Case Study · The Guilbault Line

The Voyageur Years

Three generations, one ledger account,
and the merger that ended the evidence

Three men in this family were named Gabriel Guilbault. Three were called voyageur, in three different decades, on three very unequal kinds of evidence. One account survives in the North West Company’s general ledger. This is how you determine whose it is — and what stays undetermined when the archive stops mid-question.

1 7 8 9 – 1 8 6 1

3 Men named Gabriel Guilbault in the candidate set

1 Ledger account to assign

2 Wage years with no evidence either way

Revised 16 August 2026 — this page replaces an earlier version

The April 2026 version of this page attributed the North West Company ledger account at F.4/32, folio 414 to Gabriel Guilbault père and described a continuous five-year career, 1816 to 1821. That attribution has been withdrawn. Re-examination of the plates showed that the 1816 start date came from the volume’s index rather than the account, that two of the volumes cited as independent confirmation are a fair copy and its original, and that the parish registers place Gabriel père physically in Lower Canada during the Athabasca outfit year.

Seven claims are set out in the Researcher’s Notes below, each with the evidence that unsettled it. Nothing has been quietly revised, and the research log carries the full correction and retraction series.

Map of the voyageurs' highway, the canoe route from Montreal to Fort Chipewyan

The Problem

A fur-trade account book gives you a name and a number. It gives you no birth year, no parish, no father’s name. When a family recycles its given names across three generations, that is not enough to identify anyone.

The North West Company’s general ledger carries an account for Gabriel Guilbault at folio 414, with annual wages running 1817 through 1821 and a transfer into the Athabasca department. There is no biographical detail of any kind attached to it.

Three Gabriels in one line

Gabriel Guilbault père was born at L’Assomption in 1762. His son Gabriel was born in the pays d’en haut about June 1790. That son’s own son Gabriel was born at Rigaud on 20 December 1824. All three were called voyageur in at least one document.

And more Gabriels beside them

The candidate set did not begin with three. A first cousin, Gabriel, was baptized at L’Assomption on 4 March 1764 — same parish, same generation, same name, and of an age to have worked the ledger years. Cousins, uncles and nephews had to be separated one at a time before the field narrowed. Being separated as a distinct individual is not the same as being excluded from the fur trade; several of these men remain untested.

The archive’s own boundaries

The HBCA name index to the North West Company account books covers 1795 to 1827. Gabriel père’s documented interior decade began about 1789 — before the index opens. The Company’s lists of servants run 1774 to 1841, but the columns that would identify a man, including parish of residence, stop at 1819. Servants’ contracts before 1818 are bound chronologically rather than filed by name.

So the years when this family’s voyageur career is best documented in the parish registers are the years the fur-trade archive covers least well — and the later generations fall off the other end.

Spelling is noise

The same surname appears across these records as Guilbeau, Gilbeau, Gibeau, Guilbeault, Gulbiau and more. Databases standardize these to a single form, which is an editorial judgment rather than a statement in the record. Spelling could not be used to distinguish anyone. Figures could.

Detail from a North West Company dissolution payment list, Hudson's Bay Company Archives series F.4

The Method

Figures separate accounts in the ledgers. The wife’s name separates men in the registers. And presence is credited only where the priest wrote it down — never inferred from a man being named.

1 · In the ledgers, follow the money

Two accounts under one name, resolved by their balances Resolved Two men called Paul appear in the same dissolution records under different spellings. One account carries wages of 500 and a balance of 617.14; the other 350 and 96. Each figure matches that account’s own documents elsewhere in the archive. The spellings are inconsistent; the figures never are. Note the limit: this separates two accounts. It does not name the men behind them, and one of the two remains unidentified.

2 · Check whether two sources are one source

Duplicate volumes are not independent witnesses Verified Two pairs of the 1821 volumes proved to be a fair copy and its original — line for line identical, differing only in pagination. Agreement between them confirms nothing. This is what retired the claim that a wage figure had been “independently confirmed.”

3 · Read the index date as a fact about the volume

An index date is not a service date Verified The index entry reading 1816 describes the volume, not the man. The folio itself shows wages beginning in 1817. A career length was being read off a finding aid.

4 · In the registers, the wife is the discriminator

Occupation cannot separate these men; the household can Proven Two baptisms at Rigaud three months apart in 1826: on 30 April, Gabriel Guilbault journalier, wife Josephte Lapensée — the father. On 29 July, Gabriel Guilbault cultivateur, wife Madeleine Rocbrune — the son. Same parish, same year, two trades, two men. Occupation moved; the wife’s name never did.

5 · Named is not present

Presence is credited only where the register states it Standing rule A baptism naming a man as father is not evidence he was in the parish; it is a statement about the child’s legitimacy. Presence is credited only where the act carries the clause — ainsi que le père présent — or records him acting as a party. Every appearance was tested on that basis and logged in one of three states: attested, not established, or not yet examined.

