📄 Ruszkiewicz, Dominika. 2021. Love and Virtue in Middle English and Middle Scots Poetry. Studies in Medieval Language and Literature, 58. Berlin: Peter Lang. Pp. 221. ISBN 9783631861738.

Summary

In this two-page review (SELIM 28, 2023), Nicola Royan assesses Dominika Ruszkiewicz's 2021 monograph on how Older Scots poets handle the tension between love and virtue. The book also asks how that tension reflects the relation between natural law and divine law. Its method is comparative. Each chapter pairs an Older Scots poem with a Chaucerian model, and Boethius serves as the shared philosophical framework. Royan notes that most of the pairings are familiar. She still credits the sustained Boethian reading as new and finds several interpretations convincing:

• The Kingis Quair engages Boethius more deeply than is usually assumed.
• Henryson gives Christ-like traits to Cresseid, while Chaucer lifts Troilus above earthly suffering.

The book's overall argument traces a continuity in Older Scots verse, from the Quair to The Thrissil and the Rois. In this line, virtuous conduct and lived experience become the way to endure mortal uncertainty. Royan's reservations concern dated or unusual text editions and a narrow, well-studied corpus. This is literary scholarship, not NLP, so its value for us is mainly methodological.

Key points

• The book is volume 58 of Studies in Medieval Language and Literature (Peter Lang, Berlin), 221 pp., ISBN 9783631861738.
• The chapter structure is:
  • Ch. 1: background on Boethius, vernacular and Latin traditions, and classical versus Christian views.
  • Ch. 2: The Romaunt of the Rose with The Goldyn Targe, on reason.
  • Ch. 3: Chaucer's Boethian poems with The Kingis Quair.
  • Ch. 4: The House of Fame with The Palice of Honoure.
  • Ch. 5: The Thrissil and the Rois read alongside The Parliament of Fowls, a pairing Royan judges the most original.
  • Ch. 6: Troilus and Criseyde with The Testament of Cresseid.
• Chapter 4 contrasts two views of fame and honour. Chaucer treats fame as earth-bound and fleeting. Douglas places honour within an exemplary order tied to good government.
• In Chapter 5, the language addressed to Margaret Tudor overlaps with Marian devotional verse. This casts the queen as a model of mercy, while the king acts as judge and lawgiver.
• Royan's quibbles include:
  • Denton Fox's 1981 Oxford Henryson remains the standard edition.
  • Bawcutt's Douglas edition was revised in 2002.
  • Kendal's Eneados is idiosyncratically modernised.
  • She suggests extending the analysis to less familiar texts such as Hary's Wallace, since the comparisons with Gilbert Hay already proved useful.

Interesting ideas

• I1. Reading an adaptation side by side with its source shows what the later author deliberately changed, rather than treating each text alone.
• I2. A single fixed thematic axis (love versus virtue) is applied consistently across many works, which keeps comparisons commensurable.
• I3. The same moral attribute can move between characters across versions, as when Christ-like suffering goes to Cresseid in Henryson but Troilus is elevated in Chaucer.
• I4. Borrowing the vocabulary of one register (Marian devotion) to address someone else (a queen) silently assigns that person a role.
• I5. The reviewer warns that conclusions depend on which editions are used and that an over-familiar corpus can bias findings.

For Mova

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

• I1 (learn): Build paired evaluation sets of an original fable and its retelling, then check that Mova's graphs differ exactly where the stories diverge and match elsewhere.
• I2 (frames): Define a small, fixed inventory of value conflicts (e.g. desire vs duty) as frame-level tags, so morals of different tales can be compared on the same dimensions.
• I3 (world): Track attributes as explicit links from attribute to entity, so a comparison across versions can report "this property moved from character A to character B."
• I4 (frames): Flag lexical register borrowing as an implicature cue, inferring an implied role or stance from words typical of another domain.
• I5 (learn): Record the exact source version of every training and test text, and add less-studied texts as out-of-distribution controls.