Skip to content Skip to navigation

Lesbian Historic Motif Project: #8 – Kowaleski 1999, “Singlewomen in Medieval and Early Modern Europe: The Demographic Perspective”

Full citation: 

Kowaleski, Maryanne. 1999. “Singlewomen in Medieval and Early Modern Europe: The Demographic Perspective” in Bennett, Judith M. & Amy M. Froide eds.  Singlewomen in the European Past 1250-1800. University of Pennsylvania Press, Philadelphia.  ISBN  0-8122-1668-7

Publication summary: 

While, no doubt, many lesbians in history made their peace with the need to accommodate marriage and family life, when designing a character who has the freedom to refuse marriage to a man, it helps to know what social and economic options would have been possible (or even normal) within your setting. There have been several excellent collections of papers (and even more monographs) on the topic of singlewomen, but I believe this was the first significant one to appear.

Kowaleski, Maryanne. “Singlewomen in Medieval and Early Modern Europe: The Demographic Perspective”

This is a very meaty article, full of statistics and complex intersectionality. But I consider it an incredibly important consideration when contemplating what sorts of lesbian characters would make sense in what sorts of historic settings. A young noblewoman in 14th century Florence has far different options for acting on same-sex desire than a middle-class woman of late 16th century London. One of the questions to ask yourself is: If my character chooses to remain unmarried (and is she able to choose to remain unmarried?) and even moreso, if she chooses to co-habit with another unmarried woman, will this be considered expected, typical, unusual, or unheard-of? What sorts of life choices will be available to her? What circumstances will provoke scrutiny? The answers may be far different from what a pop-culture vision of history would suggest. One set of data indicates that in Zurich, Switzerland in 1467, 49% of women were unmarried. Half! Why? What were they doing? How did they accommodate their emotional needs outside of marriage? Obviously one would expect most of them to be heterosexual, but the strategies they employed would be thought-provoking. (15th c. Switzerland is directly relevant to a character I'm planning for a future book. The current outline has her leaving an unhappy marriage behind, but perhaps it would be more realistic that she had never married at all. Perhaps the current "abandoned husband" role would be better filled by an "affronted brother". You see how statistics can drive character development?)

This summary only barely touches on the detailed statistics which occupy 14 pages in the book. While detailed information is only available for a limited range of dates and places, when planning a story that falls within that scope, I would highly recommend checking it out.

* * *

This is a complex, data-heavy survey of sources for the demographics of singlewomen, the overall (very complex) patterns that emerge, and an analys of the theoretical frameworks that attempt to explain those patterns. For my summary, I’ve rearranged the topics to try to focus on single variables at a time.

The ideal is to study entire populations, not just anecdotal cases which tend to focus on the elite. Comparative quantitative data is best, but little extensive data of this sort is available before the Black Death, which--on the basis of what is available--seems to mark a turning point in various trends. (Just for reference for those not familiar with the details: the Black Death hit in the mid 14th century and is estimated to have killed 20-60% of Europe’s population.) Legal documents such as censuses and tax records are useful for demographics but difficult to interpret for the purpose of identifying singlewomen. Even evidence for the age of marriage can be confusing and give wildly varying conclusions. By the late Middle Ages, data improves.  Italian data is particularly valuable, e.g., the Florentine Catasto of 1427 (60,000 households) and other similar sources. In the absence of detailed demographic data, family patterns can also give evidence. A count of married women and households can provide evidence for who isn’t being counted, using basic assumptions of population structure.


Analyzing singlewomen in the population rests on two factors: the age at first marriage, and the proportion of the population who married. Together these determine fertility levels of the population. Interpretations can be complicated by different treatments of widows and nuns. Fertility-based analyses excludes widows (as having already reproduced) and nuns (as having taken themselves out of the game), so this type of focus on reproductive potential undercounts single women.

In analyzing the relationship between personal behavior and historic trends, there can be wide variation in the identification of cause and effect. E.g., some historians attribute the “Italian pattern” (see below) to plague mortality driving the pressure to reproduce and creating a shortage of marriagable men, whereby competition among women for hubands lowered what was considered the ideal female age for marriage. But to what extent is there a difference between “pressure to marry young and reproduce” versus “disincentives to marry young, thus secondarily reducing reproduction”? Many analyses seem to erase personal decision-making.

In looking at variation along the multiple variable axes, it can help to start off with a vastly oversimplified model of two polarized patterns. (With the understanding that even within specific cultures at specific times, there can be exceptions to the general pattern.)

One model is commonly referred to as the “European marriage pattern” (although, in context, it would seem better to identify it as the “northern European marriage pattern”). Characteristics are: relatively late age at first marriage (mid 20s) for both women and men, with the delay typically involving employment outside the parental home (e.g., as servants or apprentices), in part to accumulate sufficient wealth to establish an independent household at marriage. Marriage partners are generally of similar age and both have significant autonomy in choosing a partner. There is also a relatively high proportion of never-married women (10-20%) living secular lives.

The “Mediterranean” marriage pattern typically involved women in their late teens who had never left the parental household marrying older men (typically 6-13 years older) and typically with little input in the choice. Never-married women in wealthier families generally entered religious life so the rate of secular lifelong singlewomen was lower (ca. 5%) and typically poorer, though if religious women are counted as “single” the overall rate is similar to the north.

