Moreover, through most of human history, marriage united not just two mates but two sets of families. When pairing off unites two kin groups instead of two individuals, that is much more than an expression of the biological functions of mating and reproduction. It is a transformation of those functions.
Finally, no other animals have elaborate rules about whom one should, must, or cannot marry. By contrast, notes anthropologist Meyer Fortes, marriage practices among humans “are universally subject to rules.”
4
For most of history, those rules have been much more complex and far-reaching than simple prohibitions against incest, rudimentary forms of which may also exist among some primates. The rules governing who can marry whom have varied immensely from group to group. In some societies, marriages between first cousins have been prohibited. In others, such unions have been preferred. Some societies have encouraged polygamy. Others strictly prohibit it. Such a contradictory hodgepodge of social rules could not have sprung from some universal biological imperative.
The same holds true for the wide gamut of beliefs over the ages about how marriage should be organized and what its main purpose should be. So once we get past the seeming universality of marriage and examine the tremendous variations in the role it plays in different societies, it becomes much harder to define marriage and its reasons for existence.
In 1949 the eminent anthropologist George Peter Murdock defined marriage as a universal institution that involves a man and a woman living together, engaging in sexual activity, and cooperating economically.
5
At first glance, this seems a commonsense definition. In fact, however, there are many exceptions to this kind of marriage arrangement.
For example, in many times and places, husbands and wives routinely lived in separate residences. Among the Ashanti of Ghana and the Minangkabau of Indonesia, men traditionally live with their mothers and sisters even after marriage. Men of the Gururumba people in New Guinea sleep in separate houses and work separate plots of land from their wives. The only time husbands and wives are together on a daily basis is when the main meal is being cooked and eaten.
6
In Zambia, Bemba husbands and wives traditionally do not even eat together. Men and women eat separately, as do boys and girls, in a variety of meal-sharing groups organized by gender, age, kinship, and friendship. In Austria in the eighteenth century, lower-class married couples commonly lived apart for many years as servants in other people’s houses, taking their meals with their employers rather than their spouses. All these people would be puzzled by our periodic panics about how rarely contemporary families sit down to dinner together.
If living together is not always what defines a marriage, neither is economic cooperation always the rule. Among the Yoruba and many other African societies, husbands and wives do not pool resources in a common household fund. Sometimes a couple doesn’t even share responsibility for their children’s economic welfare. The child is supported by one parent’s lineage rather than by the married couple. If the couple divorces, the child may not even be viewed as biologically related to the parent whose lineage isn’t economically responsible for him.
7
Faced with so many “exceptions” to Murdock’s 1949 attempt to define marriage, the Royal Anthropological Institute of Britain took a stab at it. The institute, focusing on marriage’s role in determining the status and rights of children, defined marriage as “a union between a man and a woman such that children born to the woman are the recognized legitimate offspring of both partners.”
8
This definition also proved to be too restrictive.
There are West African societies in which a woman may be married to another woman as a “female husband.” In these cultures, if the wife brings children with her to the marriage or subsequently bears children by a lover, those children are counted as the descendants and heirs of the female “husband” and her extended family. And numerous African and Native American societies recognize male-male marriages.
9
What about traditional Chinese and Sudanese ghost or spirit marriages, in which one of the partners is actually dead? In these societies a youth might be given in marriage to the dead son or daughter of another family, in order to forge closer ties between the two sets of relatives.
For most of Chinese history the decision to arrange a ghost marriage was made by two sets of parents, without regard to the youth’s wishes. But in the early twentieth century some women actively sought such marriages. This was common practice among female silk producers in the Canton delta, who wanted to maintain their economic independence but whose families wanted in-laws. Most parents would not allow more than one daughter to remain unmarried. So if one daughter had already declared herself a spinster, her sister had to conduct a marriage ceremony with a dead man, called marrying a tablet, to retain her independence. These women later told historians that “it was not so easy to find an unmarried dead man to marry,” so when one did become available, they vied with one another “to be the one who would get to marry him.”
10
Over the millennia the preferred form of marriage in many cultures was that between a man and several women. More rarely, marriage might unite a woman and several men. Among the Toda of southern India, a girl was married off at a young age, sometimes as early as two or three. From then on she was considered the wife not only of the boy to whom she was married but of all his brothers as well. When the girl was old enough to have sex, she usually had sexual relations with all her husbands. When she became pregnant, one of the brothers gave her a toy bow and arrow and promised her the next calf from his herd. That man was henceforth seen as the father of all subsequent children the woman bore—unless she performed the bow ceremony with someone else.
11
These forms of marriage are rare, at least in the modern world. So anthropologist Suzanne Frayser sampled sixty-two societies from around the world to calculate which functions marriage performs most frequently. On the basis of her statistical analysis, Frayser defined marriage as “a relationship within which a society socially approves and encourages sexual intercourse and the birth of children.”
12
But marriage has taken so many different forms in history that trying to define it by its most frequently encountered functions does not really help us understand what any particular society’s marriage system is or how and why such a system changes over time. We also can’t claim some groups did not have “real” marriages just because their marriage practices were not “typical.”
Three prominent anthropologists recently argued that while “there are a few exceptions to virtually any definition,” the point is that “there are commonly stable, mated relationships between females and males in every human society.”
