Every family that was enslaved in America runs into the same wall. The records that would prove a person existed were mostly written by people who did not consider them people, and the compiled family trees that circulate online fill the resulting silence with guesses that harden into fact after enough copying.
Gilbert Handy Sr. was born around 1825 in Louisiana and was enslaved. The compiled trees say he died around 1870. A newspaper advertisement from Opelousas says otherwise, and the way it says so is the whole argument for doing this work with documents instead of with inherited summaries.
The Advertisement That Refuted A Death Date
On 23 November 1872, the Opelousas Journal ran a sheriff's sale notice against him. The case is Anthony Webb vs. Gilbert Handy, Parish Court No. 1085.
A sheriff's sale is a hostile document. Somebody was owed money, a court agreed, and the property was advertised for auction to satisfy the judgment. Nobody involved in producing that notice was trying to preserve a Black man's biography. They were trying to collect.
That is exactly why it works as evidence. The advertisement is dated, it names him, and it describes property in his possession. A man who died around 1870 does not get sued in 1872, and does not have goods to seize. The compiled date is refuted, not softened, and it is refuted by a source with no motive to flatter him.
What The Notice Actually Inventories
The property listed is worth reading slowly, because it is the closest thing to a photograph of his life that survives:
- 3,000 pounds of cotton in seed
- a branded bay horse
- 50 bushels of corn
- a beef
That is not the estate of a man living at the edge of subsistence seven years after emancipation. Cotton in seed at that weight means a crop was brought in. Corn by the bushel means food stored rather than bought week to week. A horse means the ability to work land and to move. The brand on the horse means ownership recorded somewhere by somebody.
Read as a whole, the inventory says a formerly enslaved man was farming on his own account in Louisiana in 1872, at real scale, with equipment and stores. The document was created to take that away from him. It ended up being the record that proves he had it.
The Ceiling We Refuse To Cross
The honest version of this finding includes what it does not establish.
His death date is still open. All the notice proves is that he was alive later than 7 December 1872. It does not give a year, and we will not manufacture one.
There is a tempting candidate that has to be rejected. A Gilbert Handy married a Sarah Green in 1898, and it would be convenient to attach that to him. It does not survive arithmetic: that Sarah Green died in 1953 at age 69, which puts her at roughly 14 years old in 1898. Separately, the state death index carries a Gilbert Handy who died in 1939 at age 59, born around 1880. That is a third man entirely, born after our Gilbert Sr. was already a grandfather.
Three men, one name, one parish region. The discipline is to keep them apart and to leave the death date empty until a document closes it. An empty field is a finding. A guessed field is contamination that the next researcher inherits.
What Sits Around Him In The Record
He is not an isolated name. A FamilySearch marriage record from 1866 confirms him with Divine Minique and confirms children, which anchors the household in the immediate post-emancipation years.
His son, Gilbert Handy Jr., was born in 1855 in Louisiana and was enslaved on the Doucet plantation. Naming the plantation matters. It converts a person from an anonymous entry in an aggregate into someone attached to a specific place, with specific owners, whose own records can then be pulled and read.
That pairing is the shape of the method. One generation is anchored by a marriage record and a lawsuit; the next is anchored by a named plantation. Neither anchor came from a family story. Both came from documents that can be requested, cited, and checked by a stranger.
Why This Is The Product, Not A Hobby
Code Black exists because this process is repeatable and almost nobody is doing it for Black families at the level of individual, cited documents.
The standard consumer path gives a percentage breakdown by region and a set of algorithmic hints. That is a starting point and it is not the same thing as a sourced ancestor with a document behind them. The work above is closer to what a title search does for a property: find the instrument, read the instrument, record what it does and does not prove.
Every finding carries a confidence tier and a citation, unflattering results included. A refuted date, a rejected candidate, and an open field are all real outputs. Research that only ever produces good news is not research.
The Code Black service tiers, from a corridor match and research roadmap through full multi-generation reconstruction with a sourced document, are at dajai.io/codeblack.
FAQ
What is a sheriff's sale notice and why does it matter for genealogy?
It is a legal advertisement announcing that a person's property will be auctioned to satisfy a court judgment. Because it is dated, names the person, and lists property, it can place someone alive in a specific place at a specific time even when no other record survives.
How did one newspaper notice refute a recorded death date?
The compiled trees list a death around 1870, but the notice is dated 23 November 1872 and describes an active lawsuit and seizable property. A person cannot be sued and have goods advertised for auction two years after dying.
If the death date is refuted, why not just replace it with a better guess?
Because a guess entered into a tree is copied forward by everyone who finds it, and becomes indistinguishable from evidence within a few years. The field stays open until a document closes it.
What was found around Gilbert Handy Sr. in the record?
An 1866 FamilySearch marriage record confirms him with Divine Minique and confirms children, and his son Gilbert Handy Jr., born 1855 in Louisiana, is documented as enslaved on the Doucet plantation.
Filed by DAJAI Stewart for Code Black - ancestral intelligence. Every finding above carries a confidence tier and a citation. Research tiers at dajai.io/codeblack.