Cambridge University Press
978-0-521-86736-3 - To Have and to Hold - Marrying and Its Documentation in Western Christendom, 400–1600 - Edited by Philip L. Reynolds and John Witte, Jr.



Philip L. Reynolds

This anthology focuses on the agreements that marrying entailed in Western Christendom from 400 to 1600, and on the documentation of such agreements. It is appropriate at the outset, therefore, to reflect synoptically on the process of marrying. What agreements were made and who made them? What was the function of such agreements in the process of marrying? Which agreements were documented and which were oral? What other actions did the process of marrying entail, as well as agreements, and what were their functions? The following sketch is intended to provide an overarching historical and conceptual framework for the specialized chapters that follow, to explain some of the terms, concepts, and institutions that the authors presuppose, and to direct the reader to some of the pertinent secondary literature.

   Marriage brought about three kinds of social change, pertaining respectively to a core relationship, to a redistribution of property, and to a reconfiguration of family connections. First, and most fundamentally, a man (or boy) and a woman (or girl) entered into the core relationship that was marriage itself. They became, in certain respects, a social unit, forming a partnership (societas) characterized by a cluster of sexual, collaborative, parental, and familial obligations. Because the couple became a new family unit, marriage severed a son from his parents even in virilocal societies. Thus according to Genesis 2:24, a text that surely presupposed a virilocal norm, the man who marries leaves his father and mother to be united with his wife: he does not need to leave his parents’ home, but he does leave their embrace.1 Those who construed marriage as belonging to the natural law considered its chief raison d’être to be the procreation, rearing, and education of children, and they sometimes compared human marriage with sexual bonding in other animals.2 As Augustine was fond of pointing out (see Chapter 3, by David Hunter), Roman dotal instruments customarily referred to marriage as a relationship entered into liberorum procreandorum causa, that is, for the sake of begetting legitimate children or one’s own children (i.e., one’s heirs). With his own concubine and illegitimate son in mind, Augustine observed that whereas one married in order to have children, children were begotten only accidentally outside marriage, although they sometimes forced their fathers to love them.3

   Second, marriage was an occasion for the transfer (and thus the redistribution) of property. The husband might endow his wife with a marriage gift, and the wife might bring a dowry from her family into the marriage. Once the partners were married, their respective contributions might either be merged as a single resource or remain under the separate control of the husband and his wife (or their respective families). The wife might also acquire a dower interest in a portion of her husband’s property, which would support her in the event of her widowhood. Eventually, property from both sides would normally pass to the children after their parents’ deaths, so that dotation was a means by which wealth devolved from grandparents to grandchildren.

   Third, marriage rearranged interfamilial relationships and created new ones. For example, a husband might gain influence through becoming associated with his wife’s family, or he might manage real estate that she had brought into the marriage, or two families might become more closely allied as a result of their children’s marriage.

   Changes of the third sort require no further comment here (although they will feature in the chapters by Laurent Morelle, Cynthia Johnson, Martha Howell, and John Witte). But it is appropriate to comment at the outset on changes of the first and second sorts (pertaining respectively to status and to property), for these were intrinsically contractual and were therefore the subject of distinct oral or written agreements. Contracts of two sorts, therefore, attended marrying in premodern Europe. On the one hand, there were agreements to marry (i.e., to form the core relationship), whether in the future or with present effect. On the other hand, there were agreements by which one party conferred betrothal gifts or marital “assigns” on the other.

   Nothing more is meant by the term “contract” here than a binding agreement that came under the purview of a system of law (whether codified or customary, written or oral) and was thereby (at least ostensibly) enforceable. This mild use of the term “contract” does not necessarily imply that there was a codified system of contract law, or that there was a formal juridical procedure for enforcing contracts, or even that the existence of a written contract would have been the decisive factor in the resolution of conflicting claims. Although little evidence of the use of matrimonial documents in litigation has survived, their form presupposes that they were legally enforceable. Nor does this use of the term “contract” imply that people considered marriage to be a contract, for it is one thing to posit marriage as the object of a contract and quite another to construe marriage itself as a contract with some other object (a distinction of cardinal significance in the history of marriage as a sacrament).4

   In regard to the first aspect of marrying – the formation of the core relation- ship – something needs to be said, first, about the role of “consent” (consensus) in the process of marrying, and second, about the intervention of the clergy in this process. The intervention in question here is not the regulation of marriage in canon law (although that is part of the story) but rather the active participation of clergy in marrying (for example, by administering the liturgical celebration of marriage).

   In regard to the second aspect of marrying, something needs to be said here about the various marital assigns and their economic function, and about the place of dotation in the nuptial process. With all that in mind, one is in a position to appreciate how marriages were documented and the respective functions of written and oral agreements.


