Read an Excerpt
Marriage and Fertility
Studies in Interdisciplinary History
By Robert I. Rotberg, Theodore K. Rabb
PRINCETON UNIVERSITY PRESSCopyright © 1980 Princeton University Press
All rights reserved.
Emily R. Coleman
Medieval Marriage Characteristics: A Neglected Factor in the History of Medieval Serfdom
Only recently have historians come to appreciate the role which demographic factors have played in the social history of the Middle Ages. In the traditional view, medieval society remained rigidly stratified and static; change, when it came, was attributable largely to external factors playing upon the medieval world — the opening of frontiers, the expansion of trade, and the growth of towns. All of this is true, but it does not represent a complete picture of the forces working to transform medieval society. This paper examines another factor, hitherto neglected by scholars, which apparently played a major role in the social history of the Middle Ages, and, particularly, in the history of medieval serfdom: the marriage patterns characteristic of the servile population.
I have utilized one of the magnificent documents of medieval social history — the Polyptych of the Abbot Irminon, redacted probably between c. 801 and c. 820. It describes the lands and the some 2,000 families belonging to the monastry of Saint Germain-des-Prés near Paris. This polyptych of Saint Germain-des-Prés is an extraordinary example of the medieval censier or manorial extent of the estates and benefices which comprised and/or were dependent upon an abbey or church. There is a breve, or chapter, describing each part of the seigneury, relating in some detail the type and size of the elements of the demesne, the amount of arable tenanted land, the number of people on the land, the dues they owed, and information on mills and churches.
The polyptych has been the worthy object of intense and careful study for over a century. The manuscript, which is preserved in the Bibliothèque Nationale (Fonds latin, manuscript no. 12832), has been edited twice. The first edition (1844) was by Guérard, whose detailed introduction has become a starting point for all studies of the document. Fifty years later Longnon revised some of Guerard's paleographic interpretations and reduced the introduction to more manageable proportions. Since then, the document has revealed and confirmed numerous findings for its patient researchers. It is interesting to note, however, that these diligent scholars have concentrated almost wholly upon the tenurial aspect of the document. The material that it contains on early ninth-century demography has not been systematically analyzed.
The document is unusually rich in demographic information. It is not a total census for the Île-de-France, but it is far more than a tenurial document (in the sense of being concerned more with property than with people). The redactors were closely interested in the people who lived on the monastery's land and/or who owed personal dues to Saint Germain-des-Prés. Even a cursory glance at some of the brevia reveals the strict attention given in the census to the dependents on those ecclesiastical lands. Each fisc (villa) is broken down by manse (the dependent family farm), and those on the manse are denoted individually, by status and by name — including, in most cases, the children. Moreover, the census listed those peasants who lived on these lands, even if they belonged personally to other seigneuries. The monastery's interest in more than just a head count is also made clear by the attention paid to familial relationships, noting even when children had a mother different from the woman with whom they were living. If we utilize the available data, we find a wealth of information dealing with sex ratios, family size, and, especially, medieval marriage characteristics.
The working and personal relationships associated with the seigneurial regime of Saint Germain-des-Prés display a hierarchical society with subtle, but important, distinctions in rank. These range from the descendants of simple slaves to the descendants of purely free men, ,and cover most of the ground in between. In an entirely predictable manner, the 9,219 peasants composed four main categories: liberi (making up a diminutive 1.30 per cent of the total census), coloni (82.74 per cent), lidi (3.17 per cent), and servi (comprising a surprisingly small 5.02 per cent), with 7.77 per cent of the population of undeterminable status. Although these groups were in theory supposed to have their own rather distinct set of responsibilities and taxes due to their various positions in the social and economic hierarchy, a close examination of the polyptych reveals that the different types of tenures basically determined the redevances (dues) that their inhabitants, of whatever status or varieties of status, had to pay.
By the ninth century, the manse was no longer the area delegated to one family. There was no determinable relationship between the size of the plots and the services due from the peasants, and still less relationship between the size and the number of villein households that they supported. In Abbot Irminon's polyptych, the number of manses supporting more than one family ranged from about 8 per cent to about 67 per cent; 701 out of 1,726 manses had multiple households — equaling 40.61 per cent of the total population. The fact that over 40 per cent of the manses were supporting at least twice the number of people than was originally intended would be explained, at least in part, by a growth in population which exceeded the ability of the seigneury to clear new lands and create new manses quickly enough. This picture of a dense population and vigorous expansion is substantiated and reinforced by what seems to have been a definite upward demographic trend during the Carolingian period.
