Reconstructing the village a young lawyer rode into in 1850, from four sources that could not agree on its size

A biographer has to put his subject somewhere on the first day that matters, and for Richard Coke that day was in 1850, when a young lawyer from Virginia rode into a settlement on the Brazos that had been laid out as a town the year before. I needed to describe what he found. Not a paragraph of atmosphere, which any writer can produce, but the plain facts: how many houses, how many people, what was built and what was only surveyed. I assumed this would take an afternoon. It took most of a winter, because the four sources I had would not agree with each other, and the disagreement turned out to be the more interesting story.

Here is what I had, in the order I found it.

1. The town plat, filed in 1849, which laid out the streets, the public square and the lots on a grid running back from the river. The plat shows a town of several hundred lots with a courthouse square, a ferry landing and streets named with confidence. Read on its own it describes a considerable place.

2. The federal census of 1850, taken in the summer, which enumerates the households in the area by name. The census schedule for that part of the county lists a number of families you could count on your fingers and toes, with occupations that run to farmer, merchant, a physician and, by the next year, a lawyer or two. Read on its own it describes a hamlet.

3. A letter written by a settler that year to a relative back east, which survives in a family collection and which describes the town in the terms a man uses when he wants his brother to join him: a store, a tavern, a good ferry, a courthouse coming, and lots going for prices that will not last. Read on its own it describes a boom.

4. A memoir written forty years later by a man who had been a boy there, which describes a few cabins, an Indian trading post recently abandoned, mud, and a single street. Read on its own it describes a frontier camp.

Four sources, four towns. The plat gave me a town of hundreds of lots, the census a settlement of a few dozen people, the letter a boom, the memoir a camp. Each was a primary source in the sense the term is usually used. Each had been produced by someone who was there. And each was, in its own way, telling the truth about something other than the question I was asking.

The method I settled on was to ask of each source what it had been made to do, and to trust it only within that purpose. The plat was made to sell lots, so it records what the promoters intended to exist, and it is reliable about the street names and the location of the square and unreliable about everything that implies occupation. The census was made to count people present on a given day, so it is reliable about who was there and roughly how many, and unreliable about the town as a place, because a census taker does not record buildings. The letter was made to persuade a relative, so it is reliable about what existed to be bragged about, a store and a ferry, and unreliable about scale, because a man writing to bring his brother west rounds up. The memoir was made to entertain and to establish the writer as a pioneer, so it is reliable about the feel of the place and unreliable about nearly every number in it, because forty years is long enough to make any village smaller and muddier than it was.

Overlay the four with those limits applied and a town emerges that none of them describes alone. A surveyed grid of hundreds of lots, mostly empty. A few dozen households actually present, most of them on or near one street back from the ferry. A store, a tavern and a landing that were real, and a courthouse that was a promise. Mud, certainly, and a trading post lately closed. That is the Waco a young lawyer found, and it is a place a man of twenty could reasonably have looked at and decided to stay, because the plat told him what it was going to be and the census told him there was no competition yet.

The position I would defend against other biographers is that the disagreement among the sources belongs in the book, not in a footnote. The common practice is to pick the most vivid source, usually the memoir or the letter, describe the town in its terms, and cite the census in a note for the number. That produces a better paragraph and a worse history, because it hides the fact that the town in 1850 was itself a disagreement, a place where a grid on paper and a few cabins in the mud were both true at once. The subject of my book chose to stay in that disagreement, and the reader deserves to see the same four pictures he saw.

The limitation, which I state because it cost me two chapters of confidence, is that this method tells you what a source is for and not whether it is right within its purpose. A census taker can miss a house. A plat can be filed for a town that was never even surveyed on the ground. I had to check the plat against the deed index to confirm that lots were actually sold in 1849 and 1850, and they were, and I had to check the census against a tax roll to confirm that the households it lists were paying on property in the county, and most were. Without those checks the method is a way of organizing sources, not a way of testing them, and organized sources can still be wrong together.

The village at the ford is under a city now, and the square on the 1849 plat is still the square. If you stand on it with the four sources in mind you can see all of them at once: the grid the promoters drew, the handful of families the enumerator counted, the store the settler bragged about, and the mud the old man remembered. The young lawyer saw the same four things in 1850. He bet on the grid.