Kwakkel's 'one in seven' colophon figure carries no cited study, dataset, or corpus count in his own published text

colophonerik-kwakkelunverified-quantmedieval-manuscriptscodicologysourcing-audit

Erik Kwakkel's 2014 blog post states that "about one in seven" medieval scribes added a colophon, but neither the original post nor its identical repost attaches a footnote, a named corpus, a co-author, or a prior scholar's count to the figure. Kwakkel's other examples in the same post are drawn from named manuscripts spanning Paris, Holland, and Maaseik across the 13th–15th centuries, but the ratio itself is not qualified by language, region, or period, and is presented as the author's own summary impression from his career reading manuscripts rather than a reported result.

This matters for the sourcing floor described in 00-meta/specs/sources.md: quantitative claims require Tier 1–2 sourcing, and Kwakkel's post clears that bar for the fact that he said it, but not for the figure as a count of a real-world population of manuscripts — no dataset exists in the text to check the ratio against. Colophon frequency is also known to be tradition-specific rather than a universal constant (comparative codicological literature reports figures around 55% for Armenian manuscripts and roughly 30% for Syriac manuscripts, versus about 3.4% for Hebrew manuscripts — found via a single non-independently-verified academia.edu source this session, not itself promoted), which is further reason not to treat "one in seven" as a settled statistic rather than an expert's working estimate for the Western Latin corpus he studies.