Valori ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 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/ Unifying Fragments of Byzantine Scholarship: Stephanos’ Commentary on Aristotle’s Rhetoric Ugo Valori HE MANUSCRIPT Città del Vaticano, Biblioteca Apo- stolica Vaticana, Vat.gr. 1340 (Diktyon 67971) contains the three extant Byzantine commentaries on Aristotle’s Rhetoric. One of them is a fragment consisting of only seven pages in a modern edition (CAG XXI.2 323–329). Although limited in size, the fragment is interesting in many respects for the sources and compositional practices it shows. However, despite the recent surge of interest in the Byzantine commentaries on Aristotle’s Rhetoric,1 no scholarly work has dealt with this brief 1 These were first signaled by J. A. Fabricius, Bibliotheca Graeca III (Ham- burg 1793) 221–222, and further analysed by C. Brandis, “Ueber Aristoteles’ Rhetorik und die griechischen Ausleger derselben,” Philologus 4 (1849) 1–47. J. S. Vater, Animadversiones et lectiones ad Aristotelis libros tres Rhetoricorum (Leipzig 1794) x–xvi, is dedicated specifically to the Anonymous, while O. Schissel, “Stephanos 11,” RE 3A (1929) 2364–2369, W. Wolska-Conus, “À propos des Scolies de Stéphanos à la Rhétorique d’Aristote: l’auteur, l’oevre, le milieu,” in Actes du XIVe Congrès international des études byzantines III (Bucharest 1976) 599– 606, and W. Harinder, “Der Aristoteles-Kommentator Stephanos in Seiner Zeit,” in K. Belke et al. (eds.), Byzantina Mediterranea. Festschrift für Johannes Koder (Vienna 2007) 257–267, deal with Stephanos’ Commentary. Brief remarks on the two major commentaries are in G. A. Kennedy, Greek Rhetoric under Christian Emperors (Princeton 1983) 318–320, and N. G. Wilson, Scholars of Byzantium (London 1996) 182–184. More recently: C. Rambourg, “Qu’est-ce que le commentaire anonyme des CAG XXI 2?” in F. Woerther (ed.), Com- menting on Aristotle’s Rhetoric from the Antiquity to the Present (Leiden 2011) 14–40, for the Anonymous; and, above all, the useful book of M. Vogiatzi, Byzantine T UGO VALORI 155 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 text.2 This article aims to demonstrate that the Fragment was originally the final section of another Byzantine commentary on the same work, transmitted by the same manuscripts (Vat.gr. 1340 and apographs) and conventionally known as Stephanos’ Commentary (CAG XXI.2 263–322). My contention is that the two textual segments were split apart, owing to the displacement of a quire or of some folios in an antigraph of the Vatican manu- script. By way of conclusion, I offer a few remarks on Stephanos’ authorial practices. These are intended as a stepping stone for a future more in-depth analysis of the commentary itself. The commentaries on Aristotle’s Rhetoric Only three Byzantine commentaries on Aristotle’s Rhetoric survive.3 The material is very scant as compared with the large amount of exegetical texts on other Aristotelian treatises, especially the Organon, which enjoyed a wide circulation during Commentaries on Aristotle’s Rhetoric (Berlin 2019). On the Byzantine reception of Aristotle’s Rhetoric in general see T. Conley, “Aristotle’s Rhetoric in Byzan- tium,” Rhetorica 8 (1990) 29–44, “Notes on the Byzantine Reception of the Peripatetic Tradition in Rhetoric,” in W. Fortenbaugh et al. (eds.), Peripatetic Rhetoric after Aristotle (New Brunswick 1996) 217–242, and “The Alleged ‘synopsis’ of Aristotle’s Rhetoric by John Italos and its Place in the Byzantine Reception of Aristotle,” in G. Dahan et al. (eds.), La Rhetorique d’Aristote. Tra- ditions et commentaires de l’Antiquité au XVIIe siècle (Paris 1998) 49–64; P. Chiron, “Les commentaires grecs tardo-antiques et byzantins à la Rhétorique: hypo- thèses sur une quasi-absence,” in F. Woerther (ed.), Commenting the Aristotle’s Rhetoric from the Antiquity to the Present (Leiden 2018) 1–13; M. Vogiatzi, “The Byzantine Reception of Aristotle’s Rhetoric: The 12th Century Renaissance,” ByzZeit 113 (2020) 1069–1088. 2 Excepting some very brief remarks in Vogiatzi, ByzZeit 113 (2020) 1076, and P. M. Pinto, “I ‘parenti’ di Isocrate (Sullo scambio, 4 e un commento anonimo alla Retorica di Aristotele,” BollClass 20 (1999) 11–18, which deals with a different topic than does the present article. 