Scholarly attempts to account for the complex process that travels under the name of "the Christianization of the Roman Empire" have produced a rich and varied literature in the last century. 1 The extent to which women's participation in the early Christian movement enabled that process has been an important subtopic of research, especially among feminist historians of late antiquity. 2 Worthy of special consideration [End Page 227] here is the extensive literature documenting women's special attraction to particular forms of Christian belief and practice, those involving varying degrees of ascetic discipline. 3 It is within this broad framework that I have undertaken to engage the fifth chapter ("The Role of Women in Christian Growth") of Rodney Stark's recent book, The Rise of Christianity: A Sociologist Reconsiders History. [End Page 228]

This chapter is one piece of a larger mosaiclike argument in which Stark undertakes the very salutary task of accounting for the historical ascendancy of Christianity by making use of sociological theories and models. Stark's earlier work has focussed primarily on contemporary religious movements, temporally and geographically remote from the worlds of the late ancient Mediterranean. His work to bring sociological theory to bear on this new material is particularly laudable since early Christianity has often been treated as qualitatively different from many other religious innovations that have emerged in different times and places. That special treatment can usually be traced back to some form of theological privileging or related queasiness. To attempt to explain the spread of Christianity through recourse to rational explanation rather than deus ex machina argumentation is a very worthy task indeed. And although this essay will express some rather substantive disagreements with Stark's interpretations of the ancient evidence and renderings of the scholarly consensus on numerous points, it nevertheless shares considerable sympathy with the goal of providing a verifiable historical explanation for the eventual hegemony of Christianity in the ancient Mediterranean world. [End Page 229]

Questions of Method and Evidence

At the outset, some general questions about the assumptions that seem to underwrite Stark's understanding of early Christianity and its broader cultural context need to be raised. There seems to be a tendency in the chapter (and perhaps the book as a whole) toward the generalizing use of monolithic interpretive categories: "Christianity," "the Greco-Roman world," "paganism," "Judaism," and so on, operate as uninterrogated and unitary categories. In his recent book on the cultural transformation that took place between the fourth and sixth centuries, historian Robert Markus has reminded us that "the image of a society neatly divided into 'Christian' and 'pagan' is the creation of late fourth-century Christians, and has been too readily taken at its face value by modern historians. . . . 'Paganism' . . . existed only in the minds, and, increasingly, the speech-habits, of Christians." 4 And the category of "the Greco-Roman world" collapses considerable temporal and geographical differences, to say nothing of cultural and ethical ones.

"Christianity," too, is in many ways a heuristic construction, especially in the early centuries of the movements that travelled under its name. Indeed, both "Judaism" and "Christianity" in antiquity have increasingly been rendered as plurals by historians of religion in an attempt to signal our expanded sense of variety in both of these traditions at this time. Moreover, the sharp distinctions that Stark invokes between "pagan" and "Christian" have been considerably blurred by historians in recent years.
In a similar vein, the category "women" simultaneously renders visible the historical specificity of sexual difference while obscuring a wide range of differences among women. (The unidentified sculpture reproduced on page 96 of Stark's book is emblematic of the tendency to render any single image of a woman as a figure for "women in general.") Which "Christianity" is it that afforded which Christian women a higher status than that of which of their non-Christian peers?

Of course, in order to see the big picture, one must sacrifice some level of detail. However, some level of specificity is required lest one be reduced to making absurd and patently false claims. What does it mean to claim that Christian women (in general) possessed a higher status than their Graeco-Roman counterparts (in general)? At this level of generality, [End Page 230] is it a historically meaningful claim? Once one starts increasing the level of specificity with respect to class, geographical location, marital status, and so on, the picture becomes infinitely more complex: for example, the destitute widow in a rural imperial outback who professed faith in Jesus (a "Christian woman") surely did not enjoy a higher status than the Roman matron of the senatorial class who participated in civic religious festivals, devoted herself to her household gods, and acted as benefactress and patron at a local shrine (a "Roman woman"). The point is that generalizations can very often give way to the weight of significant specific exceptions, and that heuristic categories often obscure some historical realities even as they illuminate others.

On the matter of the evidence marshalled in this discussion of early Christian women, it should be noted that this chapter boldly encompasses a sweeping terrain of social, institutional, doctrinal, and women's history. A non-specialist cannot be expected to have mastered all of the relevant primary and secondary sources, all the more so when specialists have a difficult time keeping up with the explosion of work in the field in the last twenty-five years or so. However, if one were to bring this diverse body of scholarship into the conversation, the portrait of women's social and religious history in this period would certainly be enriched and deepened. 5 This work would provide a critical reconstruction of women's historical lives 6 and would provide models for assessing the rhetorical [End Page 231] strategies at work in a wide range of ancient sources written about women by men. 7 Moreover, this more recent work would raise some important and challenging questions about both the assumptions concerning Christian women's history and the narrative of Christianization upon which Stark's own sources depend.

Contemporary Outsiders' Evaluations of Early Christian Women

The chapter begins with a claim that becomes a cornerstone in the larger argument, the assertion that "although some classical writers claimed that women were easy prey for any 'foreign superstition,' most recognized that Christianity was unusually appealing because within the Christian subculture women enjoyed far higher status than did women in the Greco-Roman world at large" (95). The statement claims that most classical writers recognized that Christianity was appealing to women because women enjoyed higher status within Christian groups than outside of them. Leaving aside, for the moment, the question of the content of "higher status," let us consider the evidence for what ancient writers thought about women's attraction to Christianity. 8 The evidence comes primarily from five writers: Pliny the Younger, Marcus Cornelius Fronto (whose ideas were preserved in Minucius Felix, Octavius), Lucian of Samosata, Galen of Pergamum, and Celsus (preserved in Origen, Contra Celsum). Of these, probably the most important is Celsus who, far from arguing that women had higher status in Christian groups, consistently asserted that Christianity was not worthy of serious attention by serious people because it claimed dubious origins in illegitimate [End Page 232] birth (Mary the mother of Jesus) and the testimony of a hysterical woman (Mary Magdalene), while subsisting on the gullibility of "the foolish, dishonourable and stupid, and only slaves, women and little children" 9 who are drawn in by magic and sorcery. Celsus argues that the Christians routinely challenge the authority of the household, but this is not for him the salutary raising of women's status, but rather a fundamental threat to social order and well-being.

In her recent book, Early Christian Women and Pagan Opinion: The Power of the Hysterical Woman, New Testament scholar Margaret Y. MacDonald undertakes a project of historical reconstruction of women's history from the rhetoric of non-Christians' observations about Christian women. 10 In the process, she reconstructs a picture in which Christian women's religious involvement fell into the more dominant pattern of the broader culture, in which women's religiosity was simultaneously revered and distrusted (4). Moreover, she reconstructs the ways in which the institutional church's response to suspicious assessments of Christian women's activity shaped that very activity, reining it in in critical ways. Contrary to the claim that early Christian women were elevated to an unambiguously higher status and that everybody at the time recognized this as a fact, MacDonald illustrates how the blurring of the categories of public and private that characterized early Christian religious practice and missionizing activity created substantial cultural anxiety in the broader social context, resulting both in pagan attacks on Christian movements and in apologetic rhetorical gestures and defensively conservative institutional moves by church leaders.

Sex-Ratios, Relative Social Status, and the Problem of Comparison

The question of women's relative social status obviously occupies a central place in the chapter's establishment of its case. Stark's argument, based in part on the work of sociologists Marcia Guttentag and Paul Secord, 11 is that in societies where women outnumber men, women will [End Page 233] possess relatively greater power and freedom than in societies where men outnumber women, because men will, as a consequence of the numerical differential, treat women as "scarce goods" to be managed and controlled (101-2). (Whether this theory is correct or not remains an open question for me. Laying aside my feminist queasiness with the rather indecorous equation of women with "goods," 12 an alternative theory might suggest that women have the increased capacity to make demands precisely when they are "in short supply," that they don't have to "sell themselves short" in such circumstances. 13 ) The comparative example that Guttentag and Secord use, and that Stark cites, is that of Athenian and Spartan women. Stark points out that Spartan women were able to own property in their own names, initiate divorce, obtain equal levels of intellectual and physical education, marry relatively late, wear nonconfining clothing, and move freely in the city. "If Guttentag and Secord's theory is correct," Stark writes, "then we would have to predict that the status of Christian women in the Greco-Roman world would more closely approximate that of Spartan women than that of women in Athens."

