Case Studies › Who Was Paul Gibeau? › Methodology
Full Methodology · The Guilbault Line

Assigning a Fur-Trade Account
to a Man Who Has No Parish

Document-by-document analysis behind “Who Was Paul Gibeau?” — how the account at F.4/32 page 389 was read, how its figures tie five volumes and two companies together, and why it is not Paul Guilbault père’s

Primary Archives: Hudson’s Bay Company Archives · Archives of Manitoba  |  BAnQ Notarial & Parish  |  SHSB Voyageur Contracts  |  PRDH-IGD

Section One · Archive Structure

Two Record Series That Answer Different Questions

The Hudson’s Bay Company Archives hold the surviving North West Company papers in two series that this case requires in different ways. Reading either one as though it were the other is the commonest error in fur-trade genealogy, and it is where this study went wrong.

HBCA F.5 — Servants’ Contracts

Pre-printed engagement forms completed by a company agent. Each names the engagé, the agent, the term, the capacity — milieu, devant, gouvernail — the route and wintering post, the wages, the equipment, and the point at which the man becomes free. He signs by mark if he cannot write; witnesses and the agent countersign. F.5 answers: what were the terms, who engaged him, in what capacity?

These volumes carry no name index of any kind. Finding a man in them requires paginating the volume. Both of this man’s contracts survive because someone turned the pages.

HBCA F.4 — Account Books

Post blotters and company-wide ledgers recording wages, purchases, credits, debits and settlement. F.4/29 is the Lac La Pluie blotter; F.4/37 the Athabasca general blotter; F.4/32 the general ledger; F.4/46 and F.4/47 the 1821 settlement registers. F.4 answers: where did he go, what did he buy, what was he paid, was the account settled?

The entry point is the HBCA name index, and it is incomplete. A filter on guilb returns seven entries from 3,791; a parallel search on gib returns six more that the first misses entirely, because the clerks dropped the silent medial l. Two volumes central to this family are returned by neither and surfaced only through a commercial index of the same archive.

The methodological point

One index is never enough, and an index date is not a service date. The date attached to an account in a finding aid describes the account or the volume, not the man’s employment. In this family that distinction alone produced a five-year error in a published career.

A note on the images

The Hudson’s Bay Company Archives plates carry a printed prohibition on reproduction. Documents held in that archive are transcribed on this page rather than displayed, with full references so any reader may order the images from the Archives of Manitoba directly.

Section Two · Document One

HBCA F.4/32, page 389 Key Document

Headed Dr Paul Gibeau Cr, sitting immediately above the account of Charles Grant at page 390. It was returned by the archive’s name index under the gib prefix with no date attached.

When it was finally retrieved it did not show what this research expected. The expectation was a two-season account confirming a late-life adventure. What it shows is a decade.

HBCA F.4/32, p. 389 — full transcription
1811To Balance of Ledger D, fo. 317551.10
1812To Sundries ₔ NW Bk399.8By Wages400
1813To Sundries ₔ NW Bk222.10By Wages400
1814To Sundries ₔ NW Bk529.10By Wages400
1815To Sundries ₔ NW Bk250.—By Wages400
1816To Sundries ₔ NW Bk [fo. 78]377.10By Wages400
1817To Sundries ₔ NW Bk [fo. 292]351.11By Wages400
1818To Sundries ₔ NW Bk226.17By Wages400
1819To Sundries ₔ NW Bk121.—By Wages400
1820To Sundries ₔ Ath. Book600.10By Wages500
1821To Sundries ₔ A. Book52.—By Wages500
To Balance617.14By Lieut Franklin order100
669.14669.14
1821By Balance617.14

Why the transcription is secure

A handwritten page at microfilm resolution is a reading, not a fact — unless the page contains its own check. This one does. The account is struck annually, each year’s balance carried into the next, and the arithmetic is in livres and sols, where twenty sols make a livre.

Every block foots, independently verified
1812551.10 + 399.8950.18 =  400 + 550.18950.18
1813550.18 + 222.10773.8 =  400 + 373.8773.8
1814373.8 + 529.10902.18 =  400 + 502.18902.18
1815502.18 + 250752.18 =  400 + 352.18752.18
1816352.18 + 377.10730.8 =  400 + 330.8730.8
1817330.8 + 351.11681.19 =  400 + 281.19681.19
1818281.19 + 226.17508.16 =  400 + 108.16508.16
1819108.16 + 121 + 170.4400 =  400400
1820600.10 + 69.14670.4 =  170.4 + 500670.4
182152 + 617.14669.14 =  69.14 + 500 + 100669.14
Eleven blocks, all balancing. A misread digit anywhere in the wage or balance columns would break a footing. The figures are therefore evidence; the marginal folio references and the words are readings, and are treated differently below.

