Chu ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 ISSN 2159-3159 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/ Rhetorical Training in Byzantine Italy: The Evidence of Paris.gr. 3032 Minqi Chu CHOLARS HAVE LONG BELIEVED that, as an advanced subject in Byzantine education, rhetoric was taught from the ninth to the eleventh centuries almost exclusively in the capital or in major cities.1 In Constantinople, aside from private schools, the institution of higher education established by Constantine VII Porphyrogenitus dedicated one of its four chairs to rhetoric, and Psellos was proclaimed Professor of Rhetoric and Philosophy by Constantine Monomachus.2 In the 1 P. Lemerle, Le premier humanisme byzantin (Paris 1971) 242–246, and “ ‘Le Gouvernement des Philosophes’: Notes et remarques sur l’enseignement, les écoles, la culture,” in Cinq études sur le XIe siècle byzantin (Paris 1977) 196; A. Markopoulos, “De la structure de l’école byzantine, le maître, les livres et le processus éducatif,” in B. Mondrain (ed.), Lire et écrire à Byzance (Paris 2006) 87, and “In Search for ‘Higher Education’ in Byzantium,” Zbornik radova Vizantoloskog instituta 50 (2013) 37. 2 Lemerle, Le premier humanisme byzantin 242–260, and in Cinq études 212– 215; A. Markopoulos, “L’éducation à Byzance aux IXe–Xe siècles,” TM 21.2 (2017) 66; R. Riedinger, “Quatre étapes de la vie de Michel Psellos,” REByz 68 (2010) 41–42; F. Pontani, “Scholarship in the Byzantine Empire (529– 1453),” in F. Montanari et al. (eds.), Brill’s Companion to Ancient Greek Scholarship (Leiden 2015) 341–356. On Psellos’ rhetorical theory see E. Papaioannou, “Michael Psellos’ Rhetorical Gender,” BMGS 24 (2000) 133– 146, and Michael Psellos: Rhetoric and Authorship in Byzantium (New York 2013) 29–50; J. Walker, “Michael Psellos on Rhetoric: A Translation and Com- mentary on Psellos’ Synopsis of Hermogenes,” Rhetoric Society Quarterly 31 (2001) 5–13, and “These things I have not betrayed: Michael Psellos’ Encomium of his Mother as a Defence of Rhetoric,” Rhetorica 22 (2014) 49– S MINQI CHU 183 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 peripheral regions of the Empire, on the other hand, grammar might have occupied the highest level of the curriculum and was sufficient to access careers in local administration.3 Against this background it is surprising to find in Byzantine Italy indications of training at a higher level than grammar. The Greek rhetorical textbook preserved in MS. Paris.gr. 3032, a codex produced in Calabria in the tenth century, attests to the production and circulation of books on rhetoric in the Greek- speaking area of Byzantine Italy.4 In this regard Parisinus stands alone if one discounts the witness of some lexica.5 ___ 101; A. Kaldellis, “The Discontinuous History of Imperial Panegyric in Byzantium and its Reinvention by Michael Psellos,” GRBS 59 (2019) 693– 713; M. Champion, “Rhetoric, Philosophy, and Transformation in the Thought of Michael Psellos,” in E. Anagnostou et al. (eds.), Later Platonists and their Heirs among Christians, Jews, and Muslims (Leiden 2023) 211–231. 3 F. Ronconi, “Quelle grammaire à Byzance? La circulation des textes grammaticaux et son reflet dans les manuscrits,” in G. De Gregorio et al. (eds.), La produzione scritta tecnica e scientifica nel Medioevo: Libro e documento tra scuole e professioni (Spoleto 2012) 92. 4 D. Arnesano, “Ermogene e la cerchia erudita. Manoscritti di contenuto retorico in Terra d’Otranto,” in N. Bianchi (ed.), La tradizione dei testi greci in Italia meridionale. Filagato da Cerami philosophos e didaskalos (Bari 2011) 101. 5 The 10th-century Calabrian codex Matrit.bibl.univ. 116-Z-22 contains lexicons of Demosthenes’ Orations (fols. 281r–292v, 295r–302v), while Crypt.Z. α. XXX preserves Φιλιππικῶν ῥητορικῶν λέξεις, Ῥητορικαὶ λέξεις ἑτέραι, and Λέξεις τῶν συμβουλευτικῶν Δημοσθένους. F. G. H. Muñoz, “El Léxico demosté- nico del Matr. Comp.30,” Eikasmos 28 (2017) 248–249; A. L. Fonseca et al. (eds.), Catálogo de manuscritos medievales de la Biblioteca Histórica “Marqués de Valdecilla” (Madrid 2019) 166. A. Rocchi, Codices Cryptenses seu Abbatiae Cryptae Ferratae in Tusculano (Rome 1883) 458–459; A. B. Drachmann, Die Über- lieferung des Cyrillglossars (Copenhagen 1936) 54. In addition to the rhetorical lexicons, another 10th-century Italo-Greek MS., Vat.gr. 1834, transcribed at its end (on fol. 62r) a 27-line passage of Lucian’s Calumniae non temere credendum, but it is difficult to establish its relationship to the local training of rhetoric: see J. Irigoin, “La tradition de Lucien dans l’Italie méridionale (Xe siècle),” in A. Billault (ed.), Opôra. La belle saison de l’hellénisme. Études de littérature antique offertes au Recteur Jacques Bompaire (Paris 2001) 145–147. 