Yet, while there was an increase in population, the manors belonging to Saint Germain-des-Prés did not show a trend toward large families. The average number of children ranged from slightly more than one child per couple to slightly over three children per couple. Rarely did the households include other than those in the immediate family.
However, although familial characteristics were surprisingly modern, Saint Germain's seigneuries showed a startlingly high sex ratio, quite unlike that which one would ordinarily expect. The sex ratio is generally defined as:
the number of men to each 100 women in a population. The normal [modern] ratio at birth is about 105. However, this declines with the greater mortality of males until most [modern] populations show an equal number of men and women in the total.
The polyptych shows different characteristics. The ratios among the adult populations on the manors of Saint Germain-des-Prés ran from 110.3 to 252.9 men for each 100 women; if the entire population (where determinable) is taken into account, the ratio still ranged from 115.7 to 156.2 — both of which may be seen in Table 1 below.
The causes for such a marked predominance of men, especially in certain areas, are difficult to determine; we must assume that more factors are involved than a communal genetic freak. It is probably safe to discount the under-reporting of women, however. Polyptychs were relatively sophisticated and carefully constructed tax rolls. They gave information relevant to manorial income with a uniformity that would suggest methodical planning. It is unlikely that a monastic management possessing the ability to conceive and execute a project on so large a scale, and employing what appears to have been a definite basic formula of inquisition, would have simply neglected real taxable units in the form of some women. The redactors took care to mention those women who did not belong to Saint Germain-des-Prés when they were married to the monastery's peasants; and women not tied to any particular manse were mentioned regularly among those who paid the head tax at the end of the brevia. The importance of women in a peasant society was undeniable; they made up an important labor source, aside from their more obvious role in childbearing. Moreover, women were of especial importance in Saint Germain-des-Prés because it was they who passed on the status valuation to their children.
There are several plausible reasons for the unbalanced sex ratio. Certainly, one must not ignore the possibility of female infanticide. That, and the death of many women during parturition, would account for a certain disequilibrium in the sex ratio. Undoubtedly, ninth-century midwives and medical practices produced many widowers. This might also account for the marked inequality we find generally, or for the inequality apparent in children's sex ratios. It is possible, however, that the estates which showed a definite preponderance of men as opposed to women may have been in the process of clearing the forest for cultivatable fields. In those areas, women would have been of less use. They would have followed the men and boys to the untamed regions slowly, as conditions of life gradually softened and their skills and abilities could be more profitably utilized. In other words, the sex ratio might be an indication of an internal frontier clearance in the Île-de-France that provided more fields and food to supply the rising population of the ninth century. Yet, it is also possible that the numerical masculine superiority was the result of an immigration of men who had hoped to find work and protection on these monastic lands. Taken together, infanticide, maternal mortality, frontier clearance, and masculine mobility suggest a possible complex of factors explaining the high sex ratios displayed by the polyptych.
What is important is the indisputable fact of the high sex ratio, for whatever reason. The important point to note about this preponderance of men is that, as a result, the marriage market should have been auspicious for women. Since the number of men to each woman was so high and the economic advantages of a wife in a peasant society were so obvious, one would have expected the women to make excellent marriages in terms of both economics and status. Yet these two assumptions seem to be erroneous.
Many of the scholars who have studied the polyptych have noted, in passing, the odd marriage characteristics of the inhabitants. There was, in addition, an unusually high proportion of socially unequal marriages — among the different peasant statuses — for which no apparent reason has hitherto been suggested. In 251 out of a total of 1,827 marriages, or 13.74 per cent, in which the statuses of both spouses were determinable, that of the man and woman were different. On the individual estates, the percentages ran from 1.28 per cent to 80 per cent, as seen in Table 1.
In 61 out of 251 cases, the man married beneath himself. In 190 cases or 75.69 per cent of the unequal matches, it was the woman, who may be assumed to have been in an ideal position, who married below her status. And it is interesting to note that in these cases, the average number of children per couple ran to 2.04 as opposed to 1.76 where the man was the socially superior partner.