When scholars of marriage in Roman law or in the Middle Ages refer to marital “consent,” the term is a literal rendering of the Latin consensus, and it is not used in its usual modern sense. In modern English, the word “consent” usually connotes permission or compliance with the will of another. To be sure, what counted as consensus in medieval marriage was often, in fact, only compliance with the will of families or parents (especially where daughters were concerned), but the Latin term implied that the two parties were of one mind, and its prefix implied mutuality. What is in question here, therefore, is at least putatively an active rather than a merely passive consent. More precisely, it is the kind of agreement that creates a bond of commitment or obligation between the two parties. Furthermore, the marital consent of medieval canon law and theology was always an act of agreement – an event – whereas in classical Roman law, the marital consent that established a valid marriage did not necessarily require any such act. As long as the partners were qualified to marry and there was no serious misalliance of class, evidence that they regarded each other as man and wife or with “marital affection” sufficed, under Roman law, to establish that they were, in fact, man and wife.5

   During the early Middle Ages, the act of mutual agreement formally required for marriage was prospective: it created a contract that would be fulfilled at length when the spouses came together as man and wife. In other words, it was a betrothal.6 The usual Latin term for this agreement in the early Middle Ages was desponsatio, although the classical term sponsalia (denoting a promise to marry) was sometimes used in the same sense. Notwithstanding some equivocation about the precise function and the effect of betrothal, the notion of an act of agreement in the present tense that immediately creates a marriage did not become explicit until the twelfth century. To appreciate the function of betrothal, therefore, one needs to appreciate its relation to the process that it initiated.

   A remarkable letter that Pope Nicholas I wrote to Boris, the Khan of Bulgaria, in 866 provides us with a good point of departure for understanding the early medieval nuptial process.7 Boris was trying to decide whether his people should join the Orthodox or the Roman branches of the church, and Nicholas explained, among other matters, how people married in the West. It is a unique record, for references to marriage during this period usually presupposed a common understanding.

   The process outlined in the letter begins with betrothal (sponsalia), which Nicholas defines as an agreement to marry in the future. Next, the sponsus gives the sponsa a ring as a pledge (arrha) of his intent and as a symbol of their undivided fidelity, and he confers upon her, by means of a written agreement, a dowry that is acceptable to both sides. All of these steps may occur before the partners are of marriageable age. At length, when they are old enough, they are blessed and veiled in a church ceremony. Nicholas adds that, in contrast with Eastern practice, there is no sin if any of these formalities are omitted, for formal marriages are expensive and many cannot afford them; only the acts of agreement (consensus) are strictly necessary.8 But he says nothing about anyone’s agreement in the wedding phase of the process. Agreement is expressed chiefly in the betrothal, which he says is “celebrated with the consent both of those who are contracting the marriage and of those in whose power they are.” Here Nicholas is echoing an opinion of the third-century Roman jurist Paulus regarding consent to marry: marriage (nuptiae), Paulus says, is not valid“unless all give their consent, that is, those who come together and those in whose power they are.”9 Nicholas emphasizes mutual agreement, too, when he mentions the dowry. But he treats bride and groom at the wedding as passive recipients: they are brought (perducuntur) to the wedding ceremony and placed (statuuntur) at the hand of a priest, and they receive (suscipiunt) the priest’s blessing and the veil.

   The difference between Nicholas’s notion of marrying and our modern notion is fundamental. We are accustomed to regard marrying as an event consisting essentially of an exchange of vows with immediate effect. Before the exchange, the partners are unmarried (albeit probably “engaged”). After the event, they are married. Anyone approaching medieval or even early modern marriage with that assumption in mind would find much that is confusing or puzzling. Normative accounts of marrying by theologians and canon lawyers from the high (i.e., central) and late Middle Ages might seem largely consistent with it, but sources that are closer to practice during the period indicate that people still regarded marrying as a process or a series of steps, even when canon law defined a particular point or event that was, in itself, a sufficient condition for marriage. What has sometimes been called the “processual” view of marriage was deeply rooted in Western tradition, and the innovations of high-medieval theologians could not uproot it.10