6 · Read the priest, not just the entry

A missing clause may be the officiant’s habit Control Where a clause is absent, the neighbouring entries have to be read before anything is concluded. One priest at Saint-André records no father’s presence in any entry on the page; another, on a single day in 1798, used two different formulas for two children of the same family. Silence in a register is often about the writer.

7 · State the extent of a negative

Nothing found, and here is how far we looked Logged Surname variants were searched across the Montreal notarial engagements indexed by the Société historique de Saint-Boniface, and the decade sequences browsed in full with page-boundary evidence recorded. No Gabriel appears. A negative is only usable when its boundaries are written down.

Frances Anne Hopkins, Voyageurs in a Canoe, 1869

What Holds

Gabriel Guilbault père was a voyageur. That is proved — and not by a single company record. It is proved by where his children were born.

The first career, documented

On 10 October 1798 the curé of Saint-Paul-de-Lavaltrie baptized three children of Gabriel Guilbault in a single sitting. Each entry states that the child was born and privately baptized dans les pays d’en haut: Gabriel, aged eight years and four months; Angélique, aged six; Joseph Claude, aged sixteen months. Their mother is recorded as Josephte, an Indigenous woman. Their father is styled voyageur et maintenant agriculteur — the priest catching him mid-transition.

Three children born in the interior across roughly eight years is not a seasonal man’s record. It is a hivernant’s — a wintering voyageur who lived where he worked, not one who left in May and came home in September. The parish register documents a career the company records never captured.

The Athabasca years are not his

Rigaud, 18 and 23 June 1819 · Saint-Eustache, 20 February 1821 Attested On 18 June 1819 he stands at the baptism of his son Joseph at Rigaud, the act naming him Macon and recording him present. Five days later he is godfather at another Rigaud baptism, present again. On 20 February 1821 at Saint-Eustache he declares at the baptism of his son Charles. The brigades left Lachine in May; Athabasca is a wintering department and cannot be worked seasonally. A man in a Lower Canadian parish in June 1819 and February 1821 was not wintering in the Athabasca country.

What remains undetermined

Two wage years, and two candidates Open 1817 and 1818 carry no presence evidence for anyone, in either direction. Two men remain in the field. Gabriel fils, aged twenty-seven to thirty-one across the wage years, has no presence-attested appearance of any kind before November 1823 — a clean opportunity window with no positive evidence either way. The first cousin born in 1764 is documented in Quebec in April 1811, in the 1825 census and at his burial in April 1826, and nowhere in between.

The account may never be assigned. That is the honest result, and it is worth publishing as one.

Seven claims from the earlier version of this page have been retired. They are set out below, each with the evidence that unsettled it.

Researcher’s Note

The researcher is a descendant of Gabriel Guilbault and Marie Josephte Abitakijikokwe, through Gabriel Guilbault fils (born in the pays d’en haut about June 1790) → Évangéliste Guilbault (1845) → Elisabeth Emma Guilbault Gilbert (1883) → Thomas Eugene Hamall (1904) → Thomas Kenny Hamall (1932). Withdrawing a claim about one’s own ancestor is harder than making it. It is also the only way the rest of the work stays worth reading.

Companion Case Study

The same family produced a second unassigned account. Two men called Paul were separated in the 1821 settlement by their figures — and then one of the two turned out not to be who this research had said he was.

Who Was Paul Gibeau? →

Revision History

Researcher’s Notes

A researcher’s note records a claim this page once made, the evidence that unsettled it, and the date it was withdrawn. Nothing is deleted. A page on which nothing has ever failed is a page nobody has tested.

Researcher’s Note W-1 · 16 August 2026

The ledger account at folio 414 belonged to Gabriel Guilbault père Withdrawn

This page identified the account as his and described a continuous five-year career in the Athabasca. The identification is withdrawn on direct presence evidence: he stands at his son Joseph’s baptism at Rigaud on 18 June 1819, is godfather at another Rigaud baptism five days later, and declares in person at Saint-Eustache on 20 February 1821. Both fall inside the account’s own wage years, and the Athabasca is a wintering department.

His voyageur career is not withdrawn with it. That rests on where his children were born, and it stands.

Evidence: registres paroissiaux, Rigaud, 18 and 23 June 1819; Saint-Eustache, 20 February 1821; HBCA F.4/32, fo. 414.

Researcher’s Note W-2 · 16 August 2026

“A five-year career, 1816 to 1821” Withdrawn

The 1816 start date was read from the volume’s index rather than from the account. The folio itself credits wages beginning in 1817. A career length was being taken off a finding aid, and the claim is withdrawn.

The control it produced is worth more than the year it cost: an index date describes the volume or the account, never the man’s service.

Evidence: HBCA F.4/32, fo. 414, and the volume index.

Researcher’s Note W-3 · 16 August 2026

“Independently confirmed in two volumes” Reconsidered

Two of the 1821 volumes were cited as independent corroboration of a wage figure. They are a fair copy and its original — line for line identical, differing only in pagination. The claim of independent confirmation is withdrawn.