What the page establishes

  • Service before 1811. The first line is a balance brought forward from another ledger. Whatever this man did before 1811 is recorded somewhere else, and the page tells you where to look.
  • Ten consecutive years of wages. Flat at 400 through 1819, rising to 500 for 1820 and 1821. Real engagements vary; a flat rate over eight years reads as a standing wage for a man who stayed.
  • Annual charges. He is buying goods on the company’s books every single year. This is not a man who came home between seasons and bought his shirts in Montreal.
  • One department change. His purchases are charged to the NW Book through 1819 and to the Athabasca Book from 1820. That is the only movement the page records in a decade.
What the figures do not certify

The three marginal cross-references — Ledger D fo. 317, and fo. 78 and fo. 292 in the NW Book — sit outside the arithmetic. Nothing on the page confirms them, and the hand’s 2s and 3s are not always distinguishable at this resolution. They are recorded as read and flagged for confirmation, because fo. 317 is now the highest-value outstanding document in the case and a wrong digit would send an order to the wrong folio.

Section Three · Corroboration

Five Volumes That Interlock Documented

A single account proves that a man existed and was paid. It does not prove that the reading of it is right. The test is whether independent volumes, kept for different purposes, carry the same figures.

VolumeWhat it recordsWhere it meets page 389
F.4/37
Athabasca general blotter, ref 104
1820 By Balance 69.14; To Sundries at Lac La Pluie 16; small charges 36; By Lieut Franklin 100; 1821 By Wages 500; To Balance 617.14; footing 669.14. Settled. The ledger’s 1820 closing balance 69.14 is the blotter’s opening credit. The ledger’s 1821 charge of 52 is the blotter’s 16 + 36.
F.5/3 fo. 16
Engagement, 22 May 1821
Cinq cents livres courant du Grand Portage. Matches the ledger’s 1821 wage of 500 exactly.
F.4/46 pp. 8–9
1821 settlement register
Gibeau Paul — 500 — Atha — Cr 617.14. Matches the wage and the closing balance.
F.4/47 p. 7 The same list, same order, same figures, different pagination. Not an independent witness — see below.
F.4/56 p. 5
NWC claim against the Franklin expedition, 1820–23
Fort Chipewyan, receipt no. 6, 18 July 1820: Paid Paul Gibeault 100 livres N.W. Coy per Lieut. Franklin’s order — £8.6.8. The company’s side of By Lieut Franklin order 100, dated and placed.
B.239/g/1 fo. 48
HBC abstract of servants’ accounts, 1821–22
Gibeau, Paul — Athabasca — Cr 1235.8; wages 1000; book debts 12; Cr 2223.8. Exactly twice the closing balance 617.14 and the wage 500 — the currency conversion. See below.
Duplicate volumes are not independent witnesses

The 1821 settlement records survive in duplicate pairs. F.4/46 and F.4/47 carry the same names in the same order with the same figures, differing only in pagination; F.4/43 and F.4/45 do the same for the cash payouts. A copy agreeing with its original tells you the copyist was careful and nothing else.

This matters practically. Because the pagination differs, the two Pauls of this family fall on separate pages in one volume and the same page in the other. “Both men on one page in two volumes” is one document seen twice, not corroboration.

The corroboration that survives is the general ledger against the post blotter. The general ledger is not the settlement register’s fair copy, so page 389 is a genuinely independent attestation of the same figures. The Franklin claim and the Hudson’s Bay Company abstract are independent in the same way: each was kept for its own purpose, and neither copies the settlement register.

The doubling is a currency

The Hudson’s Bay Company abstract does not repeat the North West Company’s figures. It doubles them. That could have been a different man with a coincidental balance, or a new engagement at a new wage. It is neither, and the proof is in the North West Company’s own book.

One North West livre = two livres of Quebec currency

F.4/56 prices its payments in sterling and states both rates on its own pages. Paid Paul Gibeault 100 livres N.W. Coy … £8.6.8 is twelve North West livres to the pound. Livres 364 Quebec cy … £15.3.4 is twenty-four Quebec livres to the pound. The company kept its books in a livre worth half the Montreal livre, and when its servants’ accounts were transferred, every figure was restated.

