Fries ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ A Planudean Edition of Pindar? The Evidence of Parisinus gr. 2403 Almut Fries HE QUESTION whether Maximus Planudes (ca. 1255– 1305) prepared an edition of Pindar’s Epinician Odes has been debated since Irigoin ascribed to him the so-called ‘Paris recension’, represented by the late-13th- and early-14th- century manuscripts Par.gr. 2774 (C), Ambr. E 103 sup. (N), Leid. Voss.gr. Q 4 B (O), Vat.gr. 915 (Ø), and Par.gr. 2403 (V).1 Prima facie it is likely that Planudes edited or at least worked on Pin- dar. The Epinicians (especially the Olympians and Pythians) were a mainstay of the Byzantine higher school-syllabus, Planudes was the leading scholar of classical literature of his time, and his two most distinguished students and/or associates, Manuel Mos- chopoulos (born ca. 1265) and Demetrius Triclinius (ca. 1280– 1335), produced influential annotated editions of the Olympians and all four books of Epinicians respectively.2 Irigoin’s theory was accepted outright by Hunger, whereas 1 Detailed descriptions of these codices are provided by J. Irigoin, Histoire du texte de Pindare (Paris 1952) 257–266. None of them preserves a complete text of Pindar. V and C go furthest, breaking off at Nem. 6.44 and Pyth. 5.51 respectively. NOØ contain only (part of) the Olympians, with changing fam- ily allegiances in N and O (see n.26 below). The mutilation goes back to the common source (ζ) of the ‘Paris recension’ and two collections of Pindaric gnōmai, one found in Ø (fol. 47r), the other in Barb.gr. 4 (fols. 56r–64v). See Irigoin 247–248 with nn.5, 6. 2 On the evidence for an association between Planudes and Triclinius see N. G. Wilson, “Planudes and Triclinius,” GRBS 19 (1978) 389–394 and “Miscellanea Palaeographica,” GRBS 22 (1981) 395–397. T ALMUT FRIES 709 ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 Wilson remained sceptical, though not unsympathetic.3 How- ever, it was severely criticised by Günther with the following three arguments:4 a) The text of CNOØV shows no signs of systematic philological and metrical revision. b) Many correct readings in the Olympians are paralleled in the codex Ambr. C 222 inf. (A), the only extant representative of a separate strand of transmission, and thus likely to be genuine paradosis.5 c) There is no evidence in the form of a commentary that Planudes worked on Pindar in depth. While none of these arguments can be dismissed, they do not preclude the possibility that Pindar was studied by Planudes and his circle in a more informal or preliminary manner.6 It has been thought that Planudes oversaw the production of Vat. gr. 915 (Ø), which embeds its small portion of Pindar (Ol. 2.43– 7.29) within a large collection of poetic works commonly read in Byzantine schools.7 Further, and perhaps more secure, evi- dence for Planudes’ involvement with Pindar comes from Par. gr. 2403 (V). The manuscript, written on oriental paper, was dated by 3 H. Hunger, Die hochsprachliche profane Literatur der Byzantiner II (Munich 1978) 68–69; N. G. Wilson, Scholars of Byzantium (Baltimore 1983) 238. 4 H.-C. Günther, Ein neuer metrischer Traktat und das Studium der pindarischen Metrik in der Philologie der Paläologenzeit (Leiden 1998) 64–67. 5 Cf. Wilson, Scholars of Byzantium 238. Irigoin (Histoire 244, 251–252) recognises the influence of A on the ‘Planudean edition’, which he attributes to its model, that is ζ (cf. n.1 above) or one of its descendants. In addition, however, some of the readings he quotes as metrical corrections of Planudes (Histoire 249–251) also occur in A: e.g. Ol. 4.21 ἅπερ [καὶ] Κλυµένοιο παῖδα, 7.12 παµφώνοισί τ᾽ ἐν ἔντεσιν αὐλῶν (ἐν om. codd. pler.). These could still be independent conjectures, but the bulk of the evidence suggests paradosis. 