Quebec Marriage Contracts: Why the Notary's Copy Says More Than the Parish Register

You have found the marriage. It is in the parish register, in a priest's careful French, and it gives you a date, two names, the parents, and a short line of witnesses. For most countries that is where the marriage ends as a source.

For a French-Canadian family it may be only half of it. The same couple may also have sat in front of a notary and signed a contract. That document never appears in a parish index, because it was never a church record. It sits in the notary's own files, filed by date among his wills, land sales and debts, and you will only find it if you go looking for the notary rather than the name.

It is worth the search, because a marriage contract records what a parish register never tries to: what each side was bringing, what the bride would be entitled to if she was widowed, and how the families had arranged things around children from an earlier marriage.

The short answer. In New France and later Quebec, a marriage contract was a notarial act written against a legal default. From May 1664, when a royal edict made the Coutume de Paris the sole source of law in the colony, a couple who signed nothing were married in community of goods, managed by the husband, and the widow had a customary dower: the use, for her lifetime, of half of the property her husband had brought into the marriage as his own. A contract could keep that default, fix a specific dower instead, add a preferential share for the survivor, or separate the spouses' property altogether. So every clause that differs from the default is evidence of a choice someone made, and the contract also names the parties, their family ties, their trades and where they came from. To find one, identify the notaries practising where the couple married, then search their records, now largely digitised by Bibliothèque et Archives nationales du Québec (BAnQ) up to 1910.

The marriage left two records, and only one of them is indexed by name

The parish register and the notarial contract were made for different authorities and survive in different places, and that is why so many researchers find one and never look for the other.

The parish register was the Church's record of a sacrament. It is what the big indexes are built from, and it is what you already have.

The contract was a private legal agreement, drawn up and kept by a notary. BAnQ describes the notarial archives it holds as running from the time of New France to the present, and lists what a notary's files contain: on the family side, marriage contracts, wills, inventories after death, donations, and guardianships; on the land side, sales, concessions and leases; and on the economic side, receipts and obligations. A notary's collected acts are called his greffe, and they were kept as a working file, not as a register of the population.

That is the whole difficulty. A parish register is organised by parish and date, so a name search finds it. A greffe is organised by notary and date, so a name search finds it only if someone has indexed that notary's acts by the names inside them, and the New England Historic Genealogical Society, writing on its Vita Brevis blog on 1 February 2016, put the problem bluntly: indexes do not exist for every notary's records. Its advice was to start by establishing which notaries practised in the judicial district where your ancestor lived, and to remember that a family may have used more than one.

The default came from Paris, by royal edict, in 1664

To read a contract you have to know what the couple would have had without one, and that was set by law long before your ancestors married.

The Répertoire du patrimoine culturel du Québec, the Quebec government's heritage register, sets out the sequence. In May 1664 Louis XIV issued the edict creating the Compagnie des Indes occidentales, and article 33 of that edict made the Coutume de Paris the only source of law in the colonies. The Coutume itself had been written down around 1510 and reformed in 1580, when it ran to 362 articles, among them articles on matrimonial agreements, property and wills. The colony's Conseil souverain adjusted it in 1667, 1678 and 1685.

The Conquest interrupted it but did not end it. The same heritage register records that after 1763 the British Crown barred the use of the Coutume in favour of English common law, and that the Quebec Act of 1774 restored French customary law in civil matters while keeping British criminal law. It stayed in force, with amendments, until a codification commission set up in 1857 by George-Étienne Cartier produced the Civil Code of Lower Canada in 1866.

So a marriage contract signed in 1700, in 1800 or in 1850 was written against essentially the same background rules. That is what makes its clauses readable.

What a couple got without signing anything

The clearest modern account of how these rules worked in the colony is an article by France Parent and Geneviève Postolec, "Quand Thémis rencontre Clio : les femmes et le droit en Nouvelle-France", published in Les Cahiers de droit in 1995. Four rules from it matter for reading a contract.

Community of goods was the default. Under the Coutume, Parent and Postolec explain, the favoured matrimonial regime was community of goods, applied whenever no other agreement was set out in the marriage contract. The couple's common property was owned equally and managed by the husband.

The widow had a dower. The douaire was a gain of survival for the widow: the usufruct, meaning the use and income, of half of the husband's propres, the property he held as his own rather than as part of the community. It belonged jointly to the widow and her children, with the widow holding the usufruct. Crucially, it came in two forms: fixed in the contract, the douaire préfix, or by default under the custom, the douaire coutumier.

