Unmarriages

Traditional marriage was not the only option for couples in medieval Europe. Alternative forms of union could make lives precarious but also provided a degree of flexibility. The study draws on a wide geographical and chronological range of examples in order to illustrate local difference while bringing out broad patterns.

Unmarriages
Women, Men, and Sexual Unions in the Middle Ages

Ruth Mazo Karras

2012 | 296 pages | Cloth $49.95 | Paper $27.50
History | Women's Studies/Gender Studies
View main book page

Table of Contents

Introduction: Marriage and Other Unions
Chapter 1. The Church and the Regulation of Unions between Women and Men
Chapter 2. Unequal Unions
Chapter 3. Priests and Their Partners
Chapter 4. On the Margins of Marriage
Conclusion

Notes
Bibliography
Index
Acknowledgments


Excerpt [uncorrected, not for citation]

Introduction
Marriage and Other Unions

Histories of marriage are inevitably teleological: they put "marriage" as we know it at the center, and they evaluate all other forms of union in terms of that model. It is understandable, of course, that people want to know how an institution that is so important in contemporary society came to be the way it is. Given the contested nature of marriage today—between groups who think that it is primarily a bond between two people who love each other and should therefore be available to all such couples, and groups who think that it is primarily a way of creating a family environment in which to bear and rear children and should be limited to opposite-sex couples—tracing the history to see how we got to where we are can be very useful. Of course, history may be more relevant to those who base their claims on "tradition" than to those who argue that cultural change necessitates changes in marriage as well. However, even people who do not wish to see a return to "traditional marriage" can benefit from understanding the history of the institution and alternatives to it, if only to be able to identify where claims from historical truth are distorted or tendentious. Only by historicizing marriage can we see the inherent illogic of claims that there is only one "real" form.

Stephanie Coontz's Marriage, a History is a good example of a book that looks at marriage in the past with an eye not to the unchanging elements of "tradition" but of what changed and why over the centuries. But it is still concerned with the present, using the past to make clear what is distinctive about today's marriages. Other work, on medieval marriage specifically, has regarded marriage largely as a legal contract or as a sacrament. It remains both these things for many people, and since many of both the theological and contractual elements of marriage can be traced to the medieval period, it makes sense to study them, especially for those who find the medieval views important and binding as precedent.

But if the history of marriage is the history of how we got to where we are today and focuses on those elements that are seen as important today in constituting marriage—the exchange of a binding vow, the blessing by a clergyman, the sexual union—we lose sight of the elements that fell by the wayside. History's blind alleys—the customs and practices that did not continue, or that continued but were not deemed important or mainstream—were a part of the medieval experience as much as those aspects that became the roots of contemporary institutions. If we consider marriage a legal contract, we will follow one trajectory in tracing its history; if we consider it a sacrament, another trajectory; if we consider it a personal commitment, a third; and if we consider it an avenue for channeling sexual activity, yet a fourth. But even if we could agree on what marriage is today, looking for the roots of that institution would exclude relationships that were of central importance in their own societies but look very different from modern Western marriage.

The question "What is marriage?" is being asked today in a way that it never has before. Biblical texts that are normative in Christian and Jewish traditions take marriage as a given and do not explicitly define it. There are some contexts in which medieval people discussed the question of what made a marriage, but for the most part, the line between what was marriage and what was not was not sharply drawn. I do not propose to sharpen it, to impose categories on medieval society that it did not impose itself, but to demonstrate its fuzziness and the different ways in which various sexual unions were understood by different groups of people and defined in different discourses. But we need to remember that even though medieval people did not always define the line sharply between what was marriage and what was not, they persisted in the belief that such a line did exist. There was indeed a variety of statuses. Cordelia Beattie suggests that we should envision a variety of statuses approaching marriage as a continuum of singleness. It could also be seen as a continuum of types of pairings. However, either way, a continuum implies that there are endpoints—in this case, marriage and singleness—and that other types of arrangement are in a straight line, closer to one end than the other. A more accurate mathematical image would be a multidimensional graph that would include axes of formality, sexual exclusivity, sharing of resources, emotional involvement, dissolubility, and so forth. Yet while people may have been quite content to leave vague where a particular union fell among the options, both the law and cultural expectations were constructed such that firm judgments sometimes had to be made.

Some scholars working on the history and anthropology of marriage define it in a broad way so as to make the definition valid cross-culturally. This means going well beyond contemporary definitions. In the modern United States, it is clear what is a marriage and what is not: each state defines which individuals are authorized to perform a valid marriage and what they are required to do to make it valid. Even so, of course, there is room for disagreement: a marriage recognized as valid by one state may not be recognized by another, and a marriage deemed valid by all fifty states may not be deemed valid by particular religious institutions (as, for example, between a Jewish man and a Catholic woman). None of the state laws creates a cross-culturally valid definition, and even to say that one must have gone through some particular ritual or satisfied some state- or church-established requirement is not true across all cultures. On the other hand, if we make the definition so broad as to encompass multiple cultural traditions, as with anthropologist Kathleen Gough's classic definition, "a relationship established between a woman and one or more other persons, which provides that a child born to the woman under circumstances not prohibited by the rules of the relationship, is accorded full birth-status rights common to normal members of his society or social stratum," it may be so broad as to not be very useful. Many societies do not distinguish among children according to the status of their parents: does that mean that any fertile sexual relationship is a marriage, under Gough's definition?

In other ways, a definition based on the status of the children may be too restrictive: it privileges reproduction in a way that may not reflect all societies and may also obscure a situation in which a woman has fewer rights than another woman who enters another kind of union, even though their children may have the same rights. A more useful approach to a cross-culturally valid definition might be that whatever pair relationship is most privileged in a given society is ipso facto what that society considers marriage. But there might be several different types of unions in a given culture that scholars might choose to call "marriage." Most of these would likely be sexual unions, but that does not necessarily have to be the case: a marriage in the Middle Ages, at least under canon law that emerged in the twelfth century, did not have to be consummated in order to be valid. Nevertheless, medieval Europe shared with other societies a close association among sexual activity, fertility, and long-term unions. The absence of contraception meant that the onset of sexual activity was often closely followed by the onset of childbearing, and the decision of a couple about whether to recognize a permanent bond often was prompted by the arrival of offspring. Even spiritual marriages that were metaphorical unions with Christ were described as fertile in their production of spiritual fruit. In practical terms, the presence of offspring is useful to the historian in determining whether a sexual relationship can be considered a long-term union rather than a casual liaison: while pregnancy can result from the latter, for the man to acknowledge the child as his implies a confidence in his paternity that suggests a meaningful bond between the partners.