184 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 In this article I focus on what we can learn from the Parisinus about rhetorical training from the ninth to the eleventh cen- turies, before the Norman conquest, in this western border of the Byzantine Empire. After contextualizing this discipline within the educational system of Greek-speaking Byzantine Italy, I characterize local rhetorical training through the philological, palaeographical, and codicological analysis of the Parisinus. Rhetoric in the educational system of Byzantine Southern Italy The state of education in Byzantine Italy, especially in the Greek-speaking regions of Calabria and Sicily, remains un- clear.6 Local hagiographies from the ninth–eleventh centuries usually indicate in general terms that their saints received only a sacred education during childhood. Elias Speleotes, Phantinos, and Neilos of Rossano studied τὰ ἱερὰ γράµµατα.7 Vitale learned sacras litteras et divinas from his parents, as did Filaretos of Ca- labria.8 Sabbas the Younger was taught τὰ θεῖα µαθήµατα by some διδάσκαλοι.9 Only the Life of Neilos explicitly mentions that one of the saint’s followers named Proklos received an ἐγκύκλιος παίδευ- σις.10 The hagiographical text, however, does not specify this secondary education, which must have included a significant amount of secular knowledge. The Life of Filaretos, composed in the eleventh century, helps us to understand this silence. It opens with the praise of ancient Sicily as the cradle of poets, 6 P. Degni, “Literary and Book Production in Byzantine Italy,” in S. Cosentino (ed.), A Companion to Byzantine Italy (Leiden 2021) 752. 7 Vita S. Eliae Spelarote, AA SS. Septembris. III 849B; E. Follieri, La Vita di San Fantino il Giovane (Brussels 1993) 402–404; R. L. Capra et al., The Life of Saint Neilos of Rossano (Cambridge [Mass.] 2018) 6, 8. 8 Vita S. Vitalis Siculo, AA SS. Martii IX 26D–E. Vita S. Philareti Monachi, AA SS. Aprilis VII 607. 9 G. Cozza-Luzi, Historia et Laudes SS. Sabae et Macarii (Vatican 1893) 7. 10 Life of Saint Neilos of Rossano 128. MINQI CHU 185 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 rhetors, and philosophers, but the narrator scorns this secular wisdom and its glory.11 This negative attitude is typical of hagi- ographies. For instance, the Life of Leone Luca explicitly declares that its hero never received otiosa mundanae philosophiae studia.12 The rejection of secular knowledge in local hagiographies does not of itself imply the absence of training in these subjects. The eulogy at the beginning of the Life of Filaretos seems to acknowledge the existence in Sicily of poetry, rhetoric, and philosophy until the author’s time. Sabbas, Neilos, and some of his followers, like Bartholomew of Grottaferrata, were bilingual. Sabbas translated the Greek dialogue Μέλας Βυθός into Latin.13 The Nilian monks are believed to have translated patristic works from Greek to Latin.14 This suggests their mastery of grammar, if not rhetoric. Furthermore, the use of rhetorical techniques in the Life of Neilos, written by one of his disciples, indicates that the author had a certain level of knowledge and understanding of rhetoric.15 In addition to these local hagiographies, there is other evi- dence to verify the spread of the ἐγκύκλιος παίδευσις in Hellenic-speaking Byzantine Italy during the ninth–eleventh centuries. According to the Life of Methodius the Patriarch (whose composition is considered rather late), he was born in Syracuse 11 AA SS. Aprilis VII 607: sed quid ego oratorum ac philosophorum, quid vatum historicorumque, lumen splendoremque orationis commemoro? cum haec verborum lenocinia nihil apud nos sint, qui caelestia ac divina profitemur, neque eorum placita ullo pacto sequi velimus, cum a rationibus nostris longe sint aliena, neque apta ad commendandum, quod apud nos est, bonum. 12 Vita S. Leonis Lucae, AA SS. Martii I 98F. 13 Cozza-Luzi, Historia et Laudes SS. Sabae et Macarii 92. 14 F. Ronconi, “Graecae linguae non est nobis habitus. Notes sur la tradition des Pères grecs en Occident (IVe–IXe s.),” in M. Cutino et al. (eds.), Transmission et réception des Pères grecs dans l’Occident, de l’Antiquité tardive à la Renaissance (Paris 2016) 351. 15 A. Garzya, “Note sulla lingua della Vita di san Nilo da Rossano,” in Deputazione di storia patria per la Calabria (Naples 1969) 80. 186 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 and received an education, probably in his hometown, that included the following subjects: πᾶσαν γραµµατικῆς τέχνην καὶ ἱστορίας, ὀρθογραφίαν τε καὶ ὀξυγραφίαν.16 Significantly, this pro- gram seems not to have included rhetoric. The experience of Constantine the Sicilian also draws our attention. He travelled to Constantinople from Sicily to further his studies under the renowned scholar Leo the Mathema- tician.17 This implies that Constantine had received in his homeland the education necessary to continue his studies in Constantinople. If so, his training in Sicily may have included grammar and even basic rhetoric. Alongside historical records,18 local manuscripts provide a more precise picture of secondary education in the Greek language in Byzantine Italy. In the tenth and eleventh centuries several grammar textbooks were produced in this region: four manuscripts, Monac.gr. 310, Leid.Voss.gr. Q. 76, Crypt.Z. α. III, and Vat.gr. Pii II 47, copied and used intensively in Nilian mon- asteries;19 Paris.gr.suppl. 920 and Messan.S.Salv. 156, with frag- ments and epitomes of Herodian;20 some fragments discovered in the Biblioteca Diocesana di Acerenza containing the Pro- 16 Vita S. Methodii Patriarchae, PG 100 1245B; V. Prigent, “À l’ouest rien de nouveau? L’Italie du sud et le premier humanisme byzantin,” TM 21.2 (2017) 152–153. 