Where this comes from

Brice wrote within living memory of the people he was writing about, and that is both what makes the book valuable and what makes it hard to use. Much of it rests on what old residents told him, named sometimes and unnamed often. His account of Harmar's defeat credits "the vivid recollections he gave, years ago, to early settlers here" and names nobody at all.

The book also disagrees with itself. At page 129 a footnote says the indentations on either side of the Maumee still show a stranger where the ford was. At page 287 the same book says both common fording places are now most entirely obliterated.

One book with several hands in it, and the hands matter. T. B. Helm signed the preface as editor. The township chapters carry the byline BY L. H. NEWTON. The physical geography and the account of the mounds are by Colonel R. S. Robertson, and the preface says as much.

A subscription history was paid for by the people in it. The biographical sketches were written from what their subjects told an agent who came to the door, and they should be read as that: a man's own account of himself, printed.

Plat maps by township, with a written history bound in. Measured against the history published the same year by the same firm, 34.2% of its words are verbatim reuse. That figure is a floor. The two are one source, and this archive does not cite them as two.

Where it writes about what happened after 1868 it is a real second source and often the only one. Where it writes about what happened before, it is frequently Brice with the grammar tidied. Its paragraph on Richardville at Harmar's Ford is Brice's paragraph, near word for word, with one clause shortened.

Whole book against Brice, the measured reuse is 8.8%. That number is the reason the measurement is never the end of the question: a book can copy one page outright and still score low across five hundred.

Slocum wrote the river rather than the county, which is the right unit for anything before the county existed and the wrong one for most things after.

Griswold worked from records rather than from Brice: measured against the 1868 history, verbatim reuse is 0.5%. The plates are the reason to go to him, and they are drawings as often as photographs. A drawing is never described here as a photograph.

He is also capable of being wrong in his own voice. His caption on Harmar's Ford folds the campaign's whole toll of killed and missing into the casualties at one crossing.

The garrison orderly books and the Indian agency account book, printed in full. This is the closest thing in the county's early record to a document that was not written to be read by posterity: it was written to run a post.

It is a printed transcription and it is quoted as one. The three orderly books were at the Burton Historical Library in Detroit in 1927. The account book was in private hands and had been partly cut up before it was rescued. Nothing here has been checked against the manuscript, and every piece drawing on it says so.

Plat maps with landowners named on the sections. An atlas is a record of who held what in the year it was surveyed, and is good evidence of that and of nothing else.

A newspaper is the only source here written the week the thing happened, and the only one that prints a man's own account over his own signature. It is also a partisan paper that disagreed with its rivals about nearly everything, and it sets type badly: the standing head over the 1843 sketch of Harmar's campaign reads FORT WANYE IN OLD TIMES.