Robortello's Edition of Scholia to Aeschylus LANGWITZ SMITH, OLE Greek, Roman and Byzantine Studies; Spring 1971; 12, 1; ProQuest pg. 101 Arsenios and Parisinus Graecus 2070 Ole Langwitz Smith I Robortello's Edition of Scholia to Aeschylus I N AN earlier article1 I tried on the basis of the scholia to the Sup­ plices to argue for the theory that Robortello for his edition of the scholia to Aeschylus 2 took his text of the scholia to the plays out­ side the triad from a gemellus to the Paris MS, Ancien fonds grec 20703 written by Arsenios of Monembasia (Aristoboulos Apostolides). Since I wrote that paper I have had opportunity to study the text of Par. 2070 in the Oresteia-part also.4 This study both confirms and modifies my earlier view. On the one hand I have found more differ­ ences than the single one I found in the scholia to the Supplices, on the 1 C1Med 28 (1970) 77f. 2 Scholia in Aeschyli Tragoedias omnes ex vetustissimis libris manuscriptis col/ecta atque in hoc corpus redacta a Francisco Robortello Utinensi. Venetiis ex officina Erasmiana Vincentii Valgrisii MDLII. 3 This MS contains on ff.1-160 Planudes' translation of Cicero, Somnium Scipionis, with Macrobius' commentary. On ff.161-244v we have Arsenios of Monembasia's recension of the scholia Medicea to Aeschylus. I have not been able to identify the scribe offf.1-160, but this man also wrote ff.1-47 in Par. 2151. Par. 2070 consists of three different parts, each of which has on its first folio the owner's note of). Fr. Asulanus; see ff.lr, 161r, 189r. The water­ marks indicate a date ca. 1520. As Arsenios was in Florence in 1518, the MS would seem to have been written at that time. The end of the scholia to Eumenides has been lost. Three leaves were lost from the sixth quire in the third part of the MS. This quire now consists of only 5 leaves. The loss probably occurred after the volume had been bound at Fontainebleau, as ff.229, 230 and 231 have been repaired. On f.23F the scholia to Eumenides end after 866 ofL6cpvAov. Headlines are in red ink, while the scholia have been written in brown. The underlinings until f.162V are also in red; later they are in the same ink as the text. On f.189r, the first folio of the third part of the MS, there is the note "comento di Aeschilo." This is probably by Asulanus. Another hand wrote the verse Ter. Andr. 309 on f.193 r • This same hand is found in another Arsenios MS in Paris, Ancien fonds grec 3058 (the ']wvta); so possibly this was Arsenios himself. On the preliminary folio we have the index of the volume in Palaeo­ kappa's hand. 4 Through a generous grant from the University of Aarhus I have been able to study the MS in the original at the Bibliotheque Nationale. I am deeply grateful to Professor H. Friis Johansen and the governing body of the University. 101 102 ARSENIOS AND PARISINUS GRAECUS 2070 other I think that I have discovered decisive proof for the connection between Arsenios' recension and the edition of Robortello. As in the scholia to the Supplices, we also in the Oresteia-part find that not only the scholia proper but also the lemmata are in the main identical. Some of the more puzzling errors in Robortello can be ex­ plained only on the theory that they originated from a faulty under­ standing of the readings of a MS similar to Par. 2070. I mention the following instances: (jco Cho.649 , Par. 2070 SWJLd-rwv Rob. (uJLa'TWV Cho.681 '.