📄 Worcester, William. 2023. The Boke of Noblesse and the English Texts from its Codicil. Edited by Catherine Nall and Daniel Wakelin. Early English Text Society O.S. 362. Oxford: Oxford University Press. Pp. 350. ISBN 9780198894513.

Summary

Laura Poggesi reviews a new critical edition of William Worcester's Boke of Noblesse, edited by Catherine Nall and Daniel Wakelin (EETS O.S. 362). Worcester wrote this political treatise after the Hundred Years' War. It laments the loss of Lancastrian lands in France between 1448 and 1451 and argues for winning them back. The only earlier edition, John Gough Nichols's of 1860, did not recognize Worcester as the author. This edition is the first to print the treatise together with the English part of its Codicil, a companion collection meant to supplement it. It also adds three Latin and macaronic Codicil texts that Stevenson's 1864 edition left out. A long Introduction covers:

• the manuscripts, codicology and palaeography;
• Worcester's own revisions and his sources;
• his prose style and his late Middle English language.

The volume also has explanatory notes, a glossary and an index of names. The reviewer finds the text reliable and consistent with its stated principles. She judges it a key resource for historians, philologists, literary scholars and linguists, partly because it places a fifteenth-century work firmly in its context.

Key points

• Each text survives in a single witness: the Boke in British Library MS Royal 18.B.xxii and the Codicil in Lambeth Palace Library MS 506. Both are fair copies by different scribes, revised by Worcester himself.
• A colophon in Worcester's hand (Royal MS, f. 42r) dates the end of his revision to 15 June 1475. The editors detect two separate phases of marginal annotation, written in different inks.
• The volume runs to 350 pages:
  • Introduction (xiii–cviii)
  • Appendix (107–23)
  • Explanatory Notes (124–255)
  • Glossary (256–76)
  • Index of Proper Names (277–88)
• Worcester drew on a wide range of sources: Cicero, Cassiodorus, Ranulph Higden, Ralph de Diceto and Geoffrey of Monmouth, as well as Christine de Pizan and Alain Chartier.
• His style draws mainly on administrative "curial" prose and, to a lesser degree, on Latin periodic style. The editors describe its morphological, syntactic, orthographic and lexical traits as typical of late Middle English.

Interesting ideas

• I1. The editors classify the author's changes by type (erasures, interlinear or marginal additions, corrections, glosses) and separate them into dated layers by ink, so a text becomes a record of successive revision phases.
• I2. Annotations by later readers go into textual footnotes, kept apart from the author's own revisions. The edited text therefore never mixes witnesses of different authority.
• I3. The editors work out how Worcester composed the Boke by following his reading through the manuscripts he owned and annotated. This links each passage of the output back to the inputs that shaped it.
• I4. The edition pairs the text with a glossary and an index of proper names as structured companions. The glossary handles period-specific word senses and the index handles named entities.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

• I1 (world): Each fact in the text world could carry a revision or source layer, so the reader can tell which version of a state was asserted when.
• I2 (learn): Silver data could store corrections from the LLM teacher, from witness parsers and from repair rules as separately labelled layers. That would let us audit or roll back any one source on its own.
• I3 (answer): Answers could cite the exact sentences and compiled-knowledge entries they rely on, which supports the requirement that every step be explainable.
• I4 (tokens): A small period or register lexicon of archaic spellings and their lemmas would help the en crate lemmatize older English such as fables and tales, instead of treating those forms as unknown.