Bridewealth and dowry

Why does wealth flow toward the bride's family in some societies and toward the groom's in others?

Two societies, both marrying off daughters, and the money runs in opposite directions. In one, the groom's kin hand over cattle, hoes or brass rods to the bride's father, sometimes over years. In the other, the bride's family assembles land, jewellery, linen and cash and sends it with her.

The tempting reading is that the first values women and the second does not, or the reverse. That is satisfying and explains nothing: it does not tell you which societies do which, and makes no prediction that could fail. So ask a duller question. What is each payment for, and what must be true of a society for that purpose to arise?

Start with what a marriage moves: a person's labour, her residence, and — carrying most weight — the group membership of her children. Where descent groups are corporate, holding land and cattle and answering for each other's debts and killings, her children belong to one lineage and not the other. A marriage is a transfer of productive and reproductive capacity between corporate bodies.

When does such a transfer have to be paid for? When what moves is genuinely valuable to the group losing it. So look at where women's work sits in production. In hoe agriculture, and in herding societies where women do the milking and processing, a woman's output is substantial and her absence felt: her natal group has lost something. The compensation is bridewealth, and it does a second job at once, since her kin hold it as a stake in the marriage continuing — if she is mistreated and returns home, they must give it back.

That predicts several things that hold: polygyny common, since wealth converts into wives and so into children and labour; the payment raised by a group rather than an individual, which makes a young man's marriage his lineage's business; divorce complicated, because unwinding it means unwinding the transfer. It also explains why bridewealth is rarely one payment but instalments completing only when children are born — what you would expect if what is bought is rights in offspring rather than a person.

Now change the productive system. Bring in the plough: heavy work with a heavy animal, and where it dominates, field labour shifts markedly toward men and women's direct contribution to the staple falls — Ester Boserup's observation, pressed further since. A woman leaving her natal household is now a smaller economic loss and a larger cost to feed in her new one, and the compensating payment has no reason to exist.

But dowry is not simply the absence of bridewealth. Something still moves, the other way. What?

Jack Goody's answer is the one to reason through: dowry is inheritance, paid early. Where property is heritable, stratified, and passes to daughters as well as sons, the dowry is the daughter's share of her natal estate, handed over at marriage and forming the conjugal fund of the new household. That is why it is found alongside monogamy, class stratification, plough agriculture and heritable land — and why its contents look like a household's founding kit rather than a fee.

Dowry does a second thing, easily confused with the first. In a stratified society marriage is how families place daughters, and a woman marrying up competes for a scarce well-placed groom, so the dowry can escalate into a bid. That reading — dowry as groomprice — explains what inheritance does not: dowry inflation, where demanded sums outrun wealth and the payment goes to the groom's family rather than staying with the couple. Both are real, and which dominates depends on how competitive and hierarchical the marriage market is.

Three corrections.

First, this is not a claim about how women are valued. Bridewealth attaches to societies where women's labour is economically central — a statement about the production system, not about respect, and there are bridewealth societies where women's position is poor and dowry societies where it is not.

Second, this is a different question from the one cross-cousin marriage answers, though the systems interlock. Alliance theory asks which group a spouse comes from; this asks which way the goods travel once she does. The fixed difference: alliance is set by descent-group structure, payment direction by production and inheritance, so one can change while the other does not.

Third, the evidence base is a coded one. The cross-cultural claims rest on the Ethnographic Atlas and its successors: one row per society, coded by outsiders from ethnographies of very uneven quality, describing an "ethnographic present" often reconstructed as things were said to be before colonial disruption. Its societies are not independent — neighbours share history and are counted separately, which is Galton's problem, and it inflates every correlation drawn from the table. I recall bridewealth of some kind in a large majority of coded societies and dowry in a small minority, but I am recalling proportions rather than citing them, and the coding categories are disputed, so I will not give figures.

What would refute the account? It predicts that payment direction tracks production and inheritance — women's contribution to the staple, corporate descent groups, whether property descends to daughters, degree of stratification — rather than religion, language family or region. So the refuting observations are specific: bridewealth flourishing where women contribute little labour and no corporate lineage has an interest in the children; dowry in unstratified monogamous societies with nothing heritable for a daughter to be given a share of; or payment direction reversing when a religion changes while the farming system stays put.

• Why dowry inflation accelerates in some societies while bridewealth rarely runs away in the same fashion.
• What happens to both when wage labour lets a young couple found a household without either family's transfer.

Every term the collection defines is gathered in the glossary.