There are problems with this kind of comparison at a number of levels: first of all, social status is conferred in varying degrees by both explicit legal license and constraint, on the one hand, and often more implicit factors like social pressures, ideological constructions, and class-based social values, on the other. Status also, I would argue, depends to some degree on how one experiences one's social and political circumstances. Unfortunately, we have virtually no evidence testifying to the question of how women--Spartan, Athenian, Roman, of whatever religious affiliation--felt about their circumstances. Spartan women and Athenian women were governed by both the laws and customs of their particular city-states. Christian women in the Roman Empire were governed by Roman law and by varying combinations of Roman and local custom, depending upon where they lived and their class status. To [End Page 234] attempt to compare "Christian" women with "Spartan" or "Athenian" women may obscure both differences and similarities, since one would not necessarily be comparing analogous institutional constraints and freedoms, nor comparable experiences of social pressure or ideological frameworks. In the case of the right to own property, for example: what should be compared is not "Spartan" women and "Christian" women, but Spartan property law and Roman property law. Aristocratic Roman Christian women's ability to control their own property was a capacity established for all Roman women in similar circumstances by Roman law, not by Christian religion. 14 And it was a control that was controversial and contested, defended by church fathers not in order to bolster wealthy women's social autonomy, but in order to assure the generous patronage of church projects by the women in question. Indeed, it was concern over the church's rather ambitious attempts to secure bequests and gifts of property and capital--not concern over women's religious conversion--that motivated the order (to which Stark refers in the opening sentence of his chapter) by Valentinian I (and Gratian) for Christian monks and clerics to desist from their calling on the homes of upper-class Roman widows and female wards. 15

But what if the kind of comparison Stark suggests was made? Stark enumerates Spartan women's notable freedoms with respect to property, divorce, education, late marriage, dress, and freedom of movement. He implies that Christian women would have a similar set of freedoms, though when he actually turns to the evidence for Christian women, he does not address most of these areas of female existence but turns, rather, to women's status within the family and within the religious community--but not within public institutions or legal situations where, I would argue, one would be hard-pressed to provide evidence for Christian women's higher status. Even on what might seem the relatively trivial private question of women's proper dress (which Stark identifies as one of the markers of Spartan women's relative liberation), church fathers and councils spilled a substantial quantity of ink in often rather radical attempts to discipline and domesticate attire and coiffure. As with many other aspects of female existence, arguments about clothing [End Page 235] intersected with a range of complex concerns about institutional authority, gender differentiation and hierarchy, systems of honor and shame, and theological propriety. In order to be understood in all their fullness and complexity, each of these individual questions of relative status requires careful material and ideological contextualization.

The most viable comparison that might be drawn would be between non-Christian Roman women and Christian Roman women, and here Stark's engagement with the evidence for non-Christian Roman women's lives consistently slants toward the harshest--and, I would argue, not always the most accurate--reading of Roman circumstances while tending to give early Christian sources a consistently generous benefit of the doubt. Let us look at each of the arguments in turn.

Stark argues that the unbalanced sex ratios that existed in Roman antiquity can be traced in large measure to two factors: high rates of female infanticide, and mortality due to unsafe contraceptive and abortion practices. That exposure and infanticide were practiced among the Romans is not a matter of scholarly debate, although the degree to which they were practiced and their ultimate impact on population as a whole and on the female population in particular has been the subject of some considerable controversy. 16 Stark invokes the oft-quoted first-century Greek papyrus letter found at Oxyrhynchus in Egypt from a husband instructing his pregnant wife to dispose of their soon-to-be born infant if it is a girl but to preserve it if it is a boy; 17 but he does not [End Page 236] substantially engage the scholarly debate among demographers and historians in order to situate this one piece of documentary evidence within a broader historical framework. Even if one concludes that there is compelling evidence that female infants were exposed or were killed disproportionately in the Roman context, demographers of ancient Roman culture have argued that the conclusion does not necessarily follow that this practice significantly affects sex-ratios. 18 We simply have too little statistical data to make definitive arguments or, where there is significant statistical data, it can be interpreted persuasively in a range of alternative ways. 19

When Stark sets the early Christian polemics against exposure and infanticide in opposition to the Roman practice, he trustingly takes the Christian sources at face value and does not engage the significant interpretative issues involved in using texts that are largely apologetic in nature as straightforward documentary historical sources. 20 Further, he seems to assume that the transition from the widespread practice of exposure and infanticide to their disappearance was smooth, unambiguous, and linked to a clear-cut and monolithic Christian ethic. Several cautionary arguments suggest themselves here. One cannot necessarily assume that all the people who called themselves Christians in the early centuries of the church necessarily agreed with the church fathers' teachings on these or other matters, or put them immediately or unproblematically into practice in their own lives. Indeed, a disproportionate body of early Christian literature is addressed precisely to dealing with the problems attendant to the fact that far too many people have failed to adhere obediently to the church fathers' version of reality. Moreover, Christian women's historical existence is frequently reconstructed [End Page 237] through careful rhetorical analysis of the strongly worded negative proclamations of church leaders. This may mean, not that Christians did not continue to engage in these practices, but rather that they did. In the specific cases of exposure and infanticide, widespread evidence from medieval, early modern, and nineteenth-century Christian Europe for their continued practice militates against such an unequivocal reading of the history of these practices. 21

If the practices of exposure and infanticide may be open to a more complex reading than Stark gives them, all the more so are the ancient practices of contraception and abortion. Here, Stark privileges marginal (and hardly informed) ancient evidence and does not cite the most important Roman source, Soranus' Gynaecia. 22 Soranus is a particularly important source, not only because he was an educated physician of considerable renown, but also because much of what he includes in his handbook comes from the reportage and experience of midwives, who were the predominant attendants to women's reproductive needs in antiquity. And although wittily (if not altogether accurately) dismissed by one scholar as so many "ancient equivalents of gin, hot baths, and jumping off the kitchen table," 23 the pharmaceutical treatments prescribed for the termination of unwanted pregnancies in Soranus' second-century handbook are largely noninvasive and effective--and they are only prescribed after apparently generally effective contraceptive methods [End Page 238] had either not been used or had failed. 24 Soranus omits descriptions of techniques for invasive surgical abortion, unlike the Christian polemicist Tertullian (with no medical training or experience, but a rather large ax to grind), whom Stark does quote. Indeed, one gets the sense from Soranus' medical handbook that the larger difficulty faced by ancient women was not avoiding or terminating unwanted pregnancy but rather getting and staying pregnant when they wanted to. As Gillian Clark puts it, "Soranus' list (1.14) of causes for [spontaneous abortion] makes one wonder how anyone ever managed to have a baby." 25

Beyond this assessment of the medical sophistication available to ancient women, Stark argues that the patria potestas gave men unilateral control over the fate of their wives' unborn children and implies that women had no appreciable agency in the process. The more likely historical scenario is that a range of impulses and constraints must have governed values, decisions, and practices in the ancient Roman context. 26 For one thing, marriage with manus was in sharp decline during the imperial period, which meant that married women were not under the legal control of their husbands but of their fathers, or, if their fathers were dead (which was statistically highly likely, given the age differences between fathers and children and ordinary life expectancies), of themselves. Although it is certainly possible that husbands could exert some psychological or moral pressure on their wives, in many cases, they would not have been able to exert pressure grounded in patria potestas. For another thing, when Augustus sought to expand the population by creating incentives for increased reproduction, he included in the lex [End Page 239] Julia (18 b.c.e.) and the lex Papia Poppaea (9 c.e.) the so-called ius liberorum which awarded legal autonomy from the existing constraints of the tutela to women if they had three children (or, in the case of a freedwoman, four). 27 That Augustus pursued his project of social engineering by this means suggests that, prior to the creation of the "privilege of children," women had (to some degree, at least) controlled their own fertility and would be willing to stop doing so in exchange for (presumably desirable) increased liberty in the public sphere. Granted, this reform had its primary impact on the upper classes, but its strategy nevertheless suggests that some women were agents in their reproductive lives.

Stark's portrait of high mortality among Roman females because of exposure, infanticide, contraception, and abortion is very severe indeed. It extrapolates a historical reality from sources that can be--and have been--interpreted rather differently by other scholars; this alternative evaluation of a broader range of sources generally produces a more balanced view of the circumstances and constraints that characterized Roman women's lives. But even if one were to accept Stark's portrait as it stands, the question remains whether such circumstances were amenable to the sort of radical social change that Stark argues Christianity inaugurated.

Stark's argument has several parts. The first part of it involves the claim articulated by Guttentag and Secord that demographic imbalance creating a shortage of women means that women have lower status than men whereas a surplus of women implies higher status, a claim I have already questioned above. From here, the argument proceeds to a series of claims: that women were disproportionately represented within the early Christian communities, that Christian women enjoyed higher status, and that their intermarriage with and conversion of pagan husbands combined with their higher levels of fertility contributed [End Page 240] significantly to the spread of Christianity in the empire. Although I certainly agree that women were prominent in some segments of the early Christian movement, I do not see that the evidence supports the demographic, status, intermarriage, or fertility arguments that Stark advances. What follows elaborates on this disagreement.

The Demographics of Early Christianity

Admitting the absence of reliable demographic statistics, Stark nevertheless takes as a historical fact the assertion that there were numerically more women than men in the early Christian movement, using both comparative and actual evidence. The comparative evidence he draws upon involves the predominance of women converts among nineteenth- and twentieth-century new religious movements in the United States and Europe, and recent evangelical Protestant conversion in Latin America (100). Whether women always behave the same way in relation to religious innovation remains for me a crucial and unanswered question. Without careful, historically differentiated gender analysis, I would suggest that we cannot simply assume that women always respond in the same way across such considerable historical and cultural variation. Nor should we assume that apparent cultural parallels between disparate examples necessarily embody actual structural parallels.