17 Prigent, TM 21.2 (2017) 152. 18 Nothing is known about the education in their own homeland of others who were originally from Sicily or southern Italy. Such are John Sikeliotes and the grammarian Theognostus who wrote Περὶ ὀρθογραφίας and dedi- cated it to the Emperor Leo (V or VI). See Pontani, in Brill’s Companion 321; Prigent, TM 21.2 (2017) 150–151. 19 F. Ronconi, I Manoscritti miscellanei. Ricerche su esemplari dei secoli IX–XII (Spoleto 2007) 150–154, 173, 180; S. Lucà, “Testi medici e tecnico-scien- tifici,” in La produzione scritta 563, 567. 20 Lucà, in La produzione scritta 563; P. Degni, “I testi scientifici nell’Italia meridionale bizantina: forme, modelli, circolazione,” in La conoscenza scien- tifica nell’alto medioevo (Spoleto 2020) 832. MINQI CHU 187 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 legomena et scholia in Theodosii Alexandrini canones isagogicos de flexione verborum of Georgios Choiroboskos;21 and some Italo-Greek palimpsests whose lower writings were grammatical treatises.22 Treatises on classical grammar, particularly the Ars grammatica of Dionysius Thrax, played a prominent role in these textbooks, although the later works of Theodosius of Alexandria and Georgios Choiroboskos and other practical excerpts were also transmitted.23 However, in the center of the Empire, in place of the classical works, grammatical training used almost exclu- sively ‘modern’ and practical extracts (such as tables of inflec- tions, collections of epimerismi, and schedography developed especially in the eleventh century).24 It seems, therefore, that southern Italy did not follow the pedagogical tools and methods prevalent in the center of the Empire. 21 S. Lucà and A. Vena, “Resti di un codice grammaticale greco ad Acerenza, in Basilicata,” Nea Rhome 11 (2014) 127–131. 22 For example, the lower text of Vat.reg. Pii II 35 was copied in 10th- century southern Italy: Lucà and Vena, Nea Rhome 11 (2014) 131–132. S. Lucà and F. Ronconi have rejected the Italo-Greek origin of four other manuscripts long believed to have come from southern Italy in the 10th–11th centuries that contain grammatical and orthographic treatises: Rom.Vallic. E 11, Vat.barb.gr. 70, Paris.gr.suppl. 388, and Oxon.Barocci 50. See J. Irigoin, “L’Italie méridionale et la tradition des textes antiques,” JÖB 18 (1969) 49, and “La culture grecque dans l’occident latin du VIIe au XIe siècle,” in La cultura antica nell’Occidente latino dal VII all’XI secolo II (Spoleto 1975) 441; P. Canart, “Le livre grec en Italie méridionale sous les Règnes Normand et Souabe,” Scrittura e Civiltà 2 (1978) 103–162, at 141–142; S. Lucà, “La pro- duzione libraria,” in R. Lavagnini et al. (eds.), Byzantino-Sicula VI La Sicilia e Bisanzio nei secoli XI e XII (Palermo 2014) 162; F. Ronconi, “Il codice Parisino Suppl. gr. 388 et Mosè del Brolo da Bergamo,” Italia medioevale e umanistica 47 (2006) 22–23, “Bodleian Library ms. Baroccianus 50: annotazioni codicolo- giche su un manoscritto miscellaneo,” in B. Atsalos (ed.), Actes du VIe Colloque International de Paléographie Grecque (Athens 2008) 639–655, and in La produzione scritta 81–90. 23 Ronconi, in La produzione scritta 101, 107; Lucà, in La produzione scritta 563. 24 Ronconi, in La produzione scritta 106–107. 188 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 Evidence that subjects of a level higher than grammar were taught in this province is extremely rare. A sign of local interest in logic, but insufficient to envisage training in this discipline during the Byzantine period,25 is the transmission of John of Damascus’ Dialecticae and the Supplementum ad dialecticam brevem (perhaps by the same author), both copied on fols. 31r–32r of Vat.gr. 1257 in Calabria at the end of the tenth century.26 The situation of rhetoric is more complicated. Messan.S.Salv. 118, kept in the monastery of S. Salvatore before entering the Biblioteca Regionale Universitaria ‘Giacomo Longo’, comprises ancient commentaries on the works of (Ps.‑)Hermogenes.27 Its transcription had been placed in eleventh-century southern Italy, but S. Lucà gave it instead a Greco-Oriental origin.28 On this view this rhetorical manuscript probably arrived in Sicily in the Norman period, after the twelfth-century foundation of the monastery by St. Bartholomew of Simeri under the patronage of Roger II.29 Messan.S.Salv. 119, another Italo-Greek rhetorical 25 Paris.gr. 2064, another Greek manuscript, contains commentaries on Aristotle’s De Interpretatione and Prior Analytics, whose origin in Byzantine southern Italy alleged by Irigoin has been definitively rejected by Lucà. See J. Irigoin, “L’Italie méridionale et la transmission des textes grecs du VIIe au XIIe siècle,” in N. M. Panagiotakes et al. (eds.), L’Ellenismo Italiota dal VII al XII secolo (Athens 2001) 93; S. Lucà, “Dalle collezioni manoscritte di Spagna: Libri originari o provenienti dall’Italia greca medievale,” RSBN 44 (2007) 63, and “Il libro bizantino e postbizantino nell’Italia meridionale,” Territori della Cultura 10 (2012) 34. 26 Ronconi, I Manoscritti miscellanei 185–191. 27 M. T. Rodriquez, “Su un codice di Ermogene del S. Salvatore di Mes- sina,” Nea Rhome 14 (2017) 81. 