Script change and dating

Change without a single ladder

Scripts change through training, region, institution, medium, genre and contact. A formal book hand and a rapid documentary hand may coexist; one is not merely an older stage of the other. In Latin, Caroline and later protogothic features can appear in the same manuscript as main text and addition, as the Harley MS 2830 record shows. Dunhuang evidence distinguishes manuscript types and changes in brush and paper. Treat these as situated observations, not a universal sequence in which every sign inevitably becomes simpler.

Describe a proposed development by specifying which feature changes, in which dated and located witnesses, and what competing explanation—register, medium or writer—has been considered. A single unusual form can survive as an archaism or be produced by an individual habit. A chart of idealized alphabets is a finding aid, not an independent chronology.

Date as an argument

A document may carry an internal date, an archaeological context, a dated colophon, or only comparative handwriting evidence. These sources differ in what they date: composition, copying, burial, reuse, or later annotation. For an undated hand, compare several independently dated examples from the same graphic stream and region. State a range and show both early and late comparanda. Do not use the manuscript’s proposed date to prove the very letterform sequence on which that date depends.

The debate over P.Ryl. iii.457 (P52) is a useful caution. Brent Nongbri argues that a commonly repeated date around 125 CE is much narrower than handwriting comparison supports, and that plausible comparisons extend into the third century. The point for a student is methodological: a palaeographic attribution is neither a calendar year nor a direct measurement of when ink touched the sheet. Other evidence can narrow or revise the range.

Practice

Choose two dated manuscript records in one script tradition and an undated record. Build a table of three graphic features, provenance, material and basis of each date. State a range for the undated hand, including one reason each comparison might fail. If the two “dated” records are themselves dated only by palaeography, find an independent anchor before claiming a calibrated chronology.