The same doubling holds exactly on the neighbouring accounts in B.239/g/1: Gibeault Belonie, 350 and Dr 198.15 in the North West settlement register, becomes 700 and Dr 397.10; Guillotte dit Giroux, 500 and Dr 171.12, becomes 1000 and Dr 343.4. And this account foots in the new book: 1235.8 + 1000 − 12 = 2223.8.

Two practical consequences follow. A contract written in the interior is in North West livres and needs no conversion; a contract passed before a Montreal notary is in Quebec currency and must be halved before it is compared with any North West account. And any figure carried into the Hudson’s Bay Company’s books in 1821 must be halved before it is compared with the settlement register. Assigning by figures only works if the figures are in the same currency.

What the Franklin receipt adds, and what it does not

It adds a date and a place. The receipt is no. 6 of the Fort Chipewyan section, dated 18 July 1820, the day the expedition left the fort, and it sits among the tents, nets, canoes and sail of the party’s outfit. The same book lists the men who accompanied Lieut. Franklin, with the balances they owed the company, and this man is not among them. He was at Fort Chipewyan that day and did not go north.

It does not say what the hundred livres were for. Franklin requisitioned men at Fort Chipewyan and discharged seven before leaving, which would produce this entry. But the same book pays men on his order for single pieces of work — a hundred livres for bringing the party from Île-à-la-Crosse to Fort Chipewyan, sixty for the leg from Portage la Loche — and credits a hundred to another man for services rendered the Expedition. A man paid for a task and left behind fits the entry at least as well as a man engaged and discharged. The entry is recorded and not interpreted.

Section Four · Documents Two and Three

The Two Engagements

F.5/1 folio 115 — 19 July 1798 Attribution open

Transcribed Engaged voluntarily to the Société du Nord Ouest and its agents, Mr Rod. McKenzie here present and accepting for the Company, for the space of one year, in the capacity of milieu, to travel between Lac la Pluie and the Postes d’Athabasca and to winter au Grand Lac. Wages quatre cent cinquante livres courant du Grand Portage, with one equipment. Not free until his return to Montreal. Passed at Lac la Pluie the nineteenth of July. Having declared that he could not write, he made his ordinary mark: Paul × Gibeau.

Roderick McKenzie, cousin of Alexander Mackenzie, was among the senior partners of the company and closely involved in the Athabasca through the 1790s. An archivist’s slip records that folio 114 is missing. The volume also holds a blank pre-printed form at folio 81 bearing only the name Paul Gibeau.

This contract is attached to page 389 by inference

Nothing internal to folio 115 connects it to the ledger account. The attachment rests on the given name, a surname variant, the capacity milieu, and the department — reasoning of exactly the kind section seven tests and rejects for Paul Guilbault père.

It is recorded here as an open attribution rather than a fact. What supports it: the ledger opens in 1811 on a balance from an earlier ledger, so service before 1811 exists and requires a source; both documents give milieu; both men sign by mark. Until Ledger D folio 317 is read, the two engagements should not be described as one continuous career.

The blank form at folio 81

The form bears the name and nothing else: no agent, term, capacity, wages, date, mark or witnesses. It is not evidence that the company knew this man before terms were settled — a clerk writes the name he is handed — and the printed text supplies the explanation: this form stipulates avec double Equipement, where the contract actually signed provides one equipment. A clerk who began on the wrong form and started again accounts for everything visible, including its survival thirty-four folios earlier in the same volume.

F.5/3 folio 16 — 22 May 1821 Documented

Transcribed Engaged to the merchants trading under the name of the Compagnie du Nord Ouest, S. Black Esqr accepting, for one year, in the capacity of milieu et de piêcheur, to travel between Lac de la Pluye and the Postes d’Athabasca and to winter au Lac d’Athabasca, for cinq cents livres courant du Grand Portage with the ordinary equipment, free at Montreal. Passed au Ft Chipyn — Fort Chipewyan — the twenty-second of May 1821. Not knowing how to write he made his mark: Paul × Gebeault. Countersigned Sam Black; witnesses William McGillivray Jr and James Hill. A closing clause, under the signatures, exempts him from carrying quand il sortira — when he goes out — followed by a few words not yet read.

