Two Jars

The index card read only: “June 14—rain, two jars.” I found it in the bottom drawer of our community garden shed while looking for twine. For thirty years, volunteers had recorded harvests in a code nobody had thought to explain. Some cards listed pounds of tomatoes; others held stars, circles, or a name followed by a question mark.

I first treated the drawer like a puzzle I could solve alone. I sorted dates, built a spreadsheet, and invented theories about rainfall. Then Mrs. Alvarez laughed at my neat categories. Two jars meant tomato blight samples, not preserves. A star marked seeds saved for spring. The question marks recorded vegetables donated before anyone had weighed them. My dataset was orderly because I had erased everything I did not understand.

I began interviewing the gardeners beside the plots they remembered. We scanned the cards, added names only with permission, and recorded short audio explanations that preserved the pauses missing from the handwriting. When two memories conflicted, the archive kept both. The goal shifted from decoding a final answer to showing how the garden's knowledge had been made and revised.

The project changed how I understand research. A record is not neutral simply because it is old; someone chose what fit on the card, and someone else decides what counts as data. I want to study how institutions preserve community knowledge without flattening the people who made it. I still love an orderly table, but now I look first for the story that its columns could not hold.