6 On the possibility that Planudes did not live to produce a complete edition of Pindar see 713 below. 7 Homer, Hesiod, Lycophron, Dionysius Periegetes, Theocritus. Only the dramatists are conspicuously absent. On the possible connection with Pla- nudes see Wilson, Scholars of Byzantium 236–237. 710 A PLANUDEAN EDITION OF PINDAR? ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 Irigoin to the end of the 13th century, without any supporting explanation.8 Like Ø, it contains several works that belonged to the Byzantine higher school-curriculum: Aratus Phaenomena, Lycophron, Nicander Theriaca and Alexipharmaca, Pindar Olym- pians, Pythians, and Nemeans 1.1–6.44,9 and nearly the entire Odyssey (1.1–24.309). In addition, there are an anonymous work on cosmography, Proclus’ Outline of Astronomical Theories, John Tzetzes’ verse-treatise on the genres of poetry, and various small pieces added by different hands.10 The text of Pindar (fols. 116r–172r) is of a composite nature. The Olympians and Pythian 1 are copied from Vindob.suppl.gr. 64 (Vi). This book, which Irigoin named ‘manuscript of Ger- manos’ after an otherwise unknown scholar mentioned in a small set of scholia written by the main scribe,11 is a palimpsest where Pindar stands on top of a 10th-century sticherarion. Dated to ca. 1260–1280, Vi bears marks of having been intended for educational purposes, such as careful orthography, extensive punctuation, and annotation. It thus testifies to the value at- tached to higher education in the period of restoration which followed the return of the Imperial court from exile in 1261.12 The manuscript is now mutilated, breaking off after Pyth. 5.55. The remaining Pindar odes in V come from an exemplar related to the ‘Paris recension’, that is, the supposed edition of 8 Irigoin, Histoire 265: “Le manuscrit V est à peu près contemporain du manuscrit C; faute d’éléments de datations précis, je dirai: fin du XIIIe siècle.” 9 On the absence of the remaining Nemeans and the Isthmians in the ‘Paris recension’ see n.1 above. In addition, V has a lacuna between Nem. 4.68 and 6.33, owing to the loss of two folios in that codex (Irigoin, Histoire 265). 10 See Irigoin, Histoire 264. 11 Irigoin, Histoire 217. 12 Cf. A. Fries, “For Use in Schools. Prosodical Marks in Two Pre- Palaeologan Manuscripts of Pindar,” GRBS 57 (2017) 745–770, at 766–767 (with n.53 for other contemporary ‘inverse palimpsests’ of classical school authors). Images of Vi can be consulted at https://www.onb.ac.at/digitaler- lesesaal or via https://pinakes.irht.cnrs.fr/notices/cote/71527/. ALMUT FRIES 711 ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 Planudes. The text is close to C, but since this is the only other extant member of the family which attests more than the Olym- pians (see n.1 above), the observation is of little evidential value.13 However, the two parts of the Pindar text show clear, if limited, signs of scribal contamination, stronger perhaps in the direction from the ‘Paris recension’ to Vi, and some of the interventions may even point to one or more external sources (see n.30 below). I provide select evidence and further ex- planation in the Appendix, expanding on the very sparse data given by Irigoin. The educational background of V’s exemplar in the Olympians and Pythian 1 and the relatively small, but visible, philological effort expended on the composite text point to a learned milieu like Planudes’ circle, but do not substantiate this particular connection. A stronger hint comes from the presence of three cosmographic or astronomical works in V. Triclinius attests to Planudes’ interest in astronomy, and a Planudean autograph of Cleomedes’ Κυκλικὴ θεωρία and Aratus’ Phae- nomena (both used as astronomical textbooks in Byzantium) has been identified in the codex Edinburgh, Advocates Library 18.7.15.14 But by far the clearest indication that V originated in the environs of Planudes is provided by its script. The hand of V’s principal scribe resembles those of manu- 13 Cf. Irigoin, Histoire 266. 