The problem of studying pair bonds in medieval Europe is both less and more complex than in cultures more distant from our own. We can identify particular terms that may quite comfortably be translated as "marriage" because the system of laws regulating marriage in medieval Europe is ancestral to our own and there is so much that seems familiar in it. But this is precisely where the danger of teleology enters in: we focus on the aspects that led to contemporary arrangements, and we risk falling into a history of what we today call marriage rather than the full variety of pair bonds.

This book focuses on sexual unions between women and men, thus omitting two categories of relationship on which there has been considerable recent scholarship. There were undoubtedly many cases, if not of the ritualized same-sex partnerships suggested by John Boswell and doubted by other scholars, then simply of male couples living together. There were also women who chose to preserve their chastity and enter a spiritual union with another person or with Christ. An unknown but considerable number of medieval people engaged in one of these two types of union instead of, or alongside of, marriage. Here, however, we consider the majority of people who lived as couples, those involving a man and a woman in a sexual relationship. To get a full picture of the history of male-female unions—the dead ends as well as the characteristics that continued to the present day—we need to look at a variety of unions, some of them considered marriage, some of them not, and some of them very much in question. There is excellent scholarship available on various aspects of medieval marriage—James Brundage on legal aspects; Dyan Elliott, David d'Avray, and Christopher Brooke on theology; Georges Duby on contested definitions of marriage in one particular geographical and chronological setting—but most of it does not look at the context of a variety of pair bonds. This book begins from the assumption that sexual pair bonds between women and men were a dominant social form in medieval Europe. They were undoubtedly more common in lived experience—and arguably more important in the cultural imaginary—than same-sex unions or lifelong celibacy. Even while avoiding privileging only those types of unions that led to modern marriage, a history of medieval pair bonds inevitably comes circling back to medieval marriage. The important question to ask is not "What unions should modern scholars and social analysts consider to have been marriages?" but "Where, and why, did medieval people draw the line between what was, and what was not, marriage?" The line was often important because it determined the transmission of property between and within families. My goal is to analyze pair bonds without privileging marriage, while still recognizing that medieval people did, in fact, privilege marriage.

Here we look at a range of sexual unions in the European Middle Ages, some of which people at the time considered formal marriage, some that they clearly did not, and still others that were at the margins. Those whose official nature was disputed or denied, or never proposed in the first place, are nevertheless part of the history of pair bonds. The book will consider, in the later sections of this introductory chapter, models that medieval Europe received from its biblical, classical, and "Germanic" heritage. It will then turn in Chapter 1 to the process by which the church effectively imposed its control over marriage: the variety of types of unions in the earlier Middle Ages, including disputes (both at the time and among modern scholars) as to whether they qualified as marriage or not, the elaboration of the consent theory, and the triumph of a centralized regulation of marriage by the papacy. Chapter 2, dealing with the period after the church managed to articulate and enforce a fairly consistent set of definitions of types of heterosexual union, looks at groups who could not legally marry each other, such as Christians and Jews, or free and enslaved people, or could not make their sexual relationship socially acceptable as marriage because their status was so different, as well as at individuals whose formal marriages were determined by political considerations and formed supplementary unions. One particular group who was unable to marry from the central Middle Ages onward, the higher clergy, forms the subject of Chapter 3. Chapter 4 looks at the everyday life of the people of fifteenth-century Paris, providing insight into various reasons that couples could not marry—including those trapped in unhappy marriages but to whom divorce was not available—and also examining couples who could have married but chose not to. The book examines a series of moments in medieval history rather than providing a chronologically and geographically complete account.

Throughout the variety of different cases from different times and places across the western European Middle Ages that this work considers, we find a number of common threads. One is that the legal and social status accorded a union—including whether it was considered a marriage—was determined in the first instance not by what legal or contractual arrangements were made but what the status (social or legal) of the partners was. This is a bit jarring to someone working with a twenty-first-century model of marriage, in which as long as the partners are of age and mentally capable and not already married to others, they may marry in a civil marriage (various religions, of course, have their own rules). But we still have vestiges of a system for which the legal status of a union depends on who the parties are: for example, in most U.S. and European jurisdictions as of this writing, the parties to a valid marriage must be a woman and a man. In the Middle Ages, not only the sex of the parties but also their religion, clerical status, freedom or unfreedom, family prominence, and previous sexual behavior were relevant. Because clandestine marriages, performed by the partners themselves rather than by a priest at church, were valid, and parish records were not kept, it is not possible now and was not possible at the time to say that a particular union was not a marriage because of the lack of a formal ritual. Nor was a union legally not a marriage if there was no dowry or other transfer of wealth. Of course, most unions that were intended to be taken as marriages were publicly performed and did involve a transfer of wealth, such that any relationship that did not involve these elements was likely to be considered a different type of union. What led parties to choose to do without public ceremony and dowry, though, was their social standing. If both were poor, there might be no dowry available; if the woman were of lower social status than the man, the dowry would likely not be sufficient (although a large dowry could make it possible for a merchant's daughter to marry into the landed nobility or gentry). If the parties were of different religions, or one had taken religious vows, there could be no ceremony, since the only ceremonies available were religious. And, of course, if one partner was married already, even with a ceremony and a dowry it was not considered a marriage.

Tying the status of the union to the relative social status of the partners had an especially harsh effect on women. In partnerships between two people of different status, it was usually the woman who was of lower status, although this was less so in the case of formal unions, in which an elite daughter could be married to a promising young man. But being involved in a sexual relationship that was not considered a marriage could be deleterious to a woman's reputation (more than to a man's), thus creating a vicious cycle. It is a commonplace that a man's honor in the Middle Ages could depend on a wide variety of things—his reputation for honesty, his physical and military prowess, his control over his household—but a woman's depended only or mainly on her sexual reputation. Yet because the status of a union depended on the relative social status of the parties, a woman of lower social status was likely to be considered a concubine or prostitute rather than a wife. Social dishonor thus was intimately tied to sexual dishonor. The status of concubine, in other words, was less than honorable, not only because it was sexually suspect but also because it implied that the woman was of low rank. This was true at any level of society: from the workingman who had a sexual relationship with his slave to the great lord who had a mistress from the lower gentry, a woman of lower rank was not respectable enough to marry. The two types of dishonor fed on each other, and it is not possible to say which came first, the sexual suspicion or the social disadvantage. Even though nonmarital unions were seen as routine and normal in many segments of society, this attitude was compatible with one that saw the women involved as dishonorable. (Unmarried men tended to get a pass on the honor issue.) Since women of higher status rarely married men of lower status, the result was that the only honorable status for a woman was a relationship with a social equal.

