📄 Wonder, beauty, ability and the natural world: The experience of wonder as a positive aesthetic emotion in Old English verse

Summary

The paper looks at how Old English poetry expresses wonder as a positive aesthetic emotion. Aesthetic emotions in Old English have received far less systematic attention than utilitarian ones such as anger or fear. The author uses the model of Fingerhut and Prinz, which places WONDER in the AMAZEMENT family together with AWE and the sublime. He also draws on literary accounts of wonder by Walker, Brewer and Dailey.

He sorts attestations of the wonder vocabulary (wundor, wundorlīc, wundrum, wundrian) from the DOEC poetic corpus into fine-grained categories. The paper then analyses three of them: wonder at beauty, wonder at skilled craftsmanship, and wonder at the natural world. Each passage is read for its appraisals, mainly perplexity, sensory captivation and reverence, along with the awe features of vastness and need for accommodation.

The analysis finds that the emotion can be triggered by perceptual appraisals, cognitive ones, or both. In these contexts it is mostly positive and sits close to AWE and to the EXPERIENCE OF BEAUTY. It also does real work in knowledge-seeking, especially in the Riddles. This matters for the history of emotions and for how early audiences engaged with poems.

Key points

• The framework is Fingerhut and Prinz (2019; 2020), who describe three dimensions of wonder: cognitive (perplexity), perceptual (captivated senses) and spiritual (reverence). Awe appraisals come from Keltner and Haidt (2003).
• Brewer (2016) gives five features of wonder triggers: novelty, excitement, lack of explanation, a drive to understand, and weakening with familiarity. The paper's material confirms all of these.
• The paper separates two patterns:
  • wundorlīc acts as a secondary flavouring on beauty terms (Christ, Met 21.31);
  • wonder is the core response and beauty lexis only adds detail (Phoenix 230, 331, 341; Judith 7).
• Natural-world wonder without religious meaning is rare. It comes almost entirely from the Meters of Boethius and the Riddles, and it is limited to heavenly bodies, sun, moon, water and fire.
• The Meters deny wonder status to obvious facts, for example in Met 17.7 and Met 20.79. Awe-like cases often lack vastness, threat or supernatural causality. Supernatural wonder is left for future work.

Interesting ideas

• I1. The paper asks whether the emotion word is the head of the evaluation or only a modifier on another evaluative term, and treats the answer as deciding which emotion is primary in the passage. This is a structural cue rather than a lexical count.
• I2. An emotion is described as a bundle of appraisals (vastness, need for accommodation, perplexity, supernatural causality), not as a single label. This makes it possible to explain borderline cases, such as wonder versus awe, by naming which components are missing.
• I3. Texts that explicitly say something is not a wonder because it is common knowledge work as natural negative evidence for the concept.
• I4. Wonder is modelled as the opening step of a sequence: wonder, questioning, investigation, discovery. Riddles use wonder to drive the identification of a hidden object, and the wonder dissolves once the answer is found.
• I5. Wonder weakens with repeated exposure, so novelty relative to what has already been seen shapes emotional intensity.

For Mova

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

• I1 (frames): when building evaluation frames, prag could use the UD relation (amod vs. head noun/obj) to decide which evaluative predicate is primary and which merely modulates it.
• I2 (world): a character's emotional state could be stored as an explainable set of appraisal features, so the system can justify why a state counts as "awe" rather than "wonder".
• I3 (learn): such explicit denials could be mined as negative controls when evaluating whether an emotion or event classifier fires only on genuinely novel or unexplained situations.
• I4 (answer): riddle and fable QA could treat a "wonder/marvel" cue as marking an open knowledge gap that the reader must resolve, and track when the gap is closed.
• I5 (world): an event-order model could weight an entity's reaction by whether it has already met that phenomenon earlier in the text, to predict a fading or intact reaction.