14 Triclinius’ comment, which describes Planudes’ substitution of factually incorrect lines in Aratus with verses of his own, is found in two 15th-century codices of Aratus: Oxford, Bodleian Library, Barocc. 109 (fol. 167r) and London, BM Add. 11886 (fol. 26r). The slightly differing text of both manuscripts can be consulted in J. Martin, Scholia in Aratum vetera (Stuttgart 1974) XXX–XXXI. The veracity of the note has been confirmed by Planudes’ treatment of Aratus Phaen. 481–496, 501–506, and 515–524 in the Edinburgh codex. See I. C. Cunningham, “Greek Manuscripts in the National Library of Scot- land,” Scriptorium 24 (1970) 360–371, at 367–368, and A. Turyn, Dated Greek Manuscripts of the Thirteenth and Fourteenth Centuries in the Libraries of Great Britain (Washington 1980) 57–59 with pl. 41–42; cf. Wilson, Scholars of Byzantium 232. 712 A PLANUDEAN EDITION OF PINDAR? ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 scripts known to have been produced in Planudes’ circle. It is defined by the following characteristics:15 a) large circular letters (e.g. β, ο, σ, υ, φ, ω, and the ligatures for ερ or ος), b) the frequent use of oversized uncial γ and minuscule δ and τ, all with extended and slightly curly horizontal strokes, c) a long wave for the abbreviation of -ων. All three of these features appear in Planudes’ own script, albeit to a much lesser degree, as can be seen, for example, in Marc. gr. 481, the partial autograph of the Planudean Anthology.16 The best individual parallel for V is the hand of scribe G of Ambr. C 126 inf.17 This codex of ca. 1295–1296 contains Plutarch’s Moralia 1–69 “in the order and text recension of the so-called Corpus Planudeum.”18 It was written by ten scribes (A–K), including Planudes himself (B), and provides “a significant illu- stration of the variety of scholarly scripts from the Planudean milieu.”19 Even more spectacular examples of the large τ are exhibited by its scribes A and C.20 It is also worth pointing to scribe B in Marc.gr. 481 of the Planudean Anthology (fols. 16r.5– 19r, 20r–22r.11, 22r.17–23v).21 15 Digital images of V are available at https://gallica.bnf.fr/ or via https://pinakes.irht.cnrs.fr/notices/cote/52035/. 16 See A. Turyn, Dated Greek Manuscripts of the Thirteenth and Fourteenth Centuries in the Libraries of Italy (Urbana 1972) II pl. 71 (fol. 55v). Planudes wrote most of this manuscript, but two other scribes were involved. For the (possible) distribution see E. Mioni, Bibliothecae Divi Marci Venetiarum. Codices graeci manuscripti II (Rome 1985) 276. Digital images of the entire codex are accessible via http://www.internetculturale.it/it/41/collezioni/29635/ or https://pinakes.irht.cnrs.fr/notices/cote/69952/. 17 Turyn, Dated Greek Manucscripts … Italy II pl. 65; cf. https://pinakes. irht.cnrs.fr/notices/cote/42458/. There are no digital images publicly available yet. 18 Turyn, Dated Greek Manuscripts … Italy I 81. 19 Turyn, Dated Greek Manuscripts … Italy I 84. 20 See Turyn, Dated Greek Manuscripts … Italy II pl. 59, 61. 