The pattern of higher-status men in sexual partnerships of various kinds, but mainly not marriage, with women of lower status is hardly limited to medieval Europe. Many of the unions described in this book will look very familiar to students of other historical periods; indeed, there may be as much similarity between patterns in one part of the Middle Ages and those in the nineteenth century than between different periods or places in the Middle Ages. The importance of the medieval moments I describe is that they came during an era that saw the emergence and elaboration of Christian marriage; thus the relationships were in negotiation and counterpoint with marriage rather than simply outside it. The range and variation of unions comes as a surprise to those who regard the church as having been dominant (and unified) in the Middle Ages. At each particular moment, local context mattered greatly. Men in many societies have had sexual relations with the women who worked for them, but only in some were the women treated as property with no volition; unmarried couples have had ongoing sexual relations in many societies, but only in some could those unions be adjudged to create a lifetime bond without any formal ceremony.

Medieval people would have recognized the sexual unions discussed in this book as resembling marriage because they were exclusive, or long-lasting, or involved the formation of a joint household. Given the nature of the sources, it is not always possible to tell just how casual or committed a given union was, but there was a rich variety of arrangements that were assimilated or analogized to marriage or contrasted with it to various degrees. It is difficult to find appropriate language to refer to the broad range of unions I talk about here. "Quasi-marital union" was my first attempt, but it assumes marriage as the model that other unions only approached, an assumption I do not wish to make. "Domestic partnerships" would leave out people who were in long-term relationships but did not live together. "Concubinage" had very specific meanings at various points in the Middle Ages, so cannot be used as a general term. "Heterosexual unions" works if one takes "heterosexual" as a synonym for "opposite-sex" or "involving a man and a woman" but not if one thinks of "heterosexual" as describing the people involved; the concept of "heterosexuality," like that of "homosexuality," is anachronistic for the Middle Ages. "Opposite-sex unions" describes what I want to talk about, as a parallel to John Boswell's "same-sex unions," though I have tried to minimize its use because of its awkwardness.

Many couples in the Middle Ages could not marry or chose not to marry when they could have. "Many" is a vague word, but as with so much of medieval demography, the sources just do not exist to provide accurate numbers. Scholars have assumed that the formally married were the majority, and there is no evidence to prove otherwise. But marriage could be a social expectation without being the only alternative, or even an option, for many people; celibacy and virginity were held up as social goods and high goals without any realistic hope that they would be widely adopted, and formal marriage may have functioned in a similar way. We often do not know how formal the unions of the poorer strata of society were, when they were not accompanied by the property exchanges that created new households among the more affluent. This book is in part an exploration of that question, but it cannot answer it quantitatively.

The book also demonstrates that different types of pair bonds as alternatives to marriage did not emerge full-blown from the social revolution of the 1960s and 1970s. I do not suggest that the variety of unions into which medieval people entered were in a direct line of development with contemporary ones, but rather that the process of pair bonding in the Middle Ages was more complicated than a history of marriage alone would suggest. The Latin Middle Ages are often viewed as the time when the teachings of early Christianity were codified and institutionalized—to good or to bad effect, depending on one's religious views. When people speak of "traditional marriage," they do not mean a world where plural marriage was common, as in the Hebrew Bible, or where divorce was easy, as in ancient Rome, or where marriage was suitable only for those who could not be continent and unnecessary for believers since the kingdom of God was soon to come, as in the New Testament. They mean a world in which marriage is monogamous, indissoluble, and nearly universal, a world that they imagine to have existed in the Middle Ages. This book will demonstrate that, while those were indeed the norms and expectations of marriage, the lived experience was considerably messier.

Because the contemporary movement for marriage equality for same-sex couples first started me thinking about the history of marriage and other pair bonds and the ways that they have been reinvented over the centuries, it is ironic that the book does not have much to say about unions between two men or two women. There are plenty of cases we can point to of two men or two women sharing living quarters. We do not know what their physical relationship was, but we do not demand proof of sexual relations to assume that two people of opposite sex living together were, in fact, a couple, so why should we demand it for two people of the same sex? Nevertheless, we do not find much evidence about the domestic lives of these same-sex couples. For all the medieval texts that have been "queered" by modern literary critics, for all the tales of cross-dressing, and for all the claims of a type of same-sex ritual that has some resemblances to marriage, medieval writers rarely raised the possibility that two men or two women could marry, even to reject it. When they did mention it, it was only for the sake of a very medieval systematic completeness, not because they were really considering it. Jean Gerson, the fifteenth-century Paris theologian, wrote that the argument that there could be a marriage between two men or two women was merely a frivolous objection to the (by his time universally accepted) theory that marriage was created by the consent of the parties. Same-sex couples were not on the boundaries in the same kind of way as opposite-sex couples: there was no confusion between their unions and legal marriage. They were rarely hauled into church courts and fined for their relationships; indeed, they are rarely mentioned anywhere. But this book's limitation to opposite-sex unions should not be taken as in any way a denial that same-sex couples entered into many of the same kinds of domestic partnerships, and no one will be more delighted than I if another scholar can demonstrate that I was too pessimistic about the evidence for this.

The sources are nearly silent on same-sex unions, but they are not much better on the women involved in opposite-sex unions. It is much easier to know what these unions meant for the men and their families than for their female partners. Throughout this book are scattered portraits that explore individual cases of women involved in various kinds of domestic partnerships, some well-known historical figures, some fictional, some real but no more than a name to us. These portraits, while based, like the rest of the book, on available sources, are set apart because they are slightly more speculative: determining what women may have been thinking or feeling in different situations can be a chancy task, given the level of information we have. Through the portraits, I hope to balance somewhat the masculine-inflected provenance of the sources. Nevertheless, some aspects are simply unrecoverable. Of the twelve women represented in the ten portraits, Heloise is the only one who expresses what we would recognize as love. That does not mean that it was absent in most medieval unions, but we must remember that, like marriage, love is not a human constant; expectations and experiences of different kinds of love were shaped by the societies in which people lived, and the contemporary notion that love is the main reason for couples to choose to form a domestic union does not necessarily apply.

