daughters would not inherit family property and were not considered a significant part of the family.
Definitions of various population groups – those who were “working” and those “not working,” for example – were based on the male experience, so that a married woman who raised chickens, did laundry, and took in boarders was defined as “not working” in census records. Who was “married” and who was “not married” was dependent on the judgment of those keeping the records. Thus consensual unions not officially recorded as “marriage” may well have been regarded as such by the people involved and their neighbors. For example, since the Catholic Church charged for marriage ceremonies, as for other sacraments (baptism, first communion, funerals, etc.), poorer people in the Catholic lands of Latin America often cohabited for years and only married when the need arose to legitimize offspring.
Some of the earliest quantitative records have been used with other types of sources to examine large‐scale family patterns. For example, historians have identified a marriage pattern that emerged in northwestern Europe (England, the Netherlands, parts of France, Germany, and Scandinavia) in the fourteenth and fifteenth centuries in which couples waited until their mid‐ or late twenties to marry, long beyond the age of sexual maturity, and then immediately set up an independent household. Husbands were likely to be only two or three years older than their wives at first marriage, and a significant number of people never married at all, in some places perhaps as high as 25 percent (Seccombe, 1992). This contrasted with the standard premodern marriage pattern in other parts of the world where marriage was nearly universal, generally between teenagers who lived with one set of parents for a long time, or between a man in his late twenties or thirties and a much younger woman, with households again containing several generations.
Exactly why the distinctive northwestern European marriage pattern developed is not clear, but at its core was the idea that couples should be economically independent before they married, with some money and skills developed during long periods as servants or workers in other households. Religious and political leaders who valued male‐headed marital households passed laws forbidding unmarried people to live on their own, sometimes limiting these to women, but sometimes including men as well. More women than men were punished for, as the phrase went, “having their own smoke,” but there are a few records from Europe and the European colonies of North America of unmarried men actually being punished for living on their own. What this late age of marriage actually meant for women or men is more difficult to discern than the pattern itself, with some scholars arguing that it meant women were more likely to have a voice in choosing their own spouse and others that it made women desperate to marry and to accept anyone who was available.
More recent statistics have provoked similar disputes. Rates of divorce went up in many parts of the world in the twentieth century, with nearly half of all first marriages in the United States in the 1990s ending in divorce and one marriage in four in the Arabic world. Is this “decline” or “liberation”? If the latter, is this pattern more liberating for men or women? Or is it, as some have posited, that the ease of divorce and remarriage makes it possible to correct earlier decisions made in error? In contrast to disputes about the premodern period, every side in debates about the recent past can also be buttressed by additional quantitative sources, such as those showing that men have longer life expectancies when they are married, women when they are single, or that divorce increases men’s standard of living and decreases women’s.
The problems of scarce or skewed sources in family history do not disappear when we begin to use non‐quantifiable materials. Written sources from or about European peasant families before 1800 are rare, as is documentation on families in much of Africa, the Americas, Southeast Asia, and the Pacific. Historians and anthropologists use a variety of means to study kinship organizations, marital patterns, living arrangements, and other aspects of family structure in these places: oral history and traditions; later written records; reports of outsiders, such as Muslim traders in Africa or Christian missionaries in America and the Pacific islands; direct interviews with living individuals; archaeological remains of dwellings and domestic artifacts; and linguistic analysis of words denoting family and kin. While these provide evidence about families in the past, scholars warn of their limitations: outside observers bring their own gender and ethnic biases, archaeological remains are difficult to interpret, and oral history (like all history) represents a specific perspective. Family patterns are not static, and intentional or unintentional mythologies of the past shape notions of the “traditional family.”
Even if there are mountains of numbers available for studying families, quantifiable data can only go so far. Many family historians have relied less on statistics, instead basing their research on public documents, private written sources such as diaries, family chronicles, account books, and letters, as well as material sources such as paintings, furniture, household objects, toys, and architecture. These sources are skewed toward cultures in which written records became common quite early, and toward the minority among the population that was literate, thus being skewed by gender, as men in most cultures were more literate than women. We have many letters from Martin Luther to his wife Katharina von Bora, for example, and none of hers to him. Historical artifacts from women or members of non‐elite groups are so scarce that the few examples that have survived achieve a kind of iconic status. The reflections on family relationships in such rare sources are even fewer, and we must guard against assuming that all families were like the ones we know most about. The Pastons, a gentry family in fifteenth‐century England, left a huge number of letters detailing many instances of family conflict, and for a long time these were used as proof that all families were cold and bitter toward one another at this time, rather than that the Pastons were unusually dysfunctional.
Family Structures and Functions
Multigenerational social structures of related individuals are the oldest form of social organization. Scholars of early human evolution increasingly argue that the complex relationships that developed within families, similar to tool‐use, are a crucial driver in the evolution of the human brain. New technologies and ways of using them were not simply invented to solve problems or address material needs, but also to foster social activities, such as communal childcare, that would allow more children to survive (Hrdy 2009). Families in the Paleolithic lived by foraging, gathering plants and insects, trapping and hunting, and then cooking and eating in groups. About 15,000 years ago, as the earth’s climate entered a warming phase, humans began intentional crop‐raising and domesticated plants and animals, modifying them by selective breeding so as to better serve human needs. Thus people were “domesticated” – settled into sedentary co‐residential groups – before plants and animals were, not the other way around. Paleolithic and Neolithic people divided labor by gender, age, and perhaps other factors, but sources for this and other aspects of family life are fragmentary and difficult to interpret, and there may have been significant variety (see Chapter 11 by Marcia‐Anne Dobres in this volume).
In more recent eras, the structure, function, and even the definition of “the family” have varied tremendously from culture to culture, and for different social groups and genders within each culture. Some groups practiced polygamy and others monogamy; for some, the most important unit was the nuclear family of a man, a woman, and their children, while for others the extended kin network was most important; in some groups, the family was primarily a unit of reproduction, while in others it was primarily a unit of production; in some groups, married couples lived with the husband’s family (patrilocality), in others with the wife’s (matrilocality or uxorilocality), and in others in their own household (neolocality). In some groups non‐related individuals such as slaves or servants were considered part of the family, and in others they were not; in some groups adoption or godparentage created significant kinship‐like ties (termed fictive or spiritual kinship) while in others only blood mattered; in some groups marital partners were chosen by parents or the family as a whole and in others by the individuals themselves. In some groups a woman brought a dowry in goods or money to her husband on marriage while in others a man gave goods or money to his
