How arranged marriage actually works

The phrase suggests a decision taken over two people's heads. What actually happens is a sequence: an enquiry in both directions, an intermediary, a first meeting, a list of terms, an engagement, a wedding. Each stage can end it, and the two people it is supposedly about hold a stop button at every one of them.

Calling it a decision compresses the interesting part into nothing. The interesting part is the negotiation — who opens contact, who carries the messages, what gets checked, and who is allowed to say no. That is also where the variation lives, far more than in the question of whether the couple knew each other beforehand.

It is a process, not a verdict

The standard sequence runs roughly like this. A household decides that a child is of age; relatives and networks are told; candidates are suggested, usually from within the same community and often from the same narrower circle within it. Both sides then investigate. The enquiry is not casual. Household composition, whether the family owns or rents, the occupations of the parents, the stability of the son's employment, the health of the family, the horoscope, and reputation — meaning how this household has behaved in previous alliances.

The woman's side is not a passive party waiting to be chosen. It runs the same checks in the other direction, and adds a question the man's side rarely asks: is this a household a daughter can live in — who cooks, who holds authority, whether the mother-in-law has a reputation for making life difficult. That is a mechanism rather than an attitude, and it is one reason refusal so often comes from the bride's side at the final stage.

Terms follow inspection. Where the couple will live, whether the wife will work, how often each side is visited, what each household contributes to the wedding, and what happens to parents in old age. These are stated, argued over and settled before the engagement, which is the point at which the arrangement becomes public and expensive to reverse.

Who opens the contact, who can refuse

The variable that distinguishes one arrangement from another is not the presence of feeling. It is who makes first contact, and who holds the right of refusal. In the classic form the families open contact, and the principals meet once, in a room with relatives present, for perhaps half an hour, and are asked separately afterwards.

That veto is real but narrow. It is negative power: neither principal usually selects, both can block. And it is exercised once, in private, usually as a flat no rather than a negotiation, because a qualified yes can be heard as a demand and a demand is an insult. That asymmetry produces a process which optimises for acceptable-to-both rather than preferred-by-either — which is what a contract between households is for.

What has changed is the amount of information available before the veto point. The first meeting increasingly begins on a messaging app and runs for weeks before any parent meets anyone. The structure is unchanged; the principals simply arrive at the meeting already knowing how the other person writes and what they sound like. Fewer refusals happen at the meeting, and more happen earlier.

The semi-arranged variant

In the urban middle class the order of the steps is rearranged. Parents build a shortlist, or a matrimonial site builds it with filters; the two people talk directly; the families meet later; the decision is announced jointly. This is often described as the couple choosing. It is more accurate to say that the search function has been reassigned and the approval function has not.

The website is the clearest artefact of the system, because it is a database of households rather than of individuals. Its filters — community, mother tongue, region, education, income, diet — are product features, not prejudices bolted on. They exist because the enquiry they replace was doing exactly that work, and a product that did less of it would be less useful to the people buying.

The consequence is a longer list of shorter attempts. Saying no is cheaper when there are more candidates and when the refusal happens in a chat window rather than a drawing room, so introductions multiply and each one lasts less time. The end point is the same document, arrived at through more tries.

What is being negotiated, and who pays

The object of the negotiation is not romance. It is the terms on which two households will be connected for decades: residence, obligation, visits, support in illness and old age, and the reputational stake both families hold in the match lasting. That is why the process looks bureaucratic. It is a contract between organisations, executed by two individuals who are parties to it rather than authors of it.

Money moves, and the direction is not uniform across the subcontinent. Where the transfer runs from the bride's side it is called dowry. The Dowry Prohibition Act 1961 prohibits giving or taking it, and defines it as property given as consideration for a marriage. The definition is where the gap opens: gifts that are voluntary and undemanded fall outside the Act, and almost any transaction can be described that way if both sides prefer the description.

Enforcement is the second gap. Cases are registered, convictions are far fewer, and the transaction is undocumented by design — a demand made in conversation leaves no evidence unless somebody chose to record it. So the law changed what is legal without changing what is expected, and most of the scholarship on dowry is an argument about that distance rather than about the practice itself.

What the numbers can and cannot show

The surveys give ages. Median age at first marriage is about 21 for women and about 25 for men; 23.3% of women aged 20-24 were married before 18. Those are outcomes. They say nothing about procedure — who spoke first, how many proposals were declined, what was transferred, or which of the two households was keener.

On procedure the evidence is ethnographic: household studies, kinship studies, village monographs. Rich on mechanism, useless for aggregation, which is why every national claim about arranged marriage is really a claim about one place generalised. The honest version is that the sequence is remarkably stable and the content of each stage is not.

Which is why the moral vocabulary usually misfires. It is not a decision imposed on two people, and it is not a free choice either. It is a negotiation between households in which the principals hold a veto and almost never hold the initiative — and the ratio between those two powers is what varies, by class and region and decade, far more than the presence or absence of affection ever does.

This is a piece in Everyday Life, part of About India.