Three Documents, Three Record Types, and the One Story They Tell

How AI connected a 1785 land deed, a tax list, and an 1812 estate file into a forty-four year story

From Evidence to Finished Chronicle — Week 3 of 10

Welcome to Chronicle Makers. I’m Denyse, and I help family historians turn decades of research into finished stories their families will actually read.

This is the third piece in a 10-week series walking through exactly how to use AI to go from your first document upload to a published chronicle. If you find this useful, share it with someone who has a box of research and no idea what to do with it. Every previous post is archived here.

---

This week I’m showing you what happens when you stop looking at records one at a time and start reading them together.

A land deed from 1785. A tax list from Williams Township. And an estate file from 1812 that inventoried everything Stephen Crumrine owned when he died — down to the shoes at the shoemaker’s that were never finished.

Three documents. Three record types. Forty-four years of one man’s life.

Three Records, Three Windows in a Life

A census tells you someone existed in a place. That’s confirmation. Useful, but flat.

Different record types give you different dimensions of the same life. A land deed tells you what someone could afford. A tax list tells you what they built. An estate file tells you what they left behind.

Put all three together and a person starts to emerge.

The Land Deed: April 14, 1785

Stephen Krumrein purchased 200 acres and a saw mill in Williams Township, Northampton County. He paid £1,900 in gold and silver to Joseph Briggs and his wife Abigail.

That number matters. In 1762, Stephen had bought 245 acres in Berks County for £700. Twenty-three years later he spent nearly three times that on a smaller tract. But this one came with a saw mill. Stephen wasn’t just farming anymore. He was running a business.

He was forty-eight years old. A former militia captain who had served during the Revolution. A father of seven. And he had enough gold and silver coin to pay £1,900 outright.

The deed calls him a yeoman. Not a farmer. A yeoman. In the eighteenth century, that word meant an independent landowner of substance. The kind of man who didn’t work someone else’s land.

One document. And already I know more about Stephen’s economic life than ten census records could tell me.

The Tax List: Williams Township, Northampton County

The tax assessment for Williams Township shows Stephen Krumrein with 200 acres, five horses, and eight cattle.

The acreage matches the deed exactly. But the tax list adds the livestock. Five horses on a farm with a saw mill tells you this operation needed draft power. Eight cattle means the family had milk, butter, and meat. This wasn’t subsistence farming. This was a working estate.

His sons George and Adam appear on the same township tax rolls. The family put down roots here. Multiple Krumreins paying taxes in the same township across multiple years.

The tax list also gives me a surname variant I didn’t have before. The assessor wrote “Krumreine, Stephan.” Different ear, different spelling. That’s now the twelfth variant I’ve documented for this family (when I get to thirteen, I get a mystery record — just kidding).

The Estate File: April 1812

This is the one that stopped me.

On April 18, 1812, Catherine Krumrein walked into the Centre County courthouse and renounced her right to administer her husband’s estate. She handed that responsibility to her sons Henry and Adam.

Three days later, on April 21, letters of administration were granted. By April 25, two men named John Keen and Jacob Kryder had inventoried everything Stephen owned.

And what they inventoried tells you things no other record can.

A sword. The militia captain kept his sword for thirty-five years after the Revolution. Two cows, a stage wagon, four beehives, a clock and case. A large Bible valued at £3 — likely the German family Bible, possibly with vital records written inside. (It crushes me every time I see one of these in an inventory.)

Shoes at the shoemaker’s, not finished. Stephen had shoes being made when he died. A life interrupted mid-stride.

And bonds. His sons owed him money. Henry owed £45. Adam owed £55. Michael owed £5. The father was creditor to his own children. That’s the economic relationship inside a family, visible in no other record type. Stephen had loaned his sons money to get started — the same way he had started, buying land — and they were still paying him back when he died.

The estate took eight years to settle. Jacob Kryder traveled thirty days to Lehigh and Northampton counties to recover outstanding bonds. Catherine received her widow’s dower: $1,109.64, one-third of the final balance.

The DAR application said Stephen died April 10, 1812. Catherine’s renunciation came eight days later. The timeline fits.

Remember the death date problem from Week 2? The SAR application said Stephen died in 1792. But he appeared on both the 1800 and 1810 censuses. Now the estate file confirms it. Stephen died in April 1812 at age seventy-five. The 1792 date was wrong by twenty years.

Three independent record types. Census records, land deed, and estate file. They all point to the same conclusion. That’s how you build a case in genealogy. Not one record proving a fact. Multiple records from different sources converging on the same answer.

What Emerges When You Read All Three Together

Individually, each record answers a narrow question. When did he buy the land? What did he own? When did he die?

Together they tell a forty-four year story.

1762: A twenty-five-year-old buys 245 acres in Berks County for £700. He’s starting out.

1777: He serves as a militia captain during the Revolution. Berks County, First Battalion.

1785: At forty-eight, he buys a 200-acre plantation with a saw mill for £1,900 in gold and silver. He’s arrived.

1790: Census shows seven people in his household. Williams Township, Northampton County. His children are growing up around him.

1806: At sixty-nine, he sells the Williams Township property for $4,000 to Theobald Albrecht. More than double what he paid. He and Catherine move west to Centre County.

1810: Census shows six people in Haines Township, Centre County. A younger couple and two small girls living with Stephen and Catherine. Someone is taking care of them.

1812: Stephen dies. Catherine hands the estate to their sons within the week.

That’s a life. From a young farmer buying his first land to a militia captain to a prosperous yeoman to an aging patriarch being cared for by his children. The arc is there in the records. You just have to read across record types to see it.

How AI Connected the Dots

I didn’t piece this together manually. I put the land deeds, tax records, census data, and estate file into my Crumrine research folder. Then I asked Claude Cowork to read everything and build a chronological timeline with source citations for each event.

What came back was the arc I just described. Claude didn’t invent it. Every date, every number, every location came from a document in the folder. But Claude read across record types in a way that would have taken me hours to do by hand. The £700 to £1,900 to $4,000 progression? I hadn’t noticed it until Claude laid the transactions side by side.

The death date resolution was already in my notes from Week 2. But the estate file locked it down. Claude flagged the eight-day gap between Stephen’s reported death and Catherine’s renunciation as corroborating evidence. That’s the kind of connection you want AI to make.

Try This With Your Own Records

Pick an ancestor you have at least three different record types for. Not three census records. Three different kinds of records. A deed and a will. A military record and a tax list. A church record and a probate file.

Put them all in a folder. Open Claude Cowork. Give it this prompt:

“Read all the documents in this folder. Build a chronological timeline for [ancestor name]. For each event, note which document it comes from and what it reveals about this person’s life that the other documents don’t.”

Watch what comes back. Not just dates and facts. A life taking shape across different kinds of evidence.

Then ask: “What story emerges when you read these records together? What changed for this person over time?”

That’s the moment research starts becoming narrative.

If you want to learn the full process from research to finished chronicle, Chronicle Makers teaches every step with AI built in, no experience in AI needed. Join Chronicle Makers

Happy Chronicling!

—Denyse

P.S. Next week, we go deeper into the death date problem. I’ll show you how to build a formal proof argument when your sources disagree. Stephen’s SAR says 1792. His estate file says 1812. One of them is wrong, and the evidence makes the case. That’s Week 4: Building a Proof Argument.

P.P.S. The Chronicle Sprint starts April 15. Ten days, one finished chronicle. We walk through every step together, using Chat or Cowork, your choice, and have loads of fun together. All the details → https://10daychroniclesprint.com