28 Rodriquez, Nea Rhome 14 (2017) 83–84; S. Lucà, “Il Diodoro Siculo Neap. B. N. gr. 4 è italogreco?” BBGG 44 (1990) 73–74, and in Byzantino-Sicula VI 148–149. 29 C. Torre, “Un Intellettuale greco di epoca normanna: Filagato da Cerami e il De Mundo di Aristotele,” Miscellanea di Studi Storici 15 (2008) 80– 81. Books brought from the center of the Byzantine Empire to southern Italy and Sicily grew more common in the Norman period. According to the Life MINQI CHU 189 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 manuscript, was certainly copied in the twelfth century rather than in the Byzantine period.30 R. Romano doubtfully iden- tified Neilos of Rossano as the author of a commentary on Hermogenes’ De statibus, transmitted by a tenth-century monk named Neilos in Paris.gr.suppl. 670.31 The Parisinus is therefore the only sure surviving witness to the production and circulation of rhetorical textbooks in the Greek- speaking area of Byzantine southern Italy. It alone proves the existence of local instruction in this advanced discipline. The compilation of Theodosius of Syracuse and rhetorical training in Byzantine Southern Italy in the ninth century This small parchment codex (132 x 105 mm) was copied in the second half of the tenth century by an anonymous scribe, probably in Calabria, using the typical Italo-Greek script en as de pique.32 The main decoration appears at the head of fols. 1r, ___ of Bartholomew of Simeri, this saint travelled to Constantinople in search of sacred books. After he received many from Emperor Alexius I Comnenus and his wife Irene Doukaina, he stored them in the libraries of his mon- asteries. At the time of William I a Sicilian embassy headed by Henry Aristippus received from the Byzantine emperor as diplomatic gifts (largitione susceptos imperatoria) books like the Erythraean Sibyl’s prophetic collection and Ptolemy’s Almagest (MS. Marc.gr. Z.313). See C. H. Haskins, “Further Notes on Sicilian Translations of the Twelfth Century,” HSCP 23 (1912) 156; Canart, Scrittura e Civiltà 2 (1978) 149; Canart (133–134) has shown that a more significant number of contemporary works were transmitted from the center of the Empire to southern Italy during the Norman period. 30 J. Irigoin, “La tradition des rhéteurs grecs dans l’Italie byzantine (Xe– XIIe siècle),” SicGymn 39 (1986) 78–81; Lucà, in La produzione scritta 572. 31 R. Romano, “Il commentario a Ermogene attribuito a S. Nilo di Ros- sano (Par. suppl. gr. 670, ff. 1–179v),” EpetByz 47 (1987–89) 253–274; F. Woerther, Hermagoras, Fragments et témoignages (Paris 2012) 274–275. On the monk Neilos and Paris.gr.suppl. 670 see M. Patillon, Eustathe, explication des états de cause d’Hermogène (Paris 2018) XL; Rodriquez, in Nea Rhome 14 (2017) 99; Lucà, in La produzione scritta 571. 32 C. Förstel and M. Rashed, “Une rencontre d’Hermogène et de Cicéron dans l’Italie médiévale,” Nea Rhome 3 (2006) 361; Lucà, in La produzione scritta 572. 190 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 100r, 111r, and 150r. It consists of elaborately intertwined illum- inated bands with large palmettes at either end33 and it serves to divide the codex into four sections. The first (fols. 1r–99v) comprises six classical handbooks of rhetoric: the Prolegomena of Anonymous, Books 7–12 of Aph- thonius’ Progymnasmata, Hermogenes’ De statibus, some rhetorical excerpts, and Ps.-Hermogenes’ Progymnasmata and De methodo sollertiae. At the end of this section (99v) appear two diagrams on the anonymous prolegomena to Dionysius of Halicarnassus’ De compositione.34 The second section (100r–110v) contains two other late- antique rhetorical treatises, one by Maximus of Ephesus and the other by an anonymous author. These are no longer related to (Ps.-)Hermogenes. At the end appears a diagram on Maxi- mus of Ephesus’ De objectionibus insolubilibus (110v). These two sections represent a variant of the Corpus Rhe- toricum, a rhetorical collection assembled by an anonymous teacher at the end of the fifth century.35 In the ninth–eleventh centuries this corpus was transmitted by several Greek manu- scripts and was considered the main textbook for rhetoric in the Byzantine Empire.36 33 Irigoin, SicGymn 39 (1986) 75. 34 M. Patillon, Corpus rhetoricum I (Paris 2008) xxxvi. 35 The complete version of this corpus comprises Anonymous’ Prolegomena, Aphthonius’ Progymnasmata, and four works of (Ps.-)Hermogenes: De statibus, De inventione, De ideis, and De methodo sollertiae. Added sometimes at the end were the De insolubilibus obiectionibus of Maximus of Ephesus and the Methodus ad prouinciarum praefectos oratione accipiendos from an anonymous author. See Patillon, Corpus Rhetoricum I x. 36 Patillon, Corpus Rhetoricum I v–lxxxvi. See also V. Valiavitcharska, “Rhetoric in the Hands of the Byzantine Grammarian,” Rhetorica 31 (2013) 238 n.1. Aside from the Parisinus, other manuscripts in the tenth–eleventh centuries that transmitted this corpus are: Ambros. III 66 Sup., Basileensis 70, Laur.plut. 60.15, Paris.gr. 1983, Paris.gr. 2977, Paris.gr. 2983, Paris.gr. 2923, Vat.gr. 104, and Vat.urb.gr. 130. See Patillon, Corpus Rhetoricum I v, x–xi, xxxiv. On Byzantine rhetorical theory in general see G. Kustas, Studies in Byzantine MINQI CHU 191 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 The third section (111r–149v) comprises