The actual evidence Stark musters in support of his claim is rather scanty: Paul's greetings to fifteen women (and eighteen men) in Romans 16; 28 the inventory of the contents of a fourth-century house church in [End Page 241] North African Cirta, where a disproportionate number of articles of women's clothing are recorded; 29 a strategic misquotation of Robin Lane Fox's claim that "Christian women [were] prominent in the churches' membership and recognized to be so by Christians and pagans"; 30 and Adolf von Harnack's 1908 assertion that the ancient sources "simply swarm with tales of how women of all ranks were converted in Rome and in the provinces." 31 (Virtually all claims that women were prominent or predominant in early Christianity can be traced back to Harnack's study.) Compelling contrary evidence is neither discussed nor cited, including the prosopographical and epigraphical evidence interpreted by Michele Salzman in 1989 32 (which I discuss later in this essay) and the Christian grave inscriptions from fourth-century Asia Minor published twenty years ago by Evelyne Patlagean, in which Patlagean shows a striking underrepresentation of females and a surprisingly high number of deceased children. 33 [End Page 242]

Reconstructing Early Christian Women's Social Status

Is it the case that Christian women in the Roman Empire enjoyed far higher status than their non-Christian peers? This is an extremely difficult question to answer, in part because of the problem of what counts as "high status." Insofar as status is linked to one's relationship to the dominant legal system in place, Christian women and non-Christian women were governed by the same laws with respect to marriage, divorce, inheritance, rights to hold property, and so on. But I have already suggested that status involves a complex blend of legal license and constraint, often highly varying degrees of social pressure and local custom, class status, ideology, and perhaps also some significant variation dependent upon individual or group self-consciousness or self-understanding. The intersection of all of these factors determines the matter of relative social status and invites considerable methodological caution in the interpretation and assessment of different kinds of evidence.

Age at Marriage

One important piece of the argument is the claim that Roman women married quite young and, by contrast, that Christian women married relatively later. This argument depends on Keith Hopkins' 1964-65 article arguing for the relatively early age of Roman girls. 34 Historian Brent D. Shaw has challenged Hopkins' findings in a carefully argued essay published ten years ago. 35 In this essay, Shaw distinguishes between various types of evidence, especially literary evidence (which he argues is "almost useless" for drawing any conclusions about the age of most [non-aristocratic, non-elite] Roman girls at marriage) 36 and epigraphic evidence (which he carefully situates within the broader cultural, economic, and geographical contexts in which it would have been [End Page 243] produced). An important dimension of Shaw's argument is the cautious reminder that even documentary evidence--in this case, inscriptions--is conditioned by cultural, social, and economic factors; the "epigraphic habit" of certain classes in Roman imperial society produced a critical body of evidence for demographic historical reconstruction, but does not necessarily lend itself to facile extrapolation to conclusions about "the Romans" in general. 37

Important for the comparative question under discussion here, Shaw also argues that the Christian evidence does not present a difference based on religious identity (as it has routinely been interpreted) but rather a difference based on class. Although the daughters of the aristocracy likely married in their early teens, other girls generally married later, in their mid- to late teens. As he puts it, "The Christian sample [of funerary inscriptions], far from being peculiarly 'Christian,' is simply evidence of a broader lower-class pattern of marriage that typified most of Roman society, even in the earlier 'non-Christian' centuries. The Christians were in fact merely continuing in a lower-class mode of family formation that was broadly typical of most men and women in the urban centres of the western Roman empire, a pattern which, so to speak, has been hidden from our historical view only because of the determinate impact of different social customs of funerary commemoration as they developed through time." 38

Christian Marriage

Stark makes other claims concerning the special character of marriage for Christian women. He says that they had greater choice of whom they married than their non-Christian counterparts (105), though he provides no evidence to support this claim. Once married, Stark argues, Christian women enjoyed a parity with their husbands unknown to their non-Christian peers. The early Christian texts used to sustain this claim are I Corinthians 7 and Ephesians 5.22 (which Stark attributes to Paul). "The symmetry of the [marriage] relationship Paul described [in I Corinthians 7] was at total variance, not just with pagan culture, but with Jewish culture as well," Stark writes (123). [End Page 244]

There are several difficulties with this reading of I Corinthians 7. Leaving aside whether "Jewish culture" can be considered a monolithic construction, this text has a long and complex interpretative history which does not make even a cursory appearance in Stark's discussion. The discussion of marriage and celibacy in I Corinthians 7 needs to be situated very carefully within the ideological and rhetorical contexts in which it was first produced. Although Peter Brown is quite right that this text has functioned in the history of Christian interpretation as a Rohrschach test, 39 it is also the case that recent commentators (despite their many differences) have tended toward a consensus that Paul's words here do not reflect a liberationist attempt to establish a counter-cultural utopian parity within Christian marriage. Rather, they function as a specific response to very particular social and religious challenges within the Corinthian community, where a significant group of Christians have withdrawn from marriage and sexual relations altogether, claiming an alternative religious authority in the process. 40 Despite some readers' tendencies to read Paul's letters as a simple description of early Christian consensus, there is little evidence to suggest that Paul's "moderate" stance won the day in Corinth.

Moreover, the verses which appear syntactically to provide for marital symmetry (I Cor 7.3-4: "The husband should give to his wife her conjugal rights, and likewise the wife to her husband. For the wife does not have authority over her own body, but the husband does; likewise the husband does not have authority over his own body, but the wife does" [NRSV]) rhetorically mask an embodied reality in which it means something quite different for a man to have authority over a woman's body than for a woman to have authority over a man's body. When read in conjunction with I Corinthians 11, where the order of creation and the customs of the churches are used to underwrite a hierarchical gender [End Page 245] arrangement, I Corinthians 7 cannot be read as straightforward testimony to companionate and egalitarian Christian marriage.

Ephesians 5.21-33, a deutero-Pauline text probably written by one of Paul's disciples, cannot reasonably be used to sustain the claim that Christian marriage was egalitarian and symmetrical. As but one example of numerous household codes in early Christian literature, this text seeks to establish the hierarchically arranged and orderly human household as a structurally analogous microcosm of the household of God. And although conservative Christian ideology has routinely tried to argue that "submission" and "love" are reciprocal (if not altogether symmetrical) gestures, feminist scholarship on the household codes offers a clear-eyed rendering of the stakes involved in framing household arrangements in these terms. 41

Christian Prohibitions against Divorce

Stark distinguishes Roman and early Christian women's experiences of marriage further, by pointing out the Christian prohibitions against divorce, which he argues elevates women's status. (This is a bit confusing, since earlier he argued that Spartan women had a higher status than their Athenian counterparts because, among other things, they could initiate divorce.) In Rome in the late republic and the empire, divorce was a fairly common practice, and both men and women could initiate it. Roman literary sources document a sentimental preference for marital stability, 42 and they sometimes rail against unjust divorces [End Page 246] initiated by men or frivolous divorces initiated by women. The economic and social impact of divorce could vary greatly, depending on several factors including class status, whether the marriage had been with or without manus, whether there were legitimate children born into the marriage, and so on. When wealthy women left marriages and took both their dowries and their own personal property with them, their husbands who had enjoyed the use of the dowries might see their standard of living decline considerably. Divorce undoubtedly weighed more heavily on women of more modest means, though there is substantial evidence that they still saw divorce as a viable choice since they initiated it in any case. Probably the harshest consequence for many women was the loss of their children to their husbands' families. 43 The point is that, like marriage, divorce could mean different things to different women in the Roman world. Early Christian prohibitions against divorce may have protected some women against unjust or trivial dismissal by their husbands, but they might equally have bound women to abusive or neglectful men. There is no reason to believe that Christian ideology concerning marriage and divorce, generally speaking, necessarily translated into higher status or better lives for average Christian women.

Widowhood

The ambiguity of Stark's arguments concerning marriage and divorce can be seen in his treatment of widowhood. Here, he argues that the Roman pressure--admittedly intensified by the Augustan marriage legislation--toward remarriage 44 lowered Roman women's status and, [End Page 247] conversely, that the tendency among Christians to valorize widowhood and encourage the widowed not to remarry increased women's status. So, is being married better or worse for women?

Stark's claim that Roman widows lost their inheritances to their new husbands is simply incorrect. 45 And the status of widows within Christian communities is both a complicated and a contentious matter. The evidence Stark extracts from the letter of Cornelius, bishop of Rome in the mid-third century, to Bishop Fabius of Antioch, reporting that "above fifteen hundred widows and persons in distress . . . are supported by the grace and loving-kindness of the Master [i.e., by the Roman church]" 46 is evidence neither for the high regard with which widows were held nor for the choices that Christian identity afforded them, but rather merely for the existence of a significant population of destitute women who needed to be supported by the church.

The church's relationship to widows was already in the early second century both complex and ambiguous, consisting (at best) of equal measures of charity and social control. Consider I Timothy 5.3-16, which makes careful distinctions between "real widows" and "the self-indulgent," and allows only a small segment of women whose husbands have died to be counted among the widows supported by the church. Rather than being held in unambiguously high esteem, widows seem quickly to have become, in some Christian communities at least, a population some church leaders thought required considerable surveillance and control. 47 [End Page 248]

Women's Religious Leadership

If women's status in Christian marriages and families was characterized by equal measures of complexity and ambiguity, all the more so was it complicated and contested in the church as a social formation. Amassing an idiosyncratic hodgepodge of ancient evidence while effacing critical nuances and collapsing important distinctions, 48 Stark argues that "there is virtual consensus among historians of the early church as well as biblical scholars that women held position of honor and authority within early Christianity" (109). The language used is imprecise but nevertheless significant. What do "honor" and "authority" mean here? Do they mean the same thing that they meant in early Christian texts?