In addition to these polarized models, there are a number of overall patterns. Singlewomen are more common in urban than rural areas. (Towns typically had skewed sex ratios favoring women, many of whom were servants and many of whome were recent migrants from rural communities.) Becoming married was an expense (whether due to dowry customs or the need to establish a new household) so lack of money could delay marriage (e.g., spending time earning money) or prevent it entirely. Depressed levels of real wages increased the number of lifelong singlewomen. Cultural attitudes towards female chastity affected marriage patterns. The Mediterranean pattern correlates with a focus on “honor culture” and female chastity which prioritizes early marriage for women and would tend to preclude a period of economic independence for women before marriage. Upper class and wealthy women tended to marry younger and in general those groups tended to have lower rates of lifelong singlewomen, though influences like dowry inflation or the restriction of potential marriage partners to those of similar class could counter the trend. In general, age at first marriage and the proportion of singlewomen increased across the timeperiod covered by the study in all geographic areas.

North/South Patterns

As described in the oversimplified models above, there was a general pattern of earlier marriage and fewer singlewomen in the south of Europe and the reverse in the north. These patterns are various attributed to geographic and cultural factors, but with complex interaction. E.g., one approach to the Italian pattern looks at lowered marriage age for women from increased competative pressure on dowries, where “excess” women were bled off to a religious life rather than being left single. Poor women who couldn't compete with dowries and couldn't afford to enter a convent increasingly were in service. But some historians instead point more strongly to "honor" issues in marriage. Women's honor was virginity and marriage with no honorable place for single women. Younger women were associated with greater marriage potential.

This North/South contrast can be mapped in more detail as better data starts being available in the late 16th century. Here we see the extremes of later marriage in Denmark, Sweden, and England with first marriage typically around 26 or later. France shows a mixed pattern along north/south lines. By in the 18th century the distinction starts eroding with women in the Mediterranean areas shifting first marriage age to the mid to late 20s and the never-married rate can be as high as 10-15%.

Rural/Urban Patterns

Across other patterns, there is a tendency for rural areas to have women marrying earlier and leaving a smaller percentage of unmarried. For example, data from Northern England in the late middle ages suggests rural women married in the late teens but urban women in their mid 20s.  Although the article doesn’t necessarily address it from this angle, my impression is that this works in synergy with other factors. E.g., the greater economic opportunities for singlewomen in urban areas led to migration resulting not only in large numbers of “lifecycle singlewomen” in towns working on their nest eggs, but in an urban gender imbalance that then resulted in fewer local marriage opportunities. One suspects (although I saw this nowhere addressed) that rural economies may have had fewer opportunities for lifelong singlewomen than towns did, resulting either a required choice between marriage or migration. And there are some intersectionalities (especially among the very poor) where lifelong singlehood was more common in rural areas.


When comparative data is available, there is a clear trend for women from upper class and wealthy families to marry earlier and to remain unmarried at lower rates. This can skew the overall understanding in the earlier period under study, as records for the elites are often all that is available. This contrast is often overlooked by historians. For example, one study of Reims in 1422 suggests women married at 15-16, but the same data includes a large number of singlewomen as servants around the age of 24 at the same time who are clearly not being considered in this “age of marriage” statistic.

This tendency for noble/wealthy women to marry younger could be countered by other factors peculiar to their class. Dowry inflaction or closed-class marriage (i.e., inability for women to “marry down”) could leave women single. E.g., among the English peerage the percentage of lifelong singlewomen in the 16th century was 5-9%, but in the 17 th century it rose to 13-15%, and by the 18 th century it was around 26%. In comparison, the rates for never-married women of the French nobility were consistently ca. 5-10%. But this statistic evens out if nuns are courted as “single” as this option was lost in England after the Reformation.

Change Over Time

As noted above, there was a general trend for age of marriage and percentage of lifelong singlewomen to increase over the timeperiod studied here (14-18th c), even when other factors are accounted for. As always, there are exceptions and variations on the trend. This shift is perhaps most dramatic in the “Mediterranean pattern” cultures.

Economics and Opportunities

One of the interesting factors in studying singlewomen is the question of how unmarried women survived economically and socially. What circumstances enabled them to remain single or pressured them into doing so? If they were unmarried not by choice, what options existed for them? In the “southern pattern” cultures, unmarried women were pushed into a religious life, but this was only available to families with money. Poor lifelong singlewomen in those cultures typically became servants or less acceptable professions as there was not a general acceptance of “respectable” independent unmarried women. In contrast, the “northern pattern” included an expectation of economic independence before marriage (though being in service couldn’t be considered entirely “independent”), creating a greater social acceptance of continuing this state permanently. Pre-marriage independence in crafts or professions could also be extended if marriage never occurred. In the north, singlewomen were prevalent in urban areas due to migration related to greater opportunities for these jobs. But social changes in the availability of certain job classes, or in the acceptability of women performing certain jobs could affect the status of singlewomen who depended on them long term. Evidence in post-Black Death York shows not only higher numbers and proportions of women servants but more women-headed households and women testators, indicating open opportunities for female independence. But this changed for the worse in the late 15th century as economic changes pushed women out of crafts and into service. Sex-specific economic opportunities in particular locations might drive migration that created skewed sex ratios (in either direction) leaving many unmarried.

Religious Factors

As noted previously, when a religious life was an option for unmarried women, this could “bleed off” large numers of lifelong singlewomen, reducing social pressures especially in regions where other social or economic opportunities were limited. The Reformation removed this option in Protestant regions. Also, although not included generally with the data above, some studies show that Jewish women generally married relatively early, and proportionately fewer remained single. There is no discussion of how different social and legal pressures affected this statistic.


Add new comment