The docket on the outer leaf of the same contract: Paul Gibeault — 1 Year — Mid Man & Pickeurs — 1 Year — 500# — 1 [pass?] paper — for 1822.
Counting note — two contracts, not three

Film images 32 and 33 are one contract — folio 16 — and the docket on its outer leaf. Across this career there are two signed contracts, not three.

The merger did not cancel it

The contract was signed two months after the agreement of 26 March 1821 that ended the North West Company, which is why it appears on company paper at all. It was not cancelled by the merger. The F.4/46 title leaf settles it: that volume announces itself as a list of persons whose engagements for service have been transferred to the Governor and Company of Adventurers of England trading into Hudson’s Bay, and this man is in it. The engagement was carried into the reorganised company, which is exactly what the index summary’s for 1822 says. The Hudson’s Bay Company’s first abstract then carries him in the Athabasca for 1821–22 at the same wage, restated in its own currency (section three).

The exemption from carrying

A middleman written out of carrying by name, in the company’s own instrument — on the journey out, quand il sortira, rather than at every portage of the year. It is a specific accommodation for age or injury or condition, and it is the most human detail in the file. Nothing in any company document gives this man’s age, so it cannot be read as evidence of an old man. The clause is recorded and not interpreted.

It is not unique in this family’s records. A Hudson’s Bay Company engagement of May 1820 for a Joseph Guilbeau at Colville House carries a parallel clause excusing him from travelling to a named post. Two engagements, two men excused from something by name.

Section Five · The Discriminator Problem

Why Spelling Cannot Identify Anyone Here

The surname is pronounced in Quebec French with the medial l and the terminal t both silent. A clerk writing what he hears produces approximations, and the same clerk produces different ones on the same page.

The index to the Athabasca general blotter renders one family three ways in a single sequence: Gibeault at reference 93, Gilbian at 100, Guilbian at 104. Across the archive the man of page 389 appears as Gibeau in 1798 and in the settlement register, Gibeault in the blotter and the contract index, and Gebeault in the body of his own last contract.

Assign by figures, then and only then assign to a man

Two men of this name served the Athabasca department in 1820–21. The company’s own settlement register separates them, and not by spelling: it carries a department column and a wage column. One man is 500 and a credit of 617.14; the other is 350 and a debit of 96. Each figure matches an independent account elsewhere in the archive.

Every document in these volumes can be assigned by its figures, and by nothing else. Two documents were moved out of this study on exactly that basis: a general ledger page showing 446 livres of merchandise against wages of 350, and a blotter account at the same wage tracing Fort William, Fort Chipewyan, Slave Lake and Fort Providence. Both belong to the younger man.

And a standardized surname is an editorial judgment

Indexes and databases normalise spellings into buckets. That normalisation is a decision made by the editors of the finding aid, not a fact about the record and certainly not a statement about family. When a database resolves Gilbeau and Gibauche to GUILBAULT, it is telling you what its compilers concluded. Always log the surname as indexed alongside the standardized form, and reason from the former.

Section Six · The Method That Decided It

The Three-State Presence Ledger Generalisable

Testing a fur-trade attribution against a Quebec candidate requires knowing where that candidate demonstrably was. That sounds simple and is not, because the great majority of parish entries naming a man do not place him anywhere.

A marriage entry describing the bride as fille majeure de Paul Guilbault, maçon tells you his trade and that he was alive. It does not tell you he was in the church. A filiation clause names; it does not place. Treating those entries as presence would have made almost any candidate look continuously resident in his parish, and would have made the test useless.

The ledger therefore records three states, never two.

Yes

The act’s own presence language, quoted. “Ainsi que le père présent”; “en présence de”; a party appearing before a notary; a godparent asked to sign.

Not examined

No image has been pulled. The document may say anything. This is an admission of work not yet done, and it must never be counted as an absence.

Not established

An image or database card has been seen and carries no presence language. The man is named and not placed.

Why the last two must stay apart

Collapsing them into a single “absent” would count unpulled images as evidence that nothing happened. Collapsing them into “unknown” would hide which rows are cheap to close. The distinction is what tells a researcher where to spend the next hour.

In the parallel investigation of this family’s Gabriel, the two entries that decided the case sat in the not established column precisely because a database card had been seen and the register image had not. Pulling two images settled a question that had blocked three published pages.

The result for this candidate

Applied to Paul Guilbault père across forty-eight years, the ledger produced eight presence-attested appearances out of more than forty entries. Ten of his recorded occupations come from his children’s marriage filiations and place him nowhere at all.