Here, I turn again to Margaret MacDonald's impressive study, Early Christian Women and Pagan Opinion. In this work, MacDonald uses models derived from the social sciences to reconstruct the social circumstances of early Christian women. Two points are particularly relevant here. First, building on the work of numerous other biblical scholars and historians of early Christianity, as well as on models derived from comparative cultural anthropology, she examines the language of "honor" and "shame" in ancient Mediterranean societies, and explores the gendered character of this pair. She shows how male honor is sustained in public by the careful guarding of female shame in private. The scandal of early Christian practice for pagan critics was the fact that it blurred these clearly defined lines. Religion, traditionally a public/civic matter, entered domestic space in an unprecedented way in Christian observances. Meanwhile, Christian women, who (like any respectable woman in the ancient Mediterranean cultures who could afford to do so) ought to have been guarding their reputations and their "shame" at home, were apparently all too frequently visible in the world. Rather [End Page 249] than simply holding women in "honor," early Christian women's practice created significant social anxiety (both within the church and outside of it) because it complicated traditional notions of "shame." 49

Secondly, MacDonald argues that early Christian women's religious practice--practice that simultaneously evoked admiration and suspicion, both inside and outside Christian communities--did not accrue "authority" to women, but rather "power." 50 The distinction between power and authority is crucial to understanding both the nature of ancient women's religious activity and the causes of conflict within early Christian groups over women's participation and claims to religious experience. It is certainly no accident that early Christian women's religious expression often took the form of prophecy and ecstatic experience, since these modes of religiosity do not depend upon institutional sanction and support. 51 They are also modes linked, not to authority (which is mediated by office and institution), but to power (which is not). And, as a consequence, they are also modes of expression (and claims to power) that make the men who possess authority (through office) rather nervous indeed. This is why women's religious power--far from eliciting a unilaterally positive evaluation among early Christian men--created instead deeply felt ambivalences and, among other things, provoked very clear institutional attempts either to contain and domesticate it or to cast it out as heretical. 52 As MacDonald reminds her readers, along with much of their activity, Christian women's enthusiasms were viewed simultaneously with admiration and suspicion by outsiders and insiders alike. [End Page 250]

None of this is to diminish the historical claim that women were both prominent and powerful in some early Christian communities. However, neither prominence nor power were stable or unchallenged entities in these women's lives--nor, I should add, should we necessarily assume that either quality was only positively charged in the vocabularies of early Christian women. The premium placed in Christian ideology on virtues like submission, humility, weakness, and escaping notice must inject a healthy dose of paradox into any rendering of the meanings of power and prominence for the lives of early Christian women.

One final note on the question of women's attraction to Christianity in the early centuries is in order. Since Stark's argument is based so fundamentally on the consequences of heightened fertility, it is perhaps no surprise that he devotes no attention to the historical reality that a large number of women who were drawn to early Christianity were attracted principally to its ascetic forms. 53 Here, too, the question of women's status remains complex and multilayered. The ascetic life, especially the monastic life, may have provided women with a mode of escape from the rigors and dangers of married and maternal existence, with the prospect of an education and (in some cases) an intellectual life, and with access to social and economic power that would otherwise have eluded them. The ideology of asceticism was often double-edged, however, since it offered the possibility for a limited escape from the confines of gender by giving women access to masculine virtues, all the while leaving intact--indeed, reinscribing--the gender hierarchies that distinguished "feminine" and "masculine" in the first place. 54 [End Page 251]

Exogamous Marriage and Secondary Conversion

Stark argues that marriages between Christian women and non-Christian men were tolerated as a response to the oversupply of marriageable Christian women, and that intermarriage provided secondary converts in the form of non-Christian husbands as well as children who would be raised as Christians. As elsewhere in this chapter, Stark's portrait is more monolithic than the primary evidence and scholarly discussion necessarily supports. Stark argues, using evidence derived from testimonials concerning the stalwartness of Christians under persecution, 55 that "the high levels of commitment that the early church generated among its members should have made it safe for them to enter exogamous marriages" (114).

Church leaders, however, seemed to have worried quite a bit about the potential threat to the faith and well-being of Christian women who are married to non-Christian men. 56 Clement of Rome, Justin Martyr, Tertullian, and others, all expressed profound concern about the household discord and potential dangers produced by mixed marriages. The second book of Tertullian's treatise to his wife includes a sustained argument against intermarriage, claiming among other things: "If these things are so [referring to the preceding exegesis of I Corinthians 7.12-14], it is certain that believers contracting marriages with Gentiles are guilty of fornication, and are to be excluded from all communication with the brotherhood, in accordance with the letter of the apostle, who says that 'with persons of that kind there is to be no taking of food even.'" 57 Of course, such strident polemic should undoubtedly be read as [End Page 252] evidence that the practice may well have been fairly widespread in Tertullian's context, as Stark notes. But the point is that Tertullian and others did not see intermarriage as a benign endeavor, and though they hoped that it would win converts, they were less than persuaded that the risk involved was always a worthwhile one.

It is worth noting that the intermarriage of Christian women with non-Christian men presented a clearly perceived, substantive risk that the women would convert when the men in question were Jewish, as evidenced by the imperial legislation issued over the course of several decades in the fourth century. The first of these pieces of legislation, issued by Constantius II in 339, restores to employment in the imperial weaving industry Christian women who have converted to Judaism in the past ("who have been led by the Jews into the association of their turpitude") but declares that "Jews shall not hereafter unite Christian women to their villany; if they should do so, however, they shall be subject to the peril of capital punishment." 58

Michele Renee Salzman has argued, on the basis of the prosopographical and epigraphical evidence, that intermarriage between Christian women and pagan men of the upper classes was of extremely limited value in producing converts to Christianity. 59 Salzman shows that aristocratic women did not convert earlier than the men of their class. 60 Moreover, in summarizing her findings, she writes,

On the basis of this study, I do not see much evidence for the role of intermarriage as a means of conversion. On the contrary, intermarriage [End Page 253] between pagans and Christians was infrequent. There is a predominance of pagans marrying pagans and Christians marrying Christians; approximately 89 percent of the cases fit this pattern. . . . Of the six cases of intermarriage between pagans and Christians, pagans converted after marriage to a Christian spouse in only two instances. 61
She goes on to note that, on the basis of the evidence, "religion in late Roman society was transmitted across generations only when the parent and child were of the same sex." 62 Of course, Salzman's evidence is limited to the aristocracy and may not translate to the classes in which the vast majority of Christians were, in fact, found. Still, her study offers a salutary caution against assuming that women are natural primary converts and the unequivocal conveyers of culture within families in every time and place.

Christian Fertility

Men are made, not born, Christians.
Tertullian, Apology, 18.4

The claim that levels of fertility among Christians were appreciably higher than those among their non-Christian counterparts is central to Stark's argument. Here, we have virtually no demographic evidence to evaluate, and one significant study by Evelyne Patlagean of fourth-century Christian funerary inscriptions in Asia Minor presents a relative underrepresentation of females and a surprisingly high number of deceased children, militating against a simple agreement with Stark's thesis. 63

Stark argues that Christians undertook to live by the biblical injunction, "be fruitful and multiply" (116). This scriptural quotation possesses an enormously complex and contested history among early Christians who were as likely to invoke it as a prooftext for promoting the spiritual fertility of ascetic disciplines as an encouragement toward carnal procreation. Stark goes on to cite two texts from the second century--the Octavius by Minucius Felix and Tertullian's treatise To his Wife--to support the claim that "the differential fertility [between [End Page 254] pagans and Christians] was taken as fact by the ancients" (122), but he misreads them at precisely the relevant points.

The passage in question in the Octavius involves a defense against the accusation that Christians engage in incest. Contrasting Christian virtue to pagan licentiousness, Minucius Felix writes:

But we maintain our modesty not in appearance, but in our heart we gladly abide by the bond of a single marriage; in the desire of procreating, we know either one wife, or none at all. . . . We temper our joyousness [at banquets] with gravity, with chaste discourse, and with body even more chaste (divers of us unviolated) enjoy rather than make a boast of a perpetual virginity of a body. So far, in fact, are they from indulging in incestuous desire, that with some even the (idea of a) modest intercourse of the sexes causes a blush. . . . And that day by day the number of us is increased, is not a ground for a charge of error, but is a testimony which claims praise; for, in a fair mode of life, our actual number both continues and abides undiminished, and strangers increase it. 64
Stark suggests, in the way that he quotes the Octavius, that "the fair mode of life" primarily involves activities that increase fertility. Minucius Felix makes it clear, however, that the increase in Christian numbers involves conversion, not reproduction, and that the Christian virtue of celibacy is one critical feature of "the fair mode of life."

When he turns to Tertullian's treatise To his Wife in order to argue that Christians pursued "a lifestyle that could only result in comparatively higher fertility" (123), Stark takes the lines he quotes out of context and transforms their meaning in the process. The passage does not celebrate Christian reproduction, but rather argues that children are a burden and an encumbrance rather than a blessing. 65 Nor is this idea peculiar to [End Page 255] Tertullian: it is repeated frequently in the ascetic literature of the third and fourth centuries. Indeed, the relationship of early Christianity to reigning notions of the family was rather more complex than Stark allows. Very significant strands of anti-familialism were woven through the fabric of early Christian ideology, and it was the centrality of asceticism that distinguished Christianity from the dominant cultural values of both Judaism and Graeco-Roman culture far more than potentially eccentric marriage practices or the prominent presence of women in some religious settings. 66

Whether married Christians reproduced at a higher rate than non-Christians cannot be shown through recourse to the available evidence. It is certainly true that Christian writers polemicize against contraception, abortion, child-exposure, and infanticide; we cannot know for certain the degree to which Christian women abided by the restrictions these writers sought to impose. 67 (We are not surprised, in our own culture, to note the striking dissonance, for example, between the official teachings of the Catholic church on contraception, on the one hand, and the actual practice of Catholic women and men, on the other.) Moreover, there is considerable evidence for sexual abstinence within the context of marriage in the early church (albeit a controversial practice), whether fertility control was the main goal or a mere side effect of spiritual practice. 68 [End Page 256]

Stark himself, after working his way through his argument, must concede that the evidence is frustratingly insufficient. His concluding sentence--"All that can be claimed is that a nontrivial portion of Christian growth probably was due to superior fertility" (127-28)--retreats from the more daring claims that have been made earlier on in the chapter. But in the end, although it has long been clear that significant numbers of women were both attracted to early Christianity and occupied a prominent place in some of its forms, it has not been established that they contributed significantly as either missionizing wives or reproductive vessels to the Christianizing of the Roman Empire.