Phoebammon’s Περὶ σχημάτων ῥητορικῶν and some scholia of classical handbooks entitled Τετάρτη γραφὴ ἐκ τῶν στάσεων, Τεχνῆς ῥητορικῆς σχόλια, and Δευτέρα γραφὴ ἐκ τῶν στάσεων.37 At the end (149r-v) were copied two patristic extracts of St. Athanasius of Alexandria and St. Basil.38 The last section (150r–152v) is completely different: it contains four dodecasyllables and a letter to Leo the Archdeacon of Syracuse recounting the Arab occupation of the city in 878. Both were written by a ninth-century monk from Syracuse named Theodosius, whom the title of the letter calls a gram- matikos, i.e. scholar or teacher for secondary education. S. Lucà suggests that he is Theodosius the Monk and Grammatikos, who was a lexicographer and the author of Lexicon Iambicorum Canonum Iohannis Damasceni.39 Each section starts its own subjects on a new recto folio with the main decoration at the head. Except for the final section, ___ Rhetoric (Thessaloniki 1973); E. M. Jeffreys (ed.), Rhetoric in Byzantium (Alder- shot 2003) 23–37, and “Rhetoric in Byzantium,” in I. Worthington (ed.), A Companion to Greek Rhetoric (Oxford 2007) 166–184; Valiavitcharska, Rhetorica 31 (2013) 237–260, and Rhetoric and Rhythm in Byzantium: The Sound of Persua- sion (Cambridge 2013); S. Papaioannou, “Rhetoric and Rhetorical Theory,” in A. Kaldellis et al. (eds.), The Cambridge Intellectual History of Byzantium (Cambridge 2017) 101–112. 37 The treatise Δευτέρα γραφὴ ἐκ τῶν στάσεων became the sole witness to certain developments from the Neoplatonic school of Alexandria during the time of Elias and David. It postulated that an overly apparent display of logical technicalities in a speech can hinder the speaker’s ability to persuade or convince others. This was known only to Arab philosophers. See Förstel and Rashed, Nea Rhome 3 (2006) 362–364. 38 These two patristic extracts may be models for rhetorical exercises. See Lucà, in La produzione scritta 573. 39 V. von Falkenhausen, “La conquista di Siracusa (878) nella memoria storica di Constantinopoli,” in M. Palma et al. (eds.), Per Gabriella. Studi in ricordi di Gabriella Braga II (Cassino 2013) 837–839; Lucà, in La produzione scritta 576. 192 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 which is incomplete, all end with diagrams and extracts inde- pendent of the main texts. These seem to have been added after the main texts in each division had been completed and they fill the space left over before the start of the next section. It is there- fore possible that each division was originally independent. Even if the Parisinus was assembled as a unit and transcribed in full by a single scribe, without interruption or disjointedness, it may have been assembled from four separate sources. If the source of fols. 150r–152v is in fact Theodosius the Grammatikos, this suggests that he had a role in the consti- tution or transmission of this rhetorical textbook.40 He may have assembled for his teaching in Syracuse this sequence of rhetorical texts, collecting them from three independent hand- books and adding his own at the end to provide his pupils with a contemporary local model for their learning and practice. Although the original book that he compiled has been lost, it remains available to us in the Parisinus, which derives from it. In Parisinus one finds evidence that Theodosius of Syracuse mod- ified the contents of the works he compiled into his new rhetorical textbook. In the stemma of manuscripts, the Parisinus is a rather isolated variant41 which features textual adaptations that strengthen or specify transitions, explain or clarify state- ments, and replace or delete exercises. The following instance of a modification that strengthens or specifies a transition, from the sixth chapter of Anonymous’ Prolegomena, is illuminating.42 This chapter is devoted to the origin of rhetoric: Syracuse, a colony of Corinth and Sparta, received it first (Corpus rhetoricum I 27): Εἰσεδέξαντο τοίνυν πρῶτον ἁπάντων αἱ Συράκουσσαι, πόλις τῆς Σικελίας. Σικελίας δὲ νῆσος µεγίστη κειµένη περὶ τὰ ἑσπέρια 40 Förstel and Rashed, Nea Rhome 3 (2006) 365. 41 Patillon, Corpus rhetoricum I xxxiv–xxxvi. 42 For further evidence of strengthening and specifying transitions see Patillon, Corpus rhetoricum I 29, 30, 35, 39–42, and 44. MINQI CHU 193 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 µέρη. Αὕτη Κορινθίων µὲν καὶ Λακεδαιµονίων ἦν ἄποικος. The Parisinus replaces the word Αὕτη at the beginning of the last sentence with αὕται (corr. αὗται) αἱ Συράκουσσαι. This streng- thens the transition and specifies the meaning of αὕτη, making clear that it refers to the city of Syracuse, once a colony of Corinth and Sparta, rather than to the entire island of Sicily. The author of this intervention seems well acquainted with the history of Sicily and Syracuse. For explanatory or clarifying statements we can look at the beginning of the eleventh chapter of the Prolegomena, where the anonymous author argued that Aristotle’s definition of rhetoric was insufficient (ἐλλείποντα)—one of two traditional vices in definitions, the other being excess (ὑπεροχή). He then explains these faults further (I 31): Εἰ µὲν γὰρ ὑπερέχει ὁ ὅρος τῇ λέξει, ἐλλείπει τοῖς πράγµασιν, εἰ δὲ ἐλλείπει τοῖς λόγοις, πλεονάζει τοῖς πράγµασιν. The