Three of the eight fall inside the company account’s own years, and those three are the case.

Section Seven · The Exclusions

Why He Is Not Paul Guilbault père Excluded

The proposition under test: that the man of page 389 was Paul Guilbault père, born at Montreal in April 1761 and buried at St-Paul-de-Lavaltrie in January 1831.

DateThe company recordThe Quebec record, in its own words
1798 F.5/1 fo. 115, passed at Lac la Pluie 19 July; engaged to winter at Great Slave Lake, not free until Montreal. 10 October, St-Paul-de-Lavaltrie: “Le parrain a été Paul Guilbeau & la marraine Marie Olivier épouse du parrain, qui ainsi que le père présent, n’ont su signer, de ce requis.”
1820 F.4/32 p. 389: wages 500, and 600 livres 10 sols charged to the Athabasca Book. 22 July, before Notary Barthélémy Joliette at Industry: “le vingt-deux de juillet après midi… ont faits leurs marques après lecture faite.”
1821 F.5/3 fo. 16: engaged 22 May, at Fort Chipewyan, to winter au Lac d’Athabasca for 1822; engagement transferred to the HBC. B.239/g/1 fo. 48 credits him a year’s wages in Athabasca for 1821–22. 16 October, St-Paul-de-Lavaltrie: “en présence de… Paul Guilbault père… dont aucun a su signer.”

Two controls that carry the notarial half

The 1820 act is the decisive one, and it does its work only because of two distinctions that are easy to miss.

A recital is not an appearance

Notarial deeds routinely recite earlier acts and name their parties. Paul Guilbault is named as tutor inside a recital of an 1813 adjudication in a quittance of June 1816 — but he is not a comparant to the 1816 act. Counting that as presence would have been tempting and wrong. It is the notarial form of the same rule that governs filiation clauses.

Express representation makes silence meaningful

That same 1816 quittance opens by naming Laurent Belaire appearing en sa qualité de procureur fondé de Joseph Guilbault. A greffe that states representation expressly when it applies is telling you what it does when it does not. The 1820 tutelle recites no procuration for Paul, so its silence is positive evidence that the man was in the room, rather than merely the absence of evidence that he was not.

Verdict

Athabasca is a wintering department. A man engaged to it is in the country for the year. He is not at a font in Lanaudière in October, before a notary in July, or at a parish wedding the following autumn.

Paul Guilbault père is not the man of page 389. Not because a better candidate appeared, and not because new company records surfaced, but because three ordinary Quebec documents — a godparent clause, a guardianship account and a wedding witness list, none gathered for this purpose — place the candidate somewhere else.

The weakest of the three, stated

The October 1821 pairing does not carry the verdict alone, and should not be cited as though it did. A man at Fort Chipewyan in late May could in principle have reached Lower Canada by October with the outgoing brigades, and the Hudson’s Bay Company’s abstract for 1821–22 shows this man drawing only twelve livres of goods. The wage credit says he served the year; B.239/g/2 should confirm it by showing him still in the country and drawing goods in 1822–23. The July 1820 tutelle decides the question without it.

Two further men excluded

  • Paul Guilbault fils, baptized 6 October 1792, unmarried until 1846. From about 1813 there were two adult illiterate Paul Guilbaults in one parish, both signing by cross — the mark discriminates nothing and neither does occupation. He is excluded by a deed of 15 June 1820: five weeks before his father’s tutelle, he bought land at St-Paul before the same notary and made his mark on it.
  • Paul Guilbault “the Canadian,” baptized 22 January 1798, first cousin once removed of that family, later of the Pacific Northwest and dead in Oregon about 1849. He carries 350 in wages and a debit of 96 in every North West Company book that names him — the 700 livres of his December 1819 Montreal engagement, halved — and his Hudson’s Bay Company account opens at exactly twice the 96. He was thirteen years old when page 389 opens.
Section Eight · Searching for Him

Negative Searching, and When Silence Is Evidence

With Paul Guilbault père excluded, the search moved to the notarial engagements passed before Montreal notaries, which — unlike the interior contracts — state a home parish. The Société historique de Saint-Boniface has indexed roughly thirty-six thousand of them, 1714 to 1830.

The search must be run four times

The database normalises spellings into four separate buckets. No single query reaches them all, and an alphabetical browse in each decade is the only reliable method.

