The words the documents use

The Programme de recherche en démographie historique at the Université de Montréal: a reconstruction of the population of early Quebec, built by transcribing the Catholic parish registers act by act and linking the people in them into families. Everything in it is numbered, which is why citations here read PRDH couple #293884 or individual #113510 — a stable address for a specific record rather than a page reference. It is exceptionally good, and it is a transcription: someone else has already read the act and decided what it says. Under this site’s standard that makes it compiled, and a claim resting on it alone is graded no higher than Probable.

The Institut Généalogique Drouin’s microfilming of Quebec’s Catholic parish registers, now digitised — images of the register pages themselves, not transcriptions of them. Reading one is sighting the act. The images are licensed, and none are reproduced on this site: the Quebec parish acts here are described and quoted, never shown.

The Bibliothèque et Archives nationales du Québec. It holds the civil deposit of the parish registers — the registres d’état civil series, on which see greffe below — along with the notaries’ minutes and the courts’ records. The rebellion files behind Isaac Oligny’s page, the Événements de 1837-1838 (fonds E17), are BAnQ holdings.

The federal archive, formerly the National Archives of Canada. What this site draws from it is British military and colonial administration — militia rolls, governors’ correspondence, the RG-8 and MG-24 series. Where a Quebec ancestor appears in an English-language record before 1867, this is usually why.

A published index, compiled from the registers of one parish or one region, listing marriages in order with the names of the couple and their parents. Genealogical societies produced hundreds of them, and they are how an act gets found at all. What a répertoire is not is the act: it is someone’s reading of the act, condensed, and this site has published a correction that turned on exactly that difference. Compiled, graded no higher than Probable, and always superseded by the original once it is seen. A notary’s répertoire is a different object with the same name — his own running index of the acts he drew, numbered in sequence.

A greffe is the registry office of a court. Quebec’s parish registers were kept in duplicate: the curé wrote each act out twice, one copy staying at the presbytère — the parish house — and the other deposited annually with the civil authorities, where it became the greffe copy. Both are originals in the same priest’s hand, days apart, so damage to one implies nothing about the other. The trap is that not every volume filed under the greffe label is that second original; some are a nineteenth-century clerk’s transcription of the register, which is compiled however old the ink looks. How to tell them apart is a standing rule on this site, and the reason one of its central kinships is graded Probable rather than Proven. One further sense: a notary’s own accumulated series of minutes is called his greffe too, which is why a citation here can read greffe of Bénigne Basset and mean a private repository rather than a courthouse.

The standard interchange format for genealogical data — a plain-text export of a family tree, which is how this project’s own research database — some 22,300 individuals in rather over 10,600 families, and still growing — is handled when it is being counted rather than browsed. Every computed cousin degree on this site is an exact statement about that file, and only as good as the documents underneath it.

The two words that decide a grade here. A record is sighted when someone on this project has read the document itself, or a direct image of it. It is compiled when the reading was done by somebody else first — a database, a répertoire, a published genealogy, a later clerk’s transcription. Compiled never exceeds Probable, and a descent is only as strong as its weakest joint: one compiled link caps the whole chain, however many sighted acts sit above and below it. The five grades.

Codes appearing in some citation lines, pointing into this project’s own working research log — a set of Evidence Registers kept separately, where every source is numbered as it is found. They are not links; the plain description beside them is the actual citation. What they signal, and why they are unpublished.

An entry in a parish register — a baptism, a marriage, a burial — written out and signed by the officiating priest at the time, with the parents, the sponsors and the witnesses named, and each of them signing or declaring they could not. It is the foundational document of French-Canadian genealogy and the thing most often meant on this site by the record. An act is contemporary and primary; that does not make everything in it primary, since an age or a birthplace stated in a burial act is somebody’s later recollection sitting inside an original document.

The parish priest, and the man whose hand every act is written in. Where a parish had no resident curé — common on the Richelieu frontier — a priest from a religious order might serve in his place, and the act says so: a Récollet acting as curé officiated at a 1741 Chambly marriage on this site’s direct line.