The medieval women about whose feelings of love we have the most evidence are those whose love was for the divine. A great deal of the scholarship on medieval women over the last several decades has been on women who rejected marriage for spiritual reasons and focused their lives on a union with Christ or a relationship with a holy man. Either of these types of union could be, and was, discussed in marital language. These women underscore how important marriage was as a structuring image or metaphor for women's lives, even women who chose not to enter it. This book, however, focuses on unions that were sexually active and that filled the social and economic and not just the spiritual and emotional space that might otherwise be occupied by marriage. Women often chose spiritual bonds as resistance to marriage; the women discussed in this book were, for the most part, not resisting but existing in parallel with marriage.

Biblical Sources for Medieval Understandings of Unions between Women and Men

Religious as well as secular traditions shaped medieval attitudes toward sexual unions. The church had jurisdiction over Christian marriage at least since the twelfth century, and had attempted to claim it earlier. Jewish and Muslim courts had jurisdiction over marriage in their own communities. The Bible was considered by the church fathers and by medieval ecclesiastical authorities as authoritative. The Hebrew Bible, which medieval Christians called the Old Testament, however, had surprisingly little to say about marriage and other unions. It simply assumed the existence of what we can translate as "marriage" but did not define it, nor in the midst of detailed lawmaking did it give guidance on how it should be formed. Its narratives depict alliances being formed through one man giving his daughter to another, but these unions were not religious in nature, nor was it divine consecration that made a marriage. Medieval writers took as a precedent the scene in the Book of Ruth (4:11) where all the people bless the marriage of Ruth and Boaz. But the Bible never required it, remaining remarkably unconcerned about what rituals created a valid union. Marriage was mentioned a good deal less in the law-code portions of the Bible than in contracts of the same era or in Mesopotamian law codes, and medieval people took their information mainly from the narrative portions of the Bible, which discussed unions only when they were important to the story, not in a theoretical or prescriptive way.

St. Jerome, the fourth-century translator of the Latin Vulgate version of the Bible widely used during the Middle Ages, rendered ishah, the word for a woman given to a man as the prospective mother of his children, as uxor. Like the modern German Frau or the Old English wif, ishah had a semantic field that included "woman" as well as "married woman." In translating ishah as uxor ("wife") instead of mulier ("woman") in certain contexts, Jerome was making a choice, interpreting a term that had some possible ambivalence as one whose denotation was more specific. It was "Adam and his wife," not "the man and his woman" (another possible translation of the Hebrew ha-adam v'ishto) who were not ashamed of their nakedness (Gen. 2:25). Jerome's choice was likely correct in the context of the way the word was understood in his era, both by churchmen and rabbis, and a broader meaning in ancient Hebrew does not matter here. The Septuagint, the Koine Greek translation of the Hebrew Bible, has gyne, which also can mean either woman or wife.

Medieval people found guidance for the conduct of their own unions in the stories of the patriarchs and matriarchs. Old Testament stories in which the relationship could not easily be assimilated into a Christian idea about what a sexual union between a man and a woman should be are key places to look for medieval interpretations and understandings of marriage and its alternatives. For people in the Middle Ages, the larger-than-life stories of revered patriarchs were true as literal history, as prescriptive sources delineating appropriate behavior and as pointers to a deeper metaphorical or moral meaning. These patriarchs had more than one ishah. Scholars may doubt now whether polygamy was as common as it was depicted in the Hebrew Bible, but in the Christian Middle Ages, the Bible was the historical source par excellence as well as providing normative guidance and spiritual meaning.

Abraham was often understood in the Middle Ages as having been a polygamist, that is, having more than one wife, not just more than one sexual partner. Genesis 16 described the barrenness of Abram's wife (uxor) Sarai. She gave her ancilla (servant or slave; Hebrew, shifḥah) Hagar to her husband as an uxor to bear him a child. The two women were clearly not of the same status: Sarai could dictate that Hagar would be Abram's sexual partner, and after Hagar became pregnant and behaved with a lack of respect toward Sarai, Sarai had the ability to mistreat her so that she ran away. However, though Hagar belonged to Abram's wife as a slave, she also belonged to him as a wife in much the same way that Sarai did, and her child was his child. In fact, after Sarah (the renamed Sarai) bore a child herself, she asked Abraham not to allow Hagar's son Ishmael to inherit along with Sarah's son Isaac. The implication is that otherwise he would have inherited alongside him, although not first as an eldest son. Even though under most medieval legal systems, the son of a married man and his slave could not inherit, the story assumed that he might have. In both the Jewish and later in the Muslim tradition, Hagar's son Ishmael became the ancestor of the Arabs. His descent from Ibrahim (Abraham) became very important to the Muslims (Qur'an 2:127). In Muslim tradition, Hagar was not just a slave: though not named in the Qur'an, she appeared in the post-Qur'anic tradition as an enslaved princess, or a free and royal woman given to Ibrahim as a wife.

The stories of plural marriage among the ancient Israelites, particularly that of Abraham, Sarah, and Hagar, posed a conundrum for early and medieval Christians. Modern scholars long explained the Near Eastern background to the Abraham/Sarah/Hagar story in terms of the Babylonian Hammurabi's Laws, in which a man with an infertile wife could marry a second wife only if his wife did not give him a slave instead. A slave who bore her master's child could not be sold, but she did not assume equality as a wife either. Her sons, however, if acknowledged by the father, had equal inheritance rights with the sons of the first wife. The distinction between the two women was in their social circumstance, not on a particular form of ritual that they entered into. But medieval interpreters, although they distinguished between the status of the two women, considered them both wives. Unwilling to take plural marriage as a prescription for behavior, they had to explain why the Bible depicted patriarchs behaving in ways that went against God's wishes. A modern historian might simply say that they were living at a time with different customs and standards. Indeed, the Glossa Ordinaria, the biblical commentary that was widely known and used throughout Europe from the twelfth century on and provided the closest thing to a standard line in medieval theology, said something very close to that. "Abraham was not an adulterer if he was joined to a slave while his wife was alive, because the law of the Gospel about one wife was not yet promulgated." St. Augustine had written something similar in his On Christian Doctrine: "On account of the necessity for a numerous offspring, the custom of one man having several wives was at that time blameless," and the patriarchs fathered these multiple children without lust.

