📄 Terasawa, Jun 2011 : Old English Metre: an Introduction

Summary

This item is not a research article. It is Rafael J. Pascual's review in SELIM 18 of Jun Terasawa's textbook Old English Metre: an Introduction (University of Toronto Press, 2011). The problem the book addresses is that metrical scholarship has grown so complex since Sievers that most beginners' guides either skip metre or treat it poorly.

Terasawa offers two things:

• For students: a mostly Sieversian, practical introduction.
• For advanced readers: a balanced survey of open debates.

The reviewer finds both aims achieved. He praises:

• the layout of seven chapters that grow steadily harder,
• the abundant examples, mostly from Beowulf,
• the graded exercises,
• the eclectic use of other theories, such as Russom's word-foot model,
• the chapters showing how vocabulary and grammar bend to fit the metre.

He has two criticisms. First, the book wrongly implies that syllable count is unregulated. Second, it declines to call Beowulf early, despite evidence from Kaluza's law. The review matters because it positions the book as the modern successor to Bliss (1962) and as a gateway to competing theories.

Key points

• Sievers' system sorts the roughly 60,000 surviving OE verses into five basic types. The reviewer notes that it describes the verses well but does not explain why they are metrical.
• The book has seven chapters, 26 exercises with suggested answers in Appendix A, an appendix of scanning tips and a glossary. Sections are marked as intermediate or advanced so they can be skipped on a first pass.
• Chapter 5 shows that when an unstressed syllable directly preceded a compound, poets steered clear of compounds whose second element would have to begin with a resolved stress. R. D. Fulk named this tendency "Terasawa's Law" (Fulk 2007).
• Russom's word-foot theory explains how variable half-line length is still constrained: at most four extrametrical syllables may fall between feet. The reviewer says Terasawa's claim that syllable count is unregulated overlooks this.
• On dating, Fulk's work argues that Beowulf's adherence to Kaluza's law places it before about 725 if Mercian, or before about 825 if Northumbrian. The reviewer faults the book for staying neutral on this.

Interesting ideas

• I1. Each section is tagged with a difficulty level so learners can skip advanced material and return to it later; this is a cheap way to stage complexity in one artifact.
• I2. Some exercises cast the student as an editor deciding whether an ambiguous manuscript reading should be emended on metrical grounds, which trains judgment rather than recall.
• I3. Strict rules are inferred from what poets systematically avoid, such as certain compounds, morphological variants and constructions, not only from what they use.
• I4. Poetic grammar loosens prose norms under metrical pressure: weaker subject–complement concord, weak adjectives without a determiner, and blurred word order.
• I5. The review contrasts a purely descriptive taxonomy (Sievers' five types) with a principle that explains it (feet mimicking word-stress patterns), arguing the latter captures the real logic.

For Mova

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

• I1 (learn): Tag training items and repair rules with difficulty tiers, so curriculum training and evaluation can report results per tier rather than as one blended score.
• I2 (answer): Build test items where the reader must decide whether an odd input is a corruption to repair or a valid rare form, and require it to justify that choice.
• I3 (learn): Mine treebanks and silver data for structures that are expected but consistently missing, and use these gaps as negative-control evidence when proposing or rejecting self-induced rules.
• I4 (tree): Treat verse as its own UD dialect with relaxed agreement and order expectations, so the parser does not force prose analyses onto rhymed fables.
• I5 (learn): Prefer repair and classification rules that state an underlying cause over long lists of observed patterns, which keeps each step explainable.