Parisinus adds at this point the definition of Ἄνθρωπος in order to clarify the statement (I 31): Οἷον ὑπερέχων ὁ ὅρος κατὰ τὴν λέξιν· Ἄνθρωπος γάρ ἐστι ζῷον λογικόν, θνητόν, νοῦ καὶ ἐπιστήµης δεκτικόν, γραµµατικόν· ἐπεὶ δ᾽ ἐπλεόνασεν ἐνταῦθα τῇ λέξει κατὰ τὸ γραµµατικόν, οὐ πάν- τας περιέλαβεν τοὺς ἀνθρώπους ὁ ὅρος ἀλλὰ µόνους τοὺς γραµ- µατικούς· διὸ καὶ λέγεται ἐλλείπειν τοῖς πράγµασιν. ἐὰν δὲ ἐλ- λείψῃ τοῖς λόγοις, πλεονάζει τὰ πράγµατα· ἐὰν γὰρ εἴπῃς οὕτως· τί ἐστιν ἄνθρωπος; Ζῷον λογικόν, ἐπειδὴ ἐνέλειπες τὸ θνητόν, ἐπλέονασε τὸ πρᾶγµα· συνεισήνεγκεν γὰρ καὶ θεοὺς καὶ ἀγγέ- λους καὶ δαίµονας. This practice of explaining and clarifying complex statements of literary analysis effectively improves the readability of the text. The most significant evidence of modification of exercises concerns Aphthonius’ Progymnasmata. In the seventh book, on Commonplace, an original exercise was added on commonplace against the murderer, while the exercise Κοινὸς τόπος κατὰ ἀργοῦ (Commonplace against a lazy man), traditionally attributed to Niko- 194 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 laos in the fifth century, was substituted for Aphthonius’ Κοινὸς τόπος κατὰ τυράννου (Commonplace against a tyrant).43 Three exer- cises were deleted: Σοφίας ἐγκώμιον (Praise of wisdom) in the eighth book, on Encomium; Ψόγος Φιλίππου (Blame against Philip II) in the ninth, on Invective; and Σύγκρισις ᾽Αχιλλέως καὶ Ἕκτορος (Compari- son between Achilles and Hector) in the tenth, on Comparison. More- over, in the twelfth book, on Description, Aphthonius’ exercise Ἔκφρασις τῆς ἐν Ἀλεξανδρεία ἀκροπόλεως (Description of the acropo- lis in Alexandria) was replaced by a new one taken from Libanius, Ἔκφρασις κήπου (Description of a garden).44 The elimination of some exercises that were rich in ancient mythology or history, and the substitution of shorter, easier exercises for Aphthonius’ longer, more intellectually demanding ones, were designed to simplify the contents of this textbook. Evidently, these interventions were intended to make this rhe- torical textbook more readily accessible to less educated users and students. Thus, as a rhetorical textbook the Parisinus likely circulated and was used where the teaching of rhetoric was not well developed. If all these adaptations are to be attributed to a single author, Theodosius the local Syracusan scholar, its likely compiler, is an excellent candidate. His goal must have been to adapt the text of his sources to his rhetorical teaching practice in Sicily, a peripheral area of the Byzantine Empire where edu- cation remained less developed than in the center. His Syra- cusan background and his knowledge of the city reinforce this hypothesis. Thus, a brief reconstruction of the history of codex Parisinus is possible: Theodosius in the ninth century assembled three ancient rhetorical handbooks and added his own works at the end to produce a new textbook for his teaching of rhetoric in Syracuse. To make it suitable for his pedagogical practice in Sicily he adapted its content and made it more accessible. His 43 Patillon, Corpus rhetoricum I 105–106. 44 Patillon, Corpus rhetoricum I 106. MINQI CHU 195 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 textbook circulated and was re-copied in the Greek-speaking region of southern Italy down to the transcription of the Parisinus by an anonymous scribe, probably in Calabria, in the second half of the tenth century. The circulation of the rhetorical textbook of Theodosius of Syracuse in Greek-speaking Byzantine Italy The history of codex Parisinus continued after the transcrip- tion by the anonymous Italo-Greek scribe. In addition to its main contents the scribe made various annotations in the mar- gin. Subtitles and summarizing or explanatory annotations were often added in capital letters.45 A typical Italo-Greek brachy- graphic system was employed to write these majuscule notes, especially on fols. 89v–98r.46 The scribe’s fluent brachygraphic writing hints that these notes were not authored by him but merely transcribed from his source. The in-line annotation symbols on fols. 24r, 28v, 30r-v, 77r, and 81r provide further evidence to support this deduction. Identical symbols, one placed by the annotation and another within the relevant line of the text, link the marginal subtitles πε(ρὶ) πραγµάτ(ων) on 24r, πε(ρὶ) δικαί(ου) λογικῆς and πε(ρὶ) ἀντιλήψε(ως) στάσ(ις) δʹ on 28v, and περὶ µεταλήψε(ως) στάσ(ις) 45 Subtitles and notes summarize the content on fols.18v, 23v–25r, 26v, 27v–30v, 31v, 32v–33v, 37v, 38v, 39v–43r, 44r-v, 46r, 47v, 49r, 51v–53r, 54r–55v, 62v, 66v, 67v–68r, 70r, 75r, 77r, 81r, 82r-v, 83v–95r, 96r–98v, 112v, 117r, and 126v. Notes explaining vocabulary appear on 30r, 77v, 107v, and 140r. On 30r the note ἀντινοµία τὸ πταῖσµα attempted to explain the term ἀντινοµία in Hermogenes’ text (albeit incorrectly) by suggesting πταῖσµα (fault) as a syn- onym. On 77v the note τιµᾶν ἔναγχος interpreted the word ποµπεύειν. On 107v the phrase ἄντικρυς καὶ διαρρήδην was explained in the margin as σα- φῶς φανερῶς. On 140r κοντάρι(ον) was used to describe the word ἐγχειρίδιον. 