CRITICAL NOTES ON SERVIUS’ COMMENTARY ON VIRGIL (SERV. ON AEN. 11.741; ECL. 2.58; ECL. 4.4)

Abstract This article discusses three textual problems in Servius’ commentary on Virgil (Serv. on Aen. 11.741; Ecl. 2.58; Ecl. 4.4). In two notes a new conjecture is proposed; in one passage a transmitted reading, so far neglected by earlier editors, is supported.

At the beginning of the scholium to Aen. 11.741, all editors print moriturus animo, apparently taking animo as an ablative of respect ('ready to die in his soul'). But other instances of animo as an ablative of respect show that the syntagm moriturus animo is unexpected and suspect. 6 Equally suspect is the repetition of moriturus (MORITVRVS ET IPSE moriturus animo […]). Although Murgia's apparatus criticus indicates that morituri is the reading of the Servian archetype ([Ʃ]), he rejected this reading and followed previous editors in accepting moriturus found in the auctus-witness (F) and in some manuscripts of Servius. 7 But does morituri animo really not make any sense?
Given the context, morituri must be a genitive singular depending on animo, and the clear meaning of morituri animo effectively glosses moriturus: '(he spurs his horse/performs this action) with the mindset of a moriturus, of someone who is going/ready/ doomed to die'. The syntax of morituri might seem rather elliptic, whereas moriturus straightforwardly resumes the lemma at the beginning of the scholium. But this is 3  7 I agree with Murgia (n. 1) on the archetypal status of morituri, which is transmitted in three witnesses belonging to different families. In Murgia's reconstruction, the moriturus of some Servian manuscripts is either a conjectural reading or a fortunate mistake or a contamination from Servius auctus. Most likely, the reading of Servius auctus persuaded Murgia to print moriturus. The other transmitted readings are unimportant, but they do suggest that the ending of the participle was 'unstable' in the manuscript tradition.
precisely the reason why morituri should be considered a lectio difficilior: given the lemma and the subsequent moriturus non est, the easy corruption of morituri to moriturus could well have occurred independently in F, in Pc and in γ. Moreover, morituri animo seems to correspond to the structure 'affectus + genitive-singular present participle', which Servius sometimes uses to define a character's state of mind, their own 'subjective perspective', as in the case of Coroebus' dimicantis affectus (see above). 8 In the Servian corpus there is no further instance of animus + future participle, but this is not a serious obstacle. The structure is quite common in Tiberius Claudius Donatus and in Donatus' commentary on Terence, where animo + genitive-singular participle indicates the mindset with which a character performs an action. 9 The text printed so far, though suspect, is perhaps not unacceptable. However, Murgia's new evaluation of the manuscript tradition invites us to consider seriously the genitive morituri. Here is Servius' commentary on line 58 as printed in Thilo's edition (the apparatus criticus is based on a new collation of the witnesses).  , that pleasure comes in two kinds: one that is due to perception and another due to imagination, which is born from our thinking. Therefore, we must understand that Corydon has experienced this imagined pleasure, thanks to which he seemed to see his absent beloved and talk to him. But after his own reproach had led him to regain his natural clarity of thought, he was certainly deprived of that imagined pleasure, as here he says he has hurt himself by this reasoning: [Ecl. 2.56-7] 'Corydon, you are a clown! Alexis cares naught for gifts, nor if with gifts you were to vie, would Iollas yield.' Two surrounding scholia are related to this one: the scholium to lines 56-7 on the selfreproach (Serv. on Ecl. 2.56 RVSTICVS ES CORYDON arguit se stultitiae, quod eum se sperat placare muneribus, qui potest habere meliora, nam supra ait 'delicias domini') and the scholium to line 60 on Corydon's phantasia (QVEM FVGIS A. D. iterum per phantasiam quasi ad praesentem loquitur. […]).
In the scholium to Ecl. 2.58 Servius first points out a seeming contradiction: Corydon says that Alexis is going away, while at the beginning of the Eclogue Corydon is presented as being alone. 11 Then the commentator explains that the character is using his imaginatio 12 and that, after the self-reproach of lines 56-7, he comes back to criticus only for the section I analyse (quomodo … caret). his 'clarity of thought'. The structure of the scholium is very odd indeed. The opening remark (eum dicit discedere) refers not to the lemma 'heu heu, quid uolui misero mihi?' but to line 60 'quem fugis?' without any indication of a cross-reference (for example paulo post). 13 Only at the end do we find a (relatively) clear reference to the lemma (nunc sibi se offuisse dicit: contrast eum dicit discedere) and its context (56-7). 14 Apparently some exegetic material related to line 60 was conflated with that relating to lines 56-8, but a transposition of the scholium of line 58 (or also part of it) to line 60 is not possible, since at Serv. on Ecl. 2.60 there is already a note on phantasia (see above) and ubi nunc … dicit anchors the scholium to line 58 (that is, to the context of Corydon's 'rationality'). Hence the most probable hypothesis is that Servius is merging scholia related to different lines in a rather mechanical and clumsy way (as he often does), 15 with the aim of offering a general reflection on this pivotal passage (56-60). 16 In the scholium to line 60, he then presupposes this detailed note on imaginatio (iterum = 'again', that is, after a moment of naturalis prudentia).