~ 'l 'TO E ayvwc emma Par. 2070 'TO v~ayvwc Rob. Bum. 139 €7TOV P lemma Par. 2070 ., Rob. E7TOV Bum. 140 , Par. 2070 , Rob. 7Tpa 7Tpa77J Bum. 181 ,,t "l Par. 2070 7jAov Rob., l6vM Bum. 186 a7ToKErpaA' l Par. 2070 a7ToKErpaAlac Rob. I may also draw attention to the following common errors: Ag. 106-07 €C7TEpaC Cho. 23b cVYK6ifoJLa,s Cho. 413 'TpWCCE'Ta, Cho.528 Cho.813 Bum. 176 Kpavofi'Tat ., , Ol\tytKW'Ta'TOC . , 'wv Obviously Robortello's source was a MS with scholia only. These scholia had lemmata probably underlined. This can be demonstrated by referring to the numerous cases where Robortello confounded lemmata and scholia. I shall confine myself to two telling cases: Ag. 1153f runs as follows in Robortello: Avcrpa'Twt KAayyEL] acarpEt alvtYJLa'TWSEt ~ Sux 'TO SVC'TVX~ JLav- 'li ,,~" ~, '~\',I,.'~ 1I , ~ 'TEVECuaL "l ULa 'TO 7TPOELUEVat 'Ta uEtVa aVW't'EI\WC uEC7TLac 'T"lC JLaV'TLKfjc. It is evident that 6Ec7Ttac is the lemma for 'Tfjc JLaJl'TtKfjC; in the poetic text Robortello read 6ec7TEclac. Obviously he just took over what he found in his source. In this MS the underlining may have been miss­ ing, or Robortello may have overlooked it. In Par. 2070 we find the same text as in Robortello, but with 8Ec7Ttac underlined. Ii cVYKo.pwp.at is found as late as Dindorf and Kirchhoff. The reading of Laur. 32,9 is quite unmistakably KO,pWP.ru. OLE LANGWITZ SMITH 103 The other instance is Cho. 290 and 296, where Robortello exhibits: AVjLav{Uv] AVjLavfNVTa KaKW. Tap'XEcpOlvTa] Kat 0 Cwcppwv KTA. In Par. 2070 we find: Av jL av 0 E v' AVjLavOlvTa. K aK W Tap' X E cp (U VT a' Ka, 0 Cwcppwv KTA. The puzzling KaKW now appears to be the KaKWC of the poetic text line 296.6 But this is not all. At Ag. 36 the scholium on [jovc l-7Tt yAwcc7Jt in Laur. 32,9 runs: "Q' , ~ "J..QA Y" , 7J fJapoc E7TLKEtTat 7J 'f'OfJOVjLaL ~ 7JjLLav E7TLKELCOjLElI7JV jLOt. This is neither in Robortello nor in Par. 2070. Instead we find on Ag. 36 the following note: , "A ,~, 'Y 0 7TapOLjLLa E7TL TWV jL7J ovvajLEVWV 7TapP7JCLa~EC aL. Now it is hardly a coincidence that this is found in Par. 2070. As is well known Arsenios had a special interest in 7TapOtjLLaL, and this particular one is also found together with the explanation €7Tt ••• 7TapP7JcLa'EcOaL in his collection ']wv{a.7 The appearance of the same scholium in Robortello would seem to be proof of the connection.S However there is decisive evidence. Schol. Eum. 540 has been written twice in Par. 2070, in both cases at the wrong place. It has been written in the margin of f.228v , ob­ viously in order to follow schol. Eum. 527; and it is also found on f.229r after schol. Eum. 569. This state of affairs may be explained in this way. When Arsenios copied the scholia from Laur. 32,9,9 he copied 569 6 In Schiltz's edition of 1821 this KaKW' still appears. 7 Arsenii Violetum, ed. C. Walz (Stuttgartiae 1832) 143. 8 This 'scholium' was taken over by Victorius, and from there it continued to disfigure the editions. It is found as late as in van Heusde. On the composite nature of Victorius' scholia, see Fraenkel's edition of Agamemnon, I (Oxford 1950) 34. 9 Though I agree with A. Turyn, The Manuscript Tradition of the Tragedies of Aeschylus (New York 1943) 23, that Par. 2070 is a direct copy of Laur. 32,9, there are occasional addi­ tions from other sources; at PV 18 Arsenios added the scholium Dindorfp.170, 18-24. Turyn does not advance any proof, but a passage like schol. Cho. 595 shows beyond doubt that Arsenius had Laur. 32,9 before him when he wrote Par. 2070: -ric Myot" Tlc -"yo, Ka~ yvv (Kai yvv was deleted immediately) cpp~civ KT'>'. The deleted words came from the poetic text 595 Tic -"yo, Kai yvva'Kwv. 104 ARSENIOS AND P ARISINUS GRAECUS 2070 after 582 because 569 is written on top of f.l64v in Laur. 32,9 without a reference sign to the poetic text to show where it does belong. Ac­ cordingly Arsenios did not know where to place it. Having copied 569 he became aware of 540, which is the first regular scholium on this page; or perhaps he just copied it by oversight, failing to observe that it would be wrongly placed in that way. Whether he detected his error immediately or later we cannot know, but in Par. 2070 both lemma and scholium are underlined, the line under the scholia no doubt being a deletion-mark.1o Having deleted this wrongly placed scholium he added it in the margin where he thought it did belong, though still not correctly. In Robortello we find schol. 