📄 Burns, Rachel A., and Rafael J. Pascual, eds. 2022. Tradition and Innovation in Old English Metre. Leeds: Arc Humanities Press. Pp. xii + 281. ISBN 1802700250.

Summary

Daniel Donoghue reviews an edited volume that tests how far the classic Sievers–Bliss account of Old English verse still serves editors and readers. The book grew from a 2019 Oxford workshop run by the CLASP project and is dedicated to A. J. Bliss. Rafael Pascual's introduction argues that metre can help establish reliable texts and guide interpretation. Thirteen essays follow.

The essays do several kinds of work:

• They give full scansions of whole poems.
• They question editorial emendations.
• They link hypermetric passages to narrative design.
• They measure scribal word spacing.
• They rethink the five Sieversian verse types.

Donoghue welcomes the volume overall and praises the essays that connect metre to rhetoric and aesthetics. He also makes three criticisms:

• One essay defines rhyme so broadly that the category loses meaning.
• Some essays report percentages with false decimal precision.
• A few scansions contain errors.

The collection matters because it shows that metrical analysis is still active and relevant. It supports textual criticism, narrative reading and the history of how scribes perceived verse.

Key points

• The volume (Arc Humanities Press, 2022, xii + 281 pp.) comes from a 2019 Oxford workshop of CLASP, the project directed by Andy Orchard. It contains an introduction, 13 essays, appendices by Peter Lucas and Mark Griffith, a glossary of metrical terms and an index.
• R. D. Fulk criticises the metrical editing policy of the ASPR. He argues that sawl in Soul and Body II, l. 5, is a scribal spelling of disyllabic sāwul, so the line needs no emendation.
• Lucas supplies a scansion of every half-line of Exodus (lines 1–590), followed by a breakdown by verse type.
• Rachel Burns measured interword gaps in pixels across six folios copied by Scribe A of Beowulf (Cotton Vitellius A.xv). The gaps are wider after on-verses and wider still after off-verses.
• Geoffrey Russom explains the unresolved second stress in type A2k (e.g. hlēoburh wera, Beo. 1731b) with a universalist theory built on word-feet. His account cuts across the Sieversian types.

Interesting ideas

• I1. Fulk prefers to explain an apparently defective form as a spelling variant of a valid word rather than "correct" it, which keeps the source intact while still yielding a regular analysis.
• I2. Donoghue warns that counting unstressed inflections as rhyme makes rhyme appear everywhere, so a feature that always fires (like ongan + infinitive) tells you nothing.
• I3. On reporting, the review recommends dropping decimals on small samples and preferring phrases like "about half" when the counts are modest and noisy.
• I4. Burns treats measured spacing in the manuscript as indirect evidence of how the scribe perceived verse units, so layout carries structure.
• I5. Griffith reads the demonstrative in þone sælidan as pointing back to the earlier sælida addressed by Byrhtnoth, tying a definite marker to a distant antecedent. Hartman likewise shows that a change of verse form can mark a change of narrative function.

For Mova

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

• I1 (tokens): Map odd spellings to a canonical lemma through normalization rules and keep the surface form, rather than rewriting the input text.
• I2 (learn): Run each detector or repair rule on negative controls and check its base rate, rejecting it if it fires on almost all inputs.
• I3 (learn): Report each evaluation score with its sample size and rounded to a precision the data can support, so small test sets do not look falsely precise.
• I4 (tokens): Use whitespace and layout gaps (blank lines, indentation, double spaces) as weak signals for sentence and segment boundaries.
• I5 (frames): Let definite determiners drive long-distance coreference candidates, and treat stylistic shifts as cues that separate narrated events from commentary.