I’ve been staring at these documents for over twenty years.
Literally. They’ve lived in my filing cabinet—a physical one, tucked away with all my other genealogy research that had not been digitised. Photocopies from the Bibliothèque et Archives nationales du Québec in Montreal, obtained back when I was visiting distant relatives in 2002. These were French legal documents from the early 1800s. Notarial records. Court petitions. Guardian appointments.
I knew they were important; they were about my great-great-great-grandmother, Deborah Crawford. As we often find with genealogy, we tend to get bogged down researching male ancestors, viewing women primarily as bearers of their children. However, I make an effort to uncover more about the women. That old saying about every successful man having a woman behind him? Still true!
But honestly? I’d only ever managed to extract the bare bones: names, dates, the fact that she was orphaned young. The dense legal French, the archaic terminology, the complex interconnections between documents—it all just sat there, keeping its secrets.
Until now.
With the help of AI, I’ve finally coaxed these documents into telling me Deborah’s full story. And what a story it is.
AI even helped me go one step further—I used NotebookLM to create a short video overview of Deborah Crawford’s life, bringing her story to life visually for the first time.
Just to clarify one small note re this video where they mention Deborah Delisser nee Crawfords two children dying within hours of each other. They in fact died 4 days apart. Everything else in the video is spot on. AI NotebookLM did a brilliant job.
The Documents That Waited
Let me paint you a picture of what I had: baptism records, marriage contracts, guardianship petitions, estate renunciations, partnership agreements, burial records. Most in nineteenth-century legal French. All precisely documented in that meticulous way that notaries worked, but without any narrative thread to tie them together.
My Quebec relatives had transcribed them for me, but I still didn’t understand the legal jargon. I knew Deborah was born on 30 December 1796 in Quebec. I knew her parents were John Crawford, a merchant, and Deborah Cox. I knew they both died in 1803, leaving her and her siblings orphaned. I’d even found the double wedding record from 1823—Deborah and her sister Agnes both married on the same day at St Pancras Old Church in London.
But I didn’t understand the why of any of it. Why were these children orphaned with seemingly wealthy merchant parents? What happened to the business? Who were all these guardians? How did four orphaned children from Quebec end up marrying in London twenty years later?
The documents held the answers. I just couldn’t read them properly.
Enter AI: A Research Assistant Like No Other
This year, I decided to throw everything I had at an AI—Claude, specifically. Not just asking it to translate, but to analyse. To piece together the puzzle.
(Did you know Claude could translate French into English?)
I uploaded document after document. The partnership agreements from 1795. The guardianship petitions from 1803 and 1811. The death records. The inheritance renunciation from 1804. Even a newspaper clipping from Edinburgh that I’d found about her grandmother’s death.
And slowly, document by document, the story emerged.
Deborah’s Story: From Privilege to Orphanhood
Here’s what I learned:
Little Deborah Crawford was born into what should have been privilege. Her father John was a successful merchant in Quebec, partner with his maternal uncle Mathew Lymburner in a thriving business. They owned wharves, warehouses, properties along the Quebec waterfront. Her mother was Deborah Cox, daughter of Nicholas Cox, who’d been Lieutenant Governor of Gaspé.
But tragedy struck swiftly.
On 22 March 1803, when Deborah was just six years old, her mother died. She was only 26. Deborah’s father was left with four young children: Deborah (6), John Lymburner (4), Agnes Dickie (2), and baby William Nicholas, who was only six months old.
Seven months later—seven months—their father died too. Seventeen October 1803. Deborah had just turned seven.
Four orphans.
The Devastating Truth in the Documents
These children inherited nothing.
Not because their father was poor. But because of how the business partnership was structured.
You see, the documents showed that all those properties, all those wharves and warehouses? They were registered solely in Mathew Lymburner’s name. There were partnership agreements—three of them, actually, dated 31 January 1795, 2 February 1795, and 21 September 1795—that were supposed to protect John Crawford’s interests. But when he died, the business had more debts than assets.
On 9 March 1804, just months after becoming orphans, the children’s guardian had to petition the court to renounce their inheritance. Not because there was nothing, but because accepting it would have meant inheriting their father’s business debts.
Four children, aged 7, 5, 3, and 1. Protected from debt, but left with nothing.
The People Who Stepped Up
Their first guardian was George Longmore, a physician in His Majesty’s Service. The AI helped me understand something crucial: he wasn’t just a family friend. He was married to Christiana Letitia Cox—Deborah’s aunt, her mother’s sister. He was family. And he served as their guardian for eight years, until his death in August 1811.
Then came John Jones, described in the records as a “gentleman,” who took over as guardian for the next ten years. And throughout it all—for the entire eighteen years from 1803 to 1821—a merchant named Robert Woolsey served as “Surrogate Tutor,” providing continuity through all the changes.
