The Meeting Minutes Stop in March and That Is the Document

The minute book for the Hollis County Board of Education runs from 1889 to 1961, in four volumes, in a climate-controlled room in the basement of a courthouse annex. The handwriting changes six times. The March 1956 entry is one line: Meeting adjourned, no business. Then nothing until September.

Six months of a school board in a county with eleven schools, in a year in which a great deal was happening to school boards in that part of the country, and the record says nothing at all.

Historians are trained to treat gaps as evidence, and I want to argue for a narrower version of that habit than the one I usually encounter, because I think the broad version has become a way of writing whatever you already believed.

The broad version goes: the silence is deliberate, the silence conceals a decision, and the decision was the obvious one given what we know about the period. You can construct that reading in about ninety seconds and it will be publishable, and it may even be right. What it will not be is demonstrated.

Here is what the archive actually supports. The secretary who kept the minutes through 1955 was a man named Prather, whose handwriting is the fourth of the six. The September 1956 entries are in the fifth hand. There is no personnel note recording the change, but the county payroll ledgers — a different series, two rooms away, and not indexed with the school records — show Prather paid through February and a different name from August. Something happened to Prather in the spring of 1956. He may have been removed. He may have had a stroke.

That is the honest floor of what the documents give you. The interesting part is what sits on top of it: the same six-month gap appears in the minute books of two neighbouring counties, and in one of them the payroll shows no secretarial change at all. A gap that recurs across three jurisdictions is unlikely to be about one man's health, and a gap that recurs where the personnel did not change is not about personnel. Now the deliberate reading has some weight under it, and it did not have that weight when it was merely plausible.

The distinction I care about is between a silence that is evidence and a silence that is a space to put your argument in. Both look identical on the page. The difference is entirely in whether you went and checked a second series that could have contradicted you.

I labour this because the profession has a bad incentive here and I have felt it personally. A confident reading of a gap makes a better article than a careful one. It is also nearly unfalsifiable, which reviewers experience as strength rather than weakness. And it flatters the reader, who gets to feel that they can see through the record to what really happened — that the past was mostly a cover-up and the historian has lifted the lid.

Sometimes it was a cover-up. Hollis County's gap almost certainly is one, on the three-county evidence. But I have also spent a week on a two-year gap in a township roads ledger that turned out to be a fire, documented in a newspaper I had not thought to check because roads are not news.

The methodological point is unglamorous: a gap is a question, and questions are answered in other series, not in the gap itself. If your reading of an absence cannot be checked against an independent record, you have not made an argument. You have made a shape and filled it.

Prather's replacement kept the minutes until 1961 in a smaller, faster hand. In November 1957 he recorded a motion, a second, a vote of four to one, and did not record what the motion was. Somebody objected in that room and lost, and their name is not written down, and I have not yet found the second series that would tell me who they were. I have been looking for four years. Until I find it, that sentence is all I am entitled to say about it, and I have twice been asked by editors to say more.