In the end, the portrait of early Christian women that emerges from Stark's book bears an uncanny resemblance to the ideal of Christian womanhood produced so forcefully by early Christian apologetic literature and moral discourse. Feminist scholarship has made two important interventions into this discussion. First of all, it has lent us the tools to analyze the rhetorical moves of idealizing literature; it has helped us to think about how rhetoric constructs certain realities and elides others. Secondly, it has offered a crucial set of counternarratives, historical reconstructions that help early Christian women come into view with greater clarity and higher levels of differentiation. The narrative of Christianization of course includes women, but it does so with greater ambivalence and a good deal more struggle than Stark's narrative would suggest.

Elizabeth A. Castelli is Assistant Professor in the Religion Department, Barnard College.

* A version of this essay was presented at the November 1997 Annual Meeting of the Society for the Scientific Study of Religion and the North American Association for the Study of Religion in San Diego. I wish to thank James B. Rives, Elizabeth A. Clark, and Roger Bagnall, for their comments on drafts of this article.

Notes

1. Danny Praet, "Explaining the Christianization of the Roman Empire: Older Theories and Recent Developments," Sacris Erudiri: Jaarboek voor Godsdienstwetenschappen 33 (1992-93): 7-119, surveys the field and includes an extensive bibliography at 111-19. Some of the more significant essay-length literature has been helpfully collected in Everett Ferguson, ed., Conversion, Catechumenate, and Baptism in the Early Church, Studies in Early Christianity: A Collection of Scholarly Essays, 11 (New York: Garland, 1993). Of special note, see also L. Michael White, "Adolf Harnack and the Expansion of Early Christianity," Second Century 5 (1985/86): 97-127; Augustinianum 27.1-2 (August 1987), special issue on La conversione religiosa nei primi secoli cristiani: XV incontro di studiosi dell'antichita cristiana, 8-10 maggio 1986; Michele Renee Salzman, "How the West was Won: The Christianization of the Roman Aristocracy in the West in the Years after Constantine," Studies in Latin Literature and Roman History, ed. Carl Deroux, Collection Latomus 217 (1992): 451-79; and Martin Goodman, Mission and Conversion: Proselytizing in the Religious History of the Roman Empire (Oxford: Clarendon Press, 1994), which examines the complex interaction of Christianity, Judaism, and Greco-Roman religions and philosophical schools.

2. The literature on women in the early Christian movement is far too extensive to cite fully here. The groundbreaking study for feminist work in the field is Elisabeth Schussler Fiorenza, In Memory of Her: A Feminist Theological Reconstruction of Christian Origins (New York: Crossroad Press, 1983). Subsequent work is surveyed in Elizabeth A. Clark, "Early Christian Women: Sources and Interpretation," in That Gentle Strength: Historical Perspectives on Women in Christianity, ed. Lynda L. Coon, Katherine J. Haldane, and Elisabeth W. Sommer (Charlottesville: University of Virginia Press, 1990), 19-35, and in Elizabeth A. Castelli, "Heteroglossia, Hermeneutics, and History: A Review Essay of Recent Feminist Studies of Early Christianity," Journal of Feminist Studies in Religion 10.2 (1994): 73-98. See also Susan Ashbrook Harvey, "Women in Early Syrian Christianity," in Images of Women in Antiquity, ed. Averil Cameron and Amelie Kuhrt (Detroit: Wayne State University Press, 1985), 288-98; Gillian Cloke, This Female Man of God: Women and Spiritual Power in the Patristic Age, A.D. 350-450 (New York: Routledge, 1995).

Worthy of special note is the recent work by Margaret Y. MacDonald, Early Christian Women and Pagan Opinion: The Power of the Hysterical Woman (New York: Cambridge University Press, 1996).

On conversion in particular, see Hagith Sivan, "Anician Women, the Cento of Proba, and Aristocratic Conversion in the Fourth Century," VC 47 (1993): 140-57; Jan Bremmer, "Why Did Early Christianity Attract Upper-Class Women?" in Fructus Centesimus: Melanges offerts a Gerard J. M. Bartelink a l'occasion de son soixante-cinquieme anniversaire, Instrumenta Patristica, 19, ed. A. A. R. Bastiaensen, A. Hilhorst, and C. H. Kneepkens (Steenbrugis: Abbatia S. Petri, 1989), 37-47; Anne E. Yarbrough, "The Christianization of Rome: The Example of Roman Women," CH 45 (1976): 149-65; Cordula Nolte, Conversio und Christianitas: Frauen in der Christianisierung (Stuttgart: Hiersmann, 1995). The attribution to women of this important role is usually traced back to Peter Brown's groundbreaking 1961 essay, "Aspects of the Christianization of the Roman Aristocracy," JRS 51 (1961): 1-11.

Two important recent studies, in different ways, challenge the received wisdom that women played a significant role in influencing aristocratic pagan men to convert: see Michelle Renee Salzman, "Aristocratic Women: Conductors of Christianity in the Fourth Century," Helios 16 (1989): 207-20, which marshalls the epigraphic and prosopographic evidence (discussed more in depth later in this paper); and Kate Cooper, "Insinuations of Womanly Influence: An Aspect of the Christianization of the Roman Aristocracy," JRS 82 (1992): 150-64, which examines the rhetorical strategies involved in attributing such influence to women and the ideological interests they may have served.

3. Rosemary Ruether, "Mothers of the Church: Ascetic Women in the Late Patristic Age," in Women of Spirit, ed. Rosemary Ruether and Eleanor McLaughlin (New York: Touchstone, 1979): 71-98; eadem, "Misogynism and Virginal Feminism in the Fathers of the Church," in Religion and Sexism, ed. Rosemary Radford Ruether (New York: Touchstone, 1979), 150-83; Jo Ann McNamara, "Sexual Equality and the Cult of Virginity in Early Christian Thought," Feminist Studies 3 (1976): 145-58; Ross S. Kraemer, "The Conversion of Women to Ascetic Forms of Christianity," Signs 6 (1980/81): 298-307; Aline Rousselle, Porneia: On Desire and the Body in Antiquity, trans. Felicity Pheasant (Oxford: Basil Blackwell, 1984); Elizabeth A. Clark, Ascetic Piety and Women's Faith: Essays on Late Ancient Christianity, Studies in Women and Religion, 20 (New York: Edwin Mellen Press, 1986); Elizabeth A. Castelli, "Virginity and its Meaning for Women's Sexuality in Early Christianity," Journal of Feminist Studies in Religion 2.1 (1986): 61-88; Virginia Burrus, Chastity as Autonomy: Women in the Stories of the Apocryphal Acts, Studies in Women and Religion, 23 (New York: Edwin Mellen Press, 1987); Anne Ewing Hickey, Women of the Roman Aristocracy as Christian Monastics, Studies in Religion, 1 (Ann Arbor: UMI Research Press, 1987); Jan Willem Drijvers, "Virginity and Asceticism in Late Roman Western Elites," in Sexual Asymmetry: Studies in Ancient Society, ed. Josine Blok and Peter Mason (Amsterdam: Gieben, 1987), 241-73; Peter Brown, The Body and Society: Men, Women, and Sexual Renunciation in Early Christianity (New York: Columbia University Press, 1988); Jane Simpson, "Women and Asceticism in the Fourth Century: A Question of Interpretation," Journal of Religious History 14 (1988): 38-60; Susanna Elm, "Virgins of God": The Making of Asceticism in Late Antiquity (New York: Oxford University Press, 1994); and numerous essays in Vincent L. Wimbush and Richard Valantasis, eds., Asceticism (New York: Oxford University Press, 1995), including Gillian Clark, "Women and Asceticism in Late Antiquity: The Refusal of Status and Gender," 33-48; Giulia Sfameni Gasparro, "Asceticism and Anthropology: Enkrateia and 'Double Creation' in Early Christianity," 127-46; and Averil Cameron, "Ascetic Closure and the End of Antiquity," 147-61.

4. Robert Markus, The End of Ancient Christianity (New York: Cambridge University Press, 1990), 28. See more generally the chapter, "Conversion and Uncertainty," 27-43.

5. Only five recognized experts on ancient women's history are mentioned at all in the bibliography to the book: Sarah Pomeroy, Gillian Clark, Ross Shepard Kraemer, Beryl Rawson, and Bonnie Thurston. The most important of these for the history of early Christian women is clearly Ross Shepard Kraemer, Her Share of the Blessings: Women's Religions among Pagans, Jews, and Christians in the Greco-Roman World (New York: Oxford University Press, 1992), and although this book is listed in Stark's bibliography, as far I could discern, it was not cited anywhere in the text of the chapter. The others are cited once or twice. Those who are absent could fill a very lengthy bibliography indeed; for a start, see the notes preceding and following this one (nn. 2, 3, 6, 7).