But the Glossa also treated the passage allegorically, following in a long tradition that began with Paul of Tarsus, who wrote in his letter to the Galatians (4:21-31) that Hagar's son was born according to the flesh but Sarah's son according to the promise. Hagar represented the earthly Jerusalem and Sarah the heavenly Jerusalem. By the third century, Christian writers were following Paul's lead and explaining Hagar as the fleshly synagogue and Sarah as the spiritual church, among other meanings of the story. The Glossa's interpretation here, attributed, like the previous passage, to the ninth-century theologian Hrabanus Maurus, held that Ishmael, expelled from his father's household in favor of the second-born child, represented the Jewish religion that was displaced by its more favored sibling Christianity. Nevertheless, the allegorical meaning did not drive out the literal one: the idea that the patriarchs had plural marriage was not forgotten through the Middle Ages.

Indeed, although Sarah and Hagar were not of equivalent status, Abraham's grandson Jacob was described in Genesis as being in unions with two women who were of the same status—sisters, in fact. Again, Jerome introduced terminology not found in his Hebrew original: Jacob wished to marry the younger sister, Rachel, but her father said that "it is not our custom to give the younger in marriage [nuptias] first" (Gen. 29:26), where the Hebrew says "to give the younger before the elder," with no word specifically denoting marriage. Jacob's twelve sons, the founders of the twelve tribes of the Israelites, were born to four different women. Besides Rachel and Leah, the others were Zilpah and Bilhah, whom Jerome called ancilla and famula, both translating shifḥah. This Hebrew term was used elsewhere as the female counterpart to eved, "slave," but scholars have suggested that it was used especially of a woman in a sexual relationship with her owner, or her owner's husband. Jerome chose two different terms to render it into Latin, the former denoting (in Jerome's time, anyway) unfree status and the latter servant status but not necessarily unfree. Neither had a particularly sexual connotation beyond the fact that serving women were always vulnerable to sexual advances from their owners or employers. The Hebrew says that the two women, slaves of Rachel and Leah, respectively, were given to Jacob as nashim (plural of ishah) (Gen. 30:4 and 30:9), but Jerome translated the first as dedit in coniugium ("gave in marriage") and the second merely as "gave to her husband." Jerome thus associated coniugium with the famula and did not name the union with the ancilla. We should not make too much of his different vocabulary choices here; he chose different terms to translate identical language, but it is difficult to tell whether he was really trying to make a distinction because of the ambiguity of nashim.

Jerome also had another key term to translate: "Reuben went and lay with Bilhah his father's pilegesh" (Gen. 35:22). He rendered this term as concubina. Millennia of scholarship, ancient, medieval, and modern, have taken every word in the Hebrew Bible as having a very deliberate and specific meaning, and modern scholars have exercised great ingenuity creating elaborate typologies of marriage based on this and similar stories: a slave-wife was of lower status than a concubine, but because Bilhah was Rachel's slave and Rachel had died, Bilhah was automatically freed and therefore promoted to the status of concubine. Medieval exegesis, however, did not seem to be disturbed that the same woman could be called both a concubina and a famula. As we will see, even though in Roman law, concubines had to be free women, the term had very strong implications of low status. When Jerome translated the passage about King Solomon's partners (1 Kings 11:3), the seven hundred princesses (nashim sarot) became uxores quasi reginae (wives like queens), but the pilagshim were concubinae. Jerome clearly saw Solomon, like Jacob, as practicing plural marriage, not just polygyny, but not all his partners were considered wives. While Near Eastern societies had a range of kinds of unions that might or might not be considered marriage, Jerome as translator had fewer categories to work with.

Jerome and later writers would likely have known that the Jews did not leave plural marriage behind in the biblical era. In the Second Temple period in Palestine, plural marriage seems to have become less common than earlier, although it was not formally outlawed and was more common in Babylonia. It still appeared in the Talmud and was practiced occasionally in early medieval Europe. Gershom Me'or ha-Golah (the Light of the Exile), an eleventh-century German rabbi, possibly under Christian influence, eventually forbade it. Concubinage became a good deal less common than in patriarchal times as well, as far as one can tell: there are few references to it in later books of the Old Testament or in the sparse documents of practice that survive. Talmudic references to the status of the pilegesh reflected the societies in which the authors lived, but these references were few. It may be that the infrequency of mention, at least in the Palestinian Talmud, reflected the fact that concubines had little legal status or recourse, and therefore there was little need for legal comment; but given the elaborate commentary that the rabbis provided on many other topics, the paucity of comment indicates that it was likely not a major concern (see further discussion of medieval Jewish concubinage in Chapter 2). The detailed and sometimes confusing rules about different stages of the marriage process—betrothal, the transfer of the bride, and consummation—that developed in Jewish society were largely rabbinic rather than biblical. Medieval Christians got their biblical precedents from stories in which ritual and legal procedures played a smaller role.

One aspect of legal procedure that was absent from biblical accounts of unions between women and men—conspicuously so, since so many scholars have taken it as a key element of both biblical and medieval marriage—is the payment of bridewealth of any sort. Scholars who have written about ancient Hebrew marriage have generally explained that a valid marriage was accompanied by the payment of a bride price (mohar) to the bride's father by the groom. The term, however, appears only three times in the Hebrew Bible (Gen. 34:12, Exod. 22:16, 1 Sam. 18:25), and in two of these, it refers to a situation in which the man is paying a penalty for having previously raped or seduced the woman. It appears to have been interpreted as a bride price based on cognates in other Semitic languages (Arabic mahr) and on analogy with a payment in the Old Babylonian Hammurabi Code (early second century BCE) and other payments in ancient Near Eastern sources. In later Jewish usage—for example, in Aramaic documents from the Jewish military colony at Elephantine—it was a payment to the bride's representative which became part of her dowry. Jerome translated mohar twice as dos (the Roman word that originally meant a dowry paid by the woman's family) and once as sponsalia, betrothal gifts, so he clearly understood it as something that happened on the occasion of a marriage, but not as a bride price. The most famous biblical case involved not bride price but bride service: Jacob worked seven years for his uncle Laban, only to be given the elder sister he had not asked for, and had to work another seven years for the younger sister (Gen. 29:15-28). Indeed, the customary role that property transfers came to play in the formation of medieval unions came not from biblical or Germanic precedents, as many have argued, but from Roman law and custom.

