A filed town plat records a promoter's hope in ink, and the county clerk kept it whether the train came or not

The clerk unrolled it on the counter with two paperweights and stood back, and the first thing I noticed was the wide diagonal running across the grid from the southwest corner to the northeast. Every street on the sheet was named and numbered. The diagonal was labeled too, in the same confident hand, with the name of a railroad. There has never been a railroad through that town. There was never a grade cut for one. The line on the plat is the only place it ever existed, and the plat has been on file in the county since the year the promoter drew it.

A town plat, for readers who have not handled one, is the drawing a landowner files with the county when he divides a tract into lots and offers them for sale. It is a legal instrument, because deeds to the lots will describe them by block and number on this sheet, so the county keeps it forever. It is also, and this is the part a historian can use, a picture of what the person who drew it believed was about to happen. Nobody plats a town for the town that exists. He plats it for the one he expects.

I read a plat in a fixed order now, because I got fooled early on by reading the streets first and the margins last. The order goes like this.

1. Find the filing date and the name of the surveyor, which are in a corner or on the back. The date tells you the year the expectation was formed. The surveyor’s name tells you who else he was drawing for in those years, which in turn tells you which promoters were working the same county.

2. Find the public reservations before you find anything else. A block marked for a courthouse means the promoter expected the county seat to move or be created here, and there will be a fight in the commissioners’ court minutes to prove it one way or the other. A block for a depot means he expected a railroad. A block for a college means he was selling to families. These reserved blocks are promises made to buyers, and they were rarely kept without a struggle.

3. Read the street names as a dated document. Streets named for a railroad company, a railroad president or a rail town elsewhere in the state date the sheet to the year that company was surveying nearby. Streets named for delegates to a convention or for a governor date it to a political moment. A sheet with a Coke Street and a Hubbard Street, for example, was drawn by someone with a definite view of the state’s politics in the middle of the 1870s.

4. Compare the plat to the deed index. This is where the drawing meets the ground. Count how many lots actually sold in the first two years, and to whom. A plat with two hundred lots and eleven deeds is a hope that failed. A plat with two hundred lots and a hundred and forty deeds inside of eighteen months is a town, whatever the railroad did.

5. Only then look at the shape of the grid, and ask what the promoter thought would drive the traffic. A grid oriented to a river landing expected boats. A grid oriented to an existing road expected freight wagons. A grid with its widest street pointed at nothing visible expected the nothing to be filled in by a line of track, and if you go back to the corner of the sheet you will usually find the diagonal that explains it.

That fifth step is where I part company with some people who use these records. The common practice is to treat a plat as a map, to read it for what was there. I treat it as a forecast, made by a man with money at risk, and I read it for what he thought would be there. The distinction matters because a forecast that failed is still evidence. It tells you what the informed money in a county expected in a given year, which is something no diary or newspaper will tell you as plainly, and it tells you when that expectation died, because the sheet on file is never revised. The promoter who drew my diagonal believed in that railroad hard enough to give it a right of way through the middle of his town, and the deed index says he sold lots along it for three years before the sales stopped. He was wrong. He was also informed, and the year in which he was informed and wrong is a fact about the county’s history that I could not have found anywhere else.

The limitation is built into the same reasoning, and I had to learn it the slow way. A plat records one man’s expectation, not the county’s. If the promoter was a fool, the sheet records a fool’s hope, and the survival of the document in a vault gives it a dignity it did not have in its own time. So the deed index has to be read every time, because it is the check. A drawing tells you what was expected. Only the sales tell you whether anyone else expected it too, and I have seen plats where the answer was nobody, not one buyer, and the promoter’s diagonal ran through two hundred lots that never left his hands. That sheet is evidence of one man’s mind and nothing more, and a historian who writes it up as a town has written fiction with a citation.

There is also a quieter trap. Streets get renamed. A town that outlived its founding hope will often have struck the railroad’s name from the signs a generation later, so that the living town and the filed plat disagree about what a street is called. When the deed index switches from one name to the other, that switch is itself a date, and usually it is the year the town gave up.

The clerk rolled the sheet back up when I had finished and slid it into its tube, where it will stay. The railroad it promised was surveyed through the next county over, and the town that got the depot is the one with the population today. This one kept its plat.

A hope written down and filed with a clerk is still a fact about the year it was written, and the county keeps it as carefully as it keeps a deed.