540 written twice, after 527 and after 569. This disposes of the gemellus-theory: there is no reason to suppose that Arsenios would have committed the same error twice, or rather the same series of errors. There is only one way of explaining the evidence: Robortello's main source was an apograph of Par. 2070.11 It must be remembered that Robortello could not have used Par. 2070 itself; when Robortello prepared his edition, Par. 2070 was in the library at Fontainebleau.12 Besides, there are a number of features in Robortello which cannot be found in Par. 2070. So Robortello must have had access to another MS from which he drew supplementary matter.rs 10 There is a perfect parallel at schol. PV 494 on f.196r , where Arsenios wrote the scho­ Hum thus: 8alp.rm TTPOC ~80~v' £VTTp6c8£I(7'a TTPOC ~80~v 8£oLe. The deletion of the intruded lemma is indicated by the underlining. 11 When Par. 2070 was copied in the MS used by Robertello this deletion was not under­ stood, and accordingly schol. Bum. 540 was copied twice. I should like to add that the apograph used by Robertello cannot have been any of the existing apographs of Par. 2070. Of these only Mon. 91 and Vat.gr. 1464 come in for consid­ eration. These two are the only ones that have the whole text. In the case of both MSS nothing is gained by identifying any of them with Robortello's source. Where their read­ ings depart from Par. 2070, Robortello follows the parent MS. 12 In 1542 Asulanus' MSS were bought by G. Pelicier for Francis I. The catalogue of the MSS in the library of Fontainebleau prepared by Vergikios and Palaeokappa exhibits Par. 2070 as no.H. See H. Omont, Catalogues des manuscrits grecs de Fontainebleau sous FraHfois I et Henri II (Paris 1889) 6. 13 From this source Robortello took schol. Cho. 621 and Bum. 386b, apart from parts of lemmata and scholia not found in Par. 2070. Some of these additions could very well have been made from the poetic text. There are in Par. 2070 some mysterious lacunae which cannot all be ascribed to the illegi­ bility of the exemplar. I am unable to see why Arsenios hesitated at Ag. 135-38, where he wrote 7'WV ICV 8Loe, and at Cho. 51, where he did not copy ,.,.vcoe but left the blank space for five letters. On the other hand, at Pers. 41 it is difficult to read more than ot At! I(al in Laur. 32,9, and accordingly Arsenios wrote ot I(at The same thing happened at OLE LANGWITZ SMITH 105 How do these results fit with Robortello's own statement about his sources? If we are to believe him, he had at his disposal two MSS: "librum ad me pervetustum misisti in quo erant postremae huius authoris Tragoediae descriptae ... aliumque praeterea in quo tu ipse multa adnotaras." So he says in the preface to Mariano Savello. But this is about the poetic text; in the preface to the scholia he mentions only one: ccanti­ quissimum ad me misisti librum in quo tres hae manu vetusti cuiusdam ac bene docti hominis accurate descriptae Juerunt cuius ope in primis omnia vulnera tam praeclari poetae sanavimus atque omnia Jere collegimus quae adiumento omnibus erunt ad loca dijficilia intelligenda." Most scholars seem to identify Savello's "antiquissimum librum" with Laur. 32,9. But how can Savello have got hold of this MS? In his preface to L'Eschilo Laurenziano, Rostagno14 seems to imply that Laur. 32,9 was in Savello's possession at Padua in 155l. This simply cannot be true. As far as we know, the famous MS was at Florence at this time, and it is quite impossible to imagine