GUILBAULTReceives Guibault, Guilbeau, Guilbaut, Gilbeau, and the alias “Gibauche or Gilbeau”
GIBEAUReceives Gibault, Gibeault, Gibaux
GIBOINReceives Gibouin
GIBOULEAUStands alone
A negative search is evidence only when its extent is stated

Three decades were browsed across the full Gi– and Gu– ranges and found effectively empty: the 1750s facet holds 4,549 records and yields one entry in any relevant cluster; the 1760s holds 950 and yields one; the 1770s holds 474 and yields none.

Cited properly, that reads: SHSB Voyageur Contracts Database, 1750s decade facet, surnames Gi– through Gu– browsed 7 August 2026; no entry in the GUILBAULT, GIBEAU or GIBOULEAU clusters. Without the extent, the same sentence is worthless.

The 1770s absence is a weak signal in a decade holding only 474 records overall, and is not read as a break in the family’s trade activity.

The nearest match, and why it is only a lead

One Paul stands in the right cluster before 1812: Paul Gibault of Sorel, engaged at Montreal on 9 January 1789 before Notary Louis Chaboillez as a milieu for the Mississippi at 350 livres with a seven-livre advance and simple equipment.

He had been quarantined as an obvious namesake for as long as the ledger man was assumed to be a Guilbault of Lavaltrie. With that assumption gone, the reason for quarantining him goes too — but so does any temptation to promote him. The Mississippi in 1789 is the southwest trade, a long way from Rainy Lake nine years later, and one contract establishes nothing.

The search may be structurally incapable of succeeding

The database indexes engagements passed before Montreal notaries. Both of this man’s contracts were passed in the interior — at Lac la Pluie in 1798 and in the Athabasca in 1821 — in front of a company agent rather than a notary.

That is precisely the profile page 389 describes: a servant in the country continuously from before 1811, re-engaging where he stood. The best available database of French-Canadian fur-trade contracts cannot hold a record of him. His absence from it is not a failure of searching but a fact about the kind of servant he was, and it must never be cited as evidence about who he was not.

Section Nine · Genealogical Proof Standard

What This Methodology Does and Does Not Prove

1. Reasonably exhaustive research

Both HBCA indexes searched under both prefix families; the F.5 volumes paginated for the years held; the general ledger, two post blotters, both settlement registers, the company’s Franklin claim and the first Hudson’s Bay Company abstract examined; the Montreal notarial corpus browsed in four standardization buckets with three decades recorded as negative and their extents stated; the Quebec parish and notarial record of the leading candidate reviewed entry by entry. Documents known to exist and not yet retrieved are listed in section ten rather than passed over.

2. Complete and accurate citation

Every figure quoted from page 389 is transcribed with its year and its side of the account, and the footings are shown so a reader can check the transcription rather than take it on trust. HBCA plates are transcribed rather than reproduced, in accordance with the Company’s printed notice.

3. Analysis and correlation

The account, the blotter, the contract, the settlement register, the Franklin claim and the Hudson’s Bay Company abstract agree on the wage, the closing balance and the hundred livres, once the abstract’s figures are converted. That correlation establishes that these documents describe one man. It establishes nothing whatever about who he was, because none of them records a parish, a parentage or an age.

4. Resolution of conflicting evidence

The conflict between the company record and the Quebec record was resolved in favour of the Quebec record, because presence language in a parish register and a personal appearance before a notary are direct evidence of location, while a name in a company index is not evidence of identity at all.

5. A soundly reasoned conclusion

The conclusion, stated as narrowly as the evidence allows

A man served the North West Company continuously from before 1811 until the merger of 1821, wintering at Great Slave Lake and in the Athabasca, engaged twice as a milieu, unable to write, settled at 617 livres 14 sols and carried over to the Hudson’s Bay Company, which kept him in the Athabasca for 1821–22. His clerks called him Paul Gibeau.

He has not been identified. He is not Paul Guilbault père, not Paul Guilbault fils, and not Paul Guilbault “the Canadian.” The leading untested hypothesis is that he belonged to the Gibeau families of La Prairie, La Tortue, Sorel and Mascouche — men of that surname documented in the trade through the 1780s and 1790s, and never reconstructed from parish registers because there was never a reason to.

The transferable lesson

Volume of documentation and identifiability are different things. This man leaves four times the company paper of his near-namesake, and the near-namesake has a baptism, parents, a cousinage and a death in Oregon. The determining factor was not how much the fur trade wrote down. It was whether anything it wrote down could be tied to a parish.