Why would so many of the available women — almost one sixth of the married population — marry men socially inferior? It could not have been due merely to a lack of eligible men of the same status. As was noted, the vast majority of people on the seigneuries were coloni, and it was the colonae who took the major role in this social transformation when the sex ratio averaged 117.06 among adults. Even so, there was no dearth of marriageable servi. For example, in the Breve de Gaugiaco, there were four unmarried servi and only one unmarried ancilla (female servus). In the Breve de Nuviliaco, there were two to one. In fact, the overall sex ratio among servi adults was an incredible 297.56. Furthermore, it is unlikely that the unattached men of the corresponding status would be in the wrong age group. This was the least important of qualifications in an age untouched by concepts of "romantic love." Also, in some cases, the sex ratio was even more favorable among the children — which would indicate a rough similarity of age — than among adults.
We can rule out, too, any connection between personal and tenurial status as a factor in marriage. Unequal marriages appeared on free manses and servile manses. The extent of the holding made no difference, either. They occurred on manses whose territory ranged from twenty-eight bunuaria of arable land and four arpents of vineyards to eight bunuaria of arable land, one and one-half arpents of vineyards, and one and one-half arpents of meadow to two bunuaria of arable alone, with gradations in between, above, and below.
It has been suggested, by Bloche among others, that by the time when the polyptych was redacted, social stratifications had become devoid of practical significance. He suggested that the classes found within the manorial surveys were administrative anachronisms. In the day-to-day life of the individual they were meaningless, and were kept alive only by the fact that scribes and other servants of administration usually utilized terminology and categories of social analysis that were at least a generation behind the times. To an extent, this was certainly true. The dues relevant to personal service had generally fallen into disuse. Yet this unfortunately does not explain the marital mismatching that occurred on the manors of Saint Germain-des-Prés. If the solution were simply that no one really took the status hierarchy seriously, there would inevitably have been a statistically parallel number of cases of men making misalliances; and the average number of children in both groups would have been more evenly balanced. Yet this was not so — over three-fourths of these mixed marriages show that the women married below their status. Thus, while the social hierarchy had less and less meaning in economic life, it apparently retained some importance within a hierarchically-conscious society; every group had its own "pecking order" — in fact, this was of particular psychological importance when the distinctions came to mean very little in practical life.
It is nevertheless possible to suggest a hypothesis as to the cause of the peculiar situation. On the manors of Saint Germain, the status of the mother decided the condition of the children, as prescribed in the laws of the Emperors Gratian, Valentinian II, and Theodoric. Now this may not have always held true, but it apparently was a general rule, enhanced by the occasional exception. If the difference in status between husband and wife were very large, the position of the children sometimes became a compromise; this unusual situation was specifically mentioned by the redactors of the document. In the rules of inheriting status, however, Longnon agrees with Guérard that the condition of the mother seems to have been the determining factor. Their interpretation, as one would expect, is verified by an examination of the polyptych itself, in such passages as: "Frotcarius, Frudoldus, Frotbertus. These three are lidi because they were born of a lida mother; Martinus, a servus, and his wife, an ancilla, named Frotlindis, are people of Saint Germain. These are their children: Raganbolda, their daughter, is an ancilla, Faregaus, Widericus, and Winevoldus are lidi because their mother was a colona" (an example of the occasional compromise); or "These are the children of Dudoinus by another woman, and they are servi: Berhaus, Aclevertus, Dodo, Faregaus, Acleverga, Audina," or "Witbolda, an ancilla, and her sons, who are servi...."
There are many more examples. In the Breve de Murcincto, on the eighth manse, the redactors took unusual care in recording the status of the children: "Alveus, a colonus, and his wife, a colona, named Ermoildis, are people of Saint Germain; they have four children — also coloni...." The reason would seem to be that the scribe had initially mistaken the status of Ermoildis; originally he had recorded her as an ancilla, caught the error, and corrected it by replacing that word with "colona." As always, the names of the children followed their parents' names and statuses. But here the uncommon step was taken of adding "similiter coloni" so that there would be no error in calculating the children's status by misreading Ermoildis as an ancilla — because her condition devolved on her children. And, in other instances as mentioned above, the redactors carefully noted when children had mothers other than the women with whom they were living.
Excerpted from Marriage and Fertility by Robert I. Rotberg, Theodore K. Rabb. Copyright © 1980 Princeton University Press. Excerpted by permission of PRINCETON UNIVERSITY PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.