Prestige forbids translation

Claim (verbatim)

Middle English scribes translated dialect as they copied — except, this conjecture claims, when the author outranked them. Copies of Chaucer preserve alien authorial forms at rates that copies of Lydgate or anonymous romances do not, and not because Chaucer's scribes were better trained: an author who had become an auctor acquired the textual privileges of Latin classics, whose wording one copies rather than converts. Dialect translation is thus an inverse prestige meter — the more freely a scribe naturalizes the language, the lower the author stood in his eyes. If this holds, we can rank the perceived canonicity of vernacular authors decade by decade by measuring how much linguistic conversion their copyists dared, giving canon-formation a quantitative instrument that predates any explicit criticism.

Prediction clause (verbatim)

For scribes who copied both Chaucer and Lydgate (identifiable in the surviving corpus), relict non-scribal forms per thousand words will be at least 50 percent higher in the same scribe's Chaucer stints than in his Lydgate stints, in at least four of five such scribes; primary clause: the within-scribe Chaucer-over-Lydgate relict gap, which removes training and region as confounds.

Kill-dataset (verbatim)

Kill: eLALME linguistic profiles plus the Canterbury Tales Project transcriptions, with shared-scribe identifications from the Late Medieval English Scribes database.

Novelty / leakage triage

already answered in the literature

Searched scribal treatment of Chaucer's language. Horobin's The Language of the Chaucer Tradition makes this exact finding: fifteenth-century scribes carefully preserved archetypal Chaucerian spellings (against the assumption of wholesale translation), including in comparison with the Lydgate tradition. The same-scribe relict-rate quantification is a refinement of a published result.

• S. Horobin, The Language of the Chaucer Tradition (D.S. Brewer, 2003)

Predictions

No prediction registered yet.