Section Ten · Open Questions

What Would Identify Him

  • Ledger D, folio 317. Named on the first line of page 389 and never examined. Holds his service before 1811 and the best chance of a fuller account heading. The folio number is a marginal reading not confirmed by arithmetic and should be verified at higher resolution before an order is placed. Highest-value outstanding document in the case.
  • HBCA B.239/g/2, the Hudson’s Bay Company abstract for 1822–23. His account should open at Cr 2223.8. It will show whether he stayed in the Athabasca, went down, or left the service, and it is the test of the October 1821 pairing in section seven.
  • The NW Book, folios 78 and 292. Cross-referenced in the 1816 and 1817 entries. Same reading caveat.
  • The F.5 contract volumes between 1799 and 1815. No name index; finding a man requires paginating. Any engagement of his in those years would state its place of signing, and a Montreal signing would state a parish.
  • HBCA F.4/33 — the company’s Account Current of 30 November 1821, including a servants’ current struck three months after the Lachine payouts and at the close of the outfit year. Already filmed on a reel in hand, and never examined.
  • Reels 5M7 and 5M8. A substitution card partway through the reel carrying the rest of the series directs the reader to these for F.4/32 — the volume holding page 389 and every other open ledger question in this file. A researcher can work an entire reel labelled for the account books and never learn the general ledger is filmed elsewhere.
  • The Gibeau families of La Prairie, La Tortue, Sorel and Mascouche, reconstructed from parish registers. The leading hypothesis, and the least investigated line of enquiry in the case. Build the families first and map contracts onto candidates afterwards — never reason from contracts to family.
  • ADM 1/2414 at the National Archives, Kew, and Franklin’s field notebooks at the Scott Polar Research Institute. The North West Company’s side of the hundred livres is read (F.4/56, receipt no. 6). Franklin’s own accounts would hold the expedition’s side of the same receipt, and are the only place likely to say what it paid for.
A standing caution for whoever works this next

Any document bearing this surname that surfaces in these volumes must be assigned by its figures before it is assigned to a man. Two men of one name served one department in one season, three Pauls sit in the Montreal notarial corpus, and the clerks spelled all of them several ways. The wage column is the discriminator. The spelling is noise.

Complete Primary Source Documentation

Hudson’s Bay Company Archives · Archives of Manitoba, Winnipeg

  • HBCA, F.4/32, NWC General Ledger, p. 389 — account of Paul Gibeau, 1811–1821. Balance from Ledger D fo. 317, 551.10; annual wages 400 (1812–1819), 500 (1820), 500 (1821); sundries to the NW Book to 1819 and the Athabasca Book from 1820; By Lieut Franklin order 100; footing 669.14; balance 617.14.
  • HBCA, F.4/32, NWC General Ledger, p. 396 — Paul Guilbeau; debits 446; wages 350; balance 96. Another man.
  • HBCA, F.5/1, NWC Servants’ Contracts, 1798, fo. 115 — engagement of Paul Gibeau, 19 July 1798; agent Rod. McKenzie; milieu; Lac la Pluie to the Postes d’Athabasca; wintering au Grand Lac; 450 livres with one equipment; signed by mark. Archivist’s slip: fo. 114 missing.
  • HBCA, F.5/1, fo. 81 — blank pre-printed engagement form bearing only the name Paul Gibeau; printed text stipulates avec double Equipement.
  • HBCA, F.4/37, NWC Athabasca General Blotter, 1820–21, ref 104 — Paul Gibeault account: balance forward 69.14; sundries at Lac La Pluie 16; small charges 36; By Lieut Franklin 100; 1821 wages 500; balance 617.14; footing 669.14; settled.
  • HBCA, F.4/37, ref 106 — Paul Guilbeau account; Fort William transfer 179.10; charges at Fort Chipewyan, Slave Lake and Fort Providence; wages 350; balance 80.8. Another man.
  • HBCA, F.4/37 — index to the Athabasca general blotter, G run: Gibeault Paul 93 … Gilbian Gabriel 100 … Guilbian Paul 104.
  • HBCA, F.5/3, NWC Servants’ Contracts, 1815–1822, fo. 16 — engagement of Paul Gebeault, passed at Fort Chipewyan 22 May 1821; agent S. Black; milieu et de piêcheur; wintering au Lac d’Athabasca; cinq cents livres; exempt from carrying on the journey out; witnesses William McGillivray Jr and James Hill; signed by mark. The docket on its outer leaf belongs to the same contract (film images 32–33).
  • HBCA, F.4/46, NWC Account Book 1820–1821, title leaf and pp. 8–9 — register of engagements transferred to the HBC; Gibeau Paul 500 / Cr 617.14; Guilbau Paul 350 / Dr 96.
  • HBCA, F.4/47, North West Balances 1821, p. 7 — the same list copied; not an independent witness.
  • HBCA, F.4/56, North West Company Account Book, 1820–1823, “Sundries furnished Lieut. Franklin & party,” p. 5 — Fort Chipewyan, receipt no. 6, 18 July 1820: Paid Paul Gibeault 100 livres N.W. Coy per Lieut. Franklin’s order — £8.6.8; pp. 2 and 23, men who accompanied Franklin (he is not among them); p. 9, Quebec currency rate.
  • HBCA, B.239/g/1, Abstracts of Servants’ Accounts, Northern Department, 1821–22, fo. 48 — Gibeau, Paul, Athabasca; balance 1st June 1821 Cr 1235.8; wages 1000; book debts 12; balance 1st June 1822 Cr 2223.8.
  • HBCA NWC Account Books Name Index; HBCA Servants’ Contracts Index — two separate databases with different coverage.
  • HBCA, Ledger D, fo. 317; HBCA, F.4/33; HBCA, B.239/g/2; reels 5M7 and 5M8 — not yet examined.