6. The project of historical reconstruction has been one of the dominant strands of feminist work on early Christianity. In addition to Elisabeth Schussler Fiorenza's work (cited above, n. 2) and Ross Shepard Kraemer (n. 5, above), see also Bernadette J. Brooten, "Early Christian Women in their Cultural Context: Issues of Method in Historical Reconstruction," in Feminist Perspectives on Biblical Scholarship, ed. Adela Yarbro Collins (Chico: Scholars Press, 1985), 65-91; Antoinette Clark Wire, The Corinthian Women Prophets: A Historical Reconstruction through Paul's Rhetoric (Minneapolis: Fortress Press, 1990); Karen Jo Torjesen, When Women Were Priests: Women's Leadership in the Early Church and the Scandal of their Subordination in the Rise of Christianity (San Francisco: Harper, 1993); Anne Jensen, Gottes selbstbewusste Tochter: Frauenemanzipation in fruhen Christentum? (Freiburg im Breisgau: Verlag Herder, 1992), among numerous other studies.

7. Again, the literature is too widespread for an exhaustive citation here. Among some of the more interesting work in this area, see Virginia Burrus, "Reading Agnes: The Rhetoric of Gender in Ambrose and Prudentius," JECS 3 (1995): 25-46; Averil Cameron, "Virginity as Metaphor: Women and the Rhetoric of Early Christianity," in History as Text: The Writing of Ancient History, ed. Averil Cameron (London: Duckworth, 1989), 171-205; eadem, "Early Christianity and the Discourse of Female Desire," in Women in Ancient Societies: "An Illusion of the Night," ed. Leonie J. Archer, Susan Fischler, and Maria Wyke (New York: Routledge, 1994), 152-68; Elizabeth A. Clark, "Ideology, History, and the Construction of 'Woman' in Late Ancient Christianity," JECS 2 (1994): 155-84; eadem, "Foucault, the Fathers, and Sex," JAAR 56 (1988): 619-41; eadem, "Sex, Shame, and Rhetoric: En-gendering Early Christian Ethics," JAAR 49 (1991): 221-45; among many others.

8. The evidence here has been helpfully collected and carefully analyzed by Margaret MacDonald in her recent book, Early Christian Women and Pagan Opinion, cited above (n. 2).

9. Origen, Contra Celsum 3.44 (trans. Henry Chadwick; Cambridge: Cambridge University Press, 1953).

10. Published in 1996, MacDonald's book unfortunately appeared too late for Stark to take its extremely nuanced and careful arguments and amassing of evidence into consideration.

11. Marcia Guttentag and Paul E. Secord, Too Many Women? The Sex Ratio Question (Beverly Hills: Sage, 1983).

12. See Gayle Rubin's classic theoretical treatment of this problematic in her "The Traffic in Women: Notes on the 'Political Economy' of Sex," in Toward an Anthropology of Women, ed. Rayna R. Reiter (New York: Monthly Review Books, 1975), 157-210.

13. P. A. Brunt, Italian Manpower, 225 B.C.-A.D. 14 (Oxford: Oxford University Press, 1971), 152: "I would also conjecture that it was just because women were in short supply that their status was improved, that marriages cum manu which placed them in the husband's power became less common, that they acquired de facto control over their own property, that they were free to divorce their husbands without cause and to recover their dowries, unless guilty of some marital offence."

14. Jane F. Gardner, Women in Roman Law and Society (Bloomington: Indiana University Press, 1986); Judith Evans Grubbs, Law and Family in Late Antiquity: The Emperor Constantine's Marriage Legislation (New York: Oxford University Press, 1995); Antti Arjava, Women and Law in Late Antiquity (New York: Oxford University Press, 1996).

15. Codex Theodosianus 16.2.20.

16. Emil Eyben, "Family Planning in Graeco-Roman Antiquity," Ancient Society 11-12 (1980-81): 5-82, esp. 12-19, 29-32, and the accompanying bibliography for primary and secondary sources; Ruth Oldenziel, "The Historiography of Infanticide in Antiquity: A Literature Stillborn," in Sexual Asymmetry: Studies in Ancient Society, ed. Josine Blok and Peter Mason (Amsterdam: Gieben, 1987), 87-107; Pierre Brule, "Infanticide et abandon d'enfants: Pratiques grecques et comparaisons anthropologiques," Dialogues d'histoire ancienne 18.2 (1992): 53-90; P. A. Brunt, Italian Manpower, 148-54; John Boswell, "Expositio and Oblatio: The Abandonment of Children and the Ancient and Medieval Family," AHR 89 (1984): 10-33; idem, The Kindness of Strangers: The Abandonment of Children in Western Europe from Late Antiquity to the Renaissance (New York: Vintage Books, 1988), esp. 53-137; Donald Engels, "The Problem of Female Infanticide in the Greco-Roman World," Classical Philology 74 (1980): 112-20; idem, "The Use of Historical Demography in Ancient History," CQ 34 (1984): 386-93; William V. Harris, "The Theoretical Possibility of Extensive Infanticide in the Graeco-Roman World," CQ 32 (1982): 114-16; idem, "Child-Exposure in the Roman Empire," JRS 84 (1994): 1-22; Tim G. Parkin, Demography and Roman Society (Baltimore: Johns Hopkins University Press, 1992), 95-105; Roger S. Bagnall and Bruce W. Frier, The Demography of Roman Egypt (Cambridge: Cambridge University Press, 1994), 151-53.

17. P. Oxy. IV.744 (1 b.c.e.).

18. Bagnall and Frier, Demography in Roman Egypt, 108, n. 50. Roger Bagnall, "Missing Females in Roman Egypt," Scripta Classica Israelica 16 (1997): 121-38, makes the important point that infanticide and exposure need carefully to be distinguished. Exposure often did not result in death, but in enslavement.

19. Bagnall and Frier, for example, note that the census data with which they work needs to be adjusted from the raw numbers since patterns of underreporting and overreporting, variations in location (village or metropolis), and other factors need to be accounted for. This is not to say that they do not acknowledge disproportionate sex-ratios, but rather that they emphasize the narrowness of conclusions available from the evidence.

20. Some of the more important Christian texts concerning exposure and infanticide include: Justin, Apologia 1.27, 29; Tertullian, Ad nationes 1.16; idem, Apologia 9.17-18; Clement of Alexandria, Paedagogus 3.3.21.5; idem, Stromateis 2.18.92-93, 5.14; Minucius Felix, Octavius 30.2, 31.4; Epistola ad Diognetum 5.6; Athenagoras, Supplicatio 35.6; Origen, Contra Celsum 8.55.

21. Emily Coleman, "L'infanticide dans le Haut Moyen Age," AnnalesESC 29 (1974): 315-35, translated and reprinted with updated notes as "Infanticide in the Early Middle Ages," in Women in Medieval Society, ed. Susan Mosher Stuard (Philadelphia: University of Pennsylvania Press, 1976), 47-70; eadem, Enfance abandonnee et societe en Europe, XIVe-XXe siecle, Collection de l'Ecole Francaise de Rome, 140 (Rome: Ecole Francaise de Rome, 1991); David I. Kertzer, Sacrificed for Honor: Italian Infant Abandonment and the Politics of Reproductive Control (Boston: Beacon Press, 1993); and Boswell, The Kindness of Strangers, cited above (n. 13).

22. Sorani gynaeciorum libri IV, ed. Ioannes Ilberg, Corpus medicorum Graecorum, IV (Leipzig: Teubner, 1927). English translation: Soranus' Gynecology, trans. Owsei Temkin with the assistance of Nicholson J. Eastman, Ludwig Edelstein, and Alan F. Guttmacher (Baltimore: Johns Hopkins University Press, 1956).

23. Gillian Clark, "Roman Women," in Women in Antiquity, Greece and Rome Studies, 3, ed. Ian McAuslan and Peter Walcot (Oxford: Oxford University Press, 1996), 36-55, at 39. This article appeared, originally, in Greece and Rome, 2nd series, 28 (1981): 193-212. In the addendum to the reprinted article, Clark writes: "recent research suggests that Roman contraceptive medicine was more effective than you might think" (55), citing John Riddle, "Oral Contraceptives and Early-Term Abortifacients during Classical Antiquity and the Middle Ages," Past and Present 132 (1991): 3-32.

24. For a collection of the ancient sources (sixth century b.c.e. to sixth century c.e.), see Enzo Nardi, Procurato aborto nel mondo greco-romano (Milan: Giuffre, 1971); Sheila K. Dickison, "Abortion in Antiquity," Arethusa 6 (1973): 159-66, provides a summary and cautionary review of Nardi. See also Danielle Gourevitch, Le mal d'etre femme: La femme et la medecine a Rome (Paris: Societe d'Edition «Les belles lettres», 1984), 195-216; John M. Riddle, Contraception and Abortion from the Ancient World to the Renaissance (Cambridge: Harvard University Press, 1991), esp. 1-107, and idem, Eve's Herbs: A History of Contraception and Abortion in the West (Cambridge: Harvard University Press, 1997). For an assessment, on the basis of census data from Roman Egypt, of the degree to which contraception was practiced, see Bruce W. Frier, "Natural Fertility and Family Limitation in Roman Marriage," CP 89 (1994): 318-33. I thank Roger Bagnall for bringing this article to my attention.

25. Clark, "Roman Women," 40.

26. Stark, 120, makes the argument with a certain amount of equivocation: "The very high rates of abortion in the Greco-Roman world can only be fully understood if we recognize that in perhaps the majority of instances it was men, rather than women, who made the decision to abort" (emphasis mine).