that Savello, about whom we know next to nothing,15 owned the MS. The Savelli were not com­ monly known as bibliophiles or manuscript collectors, though Gia­ como Savello,16 Mariano's brother, was a very learned cardinal. Mariano himself, if we can trust Robortello, had a liking for philos­ ophy, as did so many educated men at that time. Mariano was a young man in 1551 and was probably studying at Padua, from where he sent the MSS to Robortello, who may have been at Venice. The identification can also be proved wrong in another way. As M. Dopchie17 has recently shown, the poetic text in Robortello could not the notorious crux PV 364, where he wrote TOO Il.~Ta. The lacunae Robortello filled from his second source. 14 L'Eschilo Laurenziano, facsimile ed. E. Rostagno (Florence 1896) 10. 15 I have searched in vain for information about him. M. E. Cosenza, Dictionary of the Italian Humanists (Boston 1962) s.n., only knows about him from Robortello, but does not seem to have read the prefaces. There is some material on the family in G. Moroni, Diziona­ rio di erudizione storico-ecclesiastica 61 (Venice 1853) 294ff. The few treatments of the history of this family naturally concentrate on the earlier members. Mariano Savello was bishop at Nicastro from 1554 to 1556 and at Gubbio from 1560 until his death in 1599. The Mariano Savello. to whom Simon Portius in 1521 dedicated his De mente humana disputatio, probably was the uncle of our Mariano. 16 Cf. M. L. Shay in New Catholic Dictionary (New York 1929) s.n. 17 HL'Humanisme italien et l' Agamemnon d'Eschyle," Bulletin de l'Institut Historique Beige de Rome 37 (1966) 99ff. In this paper Dopchie points out that Robortello's distinction be­ tween Agamemnon and Choephoroe probably was the work of Victorius' assistants Bartolo­ meo Barbadoro and Girolamo MeL Dopchie does not refer to the scholia, where Robortello did not distinguish between the two plays. This means that at least in November 1551 106 ARSENIOS AND P ARISINUS GRAECUS 2070 have been taken from Laur. 32,9. Robortello expressly mentions which readings he and Sophianos corrected in Savello's MS, and these have nothing to do with Laur. 32,9. Besides, the corrections very often are identical with the readings found in that MS. Now, the apograph of Par. 2070 cannot be identified with the Savello MS, as that MS contained both text and scholia. The statement of Robortello's that he took the scholia from Savello's MS cannot be true. The MS cannot have been as old as Robortello implies; I should guess that it probably was a run-of-the-mill MS from the end of the xv century, and of course a direct or indirect copy of the Medicean MS.18 The apograph of Par. 2070 cannot well have been more than 35 years old, and one is astonished that Robortello did not see that.19 The (author' of Robortello's scholia will probably replace Robor­ tello in the apparatus of future editions of the scholia to Aeschylus. In my previous article I pointed out that a number of corrections nor­ mally ascribed to Robortello were in fact made by Arsenios. This is not Arsenios' main contribution to classical scholarship; still, credit should be given to whom it is due. Robortello did not know that his Agamemnon in fact was two different plays. As Robortello at that time used Savello's MS, it is obvious that he did not distinguish between the two plays with the help of the MS, as he claims in the preface. It is therefore pretty certain that Robortello used the results of Barbadoro and Mei, of which he must have become aware during the few months between the issue of the scholia and the issue of the text. Without access to the Laur. 32,9 he could hardly have found the solution himself, nor does he seem to have seen any problem. 