Roman Legal Traditions

Medieval western European ideas about marriage absorbed much from ancient Rome as well as from the Bible. Like the Greek culture that influenced it, Roman society was characterized by a resource polygyny (wealthy men had sexual access to many women) in which only one of these women had the official title of wife. Other women could be attached to a wealthy man's household, especially as slaves. Formal marriage was expected of leading families, for reasons of alliance, but it was not expected—in fact, not available—at the lowest social levels. Roman marriage was a private matter between the families involved and did not require any action by a state or religious official to make it valid. Nor in the early period did it require, in theory, any particular property exchanges, although property exchanges were very common and came to have an evidentiary quality and eventually to be required. Although it was not publicly created, however, marriage was legally recognized and had very specific and detailed legal consequences, with regard to the legitimacy of the children and to the parties' control of property. Although the state did not validate individual marriages, it had a concern with marriage and the procreation of citizens, and from the time of Augustus (27 BCE-14 CE) took an important legislative role in the regulation of matrimony.

Rome recognized matrimonium, which we may translate as "marriage," only between citizens, who had conubium (the right to marry) with each other (or with citizens of other Latin cities, depending on the time period). If there could be no matrimonium because one of the partners was not a citizen or not freeborn, concubinage was an acceptable alternative. Unions without conubium were not legally recognized: the husband had no right to the wife's dowry, she had no right to its return if he did have control over it, and the wife could not be prosecuted for infidelity. The jurists, however, still spoke of such unions—which, up until 197 CE, included those of soldiers—using the language of marriage, although without officially recognizing them (matrimonia iniusta). Slaves could enter only into a union called contubernium, and a slave woman's children were the slaves of her owner, no matter who the father was, because the union granted the father or his owner no rights. The difference among types of union lay not in their permanence—the dissolution of even a formal marriage through divorce was not very difficult in pre-Christian Rome—but in the legal rights they conveyed, particularly with regard to the offspring.

Medieval legal systems generally did not follow the Old Testament pattern in which a man's sons from a Roman custom where the status of the mother and whether the union was a formal marriage made a great deal of difference to the inheritance prospects of the child. The sons of concubines could not be their fathers' heirs. The other fundamental difference between Roman and other concubinage was that in Rome, concubinage was not a secondary union alongside marriage: it was available only to unmarried men. A woman with whom a married man had a relationship, even an ongoing one in which he supported her, was not technically his concubine. Concubinage was typical for a young man of high status before he had a wife, or for an older man who was widowed and did not want further legitimate children who might compete with his heirs. But as the writings of St. Augustine show, the slave of a married man could also be referred to as his concubine; concubinage in this situation was condemned, but the concubinage of an unmarried man was not.

In Roman law, marital intent—particularly the man's intent—legally distinguished the formation of concubinage from marriage. The Roman law called it maritalis affectio, "marital affection," but it did not mean "affection" in the contemporary sense; rather, it meant the disposition of the male partner toward the union. Concubines were generally of lower social status than their male partners, and that status was presumed, to some extent, to determine marital intent. Indeed, Roman law suggested that a freeborn woman should be considered a wife unless the evidence proved otherwise, while a freedwoman was prima facie a concubine. This remained the case up through the time of the Christian emperors until Justinian (r. 527-65), who allowed marriage between men of high and women of low status even where it would earlier have been prohibited, but required written documents indicating that marriage rather than concubinage was intended. These documents did not create the marriage; they proved the affectio. As the old social barriers of Roman culture broke down, there must have been an increasing number of cases for which intent was in question because it was no longer unthinkable or staggeringly inappropriate to consider a union with a woman of low status a marriage. In most cases, marital intent would be clear from the way the parties treated each other, but when it came into question because of a matter of inheritance, evidence of property arrangements or of a formal ceremony could confirm intent.

Although Roman law theoretically applied across the empire, particularly after Roman citizenship was extended to all free residents of the empire in 212 CE, there was still local variation in the kinds of unions available to women and men. Roman laws on unions between relatives were stricter than those in many parts of the empire, for example, and Roman ideas about fidelity were not universally accepted, although they fit well into the rising Christian moral framework. After the Roman empire split, the laws issued by the Eastern emperors were more favorable to inheritance by children born from less formal unions (those without property exchanges, or records thereof) than those in the West, for example. In Egypt, where a relatively large amount of documentation has survived, both "written" and "unwritten" unions were recognized as marriages. As Brent Shaw notes, however, "differences in the development of the ritual and form accompanying marriage from the early to the later empire are very difficult to measure, given the fact that so little is known of marriage ceremonial in the earlier period." He suggests that we know little about marriage among the lower orders, and I would add that one thing that we do not know is how many of them chose to go through formal ceremony at all or cared whether their union was considered a marriage.

Was There a "Germanic" Law of Marriage?

As Roman practices and prescriptive law spread through western Europe, they encountered another group of legal systems that are often thought to have contributed much to the medieval understanding of forms of union: the traditions of the pre-Christian, Germanic-speaking groups that came to inhabit much of northwestern Europe and the Mediterranean region as well. As with the ancient Hebrews and the Romans, among these groups it was not uncommon for a wealthy or powerful man to have several women in his household or in separate households with whom he had ongoing sexual relationships. But because of the nature of the sources available to us—sources almost entirely filtered through Romanized and Christianized culture—it is extraordinarily difficult to discover the statuses of different women, whether it made any difference to have gone through a ceremony, or whether property transfers were necessary to form a recognized union. The problem, once again, is that documentation survives only from social levels at which property transmission was an issue. Nevertheless, scholars have made major claims for what constituted Germanic marriage in the pre-Christian era. Scholars "know" what marriage "is," so they look at the early Middle Ages and find it there, and we also find other relationships that they decide are not marriage, or are distinct types of marriage. The scanty evidence, however, points once again to a distinction not between the nature of the rituals and contracts that created different types of union, but a distinction between partners (mainly women) of different statuses.

