Cross-cousin marriage
A Socratic walk-through of cross-cousin marriage — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do many societies forbid marriage to one kind of cousin while urging it with another?
In a great many societies the two cousins are treated as opposites. Your father's brother's daughter is unmarriageable — often she is not even called a cousin but a sister, and the thought of marrying her is incest. Your mother's brother's daughter, biologically no closer and no further, is the ideal wife, and in some systems the one you are expected to marry. Same degree of relatedness, opposite verdicts. So whatever is doing the work here, it is not blood.
Reasoning it through
REASONING #If not blood, then what distinguishes the two? Trace the descent lines. Suppose the society reckons descent through fathers, and every person belongs to the descent group of their father. Your father's brother is in your group, so his daughter is in your group. Your mother's brother belongs to the group your mother came from, so his daughter is in that group.
Now add the rule that these groups are exogamous — you marry out. The puzzle solves itself instantly: the parallel cousin is a group-mate, and marrying her would be marrying inside the group, which is exactly what exogamy prohibits. The cross cousin is outside. She is the nearest relative you are permitted to marry.
Which is worth pausing on. Permitted, yes — but why preferred? Being the nearest available option is not the same as being the best one. What does a group gain by sending its sons back to the same group its mothers came from?
Think about what a marriage does between two groups that have no state, no courts, and no contract law. It creates the one durable bond available: shared children, and an in-law relationship carrying claims to help, to feasts, to protection, to bridewealth. A single marriage creates such a tie; a rule that says marry your mother's brother's daughter makes the tie renew itself every generation without anyone negotiating it afresh. The alliance becomes a standing institution rather than an event.
Claude Lévi-Strauss built his account of kinship on that observation, and drew a distinction worth keeping. If two groups simply exchange women with each other — my sister for yours, generation after generation — you get a closed pair. He called that restricted exchange, and bilateral cross-cousin marriage is its signature. But suppose the rule is asymmetric: men take wives from their mother's brother's group only, and give their sisters to a different group. Then wives move in one direction around a chain, and the chain must eventually close on itself into a circle. Group A gives to B, B to C, C back to A.
Consider what that buys. Every group is bound to two others rather than one, and none can be self-sufficient in wives. Lévi-Strauss called this generalized exchange, and argued it links a far wider society together, at the price of requiring something like trust: you give away your sisters now and receive your wives from a different quarter, so the exchange is deferred rather than settled on the spot.
He also predicted that marriage with the father's sister's daughter — the other cross cousin — would be rarer and less stable, because it reverses the direction of flow with each generation and so never accumulates a standing obligation. That prediction is contested, and the ethnographic record is messier than the model.
The analogy
THE ANALOGY #Think of two trading houses that intermarry deliberately in every generation to keep a partnership alive, versus a ring of houses in which each supplies the next and takes from the one before. The pair is simple and self-settling; the ring binds far more parties together, but only works if each is willing to give before it receives.
Merchant families choose these matches case by case and can stop when it suits them, whereas a prescriptive kinship rule assigns the category of spouse in advance, to everyone, as a matter of what kin terms mean — nobody is deliberating.
Clarifying the model
THE MODEL #Three corrections, the last of them serious.
First, these rules are usually about categories, not individuals. In many such systems "mother's brother's daughter" names a whole class of women, and marrying any of them satisfies the rule. Reading the prescription as being about one specific first cousin misstates what the system says.
Second, and following from that: Rodney Needham insisted on separating the terminology from behaviour. A society may have a prescriptive terminology while the actual rate of marriage with a genealogically real cross cousin is low. So a "rule" of this kind is often a statement about how kin are classified rather than a description of who weds whom — and the alliance account is on firmest ground when it explains the classification.
Third, the whole story assumes what needs stating: it works only where there are unilineal, exogamous descent groups. Where the pressing problem is different, the rule inverts. The best-known counter-case is the preference for father's brother's daughter marriage across much of the Middle East and North Africa — marriage inside the patrilineage, which alliance theory says should not happen. There the usual reading is that keeping land, herds and dowry within the lineage outweighs any gain from external alliance. That is not a small exception; it is the natural falsification test. If the cross-cousin pattern is about repeated alliance between exogamous groups, it should appear where such groups exist and inter-group ties are valuable, and should give way to endogamy where property retention dominates. It would be refuted by finding cross-cousin prescription flourishing in societies with no exogamous descent groups and nothing to gain from renewed alliance.
A picture of it
THE PICTURE #How to readThis is a class diagram borrowed to show groups and the directed relation between them, not objects in code. Each box is a descent group, and each arrow is the direction women travel in marriage. Start at Ego's group in the middle: it takes wives from A, which is why the mother's brother's daughter is the prescribed spouse, and gives its own daughters onward to C rather than back to A. Follow the arrows and notice they form a closed ring — that closure is the whole point of generalized exchange, since every group depends on one partner for wives and a different one for sons-in-law, and none can stand alone.
What became clearer
WHAT CLEARED #The distinction between the two cousins has nothing to do with biological closeness and everything to do with which side of a descent boundary each falls on. Once groups are exogamous, the parallel cousin is a sibling and the cross cousin is the nearest outsider — and a rule that keeps sending spouses to the same category turns a one-off marriage into a permanent alliance between groups. Reverse the underlying problem, so that holding property inside the lineage matters more than binding lineages together, and the preference reverses with it.
Where to go next
ONWARD #- Why bridewealth and dowry move in opposite directions and what each secures.
- How Needham's insistence on separating terminology from behaviour reshaped alliance theory.
Key terms
TERMS #| Term | What it means |
|---|---|
| Cross cousin | the child of a parent's opposite-sex sibling: mother's brother's or father's sister's child. |
| Parallel cousin | the child of a parent's same-sex sibling, and in unilineal systems typically a group-mate classed as a sibling. |
| Exogamy | the rule that one must marry outside a defined group. |
| Generalized exchange | an asymmetric marriage system in which wives circulate in one direction around a closed ring of groups. |
Every term the collection defines is gathered in the glossary.