18 Among the extant apographa of the Medicean MS only San Marco 222, Bon. 2271 and Guelf.Gud.gr. 88 are qualified to be considered identical with the Savello MS. San Marco 222 is the least probable candidate as its scholia are not very numerous. Bon. 2271 in 1551 was in the library of San Salvatore at Bologna (M.-H. Laurent, Fabio Vigili et les bibliotheques de Bologne [Studi e Testi 105, Vatican 1943] 270). We know nothing at present of the whereabouts of Gud.gr. 88 at that time. The non-triadic part, ff.131-193, was written ca. 1492-96 in Florence by Michael Souliardos, as I venture to conclude from the specimen of his writing in H. Omont, Facsimiles de manuscrits grees des XV· et XVI" sifc/es (Paris 1887) pU8. Souliardos worked in the Medicean library at that time; see E. Lobel, The Greek Mss of Aristotle's Poetics (Oxford 1933) 54f, and D. Young, Scriptorium 7 (1953) 17. In 1518 the MS was with Asu­ lanus when he was preparing the Aldine Aeschylus. G. F. Tomasini, Bibliotheeae Patavinae mss. publicae et privatae (Utini 1639) 114, mentions this MS as being with Nic. Trevisano at Padua. Its readings, however, do not show any close similarity to those mentioned by Robortello. 18 Compare the similar but not so glaring case mentioned by A. Dain. Histoire du texte d'Elien Ie Tacticien (Paris 1946) 318f. Franchini's Source for Aristophanes, "Pax" 948-1011 LANGWITZ SMITH, OLE Greek, Roman and Byzantine Studies; Spring 1971; 12, 1; ProQuest pg. 107 OLE LANGWITZ SMITH 107 II Franchini's Source for Aristophanes, Pax 948-1011 I T IS well-known that the passage Pax 948-1011 was not included in the editio princeps of Aristophanes by Musurus in 14981 and that these lines were restored by Antonio Franchini in his edition of 1525.2 In the preface to Benedetto Accolti, Franchini mentions the help he had from Arsenios:3 "ad hoc (i.e. the recension of Aristophanes) A rsenii Cretensis A rchiepiscopi Monembasiae magna eruditione viri acerrimo iudicio usi, dum is Florentiae adolescentulos illos, quos Leo, Pont. Max. ad reparandam graecae linguae iacturam ex media Graecia accersiverat, instituit." As Arsenios probably left Florence at Leo X's death in 1521 and went back to Crete, he must have worked with Franchini in 1521 at the latest.' We do not know the exact share Arsenios had in Franchini's work;5 perhaps this did not amount to much, as most of Franchini's text was taken from the Aldine.6 At one point, however, Franchini's text was superior to that of Musurus, as he says in the preface: "In qua qUidem recognitione hoc in primis gloriari licet, nos omnium primos in Comoedia cui titulus est Elp~J/7} ad sexaginta versiculos nostra diligentia repertos restituisse." This seems to show that Franchini took for himself the honor of having restored the text. There is evidence, however, for Arsenios having a share of it. In Franchini's edition on ff.347-48 the line €lra ft0VWO€LV EK ftTJ8€{ac (1012) has been printed twice. Why did this happen? Obviously be­ cause the missing lines were supplied from a copy of this part where lines 947 and 1012 had been added in order to show where the missing lines did belong. By an oversight the line EtTa ftOVW8€LV EK ftTJo€tac was 1 'AptCTO.pa.vovc Kwp.wtSlm €vv€a .•• Venetiis apud Aldum 1498. 2 'AptCTOCPa.vovc Kwp.wt3lat £vv€a. Mf"Ta cxoAlwv 7TOAAWV 'lTaAatwv Ka~ 'lTavv w.pf"}.{p.wv 'lTpOC TOLe 'lTpOTfPOV £VTf"TV'lTWP.€VWV, cvv 'lTlVaKt a.p{)ovoTaTwt TWV ,ba.VTWV .ole ex SEL TOV voill' 'lTPOdXEtV. Florentinae per heredes Philippi Juntae ... 1525. 3 Extracts from this preface have been printed in E. Legrand, Bibliographie helUnique ... XV' et XVI" S., II (Paris 1885) 156. Quoted in full in A. M. Bandini, De Florentina Iuntarum Typographia II (Lucae 1791) 194f. 4 On Arsenios and his work in this period see D. Geanakoplos, Greek Scholars in Venice (Cambridge [Mass.] 1962) 187 with notes and literature quoted. 