46 N. P. Chionides believed that the notes in brachygraphy originated in Calabria in the mid-tenth century. J. Irigoin dated them in turn to a time close to the transcription of the manuscript. See Irigoin, SicGymn 39 (1986) 75; N. P. Chionides and S. Lilla, La Brachigrafia Italo-bizantina (Rome 1981) 29, 46. 196 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 ιγʹ on 30v to the corresponding passages of Hermogenes’ De sta- tibus, the text under consideration (1.8, 2.5, and 2.14). On 30r a mark in the 14th line of the main text of Hermogenes 2.12 directs attention to the marginal note ἀντινοµία τὸ πταῖσµα. Two notes on 77r, ποσαχῶς ὁ κακοῦργος and ἴδιον πῶς, summarizing two brief passages (οἷον κακοῦργος πᾶς ὁ κακόν τι ἐργαζόµενος and ἰδίως δ᾽ ὁ κλέπτης) from Ps.-Hermogenes’ De methodo sollertiae (1.2), are headed by two marks placed in the 8th and 9th lines of the main text. Finally, an annotation mark in the 10th line of 81r points to the marginal summary τὸ περιπλέκειν ποῖος κ(αι)ρὸς ζηλωτ(ὸν) καὶ χρηστὸν ποιεῖ of Ps.-Hermogenes 8.1. The fact that the annotation marks were included in the line itself, between words, rather than in the interlinear spaces, and that the corresponding marginal notes were written in capital letters, strongly suggests that both annotations and accompany- ing marks were already embedded in the source and that the scribe faithfully copied them. He is therefore not the author of these notes. Notably, numerous minuscule notes that correct scribal errors are also present in the margins of the Parisinus. Apparently, these were also written by the same scribe.47 Two of these are of especial interest. The first is found on fol. 85r, a long note that is incomplete owing to the re-trimming of the right edge of this page: ὅτι τὸ ὑπ(ερ)βατ(ὸν) ὀ[…] καλ(ὸν) σχῆµα ἀλλ᾽ ἀναικαῖ(ον)· καὶ πότ᾽ […]ναγκαῖ(ον) σαφην[…] ὄργαν(ον) γίνετ(αι)· καὶ π[…]τε µακρὸν ὑπ(ερ)βα[…] ἐν ᾧ τι τ(οῦ) Ὁµήρ(ου) […]µιζόµενον ὑπ(ερ)βατ(ὸν) οὐκ ἔστι καὶ διὰ τί […]κ ἔστιν· By referring to the main text on 85r, from Περὶ ὑπερβατοῦ (About hyperbaton) of Ps.-Hermogenes’ De methodo sollertiae (14.1–5), we can reconstruct the note as follows: 47 Minuscule notes by the scribe correcting his transcription are on fols. 6r, 28r, 30v, 32v, 34v, 38r, 42v, 45v, 49r, 50r-v, 52r, 53r, 54v, 55v, 56r, 57v, 58r, 59r, 60v, 73v, 88v, 91r, 97v, 102v, 114v, 115r, 122r, 133v, 135r-v, 136v, and 140r. MINQI CHU 197 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 ὅτι τὸ ὑπ(ερ)βατὸν oὔκ (ἔστι) καλὸν σχῆµα ἀλλ᾽ ἀναγκαῖον (ante corr. ἀναικαῖον)· καὶ πότ᾽ (ἀ)ναγκαῖον σαφηνείας ὄργανον γίνετ(αι) καὶ πότε µακρὸν ὑπ(ερ)βατὸν ἐν ᾧ τι τοῦ Ὁµήρ(ου) νοµιζόµενον ὑπ(ερ)βατ(ὸν) οὐκ ἔστι καὶ διὰ τί οὐκ ἔστιν· This is a summary of Ps.-Hermogenes’ text rather than an inde- pendent commentary. On fol. 88r πανουργ(οῦ) καὶ ψεύστ(ου) was added as a note to explain the uncommon term Εὐρυβάτου, the name of a Greek mythological character. It occurred in the chapter Περὶ ἐγνωσμέ- νου ψεύσματος καὶ πότε χρηστέον αὐτῷ (About an obvious lie and its use) of the same text of Ps.-Hermogenes (19.2). The use of the word πανοῦργος to describe Eurybatus can be traced back to Aristotle’s text, as cited by Suetonius: Εὐρύβατος· ὁ πανοῦργος καὶ Εὐρυβάτης ὁ αὐτός … Ἀριστοτέλης δ’ ἐν πρώτῳ περὶ δικαιοσύνης κλέπτην αὐτὸν γεγονέναι φησίν (fr.84 Rose). This interpretation remained popular in the meso-Byzantine period and was cited in the lexicons of Hesychius, Photius, and the Suda.48 However, a search of the TLG did not yield any text that uses the word ψεύστης/ψεύστου to describe Eurybatus. Only the synonym ἀπατεών was found in Hesychius.49 Therefore, while charac- terizing Eurybatus as a knavish liar was not exceptional in the Byzantine world, the note on 88r was probably not quoted from other treatises but conceived by a reader or a copyist during the transmission of the textbook of Theodosius of Syracuse or its source material. Unfortunately, it is impossible to establish that the Parisinus’ scribe authored these annotations. Because their ink is of the 48 Hesych.: Εὐρυβάτης· πανοῦργος, ἀπατεών, κέρκωψ (ed. Latte); Photius’ Lexicon: Εὐρύβατος … Νίκανδρος Αἰγιναῖον Εὐρύβατον πανουργότατον, οὗ µνηµονεύει Ἀριστοτέλης ἐν αʹ περί δικαιοσύνης. Δοῦρις δέ ἐν δʹ τῶν περί Ἀγαθοκλέα ἀπὸ τοῦ Ὀδυσσέως ἑταίρου (ed. Theodoridis); Suda: Εὐρύβατος … Νίκανδρος· Αἰγίνεον Εὐρύβατον πανουργότατον· οὗ µνηµονεύει Ἀριστοτέλης ἐν αʹ περὶ δικαιοσύνης (ed. Adler). 49 A scholion to Aeschines’ Against Ktesiphon (ad 3.137) uses ἄπιστος to de- scribe Eurybatus: Εὐρύβατοι δύο ἐγένοντο ἄµφω πονηροί, ἄπιστοι, πανοῦργοι. 198 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 same color as the ink used for the adjoining content, we cannot exclude the possibility that that our scribe simply copied them from his model. They could therefore go back to earlier times, like the majuscule notes transcribed by the same copyist. If these notes in majuscules and minuscules do not predate the alleged compilation of Theodosius, then they must have been composed in Byzantine Italy, where this new rhetorical textbook with the works of Theodosius of Syracuse exclusively circulated. Other than Theodosius’ innovative compilation, local interest in rhetorical learning was apparently restricted to summarizing preexisting content or explaining difficult vocab- ulary. Unlike contemporary scholar-teachers in Constantinople, local annotators did not analyze or develop their own theories or compose their own commentaries.50 If on the other hand these annotations were already present in the sources assembled by Theodosius—an alternative that seems distinctly less likely to me—then both he and his succes- 50 Constantinopolitan scholar-teachers from the ninth–eleventh centuries such as John of Sardis, Anonymous, John Sikeliotes, his near-contemporary John Geometres, and John Doxopatres all left commentaries on the works of (Ps.