Quebec parish and notarial records · BAnQ and PRDH-IGD

  • Registres paroissiaux, St-Paul-de-Lavaltrie, 10 October 1798 — baptism of Angélique Guilbeau; godparents Paul Guilbeau and Marie Olivier, “qui ainsi que le père présent, n’ont su signer, de ce requis”; L. Lamotte, ptre.
  • BAnQ, CN605,S24, Notary Barthélémy Joliette, 22 July 1820 — Compte de tutelle rendu par Paul Guilbault père; estate 1442.3; balance 127.19 discharged; both parties by mark, après midi.
  • Registres paroissiaux, St-Paul-de-Lavaltrie, 16 October 1821 — marriage of Rose Guilbault and Étienne Baudry; “en présence de… Paul Guilbault père… dont aucun a su signer.”
  • BAnQ, CN605,S24, Notary Barthélémy Joliette, 15 June 1820 — sale by Jean-Bte Dalpé dit Parisot to Paul Guilbault fils, 300 livres, signed by mark.
  • BAnQ, CN605,S11, Notary François Allard, 14 June 1816 — quittance reciting the 1813 adjudication; Laurent Belaire appearing en sa qualité de procureur fondé de Joseph Guilbault.
  • PRDH-IGD #296685 (baptism 1761); #334384 (marriage 1783); #2990670 (burial 1831); #841045 (Paul fils); family record, Paul Guilbault & Marie Geneviève Olivier Milot, 23 children 1783–1811.

Notarial engagements · Société historique de Saint-Boniface

  • Voyageur Contracts Database, Guilbault name-cluster: 33 records abstracted across four standardization buckets, 1759–1815. Derivative index data; no notarial image examined.
  • Entry for Gibault, Paul, contract dated 9 January 1789, SHSB microfilm M620/1197 — Sorel; milieu; Mississippi; 350 livres. A lead, not a candidate.
  • Decade facets, 1750s, 1760s and 1770s, surnames Gi– through Gu– browsed 7 August 2026 — documented negatives with extents recorded.
  • Scope: the database indexes engagements passed before Montreal notaries only.

Published and archival corroboration

  • John Franklin, Narrative of a Journey to the Shores of the Polar Sea, in the Years 1819, 20, 21, and 22 (London: John Murray, 1823), ch. IV, pp. 142–166.
  • ADM 1/2414, The National Archives, Kew — Franklin’s contingent account. Not yet examined.
  • SPRI MS 248, Scott Polar Research Institute, Cambridge — field notebooks. Not yet examined.

This methodology accompanies the case study “Who Was Paul Gibeau?” For the documentary biography of the man wrongly identified as him — whose Quebec record is what defeated the identification — and for the one Paul of this family who has been identified, see below.

Case Study: Who Was Paul Gibeau? → The Voyageur Register → Documentary Biography: Paul Guilbault père → The Canadian: Paul Guilbault of Lavaltrie → A Genealogist’s Guide to NWC Records →
THE GUILBAULT LINE
Full Methodology · Storyline Genealogy · From Research to Story