These men weren’t obligated to do this. But they did.
The Edinburgh Connection
One of the most stunning discoveries came from a simple newspaper clipping. The AI helped me connect a death notice from the Caledonian Mercury dated 28 October 1815 to my family.
“At her daughter Mrs Longmore’s house, 18 Gayfield Square, on the 17th instant, Mrs Deborah Cox, relict of the late Colonel Cox, much and justly regretted.”
Deborah’s grandmother hadn’t died in Quebec. She was living in Edinburgh, Scotland. At her daughter Christiana’s house—the same Christiana who was married to George Longmore, the children’s first guardian.
This explained so much. Why couldn’t grandmother take in the orphaned children in 1803? Because she was 3,000 miles away in Scotland. But it also showed that there was family in Britain. Connections that would later help these children make their way across the Atlantic.
The Turning Point: 1821
In June 1821, the children’s uncle—William Crawford, their father’s brother who’d been Provincial Judge for the Inferior District of Gaspé—died intestate in New Brunswick.
The documents show that in October 1821, John Lymburner Crawford, now 22 and living in Montreal, petitioned to become guardian to his younger siblings. He identified himself and his siblings as “sole heirs” to their uncle’s estate.
Finally, an inheritance. Finally, financial independence.
And within two years, all four siblings had made their way to Britain.
A Double Wedding in London
On 2 August 1823, at St Pancras Old Church in London, two sisters married on the same day.
Deborah Crawford, age 26, married Alexander Delisser, a physician with an MD. Her sister Agnes, age 22, married Samuel James Douglas.
Imagine that day. Two orphaned girls from Quebec, who’d lost their parents twenty years earlier, who’d been raised by a succession of guardians, now standing together in a London church. Were their brothers there too? Did they think of their parents? Of the grandmother who’d died in Edinburgh eight years earlier? Of Uncle George Longmore who’d cared for them?
The documents don’t tell me. But knowing what I know now, I can imagine.
I do know that their father’s uncle, Adam Lymburner was there. He witnessed both their marriages and he played a big part in their future. To learn more about his role in their future, visit my family history archive at JenealogyScrapbook
What Twenty Years and AI Taught Me
For two decades, I had the pieces of this puzzle. But I couldn’t see the picture.
What AI gave me wasn’t just translation—though that helped enormously with the dense legal French. It was analysis. Connection-making. The ability to say “this document from 1795 explains what happened in 1804” and “this guardianship arrangement connects to that family relationship” and “here’s why these children couldn’t inherit even though their father was a merchant.”
AI helped me understand the why behind the what.
I learned that Deborah’s maternal grandfather, Nicholas Cox, had been a British Army officer who served in the French and Indian War, fought at the Siege of Quebec, and became Lieutenant Governor of Gaspé. That adds context to her mother’s family status.
I learned that her father’s business partner, Mathew Lymburner, couldn’t help the Crawford orphans because he was secretly supporting his own two illegitimate children—revealed in a document from July 1821 where he donated land to them. That explains so much about why the children had to make their own way.
I learned that there’s still a mystery about how these children got from Quebec to Britain between 1821 and 1823—no passenger lists found yet. That’s my next research challenge.
The Documents Still Speak
Those same documents that sat silent in my filing cabinet for twenty years are now the richest source I have. Not because they’ve changed, but because I finally have a tool that can help me truly read them.
Every genealogist knows the frustration of having documents you can’t fully interpret. Language barriers, legal jargon, handwriting challenges, missing context. We do our best. We transcribe what we can. We make notes. We move on to other research.
But sometimes, with the right help, we can come back, dig deeper and finally hear our ancestors’ voices.
I knew the bare facts about Deborah for twenty years. But now I know her story. And that changes everything.
If you’re sitting on old research documents that you’ve never fully understood, I encourage you to revisit them with fresh eyes—and perhaps with new tools. Our ancestors’ stories are worth the wait.


AI to the rescue, but in the hands of a skilled archivist and storyteller. That decision to try out Claude’s brilliant document analysis capabilities, was a brilliant one. Your treasure trove of French/English records and documents has now broken through a wall of mystery through your clever guidance. I imagine the delight you had in revealing all of this and so glad you shared it with us. I loved the NotebookLM video to out everything into a focussed snapshot. Brilliant!
This is the kind of story that reminds you why genealogy isn’t just a hobby. It’s Twenty years of silence shattered by one well aimed question to an AI. That’s not cheating. That’s evolution. The machine didn’t write the story. You did. It just handed you the keys to the locked room. Deborah Crawford finally got her voice back because you refused to give up. That’s what separates genealogists from data collectors. We don’t just chase names. We drag the dead back into the light.
Substack was built for this kind of storytelling. Truth, craft, and obsession fused into something alive. Absolute masterwork.