The survivor could have a preferential share. The préciput was an advantage given to one spouse, taken before the community property was divided.

A widow could walk away from a bad community. At the dissolution of the community, Parent and Postolec note, the surviving wife had an inventory drawn up and could close the community with or without renouncing it, or keep it running. Renouncing let her escape the community's debts. And the law was severe on widows who remarried: several articles of the Coutume took back the advantages of a first marriage, the dower and the préciput among them, for the children of that marriage.

One escape existed during the marriage too. A wife could seek a séparation de biens to protect her own property from a husband's mismanagement. Parent and Postolec's example from the Prévôté de Québec is a woman, Marie Varin, who asked for one because her husband was, in the court's phrase, struggling with drink, and she feared he would drink away her inheritance.

Every clause that departs from the default is evidence

This is the method, and the reasoning below is ours rather than any archive's, so read it as an interpretive approach, not a rule.

Because the default was known to everyone at the table, a contract did not need to say anything to get it. So whatever the contract does say, beyond the default, was put there on purpose. Read each clause and ask what problem it solved.

A fixed dower instead of the customary one. The customary dower depended on what the husband held as his own, which might be a lot or almost nothing. A specific sum or a specific property, written into the contract, suggests the bride's family wanted certainty rather than a share of an unknown quantity. It tells you a negotiation took place.

A préciput for the survivor. Since the préciput came off the top before division, it protected whichever spouse outlived the other from being stripped by heirs. It is a clause about what the families expected to happen if one of them died young.

Separation of property. Contracting out of community altogether is a strong signal. It implies at least one side had assets, debts or a business they wanted kept apart, and it is worth looking for the reason in the other acts of the same notary around the same date.

Arrangements for children of a first marriage. When a widow or widower remarried, the Coutume's rules about first-marriage advantages came into play. A contract that deals with the children of an earlier marriage is very often your clearest statement anywhere of who those children were and which parent they belonged to.

What each side brought. The Drouin Institute, which runs the Généalogie Québec site, notes in describing its notarial collection that a marriage contract may describe the property brought into the marriage and identify relatives of the spouses. A list of what a bride brought, down to a cow or a bed, is not trivia. It is a snapshot of a household at a datable moment.

The same discipline, reading a legal document for why it was made rather than for the family facts it happens to contain, applies to the Italian marriage packets that reach back to the 1700s, where a consent rule decided who had to be proved dead, and to what a pre-1858 English will actually proves, where a court was settling obligations rather than describing a family.

The opening of the act is where the genealogy sits

Before the clauses, a notarial act introduces the people in it, and that introduction follows a set pattern.

BAnQ's archival catalogue, Advitam, describes the pattern in its notice for one notarial sub-series, held at its Rimouski centre and covering 1871 to 1918, which runs to 10,559 minutes. Each act, it says, is drafted to a standard model: first a presentation of the persons, with their names, their marital or family ties, their trade or profession and where they came from, and then the main body of the text describing the nature of the act.

For a marriage contract, that opening is often richer than the parish entry, because it places both families by occupation and origin at a precise date. Read it slowly, then read the signatures and marks at the end, which show who put their name to the act.

How to find the contract when you only know the marriage

Work from the marriage you already have toward the notary, not the other way round.

Start with the parish and the date. You need the place of the marriage and its date, because they are the only fixed points you have for narrowing a notary's files.

List the notaries. BAnQ maintains a database, dated 11 May 2022, of biographical notes for 702 notaries whose greffes are held at its Quebec City archives, covering the judicial districts of Beauce, Montmagny, Québec, Saguenay (Charlevoix) and Thetford Mines, with each notary's years of practice and places of residence, for notaries active between 1626 and 1970. For other districts, BAnQ's own guidance is to consult Advitam or the staff of the relevant archives centre.

Search the indexes that exist. For the earliest period, the Parchemin database, compiled by the Société de recherche historique Archiv-Histo, describes itself as an almost complete inventory of Quebec's notarial minutes from the beginning to 1809, and Archiv-Histo sells subscriptions only to institutions. Gail Dever reported on her blog Genealogy à la Carte on 4 December 2020 that BAnQ had just made remote access to Parchemin free for BAnQ cardholders, which in practice means residents of the province. For later periods, Ancestry holds a collection titled "Quebec, Canada, Notarial Records, 1637-1935", and the Drouin Institute's Notarized Documents collection, which covers Quebec and Ontario, gives a figure of about 83,000 records.