Modern scholars since the nineteenth century have recognized that most of the sources for the early Middle Ages are ecclesiastical sources with a particular point of view. Their authors may have wished to label as concubines women who considered themselves, and were considered by their families and partners, as wives. Scholars tried to get around this source problem by constructing a story about what they thought to have been primitive Germanic marriage customs before church influence. For this, they turned to the so-called barbarian law codes, which were thought to codify pre-Christian and pre-Roman practice that could be traced back to an ancient common Germanic culture. They also used Scandinavian sources (sagas and law codes) that, while chronologically later than the continental laws, were thought to encode an earlier stage of Germanic development. Neither of these assumptions holds any longer, as both the continental and Scandinavian legal material has been shown to be heavily influenced by Roman or canon law.

The received wisdom went something like this: the early Germanic peoples had two distinct types of marriage, Muntehe and Friedelehe. The former involved the transfer of the guardianship over a woman from the woman's kin group to her husband. This guardianship was called Munt, from a word meaning "hand," Latinized as mundium. The transfer had at some original point in the past taken place in exchange for a bride price, which gradually was replaced by a payment known as dos, which went to the woman herself rather than to her relatives. Dos can translate as "dowry," but in this situation was a "reverse dowry," from the man to the woman. The second form of marriage, Friedelehe, did not involve the transfer of the Munt, which remained with the woman's natal family. A woman married in a Friedelehe was recognized as a wife, and she received the Morgengabe, or morning gift, a payment directly from the groom after the consummation of the marriage (and sometimes interpreted as an acknowledgment of her virginity), which was also paid in Muntehe. Friedelehe did not, however, involve a bride price and was more easily dissolved. According to this story, when the church began to claim control over marriage, it did not approve of Friedelehe, wanting to recognize only those marriages contracted with a formal ceremony and a dos. By the Carolingian era, the Friedelfrau had been relegated under church influence to the status of a concubine. Thus the imposition of church regulation of marriage put women in a devalued position.

The strange thing about this story is that neither term, Muntehe or Friedelehe, appears in any extant source. The attempt to reconstruct a common pre-Christian Germanic culture and set of institutions by tracing backward—from a term in a seventh-century Lombard law, an incident in a thirteenth-century Icelandic saga, a story in a sixth-century Frankish ecclesiastic's chronicle, an ethnographic description by a first-century Roman who never visited Germany and wrote about it in order to critique his own Roman culture—has come under serious question. Many scholars today reject the idea that a common "Germanic" past can be reconstructed from its cultural descendants. Yet while they do not accept the method that gave rise to the theory of Germanic marriage forms, and while they question the Nazi-era scholarship that presented these marriage forms as evidence of German superiority, allowing them to posit a prehistoric institution that allowed for matches based on love, too few have questioned the existence of the forms Friedelehe and Muntehe. A reexamination of the evidence suggests that they should. Suzanne Wemple, in her pathbreaking Women in Frankish Society, called Friedelehe "quasi-marriage" rather than considering it a different form of marriage, and pointed out that the evidence often adduced for its existence "actually provides a corrective to the romantic picture German historians usually present of Friedelehe" because it shows that women who arranged their own unions without family involvement and bride price had greatly reduced legal rights vis-à-vis their partners. Yet other recent and careful work that has abandoned the idea of finding common Germanic marriage practices, and has focused, for example, on Friedelehe as a phenomenon of Merovingian and Carolingian society, still accepts its existence as an institution.

The 1986 article on Eherecht (marriage law) in the Reallexikon der Germanischen Altertumskunde may be taken to represent the late twentieth-century scholarly state of play. Already at that point, Rainer Schulze recognized that the modern term Friedelehe was used to cover many different things and could not be considered a single institution of great antiquity: "Very differently formed marital relations are attributed to Friedelehe (in the narrower sense). Usually it is a matter of cases in which account must be taken of social inequality between the parties. Whether the multiple legal forms that developed in this context go back to a common Germanic origin seems doubtful." Nevertheless, Schulze was able to catalog several different circumstances under which Friedelehe took place (or, otherwise put, several circumstances that he found convenient to call Friedelehe): (1) the bride's family was socially or economically superior to the groom, and he therefore did not receive the rights over her and her property that usually adhered to a husband with marriage; (2) a woman married down but was able to keep her legal status; (3) a widow remarried but kept control over the property and status she gained from her first marriage; (4) the bride had no kin group or was unfree, and the husband therefore gained full marital rights over her without paying a bride price; and (5) a man of a ruling or aristocratic family wished to marry more than one wife but could only marry one with the bride price. For Schulze, Friedelehe was a term of convenience covering a wide variety of situations; but it still had a certain reality to it, since what tied these cases together was that the marriage took place without the exchanges of property that accompanied early medieval marriage, and without the transfer of rights over the wife to the husband—features that Schulze considered defined "standard" marriage. It is doubtful whether contemporaries would have considered these five sets of circumstances part of the same institution. Nevertheless, historians lumped them together, romanticizing Friedelehe. Because no payment to the woman's family was required, neither was their consent, and the couple themselves made the choice to enter into the relationship. Because the tutelage over the wife did not pass to the husband (remaining with her father or brother, or whoever held it previously), she had a more independent status than a woman who underwent Muntehe. Because the husband did not have legal control over the wife, she could leave the marriage if she were mistreated, or wished to for any other reason. This understanding of Friedelehe as empowering the woman sat well with scholars who wished to present "the Germans" as enlightened and morally superior to decadent Romans. Some feminists also found appealing the idea of a past age when things were more fluid and therefore better for women (before the iron hand of patriarchy, clad in the velvet glove of the church, took away their freedoms and status).

As a modern term of convenience, Friedelehe is inconvenient. It brings with it too much baggage: assumptions about a primitive common Germanic culture, erroneous ideas about the status of women, and the implication of the existence of a formally recognized institution where none existed. To use it as an omnibus term for unsanctioned, nonmonogamous or otherwise "different" long-term relationships risks losing sight of the variety of arrangements that medieval people made and obscures distinctions that some drew between an honorable and a dishonorable or unfree union. Just as the fact that lords in the high Middle Ages occasionally or even frequently had sex with their serfs does not mean that a droit de cuissage existed, so too the fact that some unions in the early Middle Ages took place without financial exchange and the consent of the relatives does not mean that a distinct form of marriage existed.