27. Richard I. Frank, "Augustus' Legislation on Marriage and Children," California Studies in Classical Antiquity 8 (1975): 41-52; Leo Ferrero Raditsa, "Augustus' Legislation Concerning Marriage, Procreation, Love Affairs and Adultery," Aufstieg und Niedergang in der Romischen Welt 2:13, ed. Hildegard Temporini (Berlin: de Gruyter, 1980), 278-339; Karl Galinsky, "Augustus' Legislation on Morals and Marriage," Philologus: Zeitschrift fur Klassische Philologie 125 (1981): 126-44; Gardner, Women in Roman Law and Society, 20; Susan Treggiari, Roman Marriage: Iusti Coniuges from the Time of Cicero to the Time of Ulpian (New York: Oxford University Press, 1991), 60-80, gives a thorough discussion of the impact of the Augustan legislation.

28. The evidence of Romans 16 is significant for the reconstruction of early Christian women's history, as the literature on the text has amply demonstrated: Elisabeth Schussler Fiorenza, "Missionaries, Apostles, Coworkers: Romans 16 and the Reconstruction of Women's Early Christian History," Word and World: Theology for Christian Ministry 6 (1986): 420-33; Caroline F. Whelan, "Amica Pauli: The Role of Phoebe in the Early Church," JSNT 49 (1993): 67-85; Bernadette J. Brooten, "Junia . . . Outstanding among the Apostles (Rom 16:7)," in Women Priests: A Catholic Commentary on the Vatican Declaration, ed. Leonard and Arlene Swidler (New York: Paulist Press, 1977), 141-44; Mary Rose D'Angelo, "Women Partners in the New Testament," Journal of Feminist Studies in Religion 6.1 (1990): 65-86; Elizabeth A. Castelli, "Romans," in Searching the Scriptures, vol. 2: A Feminist Commentary, ed. Elisabeth Schussler Fiorenza with the assistance of Ann Brock and Shelly Matthews (New York: Crossroad Press, 1994), 272-300, esp. 276-80. As important as this evidence is for the participation of women in early Christian movements, it is by itself not evidence for a disproportionate representation of women within the movement as a whole. Indeed, Romans 16 is distinctive for its inclusion of so many women's names, and scholars frequently remark that its status in this regard is quite anomalous.

29. Gesta apud Zenophilum, CSEL 26:185-97, at 187.4-10. The inventory reads as follows: "calices duo aurei, item calices sex argentei, urceola sex argentea, cucumellum argenteum, lucernas argenteas septem, cereofala duo, candelse breues aeneas cum lucernis suis septem, item lucernas aeneas undecim cum catenis suis, tunicas muliebres LXXXII, mafortea XXXVIII, tunicas uiriles XVI, caligas uiriles paria XIII, caligas muliebres paria XLVII, coplas rusticanas XVIIII." The relevant items are "eighty-two women's tunics, thirty-eight veils, sixteen men's tunics, thirteen pairs of men's shoes, forty-seven pairs of women's shoes, nineteen rustic coplae." This inventory is mentioned in W. H. C. Frend, The Rise of Christianity (Philadelphia: Fortress Press, 1984), 404, 429 n. 46, 458-60, 469 n. 82, and Robin Lane Fox, Pagans and Christians (New York: Alfred P. Knopf, 1987), 312.

30. Fox, Pagans and Christians, 308, reads: "Not only were Christian women prominent in the churches' membership and recognized to be so by Christians and pagans. . . ." (my emphasis); Stark, 98, reads: ". . . the predominance of women in the churches' membership was, as Fox reported, 'recognized to be so by Christians and pagans'" (my emphasis).

31. Adolf von Harnack, The Mission and Expansion of Christianity in the First Three Centuries, 2 vols., trans. James Moffatt (New York: G. P. Putnam's Sons, 1908), 2:73, cited at Stark, 98-99.

32. See Salzman, "Aristocratic Women," cited above (n. 2).

33. Evelyne Patlagean, "Familles chretiennes d'Asie Mineure et histoire demographique du IVe siecle," in Transformations et conflits au IVe siecle apres J.-C.: Colloque organise par la Federation Internationale des Etudes Classiques, Bordeaux 7. au 12. septembre 1970, Antiquitas Reihe, 1:29 (Bonn: Rudolf Habelt Verlag, 1978), 169-86. See also Antti Arjava, Women and Law in Late Antiquity (cited above, n. 14), 82-83.

34. Keith Hopkins, "The Age of Roman Girls at Marriage," Population Studies 18 (1964-65): 309-27.

35. Brent D. Shaw, "The Age of Roman Girls at Marriage: A Reconsideration," JRS 77 (1987): 30-46.

36. Stark, meanwhile, places a great deal of emphasis on the literary evidence, though paradoxically so, since he writes: "As to the histories, silence offers strong testimony that Roman girls married young, very often before puberty" (105). Historians rightfully give only very slight weight to arguments from silence, since the silence of sources can mean any number of undocumented things.

37. On methodological problems and alternative solutions, see Shaw, "The Age of Roman Girls," 33-39. See also Ramsay MacMullen, "The Epigraphic Habit in the Roman Empire," AJP 103 (1982): 233-46.

38. Shaw, "Age of Roman Girls at Marriage," 42. See also Brent D. Shaw, "Latin Funerary Epigraphy and Family Life in the Later Roman Empire," Historia 33 (1984): 457-97, esp. 483-84.

39. Peter Brown, Body and Society (cited above, n. 3), 48.

40. The literature is widespread here: the most important treatments of the text in recent discussions include Peter Brown's discussion, 44-57; Antoinette Clark Wire, The Corinthian Women Prophets (cited above, n. 6), 72-97; Margaret Y. MacDonald, "Women Holy in Body and Spirit: The Social Setting of 1 Corinthians 7," NTS 36 (1990): 161-81; Daniel Boyarin, "Brides of Christ: Jewishness and the Pauline Origins of Christian Sexual Renunciation," in A Radical Jew: Paul and the Politics of Identity (Berkeley: University of California Press, 1994), 158-79; Dale Martin, The Corinthian Body (New Haven: Yale University Press, 1995), 198-228. See also Elizabeth A. Castelli, "Disciplines of Difference: Asceticism and History in Paul," in Asceticism and the New Testament, ed. Vincent L. Wimbush and Leif Vaage (forthcoming).

41. Sarah J. Tanzer, "Ephesians," in Searching the Scriptures, vol. 2: A Feminist Commentary, ed. Elisabeth Schussler Fiorenza with the assistance of Ann Brock and Shelly Matthews (New York: Crossroad, 1994), 325-48; Kathleen E. Corley, "1 Peter," in Searching the Scriptures, vol. 2, 349-60 (especially 355-56, where Corley documents the role of the biblical household codes in contemporary Christian pastors' belief that women should endure domestic violence in imitation of Jesus' suffering for the Church); Linda M. Maloney, "The Pastoral Epistles," in Searching the Scriptures, vol. 2, 361-80; Mary Rose D'Angelo, "Colossians," in Searching the Scriptures, vol. 2, 313-24; Clarice J. Martin, "The Haustafeln (Household Codes) in African American Biblical Interpretation: 'Free Slaves' and 'Subordinate Women,'" in Stony the Road We Trod: African American Biblical Interpretation, ed. Cain Hope Felder (Minneapolis: Fortress Press, 1991), 206-31. See also D. L. Balch, Let Wives Be Submissive: The Domestic Code in 1 Peter, SBL Monograph Series, 26 (Chico: Scholars Press, 1981).

42. Suzanne Dixon, "The Sentimental Ideal of the Roman Family," in Marriage, Divorce, and Children in Ancient Rome, ed. Beryl Rawson (Oxford: Clarendon Press, 1991), 99-113.

43. Iiro Kajanto, "On Divorce among the Common People of Rome," Revue des etudes latines 47bis (1969): 99-113; Susan Treggiari, "Divorce Roman Style: How Easy and How Frequent Was It?" in Marriage, Divorce, and Children in Ancient Rome, ed. Beryl Rawson (Canberra: Humanities Research Centre, 1991; New York: Oxford University Press, 1996), 31-46; eadem, Roman Marriage, 435-82 (cited above, n. 27). See also Gardner, Women in Roman Law and Society, 81-95 (cited above, n. 14).

44. Roman society itself, as I have argued throughout this paper, was rarely of one mind on these questions. Remarriage was fiercely promoted for specific demographic reasons by Augustus, but many (especially upper-class) Romans continued to place a high cultural and social value on women who had only been married once. See Majorie Lightman and William Zeisel, "Univira: An Example of Continuity and Change in Roman Society," CH 46 (1977): 19-32. For a discussion which complicates the picture of univira in the Christian context, see Hagith Sivan, "On Hymens and Holiness: Opposition to Aristocratic Female Asceticism at Rome," Jahrbuch fur Antike und Christentum 36 (1993): 81-93.

45. See J. A. Crook, "Women in Roman Succession," in The Family in Ancient Rome: New Perspectives, ed. Beryl Rawson (Ithaca: Cornell University Press, 1986), 58-82; Richard P. Saller, Patriarchy, Property and Death in the Roman Family, Cambridge Studies in Population, Economy and Society in Past Time, 25 (Cambridge: Cambridge University Press, 1994), esp. Part III, "The Devolution of Property in the Roman Family," 155-224; Susan Treggiari, "Res: Property--Separation and Mixing," in Roman Marriage, 365-96.

46. Eusebius, Ecclesiastical History 6.43.11 (LCL translation).

47. The tension between widows and church authorities intensified as time progressed. See Charlotte Methuen, "Widows, Bishops, and the Struggle for Authority in the Didascalia Apostolorum," JEH 46 (1995): 197-213. See also Rosabianca Bruno Siola, "Viduae e coetus viduarum nella Chiesa primitiva e nella normazione dei primi imperatori cristiani," in Atti dell'Accademia Romanistica Costantiniana: VIII convegno internazionale (Napoli: Ed. scientifiche italiane, 1990), 367-426; and Jan Bremmer, "Pauper or Patroness? The Widow in the Early Christian Church," in Between Poverty and the Pyre: Moments in the History of Widowhood, ed. Jan Bremmer and L. van den Bosch (New York: Routledge, 1995), 31-57.