   Betrothal, then, was an agreement between the partners and between their respective parents or kinsfolk that created an obligation that would be fulfilled when the partners eventually came together. Courtship, negotiations, or pourparlers might occur before the betrothal, but once the partners were betrothed, they were contractually bound together, albeit not indissolubly (for they could dissolve their betrothal on numerous grounds as well as by mutual consent). Although the partners could become betrothed before they were of marriageable age (in theory, from the age of seven), it was chiefly through this prospective agreement that the parties expressed the consensual, contractual aspect of marriage. Clearly, this was not a norm that emphasized the genuine consent of the partners themselves, although (as Nicholas notes) their consent, too, was supposedly required. If all went according to plan, there was no need to repeat at a wedding the agreement that had already been expressed in the betrothal, although subsequent steps (such as dotation) would confirm the agreement and keep it on track. Marriage was completed at length in the coming together of the partners in a shared life, an occasion marked by the “handing over” (traditio) of the wife to her husband, when the husband was said to lead (ducere) his wife in marriage. Whereas she had formerly been his betrothed (sponsa) and a wife-to-be, she was now fully married (nupta). Their coming together was the presumptive occasion for the sexual consummation of their marriage, but prior to Gratian, it seems, there was no definitive doctrine that consummation was a formal requirement for marriage. Ideally, as Nicholas indicates, a church ceremony or at least a priestly blessing would precede or mark the occasion of their coming together, but the term nuptiae does not necessarily denote a liturgical event, despite its etymological connection with veiling.

   The Christian understanding of the nuptial process during the early Middle Ages was therefore closely akin to the Jewish one, although the betrothed woman in Judaism was arguably even more “married” than her Christian counterpart, and the process was typically quicker. Marrying among Jews began with kiddushin (betrothal), which created an inchoate marriage.11 The betrothed woman continued to live in her parents’ home, but her status in other respects was that of a married woman. (If she was unfaithful, she was in the fullest sense an adulteress.) After sufficient time had elapsed for the necessary preparations,12 the marriage was concluded in nisuin, when the partners came together as husband and wife. Thus, according to the Vulgate version of Matthew’s Gospel (Matt. 1:18), Mary was betrothed (desponsata) to Joseph when she conceived Jesus, but they had not yet “come together” (convenirent). Yet the implications of that crucial precedent were ambiguous, and attempts to resolve the ambiguity by theologians and canonists of the high Middle Ages were not entirely successful. Augustine observes that an angel called Mary Joseph’s wife (coniux) as soon as they were betrothed (ex prima desponsationis fide), even though Joseph would never “know” her.13 Is a betrothal (desponsatio) a promise to do something in the future or an act that has immediate effect?

   In the twelfth century, the canonist Gratian defined the nuptial process formally by characterizing betrothal as matrimonium initiatum: a marriage that had begun. Marriage was perfected and rendered legally valid (ratum), according to Gratian, through subsequent coitus (the “knowing” to which Augustine referred).14 Gratian presented his theory as a way of reconciling two sets of dicta from the church Fathers and later Christian authorities. On one side were the “consensual” proof texts, which, taken together, implied not only that consensus alone was sufficient for marriage but also that the partners became man and wife as soon as they were betrothed. On the other side were the “coital” proof texts, according to which a woman was not married and did not participate in the nuptial mystery of Christ and the church (Eph. 5:32) until her marriage had been sexually consummated.15 In fact, although it probably did not seem so to Gratian, the two sets of auctoritates were not evenly matched, for while the dossier on the consensual side was genuine and mostly apropos, the chief texts on the other side were spurious, corrupted, or misappropriated.16 Be that as it may, Gratian found his solution in the notion of matrimonium initiatum: yes, spouses were married as soon as they were betrothed, but only by matrimonium initiatum, and not by matrimonium ratum.

   Gratian’s theory was congruent not only with the traditional customs and expectations of his day17 but also with the key biblical premise that marriage is a union of “two in one flesh” (Gen. 2:24). Yet although the position that he advocated was not without precedent in patristic and medieval thought, it defined the role of coitus in the formation of marriage with a clarity that was quite new in continental Europe, and it therefore provoked debate. Scholars have often assumed that Gratian’s coital theory originated in Germanic law, but the evidence regarding continental Europe is wanting.18 Some twelfth-century scholars of Roman civil law maintained that the “handing over” (traditio) of the bride or her being led into her husband’s home (deductio) was the moment at which marriage was complete, whether or not consummation ensued immediately, and although this was a minority opinion (at a time when scholarly opinions became polarized between the consensual and coital theories of marriage formation), it was probably a better reflection of the traditional view.19

   There was still some tension, if not outright inconsistency, between Gratian’s theory and the teaching of Augustine, who, for complex theological and ideological reasons, had taught not only that agreement alone created a marriage but that Mary and Joseph had been married in the fullest sense.20 Gratian’s attempt to interpret Augustine in the light of his conciliatory position was astute but not entirely convincing, although it must be said that Augustine’s own intentions had been pastoral and ideological: he had never intended to resolve canonical or juridical issues regarding the formation of a valid marriage.