This one failed usefully. Checking whether two agreeing sources are actually one source is now a standing control across the whole research file, and it has since caught the same trap elsewhere. A copy agreeing with its original tells you the copyist was careful and nothing more.

Evidence: HBCA F.4/46 against F.4/47; F.4/43 against F.4/45.

Researcher’s Note W-4 · 16 August 2026

“Father, son and uncle within eighteen pages” Withdrawn

The proximity of three accounts in the general ledger was offered as evidence that the three men were related and serving together. The ledger is arranged alphabetically. Proximity in an ordered document is not evidence of anything, and the argument is withdrawn.

Evidence: HBCA F.4/32, the arrangement of the G run.

Researcher’s Note W-5 · 16 August 2026

The gross-and-net wage explanation Withdrawn

Two differing wage figures were reconciled by proposing that one was gross and the other net of deductions. The arithmetic runs backwards, and the explanation is withdrawn. Where two figures cannot be reconciled, the honest entry is that they cannot be reconciled.

Evidence: HBCA F.4/32, fo. 414, recomputed.

Researcher’s Note W-6 · 16 August 2026

The account attributed to the son Joseph Claude Reconsidered

A ledger account was attributed to Joseph Claude Guilbault, born in the pays d’en haut about 1797. That account opens in 1811 carrying more than sixteen hundred livres brought forward from an earlier ledger. He was fourteen. The attribution is withdrawn; the account belongs to an established man with years of prior service.

The reasoning that retired it has since done more work than the claim ever did. An account opening on a large carried-forward balance describes a career already under way, and recognising that shape has since reframed a second account in the same volume.

Evidence: HBCA F.4/32, p. 403, opening line; PRDH baptism, Saint-Paul-de-Lavaltrie, 10 October 1798.

Researcher’s Note W-7 · 16 August 2026

A third surname variant, and the Athabasca wage figure Withdrawn

Both were misreadings of the plate rather than inferences from it, and both are withdrawn. They are recorded here because a misreading corrected in silence teaches a reader nothing, and because the figures on these pages are the only thing in them that can be checked.

Evidence: HBCA F.4/32, fo. 414, re-examined at higher resolution.

The Hinge

The merger that solved one problem and closed the other

In 1821 the North West Company and the Hudson’s Bay Company amalgamated. The same event produced the document that separated the two Pauls’ accounts — and destroyed the records that would have resolved the two Gabriels.

What it gave

Dissolution required a single comprehensive settlement: the entire workforce, name by name, wages and closing balances side by side on the same page. That is an instrument no ordinary year of trading produces.

It is what separated the two Paul accounts. Without the merger there is no such page, and no separation — only two men of one name and no way to tell their records apart.

What it took

Every North West Company account ends in 1821, because the company ended in 1821. The series does not close around one man; it closes around everyone in it at once.

After the merger, servant records move into different classes under a different company’s filing logic. The evidence that would have separated father from son in 1822, 1823, 1824 — a renewed contract, a continuing account, a parish of residence in a servants’ list — was never created, because the employer was gone.

This is why an account ending and a life changing must not be read as the same fact. A man of this family marrying at Rigaud in November 1823 is doing what hundreds of discharged Canadien voyageurs did after 1821. It is a pattern, not a fingerprint.

One word, three generations, unequal evidence

The title of this study survives the reframing, because the word itself is what changed. Between the first generation and the other two lies the merger — and by the time the grandsons were called voyageurs, the word meant the Ottawa and the Gatineau timber trade, not the Athabasca brigades.

The man	Called voyageur where	What supports it	Weight
Gabriel père b. 1762, L’Assomption d. 1833, Saint-Benoît	Parish register, Saint-Paul-de-Lavaltrie, 10 October 1798 — voyageur et maintenant agriculteur	Three children born and privately baptized in the pays d’en haut, 1790–1797, entered in the register the same day	Record-stated, independently corroborated. A wintering career of roughly a decade.
Gabriel fils b. c. June 1790, pays d’en haut d. 1880, Saint-Placide	Census of Canada East, Saint-Placide, enumerated January 1852	Nothing else. Every parish act naming him, 1824 to 1833, calls him journalier or cultivateur, and the 1861 census calls him cultivateur again	A single enumerator’s word, thirty years after the ledger years, on a page whose age column is demonstrably unreliable.
Gabriel, his son b. 20 December 1824, Rigaud d. 7 October 1904, Maniwaki	Census of Canada East, Saint-Placide, January 1861 — with two of his brothers marked the same	Nothing else, and nothing at all between 1824 and 1861. Six months after the census he married at Maniwaki, on the Gatineau	A single census line, forty years after the merger, on the Ottawa watershed rather than the northwest.

This summary states the argument. The full methodology sets out every plate examined, the presence ledger entry by entry with the register’s own language quoted, the duplicate-volume analysis, the standing evidence controls, and the complete correction and retraction series behind the notes above.

The Voyageur Years: Which Gabriel Guilbault? → The Companion Problem: Who Was Paul Gibeau? →