford, 1913); H. Chadwick, Early Christianity and the Classical Tradition (Oxford, 1966); and E. A. Clark, Clement's Use of Aristotle (New York, 1977).

59. Reported by M. Simon, Verus Israel (London, 1996), 33-34, who in turn depended mainly on Juster, Les juifs dans l'empire romain (Paris, 1914), 1:180-209.

60. M. Broshi, "The Population of Western Palestine in the Roman-Byzantine Period," BASOR 236 (1979): 1ff.

61. II Baruch (like IV Ezra) celebrates the despair felt by some Jews after the destruction of Jerusalem; although the author finishes by reaffirming his trust in God and the Law, and the hope of eventual revenge, he also acknowledged that some Jews had deserted (41.3: "Behold, I see many of your people who separated themselves from your Law"; 85.3: "we have nothing now apart from the Mighty One and his Law. Therefore, if we direct and dispose our hearts, we shall receive everything which we lost again by many times"). In the rebellion of Bar Kokhba, Jews who had had their foreskins surgically restored, presumably in order to take part nude in hellenistic civic life, were forcibly recircumcised (Palestinian Talmud, Shabbat 15[16].9). To not much avail, at the end of the second century, the Jewish cities of Sepphoris and Lydda, for example, changed their names to Diocaesarea and Diospolis (Zeustown).

62. Luke/Acts records repeated visits by Paul to synagogues in different towns, where he taught his message to Jews and persuaded many of them. Of course, Acts is not a work of accurate history (whatever that is), but a doctrinal tract with a message. Even when it was written, towards the end of the first century, the appeal of Christianity to Jews, and the break from Judaism, was central to Luke's perception of Christian evolution, and I suspect, of contemporary Christian preoccupations. See very helpfully on Acts' historicity, R. I. Pervo, Profit with Delight (Philadelphia, 1987).

63. Stark, Rise of Christianity, 16.

64. See, for example, Acts 3-4.4, Acts of Andrew 1, Acts of John 18ff., Acts of Peter 27, Acts of Thomas 23-24, 31-38. R. Macmullen, Christianizing the Roman Empire (New Haven, 1984) considers miracles central to Christian expansion. But I'm not sure that Romans "took miracles quite for granted. That was the general starting point" (22).

65. I take seriously Justin's clear and repeated statement that Jews cursed Christians (dial. 16.4; 93.4; 95.4; 96.2; 108.3; 123.6; 133.6; 137.2). But the issue is complicated; see W. Horbury, "The Benediction of the Minim and Early Jewish Controversy," JTS 33 (1982): 19ff., and P. W. van der Horst, Hellenism-Judaism-Christianity (Kampen, 1994), 99ff.

66. Ignatius, Polyc. 6; cf. Smyrn. 8 and Philad. 7.

67. My favorite text, illustrating (to me) the balance of disagreements among rabbis, and describing the excommunication of Eliezer ben Hyrcanus, is Babylonian Talmud, Baba Mezia 59a-b; the structure of the text is made more evident in the translation by J. Neusner, Bava Mesia (Atlanta, 1990), 154-56. The same story, but with important variants, is told in the Palestinian Talmud, Mo'ed Qatan 3.1.

68. Perhaps this is overdramatic and overstated; perhaps reconciliation and tolerance of differences, which certainly occurred, leave less trace in the sources, or are less remarkable. All that said, the hostility between early Christian sectarians is notable, a symptom of their commitment. Examples are II Cor 6.14 (interpolated?); Col 2.8; Eph 4.14; I Jn 2.18; II Jn 7-8; Rev 2.6.

69. I hope this does not sound too reductionist. I am thinking here not that the small size of house cult-groups, their dispersion, and the scarcity of sophisticated literates in the early church created the importance of dogma, rather that perhaps these were the main factors which preserved and enhanced the importance of dogma. Perhaps also the importance of dogma, or of explicit belief statements, as a criterion of sect membership, is itself a figment of the surviving literary sources: sophisticated literates stressed belief, stalwart practicing Christians stressed practice. I suspect truth lies in their combination.

70. This is a bit complex to work out, and depends on several assumptions, such as the age of the new converts, whether before or after marriage, and the age distribution of preexisting members, as well as on their respective fertility. On reasonable (but not necessarily true) assumptions, exempli causa, if all converts were young adults, and older and new converts had similar fertility, then those converted within the previous ten years constituted about 40% of adult Christians, but only a quarter of all Christians including children.

71. S. J. D. Cohen, From the Maccabees to the Mishnah (Philadelphia, 1987), esp. 66ff.

72. Of course, I do not mean by this that straight line growth is an accurate description of what happened; it is merely the easiest? best? most economical way of thinking about Christian growth; it is a first approximation, which elicits correspondingly simple, structural implications and further thought.

73. The (attempted) religious unification of Iran by the high priest Kartir under four successive third-century Iranian kings is recorded in four autobiographical inscriptions, three of which are transcribed and translated with learned commentary by M. Sprengling, Third Century Iran, Sapor and Kartir (Chicago, 1953). The fourth Kartir text is given by P. Gignoux, "L'inscription de Kartir a Sar Mashad," Journal Asiatique 256 (1968): 390ff.

74. Sozomen, h.e. 2. 5.

© Экклесия 2004 Все права защищены.
Напишите нам: ecclesia@list.ru