   An alternative position arose in the schools of twelfth-century Paris, where its first proponents were theologians. (The relationship between the theory and contemporaneous jurisdiction in Paris at that time remains obscure.) Where Gratian tried to conciliate between coital and consensual theories of marriage formation, Hugh of St. Victor and Peter Lombard took the consensual theory of marriage formation to its logical extreme. With good support from Augustine, Hugh developed his theory in two works composed in the 1130s. The first was a polemical treatise on the virginity of Mary, in which he rebutted an unnamed adversary who held views similar to Gratian’s.21 He later incorporated the theory of marriage developed there into his comprehensive treatment of marriage in the De sacramentis Christianae fidei, the first of the great theological summae.22 Hugh maintained, on the one hand, that Mary and Joseph were truly married, and on the other hand, that Mary was a virgin not only in body but also in mind, which is to say that she made her marriage vows while intending not to consummate her marriage (a problematic position under medieval canon law). Hugh carefully expounded Genesis 2:24: “Wherefore a man shall leave father and mother, and shall cleave to his wife, and they shall be two in one flesh.” According to Hugh’s analysis, the dictum clearly implies that husband’s union with his wife excludes a prior bond with his parents. Therefore the husband must take something from one side to give it to the other. Clearly this cannot be coitus or the conjugal debt; nor can it be cohabitation, given that marriages are often virilocal. Instead, it is a bond of mutual intimacy. The union of “two in one flesh” is indeed sexual union, but that union is another, superadded component in marriage. Hugh shared with Gratian the assumption that only sexual union can establish the union of two in one flesh.

   Hugh’s rationale for his theory was based on the novel premise (still only some three decades old) that marriage was a sacrament,23 although he made no attempt to apply to marriage the general theory of the sacraments that he worked out elsewhere in the De sacramentis. (According to the latter theory, each sacrament contains or confers what it signifies.)24 Instead, he predicated his argument on the symbolism of marriage. Conceding that only consummated marriage could be the “great sacrament” of Christ and the church (Eph. 5:32), Hugh argued that there was a deeper, spiritual relationship in marriage that was a still “greater sacrament,” namely, that of the union between God and the soul. The greater sacrament was more valuable and could thrive in a celibate, unconsummated marriage, such as that of Mary and Joseph.

   Peter Lombard codified (and brought down to earth) Hugh’s theory in his Sentences, composed in the 1150s, construing marriage not as two sacraments but as a single sacrament that signified two aspects of Christ’s union with the church, respectively spiritual and incarnate.25 By this time, Gratian’s position was well known, and Lombard wrote in opposition to it. Peter Lombard was less interested than Hugh in the spirituality of chaste marriage, but he was more diligent about tying up canonical loose ends. Lombard deduced that agreement (consensus) alone was sufficient to create a sacramental and indissoluble marriage.

   Crucial for both Hugh and Peter Lombard was a distinction between the agreement to marry in the future and the agreement to marry with present effect, which first appears early in the twelfth century.26 Gratian was apparently unaware of this distinction. Canon law had always recognized that at a certain point in the process of marrying, a compact that had been dissoluble became indissoluble. For the followers of Gratian, that point was consummation. For Peter Lombard, the point was an agreement about the present (consensus de praesenti), and the compact could be dissolved only as long as it was merely prospective. According to Lombard, a simple agreement was sufficient to create a binding, sacramental marriage, regardless of consummation, but only if the agreement was expressed (i.e., stated orally) in words of the present tense, and only if it was a genuine (rather than coerced) expression of intent. An agreement expressed in words of the future tense, therefore, did not make a marriage at all:

The efficient cause of matrimony is agreement [consensus], not of any sort, but expressed in words; and not as to the future [de futuro], but as to the present [de praesenti]. For if they agree about the future, saying, “I shall take you as a husband,” and “I shall take you as a wife,” this is not the agreement that is effective of matrimony.27

The position that Hugh and Peter Lombard advocated was innovative not only in emphasizing the distinction between the two kinds of consent but also in positing an agreement that was not prospective at all but rather had immediate effect. It became possible to marry suddenly and casually, without any preparations, negotiations, or permission.

   Canonists as well as theologians during this period were preoccupied with the necessity of the partners’ consent to their own marriage,28 but the Parisian position was consensualist not only in the obvious sense that agreement was sufficient for marriage but also in the subtler sense that it emphasized the consent of the partners themselves. Betrothals were typically agreements negotiated between parents, families, and kinsfolk, and children could become betrothed well before marriageable age. Once the distinction between consensus de futuro and consensus de praesenti was in place, families negotiating a marriage had little to gain by jumping precipitously to the final stage, while couples wishing to marry despite the wishes of their families had nothing to gain by agreeing to marry de futuro. By requiring an exchange of vows in the present tense (or something presumptively equivalent), the clergy required persons who were marrying to speak up for themselves. Children who were the subject of a prospective marriage contract drawn up by their parents or families while they were still underage would confirm their own consent when they came of age, sometimes with the help of clergy.29

© Cambridge University Press