21 See Turyn, Dated Greek Manuscripts … Italy II pl. 73 (fol. 22v), and n.16 ALMUT FRIES 713 ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 Planudes’ students and ammanuenses clearly imitated and exaggerated the flourishes of their master’s hand, and the scribe of V seems to have been one of them. In that case it nevertheless remains possible that V was produced indepen- dently of Planudes by a scholar-scribe who was trained by him before moving elsewhere.22 But we would still have to account for the astronomical bias of the codex as, presumably, the legacy of his former teacher. On the whole therefore it seems preferable to assume that V stems from the Planudean circle. These considerations do not prove that the second part of the Pindar text in V (and its relatives CNOØ) represents a ‘proper’ edition by Planudes. Günther’s arguments, especially the one about the absence of a commentary, remain serious objections. But they corroborate our prima-facie assumption that Pindar was studied in the Planudean milieu. Perhaps Planudes was prevented by illness and/or his premature death around 1305 from completing a full edition of the Epinicians, and what we see in the ‘Paris recension’ is the remains of his preliminary work.23 One may compare the case of Eustathius, whose Proem to Pindar survives, while the commentary it was intended to accompany appears, for whatever reason, never to have been realised. Further support for this theory may come from the very fact that V has a composite text. Irigoin suggested that the reason why the Olympians and Pythian 1 were copied from Vi was that this manuscript offered metrical scholia from Olympian 2 to Pythian 1, “whereas the ‘Planudean edition’ [my inverted com- ___ above for the link to digital images. N. G. Wilson (“Maximus Planudes, the Codex Laurentianus 60.8 and Other Aristidean Manuscripts,” REG 122 [2009] 253–261, at 255) recognised the same hand in the main text of Laur. 60.8, which contains Aelius Aristides. For images of this codex see http://opac.bmlonline.it/Home.htm (via ‘Manoscritti’) or https://pinakes. irht.cnrs.fr/notice/fonds/1418/. 22 Wilson (GRBS 19 [1978] 393) envisages this possibility for the principal scribe of Laur. 32.2 and cites the Planudes student John Zarides as a prob- able example of such movement. 23 I owe this suggestion to Nigel Wilson. 714 A PLANUDEAN EDITION OF PINDAR? ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 mas] omits them.”24 It is unclear whether Irigoin considered this ‘omission’ intentional or unintentional, but the former seems unlikely, given that Planudes made (basic) metrical cor- rections in other poetic texts and that both Moschopoulos and Triclinius did the same extensively for Pindar. But if the com- mon ancestor of the manuscript family had failed to transmit the metrical scholia, their partial acquisition from another source would have been of interest to Planudes and his stu- dents. In Ambr. E 103 sup. (N) the metrical scholia to Olympians 3–12 were likewise added from a different source.25 The question whether Planudes worked on Pindar and, more specifically, curated the ‘Paris recension’ must inevitably re- main open. But the fact that two members of this family, Vat. gr. 915 (Ø) and Par.gr. 2403 (V), can independently be con- nected with him tips the scales somewhat in favour of Irigoin’s hypothesis. APPENDIX The mutual influence of V and Vi upon each other Irigoin (266) observed that the two recensions of the Pindar text that have been joined in V exercised “a slight influence” upon each other. This ‘contamination’ was largely carried out by the main scribe in the process of copying; only occasionally have readings from the respective other source been added as corrections by the first or a second hand. Irigoin quotes three examples from the part of V that was copied from Vi (Ol. 1–Pyth. 1) and three from that which belongs to the ‘Paris recension’ (Pyth. 2–Nem. 6.44). Nearly all the changes, one of which is a genuine improvement (Ol. 7.62 αὐξ[αν]οµέναν), the others Verschlimmbesserungen (Ol. 1.75 ἔειπε : εἶπε, 10.57 ἄρ᾽ : ἄρα, Pyth. 2.14 εὐα<υ>χέα, 44 ὀνυµάξε : ὀνυµάζε, 3.45 πόρε<ν>), reflect metri- cal considerations and suffice to prove that the intervention was intentional. 