Then read the notary's own index. BAnQ's digitised notarial collection gives access to the finding aids and to part of the acts themselves, for notaries who practised from New France up to 1910. A notary's répertoire lists his acts by date, so with a marriage date and a shortlist of notaries you can check the same stretch of dates in each one.

One caveat, stated honestly. The published counts for Parchemin do not agree. The FamilySearch Research Wiki gives 442,402 deeds from 1626 to 1799; Dever's 2020 post gives more than 559,168 records from 275 notaries up to the end of 1801; Archiv-Histo's own page gives no count and an end date of 1809. They may be counting different editions or different things. Do not quote any of them as the size of the surviving record.

How many couples signed one is not well known

The obvious question is how often you should expect a contract to exist, and here we have to leave a gap.

The best-known work on marriage contracts as a source in Quebec is by Bettina Bradbury, Peter Gossage, Evelyn Kolish and Alan Stewart, "Property and Marriage: The Law and the Practice in Early Nineteenth-Century Montreal", published in Histoire sociale / Social History in 1993. Its abstract reports three changes in Montreal between the 1820s and the 1840s: signing a marriage contract became increasingly the practice of a propertied minority; a rapidly growing share of those who did sign chose to keep the spouses' property separate rather than create a community, most of all when the husband was a merchant or "bourgeois", and not only among anglophones; and more wives in wealthier families appear to have gained the power to administer their own goods. We could read the abstract but not the full text, so we are not quoting a percentage, and that finding is about one city in one period. It does not tell you how common contracts were in a rural parish in 1720.

What can be said is narrower. A contract was a choice, not a requirement, since the default applied without one. So the absence of a contract in a notary's index is not evidence that the family had no property or no notary. It may mean they accepted the default, used a notary you have not checked, or used one whose acts were never indexed by name. That is a search worth recording as negative evidence, with the notaries and the date range you covered, not a conclusion.

What KleioBase does with a notarial act, and what it does not

Disclosure, since we sell a tool.

KleioBase is an AI research companion that turns scattered historical records into one connected family archive. A Quebec marriage contract is exactly the kind of record that otherwise sits in a folder, disconnected from the parish entry it belongs with.

Upload the pages of the contract, and the people, dates, places, occupations and relationships named in its opening are extracted, with the original French kept beside them so a phrase such as douaire préfix is still there to read rather than flattened into a field. Extraction goes through a Review step where every field can be corrected before anything is committed, which matters with an eighteenth-century notary's hand. The contract then links to the same people in the parish register and in the census, and the Research Companion can be asked what all of those records say together about one couple. When you search a notary's répertoire and come up empty, you can record the search that found nothing against the person, with the period you covered, so the gap reads as checked rather than unexplored. Upload and extraction cover French along with many other languages, as set out on our French records page.

Now the honest limits. Three of them.

It does not know the Coutume de Paris. It will extract a dower clause without knowing what the default would have been, so it cannot tell you that a fixed dower means a negotiation took place. That reading is yours.

It holds none of these archives. The acts are at BAnQ, on Ancestry, with the Drouin Institute and in Parchemin. Finding the notary and the page is work you do first.

Its record-gap suggestions stop short of notaries. On the Researcher plan and above, Expected records suggests census and vital-record targets for people in Canada, but it does not know that a French-Canadian marriage may have a notarial contract behind it, so it will never prompt you to look for one.

Where to start with the marriage you already have

Take one marriage from the parish register, and write down the parish, the date, and both sets of parents.

Then find which notaries were practising in that judicial district in the months before the date, using BAnQ's notary database or Advitam, and for anything before 1810, Parchemin if you can reach it through a library or a BAnQ card.

Open each candidate notary's répertoire and read the entries for the months around the wedding, starting with the weeks just before it and widening the window if nothing turns up. If you find the contract, read the opening for family ties and origins first, then read each clause against the default described above and ask what it was for.

Finally, write down every notary you checked and the dates you covered, whether or not you found anything.

The notary who drew up the contract was not recording a family for anyone's genealogy. He was writing down a bargain between two households, in terms set by a custom from Paris that the colony had made its own, so that a court could enforce it if one of the spouses died. That is precisely why it is so useful now. A bargain has to say who the parties were, what they had, and what they were afraid of.