‑)Hermogenes and Aphthonius. Kustas, Studies in Byzantine Rhetoric 21–26; C. Jouanno, “Les Byzantins et la seconde sophistique: étude sur Michel Psellos,” REG 122 (2009) 118; S. Papaioannou, “Sicily, Constantinople, Miletos: The Life of a Eunuch and the History of Byzantine Humanism,” in Myriobiblos. Essays on Byzantine Literature and Culture (Boston 2015) 271–277, and “Ioannes Sikeliotes (and Ioannes Geometres) Revisited with an Ap- pendix: Edition of Sikeliotes’ Scholia on Aelius Aristides,” TM 23.1 (2019) 659–679; P. Roilos, “Ancient Greek Rhetorical Theory and Byzantine Discursive Politics: John Sikeliotes on Hermogenes,” in T. Shawcross et al. (eds.), Reading in the Byzantine Empire and Beyond (Cambridge 2018) 159–184; C. A. Gibson, “The Anonymous Progymnasmata in John Doxapatres’ Homiliae in Aphthonium,” ByzZeit 102 (2009) 83–84; S. Efthymiadis, “De Taraise à Méthode (787–847): l’apport des premières grandes figures,” TM 21.2 (2017) 177; D. Resh, “Toward a Byzantine Definition of Metaphrasis,” GRBS 55 (2015) 754–787; B. D. MacDougall, “John of Sardis’ Commentary on Aph- thonius’ Progmnasmata: Logic in Ninth-Century Byzantium,” GRBS 57 (2017) 724–726. MINQI CHU 199 ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 sors in southern Italy would have merely reproduced and trans- mitted ancient summaries and lexical interpretations. On this view, local knowledge of rhetoric in southern Italy was sparse and derivative. In either case, one may reasonably conclude that rhetorical learning in the Greek-speaking area of southern Italy fell to rather low levels during the Byzantine period after the ninth- century production by Theodosius the Grammatikos of his new rhetorical textbook in Syracuse. The local study of rhetoric was impoverished and restricted to a narrow circle. The very small number of locals interested in learning ancient rhetoric is re- flected in the fact that throughout the Byzantine period only the scribe himself seems to have made notes in the Parisinus.51 The high quality of this manuscript supports this view. Codi- cologically, the regularity of the quires52 and the rarity of holes and untrimmed edges bear out this excellence. Its high cost is shown by the use of several colors in the decoration, par- ticularly gold on fols. 3v, 4v, 7v, 10v, 13r, 14r, 16r, 17r-v, 19r, 22v, 23r, 145v, 146r, 149v, and 150r;53 and the expensive azure on 22v, 34v, 51r-v, 100r, 101v, 111r, 122r, 127v, 128r, and 150r. A codex of this quality is quite rare in southern Italy during the Byzantine period. It may therefore have been commissioned and used by a member of the local elite interested in reading or collecting texts on rhetoric. The study of this advanced subject was perhaps exclusively accessible to this small group. 51 This manuscript transmitted other annotations in Latin and Greek written in southern Italy or Sicily after the end of Byzantine rule; see Förstel and Rashed, Nea Rhome 3 (2006) 365–370. 52 This manuscript comprises 19 regular quaternions. 53 Gold appears rarely in the Italo-Greek manuscripts produced during the Byzantine period. See J. Leroy, “Le Parisinus gr. 1477 et la détermina- tion de l’origine des manuscrits italo-grecs d’après la forme des initiales,” Scriptorium 32 (1978) 198. 200 RHETORICAL TRAINING IN BYZANTINE ITALY ————— Greek, Roman, and Byzantine Studies 63 (2023) 182–200 Conclusion In ninth-century Sicily a local scholar by the name of Theo- dosius compiled a new rhetorical textbook for his pedagogical practice in Syracuse. He assembled three ancient handbooks that were popular in Byzantium and added his own compo- sitions at the end. To suit this textbook to his local audience, he modestly modified the contents, reinforcing or specifying transitions, explicating or clarifying statements, and replacing or omitting exercises. After his innovative work, this new textbook of rhetoric circulated during the Byzantine period almost exclusively within a small circle of the elite. The local learning of rhetoric was brought to a low level that consisted at most in summarizing the content of parts of this textbook and interpreting difficult vocab- ulary. This conclusion is consistent with the impression one naturally obtains from the reading of the Siculo-Calabrian hagiographies, and with the fact that the Parisinus is the sole surviving Italo-Greek manuscript that features rhetorical texts.54 July, 2023 Sorbonne Université minqichu@gmail.com 54 I would like to express my gratitude to Vincent Déroche (Professor at Sorbonne Université) and Filippo Ronconi (Directeur d’études at the École des hautes études en sciences sociales in Paris) for their invaluable remarks and suggestions. I also thank the anonymous reviewers and the editors for their thoughtful comments and corrections.