24 Irigoin, Histoire 265: “La répartition du texte … a probablement été établie en fonction des scholies métriques qui, dans le manuscrit de Germa- nos, vont de la seconde Olympique à la première Pythique, alors que l’édition planudéenne les omet.” 25 Irigoin, Histoire 257. ALMUT FRIES 715 ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 However, further investigation seemed in order. Since Vi is mutilated after Pyth. 5.55 (see 710 above), the evidence had to be collected from Olympian 1 to Pythian 4. I collated Pythians 2–4 (510 verses) for the part copied from the ‘Paris recension’, and Olympians 1, 6, 7, 13, and Pythian 1 (531 verses) for the part copied from Vi. This selection was intended not only to expand on Irigoin and to match approximately the combined number of verses in Pythians 2–4, but also to reflect the widest possible spread of attestation in the manuscripts of the ‘Paris recension’.26 These collations have revealed a broader range of philological activity in V. We see simple corrections (e.g. Ol. 6.82, 7.37, Pyth. 4.245), the introduction or removal of dialect forms (e.g. Pyth. 1.8, 92, 4.232), and even independent intervention, presumably again for metrical reasons (Ol. 6.5). Yet the changes remain intermittent, of variable quality, and occasionally internally inconsistent,27 while a large number of opportunities for correction have been missed (e.g. Ol. 1.112 κρατερώτατον VVi : καρτερώτατον CN cett. recte; Pyth. 4.41 λυσιπόνοισι VC : λυσιπόνοις Vi cett. recte). It also appears that popular odes like Olympian 1 and Pythian 1 attracted more extensive intervention and that the portion of V that was copied from Vi was more heavily interfered with than that which stems from the ‘Paris recension’. This may suggest an editorial bias towards the ‘Paris recension’ (cf. 711 above), although one has to keep in mind that the scope for comparison in the second part of V is limited by the mutilation of Vi and the fact that much of the available evidence consists of Pythian 4, which because of its enormous length was pre- sumably less commonly read. The tables below first list the (final intended) reading of V, re- gardless of whether it is correct, followed by that of Vi. Wherever possible, the quotation of other manuscript sigla, adapted from the 26 Olympian 1 is attested in CN (O here belongs to the ‘Laurentian fam- ily’), Olympians 6 and 7 in CNOØ (N up to Ol. 7.64, O from Ol. 6.76 to 7.64, Ø up to Ol. 7.29), Olympian 13 in CNO, Pythian 1 only in C (cf. n.1 above). For details of the varying family associations of NOØ see A. Turyn, Pindari carmina cum fragmentis (Oxford 1952) X. 27 See below for the introduction of the Aeolic ending -αις to the mascu- line aorist active participle in Pyth. 1.92 as opposed to its removal in Ol. 1.60 and Pyth. 4.232 (and apparently also 230). 716 A PLANUDEAN EDITION OF PINDAR? ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 apparatuses of Turyn and of Gentili et al.,28 has been kept to the minimum. 