48. The discussion of I Corinthians 14.34-36 does not take account of the extensive literature on this passage and its problematic relationship to the rest of I Corinthians. See Wire, 229-32, for a summary of the debate and bibliography through 1990. The analysis of Romans 16.1-2 and the term diakonos quickly leapfrogs through several complex centuries of history of women's ministry and the meaning of the diaconate, collapsing important differences in the process. See Pier Giovanni Caron, "Lo status delle diaconesse nella legislazione giustinianea," in Atti dell'Accademia Romanistica Costantiniana: VIII convegno internazionale (Napoli: Ed. scientifiche italiane, 1990), 509-15; Abraham-Andreas Thiermeyer, "Der Diakonat der Frau: Liturgiegeschichtliche Kontexte und Folgerungen," Theologische Quartalschrift 173 (1993): 226-36; Whelan, "Amica Pauli: The Role of Phoebe in the Early Church," cited above (n. 27); P. Hofrichter, "Diakonat und Frauen im kirchlichen Amt," Heiliger Dienst 50 (1996): 140-58.

49. MacDonald, Early Christian Women and Pagan Opinion, 25-30, 69-70, 144-54.

50. MacDonald, Early Christian Women and Pagan Opinion, 41-47, 123-24.

51. Ross S. Kraemer, Ecstatics and Ascetics: Studies in the Functions of Religious Activities for Women in the Greco-Roman World, Ph.D. dissertation (Princeton University, 1976); Wire, Corinthian Women Prophets, passim, but see especially her collection of ancient sources on women prophets, 237-69; Christine Trevett, Montanism: Gender, Authority and the New Prophecy (Cambridge: Cambridge University Press, 1996).

52. The process of domestication can be first discerned in Paul's first letter to the Corinthians; see Antoinette Wire's reconstruction of the circumstances. The process by which women's religious practice becomes linked with the label "heresy," see Ross Shepard Kraemer, "Heresy as Women's Religion, Women's Religion as Heresy," in Her Share of the Blessings, 157-73; Virginia Burrus, "The Heretical Woman as Symbol in Alexander, Athanasius, Epiphanius, and Jerome," HTR 84 (1991): 229-48. An argument against the gender analysis of the orthodoxy/heresy problematic may be found in Paul McKechnie, "'Women's Religion' and Second-Century Christianity," JEH 47 (1996): 409-31.

53. In addition to the literature cited in n. 3, see also MacDonald, Early Christian Women and Pagan Opinion, 127-82. It should be noted that MacDonald presses the question of whether the traditional divisions historians have constructed between celibate women and married women should necessarily be so high or impermeable. Her intriguing suggestions invite further research.

54. Castelli, "Virginity and its Meaning for Women's Sexuality" (cited above, n. 3), esp. 78-88; Elizabeth A. Castelli, "'I Will Make Mary Male': Pieties of the Body and Gender Transformation of Christian Women in Late Antiquity," in Bodyguards: The Cultural Contexts of Gender Ambiguity, ed. Julia Epstein and Kristina Straub (New York: Routledge, 1991), 29-49. Verna Harrison has recently argued that, just as women's spiritual transformation was signaled by a shift in gender identification, so too was men's. However, the feminizing of men was part of a process of spiritual humbling, reinscribing the dominant notions of femininity as passivity and receptivity. See Verna Harrison, "Feminine Man in Late Antique Ascetic Piety," Union Seminary Quarterly Review 48 (1994): 49-71. Also see the argument that women's spiritual achievement, resulting in "manliness," was used rhetorically to shame men into higher virtue: Elizabeth A. Clark, "Sex, Shame, and Rhetoric: En-gendering Early Christian Ethics" (cited above, n. 7).

55. Stark refers here to Marcus Aurelius' comments concerning what Stark calls "the obstinacy of Christian martyrs." The text to which Stark may refer here (he does not provide the precise citation) is the Meditations, 9.3, where the explicit reference to Christians has been judged by scholars to be a scribal emendation. The actual quotation is not so much about obstinacy per se, but rather about the exuberance with which the people in question embrace their fate. Marcus Aurelius argues that readiness for death "must spring from a man's inner judgment. . . . It must be associated with deliberation and dignity and . . . with nothing like stage-heroics." The point seems to be that the exuberant people in question (Christians?) are making an unseemly spectacle of themselves.

56. Margaret Y. MacDonald, "Early Christian Women Married to Unbelievers," Studies in Religion/Sciences religieuses 19 (1990): 221-34. See also MacDonald, Early Christian Women and Pagan Opinion, 69-71, 113-14, 188-213, 244-48.

57. Tertullian, To His Wife 2.3 (ANF 4). See Marie-Therese Raepsaet-Charlier, "Tertullien et la legislation des mariages inegaux," Revue internationale des droits de l'antiquite 29 (1982): 254-63.

58. Codex Theodosianus 16.8.6 (Pharr translation, 467). For a broader discussion of conversion to Judaism during this period, see Louis H. Feldman, "Proselytism by Jews in the Third, Fourth, and Fifth Centuries," JSJ 24 (1993): 1-58, revised and reprinted in Louis H. Feldman, Jew and Gentile in the Ancient World (Princeton: Princeton University Press, 1993), 383-415, and see also 177-382 for discussion of the attractions of Judaism and for documentation on conversion to Judaism in earlier periods. See also Shaye J. D. Cohen, "Conversion to Judaism in Historical Perspective: From Biblical Israel to Postbiblical Judaism," Conservative Judaism 36 (1983): 31-45, and James Carleton Paget, "Jewish Proselytism at the Time of Christian Origins: Chimera or Reality?" JSNT 62 (1996): 65-103. The particular attraction of Judaism for women and patterns of conversion to Judaism by women invite further research; for a beginning discussion, see Kraemer, Her Share of the Blessings, 110-13, 121-23, and Tal Ilan, "The Attraction of Aristocratic Women to Pharisaism during the Second Temple Period," HTR 88 (1995): 1-33. For a helpful reminder that conversions in late antiquity were not all in the direction of Christianity, see Eugene Gallagher, "Conversion and Community in Late Antiquity," JR 73 (1993): 1-15, and Goodman, Mission and Conversion (cited above, n. 1).

59. Salzman, "Aristocratic Women," cited above (n. 2).

60. Salzman, "Aristocratic Women," 212.

61. Salzman, "Aristocratic Women," 214.

62. Salzman, "Aristocratic Women," 215.

63. Patlagean, "Familles chretiennes d'Asie Mineure," cited above (n. 33).

64. Minucius Felix, Octavius 31 (ANF 4). Emphasis added.

65. Tertullian, To his Wife 1.5: "Further reasons for marriage which men allege for themselves arise from anxiety for posterity, and the bitter, bitter pleasure of children. To us this is idle. For why should we be eager to bear children, whom, when we have them, we desire to send before us (to glory) (in respect, I mean, of the distresses that are now imminent) . . . ?" This passage is followed by the passage Stark quotes, a passage which, when read in context, clearly is intended as an occasion of irony. It is followed immediately by: "Burdens [children] which, finally are to us most of all unsuitable, as being perilous to faith! For why did the Lord foretell a 'woe to them that are with child, and them that give suck,' except because He testifies that in that day of disencumbrance the encumbrances of children will be an inconvenience?" This is no paean to the joys of parenthood or heightened Christian fertility! See Garth R. Lambert, "Childless by Choice: Graeco-Roman Arguments and their Uses," Prudentia 14 (1982): 123-38, which traces the continuities between Graeco-Roman and Christian patristic rhetorical topoi concerning the pains of childrearing.

66. In addition to the literature already cited at n. 3, see Elizabeth A. Clark, "Antifamilial Tendencies in Ancient Christianity," Journal of the History of Sexuality 5 (1995): 356-80; Carolyn Osiek, "The Family in Early Christianity: 'Family Values' Revisited," CBQ 58 (1996): 1-24; and Halvor Moxnes, ed., Constructing Early Christian Families: Family as Social Reality and Metaphor (New York: Routledge, 1997).

67. Andreas Lindemann, "'Do Not Let a Woman Destroy the Unborn Baby in her Belly': Abortion in Ancient Judaism and Christianity," Studia Theologica 49 (1995): 253-71, surveys the ancient literature and draws important attention to significant silences within the ancient texts on questions that have dominated contemporary American and western European debate on abortion within religious circles.

It should be noted that Stark's discussion of biblical and Christian opposition to birth control is marred by errors. The sin of Onan was not masturbation, but the refusal to marry his dead brother's wife. Romans 1.26 is not a prohibition of anal intercourse: see the impressive and exhaustive treatment of this text in Bernadette J. Brooten, Love Between Women: Early Christian Responses to Female Homoeroticism (Chicago: University of Chicago Press, 1996).

68. Evelyne Patlagean, "Sur la limitation de la fecondite dans la haute epoque byzantine," AnnalesESC 24 (1969): 1353-69. Recent scholarship on the numerous practices that travelled under the name of "spiritual marriage" in the early and medieval churches has been summarized by Dyan Elliott, "'A Place in the Middle': Intramarital Chastity as Theoretical Embarrassment and Provocation," in Spiritual Marriage: Sexual Abstinence in Medieval Wedlock (Princeton: Princeton University Press, 1993), 16-50.