The public announcement of an intended marriage, read out in church on three successive Sundays so that anyone knowing a reason the couple could not marry had the chance to say so. A couple in a hurry could ask to have one or two of the three set aside — a dispensation of two banns, which the church granted routinely and which says nothing on its own about why. Not to be confused with a dispensation for kinship, which is a different permission for a different obstacle.

A bishop’s written permission to set aside a rule that would otherwise block or delay a marriage — the banns schedule, or an impediment of blood kinship. It is recorded in the marriage act itself, in the curé’s own hand, naming which bishop granted it and, for kinship, how close the couple were. A dispensation clause is a bishop doing his job, not a family confessing to something; and because it was written down, it is now the hardest measurement this site has of how small the marriage pool had become.

Blood kinship, and the impediment it raised. The Church counted the degree as the number of generations back to the common ancestor, and a clause naming two numbers — du troisième au quatrième degré, third to fourth — is telling you the two sides stood at different distances from it: three generations on one side, four on the other. On the 1854 marriage at St-Jean-Baptiste-de-Rouville that clause resolves exactly, to second cousins once removed. Kinship beyond the fourth degree needed no dispensation at all, which is why an act can be silent about a relationship that genuinely existed.

The two sponsors named in every baptismal act, standing for the child at the font. Genealogically they are worth as much as the parents: sponsors were chosen from among kin and close neighbours, so the pair of names records who a family had, and who it trusted, on one dated afternoon. The earliest documented contact between the two families on this site is a godmother at a 1721 baptism.

Two waves of women who crossed to New France to marry, either side of one royal decision in 1663 — the earlier group unsponsored and travelling on their own account, the later with their passage paid from the King’s purse and a dowry attached. Eight of them are ancestors of both the 1912 bride and the 1912 groom. The full distinction, and why it decides how a woman was written down.

Widow — and, in an act, often the whole of what a woman is given. A remarrying widow is regularly identified by her late husband rather than by her parents, and a burial act may enter her as la veuve so-and-so with no name of her own at all. That is not an error to be corrected; it is the record doing what it did, and this site published a correction after mistaking one such entry for a name that had been lost.

In French civil practice a notary is not a witness to a signature but the drafter and keeper of the act. Marriage contracts, estate inventories, land sales and guardianships were all drawn before one, and the original he retained — the minute — stayed in his own numbered series, which is why an act from 1684 can still be read on the page he wrote it on. Notarial minutes survive in enormous numbers and record property and family arrangements the parish registers never touched.

A property agreement signed before a notary, usually within days of the wedding and sometimes years before it, settling what each side brought and what the survivor would keep. It is a separate document from the church act and frequently more informative, because it names parents, home parishes and previous marriages in order to establish who is entitled to what. Several of the origins on this site rest on a contract rather than on an act.

Inventaire après décès — a notary walking through a dead person’s house with the heirs and listing what is in it, room by room, valued item by item, so the estate can be divided. It is the closest thing genealogy has to a photograph of a household, and it establishes, in passing, exactly who the heirs were.

The guardian appointed over children who still had property and had lost a parent — tuteur a man, tutrice a woman. It was not an informal arrangement: a council of relatives assembled before a judge, elected the guardian, and the choice was then confirmed by the court. A surviving mother was regularly elected over her own children, which is how a widow with no independent legal standing to speak of ends up on the record running the estate. Marguerite Marotte was elected tutrice of four minors in 1825.

The court’s confirmation of a decision taken elsewhere — most often the family council’s election of a guardian. A tutorship homologated on a given date is one the judge has ratified, and the date is the one that enters the record. It is a useful date genealogically because it sits within weeks of a death that may itself be undocumented.

Land in New France was granted in long river-front blocks to a seigneur, who conceded strips within it to settlers in return for dues. An habitant is one of those settlers — a landholding farmer, not a labourer, and the word is a status rather than a description. Disbanded soldiers of the Carignan-Salières regiment were settled this way in 1668, which is how several of the men on this site came to hold land at all.

