We list here, with brief comments, passages where we have adopted readings rejected by Petrocchi.
3.31 We depart from the Petrocchi reading, error, in favor of orror, even though orror may be a lectio facilior; error sacrifices the powerful sensory vividness of the bristling hair.
8.111 Petrocchi’s “modal conjunction” (che, deriving, like ché, from Latin quid, and not orthographically distinguished in Dante’s day) seems a needless sophistication here, though we adopt it in 8.64 and 30.132.
11.28. The context seems clearly to imply the definite article. The traditional reading here is “de’ violenti” [of the violent]; as Petrocchi himself established, Dante’s practice is best interpreted in such cases as d’i; Petrocchi’s di seems arbitary here.
14.48. We adopt the traditional reading here (maturi [ripen]); Petrocchi’s marturi [inflict pain] seems clearly a lectio facilior and sacrifices both a striking metaphor and a very characteristic sarcasm.
16.102. We accept the traditional dovria … esser [there should be], as opposed to Petrocchii’s dovea … esser [there was to have been], which some of the early commentators explained by supposing that the Conti Guidi had once planned to build a castle at San Benedetto delle Alpi; but mille [a thousand], which they took to mean “a thousand people,” is clearly the correlative of una scesa [one descent], meaning that there is a single cataract when the river is in spate, but a thousand stages of descent when it is not.
19.45. The context clearly requires che sì piangeua [who was weeping so]—line 32 has identified Nicholas as “wriggling more” than the others—rather than Petrocchi’s che si piangeva[who was weeping].
24.69. Like other modern editors, we adopt Pietro di Dante’s emendation ad ire for the widely attested ad ira, which seems the lectio facilior; however, the meaning he asserts, “to go,” seems redundant with mosso [moved], which if complemented would seem to require an expression of ethos. We incline to the view that ire is the plural of ira [anger], though such usage would admittedly be unprecedented.
26.14. This is a vexed passage, for the manuscripts, givingfatto [made], do not provide the agreement of past participle (normal in Dante), which would resolve the difficulty (i.e.,fatte would indicate that the stairs—[scalee, feminine] had been “made” by the projections;fattiwould indicate that “we” [ne, masculine] had been “made”—pale, if iborni is accepted as a neologism based on the Latin eburneus [ivory-colored]). We follow the traditional reading, rather than Petrocchi’s die n’avea fatto iborni, which seems forced.