The textual problem at the opening of Serv. In the Servian corpus (especially in the auctus) there are many instances of quaestiones introduced by quomodo on possible contradictions in the Virgilian text. 17 The addition of sed (sed ratione non caret), found in MSS Bo H and printed so far in many editions, is not strictly necessary, but highlights very well the editors' discomfort with the conciseness of ratione non caret. Some manuscripts offer a quite different scenario: Bc M Pb σ: quomodo eum dicit discedere (discedere dicit Pb), 18 quem supra cum eo diximus non fuisse? nam quod ait 'solus montibus et siluis' ratione non caret: Epicurei enim … 13 Thilo offers no indication of this, but see already Burman (n. 10), ad loc.: 'haec nota pertinet ad vs. 60' (similarly H. Georgii, Die antike Vergilkritik in den Bucolica und Georgica [Leipzig, 1904], 226 and Setaioli [n. 12], 36). Cf. also the parallel of Serv. on Aen. 6.465 SISTE GRADVM discedere eam datur intellegi (at line 466 Aeneas asks Dido precisely 'quem fugis?').
14 With 'nunc dicit + quotation' Servius usually gets back to the lemma, but here the quotation does not go so far as to include the lemma and works only as a cross-reference to the ratiocinatio/obiurgatio of lines 56-7. 15 It is often possible to detect the merging of different scholia in the commentary of Servius by means of comparison with the auctus. The conflation in Serv. on Ecl. 2.58 may also be due (at least partially) to a rearrangement of this section in the transmission of the commentary. These readings, though perhaps conjectural, deserve serious consideration. 20 Servius uses this expression (quod dicit/ait … ratione non caret, 'the fact that he says … has a reason') to clarify seeming inconsistencies in the Virgilian text. 21 In this scholium the quod-clause would provide a subject for ratione non caret, which otherwise would be hanging in the air-hence also the sed added by many editors. While the reading of MSS Bc M Pb σ is not convincing (nam must introduce the quotation solus montibus et siluis, certainly not the answer to the quaestio itself), the reading of MS Y makes perfect sense and is very likely to be the right reading, which was then corrupted in the typical quomodo opening many scholia. Following MS Y, I would go a step further, by offering a correction that could better account for the quomodo of the other witnesses: quod modo ('the fact that now he says', in contrast with supra diximus … nam ait …). 22 Servius uses quite often the expression quod modo, also in comparisons of two (seemingly) contradictory passages. 23 Obviously, the presence of modo at the beginning of the scholium to Ecl. 2.58 seems to clash with ubi nunc … dicit at the end of it: quod modo dicit would perfectly fit with quem fugis? as a lemma, while it is less suitable to introduce a cross-reference to a nearby passage. But if, as demonstrated above, a scholium from line 60 was conflated by Servius quite mechanically here, quod modo could well be the original reading. This is a further element to consider in the issue of the odd structure of the scholium to line 58. The relative pronoun quae clearly refers to an implicit Sibylla, but the formulation is particularly awkward and hardly acceptable. If we consider ancient Virgilian exegesis besides Servius, two elements stand out. From an exegetical point of view, Servius is implicitly rejecting the interpretation of Cumaei as a reference to the Asian Cyme and Hesiod as a source of the myth of the Ages (saecula per metalla diuisit), 27 while also hinting at the existence of different Sibyllae defined on a geographical basis (quae Cumana fuit). 28 From a textual point of view, we have a confirmation that a Sibylla before quae is needed. 29 The following corrections were proposed: • As noted by Guarinus, both the meaningful gloss CYMAEI Sibyllini and the transmitted quae should be preserved. 33 Guarinus's solution is interesting for two other reasons: • dixit etiam is also found at the beginning of the sentence some lines below (dixit etiam finitis, etc.); • the interpretation of Sibyllini … diuisit as an 'autonomous' sentence is supported by the auctus-witness (Leid. Voss. Lat. O. 80), which omits the rest of the note after diuisit and provides a text that can be understood only if a genitive Sibyllae is added. 34