Two Massachusetts courts, and the distinction matters on this site because one family used both. The Probate Court of a county handled estates, guardianship, adoption, name changes and separate support; the Superior Court handled divorce. A 1916 filing and a 1921 filing by the same woman in the same county are therefore in two different courthouses, two different docket series, and are two different proceedings — which is precisely the error this site had to withdraw.

A Massachusetts remedy, filed in the Probate Court, that orders a husband to provide maintenance and settles custody of the children without ending the marriage. It was what a woman filed when she needed money and her children rather than her freedom, or when she had no ground for divorce that the law would accept. The 1916 petition on this site is one of these, and reading it as a divorce is a mistake the site made and corrected.

In Massachusetts divorce practice the petition was called a libel and the spouse who brought it the libellant — an old term of art from admiralty procedure, carrying none of the modern sense of defamation. To libel a spouse for divorce is simply to file. The libel states the grounds; a printed block of ruled lines for the specifications, the detailed particulars, may be left entirely blank.

A divorce decree that is granted but not yet final. Nisi is Latin for unless: the decree takes effect at the end of a waiting period — six months, under the printed form used in Massachusetts in 1921 — unless cause is shown in the meantime why it should not. During that interval the parties are divorced in every respect except the one that matters, and neither is free to remarry.

The day the nisi period runs out and the divorce takes effect. It is a date nobody attends and no document is signed on, which is why it is so often got wrong — the clerk’s later note closing the case file is not it. On this site the correct date came from a certificate of the court, issued the morning after the decree and filed a year later inside a New York marriage licence as proof the bride was free to marry.

The first of the two American naturalisation papers — a sworn statement that the applicant intends to become a citizen and renounces his former allegiance, filed years before the petition that completes the process. Genealogically it is often the better document of the two, because it asks for a birth date, a birthplace and a date and vessel of arrival while the applicant still remembers them. Filing one is not becoming a citizen, and whether the process was ever completed is a separate question that the declaration alone cannot answer.

The original application for a Social Security account number, first filed in 1936–37. It asks for the applicant’s name, date and place of birth, and both parents’ full names including the mother’s maiden name, in the applicant’s own words rather than a relative’s — which makes it, for the generation that filed one, a first-person statement of parentage. It is obtainable from the Social Security Administration under the Freedom of Information Act, and one of them has already overturned a claim on this site.

A second surname attached with the word dit, “called” — the colony’s answer to a small founding population producing three René Roys in one parish. It functioned as legally as the surname: children were baptised under it and land deeded under it, and after a generation or two it could simply replace the name it had been added to clarify. Four ways a family could acquire one, and one line where the takeover can be watched happening.

A barracks name, issued to a soldier on enlistment the way a uniform was issued — flowers, rivers, qualities: Larose, Richelieu, Laperle, Brisetout, Sanssoucy. Men kept them after discharge and passed them to their children, so a large share of dit names in this family are military in origin. They also make the surviving rolls hard to use, since a roll may list its men by nom de guerre and nothing else: a nickname appearing once in such a roll is evidence, and the same nickname appearing twice is nothing at all.

An illiterate family’s surname spelled by ear, by whichever clerk or notary happened to be holding the pen. Saurette, Sauret, Soret, Sorel, Sorette, Larose, Chorel — one family, filtered through however many different people wrote its name down, and the same mechanism running again a century later through American census enumerators. Variant spellings are not evidence of different families and are not treated as such here.

The name a person actually went by, as against the name on the baptismal act. The two can be unrelated: a man baptised Josué-Armorique was Alméric his whole life. Because censuses and civil records take the call-name and parish records take the baptismal one, the same person can appear under two names that share nothing but a household, and an enumerator’s misheard third version is common on top of that.

This project’s own term for the arrangement that has caused every published correction so far: two men of one name, close in place and time, in a record set that gives no unique identifier. It is the exact configuration in which genealogy software silently merges two people into one, and the only defence is to look at the generation above and below rather than at the man in front of you. Two Étienne Benoits, both ancestors of the same woman, is the worked example.