1. The influence of the ‘Paris recension’ on the text copied from Vi Olympian 1 23 βασιλῆα V cett. recte : βασιλέα Vi (+ HỊ) 23–24 κλέος / εὐανόριος Λυδοῦ VCpcN (+ AHL) : κλέος / ἐν εὐάνορι Λυδοῦ Viac u.v. (+ ÅBE) recte 26 ποσειδῶν VCN (+ BỊ) : ποσειδᾶν Vi cett. fere recte (-άν) 41 γαννυµήδης VCN (+ codd. pler.) : γάνυ- Vi (+ AB) recte 60 κλέψας VCN (+ AÅHLac) : κλέψαις Vi u.v. (+ AγρEỊLpc) recte 75 ἔειπε VCN : εἶπε Vi cett. recte 82 κέ VN (+ ABLpc) recte : καί Vi (+ C1E : δὲ C2H)29 Olympian 6 5 τε βωµῷ Vpc : βωµῷ τε Vac cett. recte 29 ποσειδάονι VCNØ (+ L) : ποσειδάωνι Vi cett. recte 82 λιγυρᾶς V et fere cett. recte : λυγερᾶς Vi Olympian 7 37 ὑπερµάκει V cett. recte : ὑπερβάκει Vi 44 προµάθεος V codd. pler. recte : προµαθέως Vi (+ ABO) 74 λίνδον τ᾽ V et fere cett. recte : λύδόν τ᾽ Vi φρένες V cett. recte : φρέναις Vi 94 ἐν δὲ V cett. recte : δὲ om. Vi (+ O) Olympian 13 52 προσπόλοις VCNOpc (+ Fs) : προπόλοις Vi cett. recte 95 ἑξηκοντάδι δ᾽ VCac : ἑξηκοντάκι δ᾽ Vi cett. : ἑξηκοντάκις Mommsen recte Pythian 1 8 γλεφάρων VC (+ codd. pler.) recte : βλεφάρων Vi (+ Cs) 26 ἰδέσθαι V (+ FglGγρHγρ) : πυθέσθαι Vi codd. pler. : προσιδέσθαι CEγρ recte30 28 Turyn, Pindari carmina; B. Gentili, C. Catenacci, P. Giannini, and E. Lomiento, Pindaro. Le Olimpiche (Milan 2013); B. Gentili, P. Angeli Bernar- dini, E. Cingano, and P. Giannini, Pindaro. Le Pitiche (Milan 1995). 29 While this variant is metrically relevant, the distribution of readings in the ‘Paris recension’ (CN), and especially the correction δέ in C2, suggest that κέ in V may be a lucky itacistic slip. 30 The fact that none of the other codices which attest the reading of V belong to the ‘Paris recension’ may indicate that the scribe had further ALMUT FRIES 717 ————— Greek, Roman, and Byzantine Studies 60 (2020) 708–717 85 οἰκτιρµοῦ VC (+ codd. pler.) recte : οἰκτιρµῶν Vi (+ Å) 92 πετάσαις VpcC : πετάσας Vac cett.31 2. The influence of Vi on the text of the ‘Paris recension’ Pythian 2 65 ἱπποσόοισιν VsVi : ἱπποσόαισιν V cett. recte 88 τηρέωντι VVi (+ G) recte : τηρέοντι cett. Pythian 3 6 τέκνον ἀνωδυνίας V2Vi (+ BHγρ) : τέκτον᾽ ἀνωδυνίας V1C (+ ÅEG2) : τέκτονα νωδυνίας G1H Tricl. recte 52 περιάπτων VVi cett. : περάπτων EHC recte 114 γιγνώσκοµεν VVi (+ GH) : γινώσκοµεν C cett. recte Pythian 4 232 ῥίψας VpcVi codd. pler. : ῥίψαις VacC (+ BG) recte32 245 πεντηκόντερον VVi (+ EG) recte : πεντηκόντορον C (+ BH) et fere cett.33 July, 2020 The Queen’s College Oxford OX1 4AW United Kingdom almut.fries@classics.ox.ac.uk ___ sources at his disposal (711 above). However, since in Pythian 1 C is the only extant member of the Paris family, we cannot be certain of that proposition. 31 This participle is metrically and semantically superfluous and was rightly excised by Callierges. Given the general absence of dialect glosses in Pindar, the majority reading πετάσας is probably original and Aeolic πετάσαις due to what Irigoin called contre-normalisation (i.e. the introduction of dialect forms during transmission as opposed to their elimination). C is particularly faithful in preserving -αις. 32 The same ‘correction’ is found in Pyth. 4.230 (τελέσ(σ)ας Vpc codd. pler. : -αις Vac (+ CB) recte), but it is impossible to verify the reading in Vi from the digital images (cf. n.12) as the word is hidden in the gutter margin. 33 This article is a ‘spin-off’ from my forthcoming edition, with intro- duction and commentary, of Pythian 1. I am grateful to Nigel Wilson for his initial suggestion to find parallels for the script of V, for reading the first version of this piece, and for supplying me with several references from his own library at a time when the public ones were closed. I also thank the anonymous referee for GRBS for pointing out ways to improve the evi- dential value of the Appendix.