13
That is certainly true. But it is only part of the picture. For example, many societies prohibit some people from marrying even if they have stable, mated relationships that produce children, but allow other people to marry even if they do not engage in sexual intercourse or cannot bear children.
Throughout history and across the globe the huge majority of marriages have been between heterosexuals, even in societies where same-sex marriages have the same legitimacy as heterosexual marriages. But in most societies not all heterosexual relationships count as marriage. Few societies in history have given heterosexuals who live together outside marriage the same legal rights as married persons, even if the cohabiting couple is in a long-term, stable relationship with several children. In fact, in societies that recognize same-sex marriages, such unions, though numerically rare, have a firmer legal standing than the relationships of unmarried heterosexuals who live together and have children.
A different approach to defining marriage is taken by anthropologist Edmund Leach, who suggests that marriage should be seen as being more about regulating property than regulating sex and child rearing. He argues that marriage is “the set of legal rules” that govern how goods, titles, and social status “are handed down from generation to generation.”
14
In most complex civilizations, inheritance rights have indeed been at the center of marriage. This meant that the definition of a “legitimate” marriage was a burning and often a disputed question. However, in some societies, inheritance rights do not depend on marriage. A child born out of wedlock among the Kachin of northern Burma was counted as legitimate if the father paid a fine to the girl and her family. Among the Kandyan of Sri Lanka, by contrast, a child’s legitimacy derived from the mother. As long as the presumed father was not from a caste lower than the mother’s, his actions, intentions, and marital status had no impact on the child’s status.
15
Another wrinkle in the relationship between marriage and inheritance is found in those Middle Eastern societies that recognized the pre-Islamic tradition of mut‘a, or temporary marriages. These were designed to allow sexual outlets for men and women under certain circumstances without subjecting them to the otherwise harsh penalties for nonmarital sex. Mut‘a was condemned by Sunni Muslims but accepted by Shi‘ites and by Babylonian Jews, who allowed a sage entering a new town to request a “wife for a day.” In these temporary marriages the man and woman had no obligation toward each other once the contract was over. But if the woman bore a child as a result of the relationship, that child was legitimate and was entitled to share in the father’s inheritance.
16
Some societies pay no attention at all to “legitimacy” in determining a child’s rights. When Jesuit missionaries from France first encountered the North American Montagnais-Naskapi Indians in the early seventeenth century, they were shocked by the native women’s sexual freedom. One missionary warned a Naskapi man that if he did not impose tighter controls on his wife, he would never know for sure which of the children she bore belonged to him. The Indian was equally shocked that this mattered to Europeans. “You French people,” he replied, “love only your own children; but we love all the children of our tribe.”
17
The concept of illegitimacy is completely foreign to matrilineal societies, such as the Navajo people of North America, in which descent and inheritance pass through the female line. But even some patrilineal societies give inheritance rights to the child of an unmarried woman. Among the LoWiili people in Africa, if a household needed more members, the head of the household might encourage an unmarried daughter to bear a “house child,” who would become a member of its maternal grandfather’s descent group.
18
Japan had no equivalent to the English word
bastard
until the Meiji Restoration in 1868. Only then did Japanese reformers adopt Western distinctions between legitimate and illegitimate children. Prior to that time, the language had a word to indicate that a child had been born to a concubine rather than a wife, but such a child was not necessarily denied inheritance rights or legal recognition. Indeed, the Taishō Emperor, who ascended to the throne in 1912, was the son of a concubine of the last Meiji emperor.
19
In societies where inheritance rights depend upon legitimacy, marriage is usually an elaborate ceremony that confers a whole package of rights and obligations on the partners—but only if all the procedures and social exchanges required by law or custom are carried out. In these cases, people have traditionally been enormously concerned to prove that their marriages were legally binding or to prove that someone else’s marriage was not. A person’s future could ride on whether the authorities declared that the marriage had been contracted in the proper manner and conducted with all the necessary rituals.
In other cultures, marriage may be nothing more than a public acknowledgment that a man and a woman have become a regular couple or are raising a child together. Among the Mbuti Pygmies of the Congo, a couple is considered married if they have lived together for two seasons.
20
In some small-scale societies, if a man and woman are seen eating together alone, they are considered married. Among the Vanatinai of the South Pacific, studied in the 1970s by Maria Lepowsky, unmarried couples may sleep together, but until they intend to move their relationship to a new stage, they do not eat together separately from their kin or other social groups. “The act of marriage consists of the new husband’s staying in the house with his wife after dawn and eating the breakfast his bride prepares.”
21
Anthropologist Edmund Leach, working in Sri Lanka, was startled to be told by villagers that a nineteen-year-old woman had already been married seven times. When he asked how that could be, he was told “that if a girl was seen to be cooking a meal for a man this was evidence that she was ‘married’ to him.” Often the corollary is that when a woman stops cooking for a man, the marriage is considered over. And among the Gururumba of New Guinea, males and females who are not married to each other never eat meals together, because eating cooked food together is considered the equivalent of having sexual intercourse.
22
Despite all these variations in the social role and meaning of marriage, through most of history marriage has generally involved a societally approved division of labor between the partners, with each sex doing different tasks. If a man went out on long hunting trips, which always ran the risk of his coming home empty-handed, it was good to have a woman gathering plants and nuts or tending crops. If a male was trapping animals, it helped to have a female manufacturing pottery and clothes. For millennia, one reason people married was that an individual simply could not survive trying to do everything on his or her own.