But to say that Friedelehe was not a distinct form of marriage is to beg the question: Not distinct from what? The point of comparison, Muntehe, although less questioned by scholars, is also problematic. The term is based largely on the Lombard laws, which used the term mundium both for guardianship over a woman and the payment for that guardianship. On this basis, it has been assumed that all Germanic peoples transferred such a guardianship at marriage and that the transfer of a bride price to the bride's family and eventually of a dos to the bride herself was a payment for this guardianship. Neither assumption, however, is tenable. If we exclude the automatic assumption that everything in any "Germanic" law code must derive from a primitive, common Germanic culture, it is not at all clear why the Lombard laws provide evidence for Frankish, Visigothic, Anglo-Saxon, or Norse social arrangements.

Since Muntehe—a particular form of union that transferred a particular bundle of rights from the father or kin to the husband—is not a common Germanic concept any more than Friedelehe, it is especially unsafe to leap from there to the conclusion that a bride price, or any other particular transfer of property, denoted this particular form, although such a leap is very common. For example, Hans-Werner Goetz, in making the point that dos and Morgengabe were required for legitimate marriage in the Germanic kingdoms generally, cited Benedictus Levita, an early collector of a forged group of capitularies, to buttress his statement that "marriage with guardianship (Muntehe) was a marriage with a dos." But what Benedictus Levita actually said was drawn from the Visigothic law: Nullum sine dote fiat coniugium; nec sine publicis nuptiis quisquam nubere audeant, "let there be no marriage without a dos; nor let anyone dare to marry without public nuptials." This did not say anything about Muntehe or guardianship. I cannot prove by an argument from silence that there was no mundium transferred in "Germanic" marriage other than among the Lombards; but one certainly would not be warranted in arguing from silence that there was.

Where did the mundium of Lombard law come from if not from an earlier, pan-Germanic legal system? Republican Roman law featured the concept of manus, also meaning "hand," which denoted the tutelage over a woman transferred from her father to her husband at marriage. Manus marriage was in disuse by the age of Augustus. Its most prominent articulation appears in Gaius's Institutes, a Roman legal textbook. The shortened version of the Institutes found in the Breviary of Alaric (a Roman law compilation, dating from 506, for the Roman subjects of the Visigothic kingdom) omits this section, as do Justinian's Institutes, which are based on those of Gaius. However, the full version of Gaius's Institutes circulated in Ostrogothic Italy and could have been known to the Lombards, since the surviving manuscript comes from fifth- or early sixth-century Verona. Munt may well have had a Roman rather than a Germanic origin, and we should think twice about extrapolating from the Lombard law to all the Germanic groups.

Lombard law is not the only problematic witness: all the barbarian or Germanic law codes were textualized under the influence of the church and placed within a framework of Roman law. A number of traits of Roman marriage resembled those found in the law codes, and undoubtedly influenced them. One is the absence of a required, formal ceremony. Germanic marriage, as scholars commonly understand it, included the stages of betrothal; the handing over of the woman, or traditio puellae; and the bedding of the couple. However, none of these were specified as requirements consistently across the various codes. Barbarian laws, for the most part, followed Roman laws in keeping marriage a private matter. When they did prescribe the payment of a dos, they did not stipulate that the marriage was invalid without it (this issue will be discussed further in Chapter 1).

What, then, do the sources tell us about pre-Christian Germanic forms of union? This question is based on a false assumption that there was one system underlying the practices of all the barbarian kingdoms in the early medieval era. The surviving sources cannot properly be made to fit an evolutionary schema, and all have been influenced by Roman or Christian culture in different ways. The one text that certainly was not influenced by the church, the Germania of Tacitus (first century CE), which has often been taken to epitomize primitive Germanic culture, is not highly reliable. Tacitus lumped all the Germans together as one group and attributed to them characteristics intended to contrast with Roman decadence. Nevertheless, it is worth noting what Tacitus did say: "Alone among the barbarians, they are content with one wife, except for a few who, not because of lust but because of their nobility, enter into several marriages. The wife does not bring the dowry to the husband, but the husband to the wife." This would seem to indicate that the bride gift was a customary part of marriage, but it may be something that Tacitus's informants particularly noticed among one group of Germans because it contrasted with Roman practice, rather than being a universal custom. Tacitus does not say that it was required, and his statement that the dos was paid to the woman contradicts the theory that the law codes represent a late mutation of a primitive bride price that went to the male guardian in return for the Munt. It is also notable that he made no reference to any other form of union; but this was not a legal treatise—it was an ethnographic and moral one, and he was concerned with contrasting German marriage with Roman marriage.

If there was no Friedelfrau who occupied a status between that of fully recognized wife and that of concubine, and a wife's status did not depend on the transfer of guardianship to her husband, and if there was a whole range of social statuses that could be held by a female partner with no clear vocabulary to distinguish between them, does it make any sense to look for a sharp line between what was "marriage" and what was not? On what basis did people assign status to different kinds of pairings? The criterion most often suggested by scholars is that of the payment of a dos from the groom to the bride's family or, somewhat later, to the bride herself. As we will see in the next chapter, the church attempted to enforce such a line.

We have seen here, however, that although various rituals and property exchanges may have been typical in marriage formation, none of the traditions that Western medieval culture inherited—Hebrew, Roman, or "Germanic"—used them to distinguish formally between marriage and other forms of union. The concept of marriage itself was rarely defined. Unions between two members of the elite that appear in the sources accompanied by property exchanges were only the tip of the iceberg; because no particular process was technically required, we do not know how people of lower status formed long-term partnerships. We risk mistaking a usual process for a necessary one.

Across several different traditions that contributed to medieval understandings of sexual unions, this Introduction has shown that a woman who was important enough relative to her partner might have certain legal rights recognized, and otherwise not. The status of a particular union was intertwined with the status of the woman in a very complicated way. The nature of the union depended on who she was; at the same time, her reputation—and her well-being, which could be contingent on whether her children inherited from their father—might depend on how the union was perceived by others. A man's position in life generally depended a great deal less on who his partner was than did a woman's. As subsequent chapters will show, while definitions tightened up considerably during the Middle Ages, various categories of union still remained blurred, and attribution of a union to one type or another was still often based on the status of the woman rather than on particular processes of formation.