6 A. Renouard, Annales de l'imprimerie des Alde3 (Paris 1834) 17, asserts that the scholia were edited by Arsenios. There is nothing about this in the preface, though the guess may be correct, since Arsenios wrote the scholia in Par. 2070. 6 See e.g. W. G. Rutherford, Scholia Aristophanica I (London 1896) xv. 108 ARSENIOS AND P ARISINUS GRAECUS 2070 printed twice because it was found both in the copy of the missing part and in the main text used for printing. As far as I know no such excerpt has yet been found. No manuscript exhibiting the lines Pax 947-1012 seems to have been recorded.' While I was doing preliminary researches for a new edition of the scholia to Aeschylus I investigated the Paris MS Ancien fonds grec 2070, which was known to contain the scholia Medicea copied directly from Laur. 32,9 by Arsenios.8 On ff.242ff this MS exhibits in a hand different from that of Arsenios9 the lines Pax 947-1012. This part of the MS seems to have escaped the notice of former investigators, and it is not mentioned in Omont.10 There are scholia to the text in the hand of Arsenios. As I am presently going to show, the excerpt is a direct copy of the Codex Ravennas 137,4A. The history of the Ravenna MS in the first quarter of the XVI cen­ tury is not sufficiently clear. In 1515 or early in 1516 the MS was with B. Giunta when he edited Thesmophoriazusae and Lysistrata. It may have been in the possession of Euphrosynus Boninusll afterwards, and we know that it was bought in Pisa in 171212 for the Bibliotheca Classensis in Ravenna. We may guess that Arsenios met Boninus in Florence in 1518, as Boninus had edited some Greek works for the Giunta press in the period 1514-17. He may still have been there when Arsenios arrived in 1518. What happened was probably that Arsenios and another scribe copied from the Ravennas in Boninus' possession 7 I have checked J. W. White's lists (CP 1 [1906] 9ft") and subsequently published informa­ tion as far as I have been aware of it. S Cf A. Turyn, The Manuscript Tradition of the Tragedies of Aeschylus (New York 1943) 23, and my edition of the scholia to Aeschylus' Supplices in H. Friis Johansen, Aeschylus, The Suppliants I (Copenhagen 1970) 150. For a description of the MS see the preceding article, section I above, n.3. 9 The hand of Arsenios is well known. Specimens of his writing in H. Omont, Facsimiles de manuscrits grecs des XV' et XVI siecles (Paris 1887) p1.5 and 6. See also M. Wittek, Album de paleographie grecque (Gand 1967) p1.43 and 46. The writing in Par. 2070 is so different from Arsenios' that it cannot possibly be his hand in the poetic text. 10 H. Omont, Inventaire somma ire des manuscrits grecs de la Bibliotheque Nationale II (Paris 1888) 187. Neither is it mentioned in Vergikios' and Palaeokappa's catalogue of 1550 pub­ lished by Omont, Catalogue des mss. grecs de Fontainebleau (Paris 1889). Our MS is no.15 in the alphabetical list. 11 Cf J. W. White, CP 1 (1906) 257; J. van Leeuwen, Aristophanis comoediae undecim cum scholiis. Codex Ravennas 137.4A phototypice editus (Leiden 1904) vi. On Boninus, cf Bandini, op.cit. (supra n.3) I.67ft'. 11 Van Leeuwen, op.cit. (supra n.11) v. OLE LANGWITZ SMITH 109 the missing part of text and scholia. As White has argued,13 Franchini did not have access to the Ravennas in 1525, and Arsenios probably made a copy of his excerpt for Franchini's use when he left Florence. This may explain the slight inconsistencies between the edition and our excerpt. That Par. 2070 is a direct copy from the Ravenna MS may be seen from the blank spaces in schol. 968, where the Ravennas is illegible. Par. 2070 exhibits the following text: 7'OV7'O ovv E1T0tOVV ot C1TEVOOV7'EC ••• CVVEL067'EC 7't JaV7'OLC •.• EKXWPOtEV 7'