3 Hugo Rabe published all three in 1896, CAG XXI.2: the Anonymous Commentary (1–262), Stephanos’ Commentary (263–322), and an anon- ymous Fragment of Commentary on Book 3 (323–329). A short fragment of paraphrasis does exist as well, published in the same volume (330–334), trans- mitted by only one MS., Paris.gr. 1874. 156 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 the Greek Middle Ages.4 Byzantine manuscripts of the Rhetoric are equally rare and, except for Paris.gr. 1741 (a 10th cent. codex including Rhetoric and Poetics within a corpus of ‘minor’ rhetori- cians),5 all the other manuscripts date from the twelfth century onward.6 Clearly, Byzantine learned audiences neglected Aristotle’s Rhetoric, despite the relevance of rhetoric as a discipline in the curriculum:7 starting in Late Antiquity, in fact, Aristotle’s Rhetoric probably was no longer taught in rhetorical schools, its place being taken by other treatises, namely Aphthonios’ Progymnas- mata and Hermogenes’ rhetorical corpus. The success of these works was due to their clarity and imitability, while Aristotle was proverbial for his “obscurity.”8 On the other hand, Aristotle’s 4 For an overview of Aristotle’s reception in Byzantium see K. Oehler, “Aristotle in Byzantium,” GRBS 5 (1964) 133–146; M. Trizio, “Reading and Commenting on Aristotle,” in A. Kaldellis et al. (eds.), The Cambridge Intellectual History of Byzantium (Cambridge 2017) 397–412. 5 H. Rabe, “Rhetoren Corpora,ˮ RhM 67 (1912) 321–357, at 337–343; D. Harlfinger and D. Reinsch, “Die Aristotelica des Parisinus gr. 1741,” Philologus 114 (1970) 28–50. 6 R. Kassel, Der Text der Aristotelischen Rhetorik. Prolegomena zu einer kritischen Ausgaben (Berlin 1971) 1–18. 7 On the role of rhetoric in Byzantine culture see S. Papaioannou, “Rhet- oric and Rhetorical Theory,” in The Cambridge Intellectual History 101–112; A. Riehle, “Rhetorical Practices,” in S. Papaioannou (ed.), The Oxford Hand- book of Byzantine Literature (Oxford 2021) 294–315, with bibliography. 8 A compelling description of this phenomenon has been given by P. Chiron, “Les commentaires grecs tardo-antiques et byzantins à la Rhétorique: hypothèses sur une quasi-absence,” in F. Woerther (ed.), Commenting on Ari- stotle’s Rhetoric (Leiden 2018) 1–13. On the Neoplatonic schools and their ‘choice’ of Hermogenes see G. L. Kustas, Studies in Byzantine Rhetoric (Thessa- loniki 1973) 5–26; G. A. Kennedy, “Later Greek Philosophy and Rhetoric,” Ph&Rh 13 (1980) 181–197. Several Byzantine authors report the difficulty of Aristotle’s prose, as noted by Trizio, in The Cambridge Intellectual History 403– 405. On the fortune of Hermogenes’ and Aphthonios’ works in Greek Late Antiquity and Middle Ages see C. Pepe, “The Rhetorical Commentary in Late Antiquity,” AION(filol) 40 (2018) 86–108; V. Valiavitchsarska, “Rhetoric in the Hands of the Byzantine Grammarian,” Rhetorica 31 (2013) 237–260. UGO VALORI 157 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 most studied works were the Organon and (even if to a less extent) the Physics, used largely in didactic contexts. The idea of Aristotle as mainly a logician and physicist, developed in the Neoplatonic schools, was uninterruptedly perpetuated up to Michael Psellos and beyond.9 The three commentaries belong to a very particular phase of Aristotle’s reception in Byzantium, namely the twelfth century. In this period we see some distinctive changes in the selection of Aristotle’s works. Michael of Ephesos, Eustratios of Nicaea, and some anonymous writers produced a large number of commen- taries on treatises previously neglected (Rhetoric, Politics, bio- logical works).10 This renaissance of Aristotelian studies could be (at least partially) linked to the patronage of princess Anna Komnene.11 Turning specifically to the commentaries on the Rhetoric, their 9 For the Aristotelian curriculum of Neoplatonic schools see R. Sorabji, The Philosophy of the Commentators, 200–600 AD. A Sourcebook III Logic and Metaphysics (London 2004) 41–43; P. Golitsis, Les Commentaires de Simplicius et de Jean Philo- pon à la Physique d’Aristote (Berlin 2008) 10–14. For the teaching of Aristotle’s Logic and Physics in Byzantium see Trizio, in The Cambridge Intellectual History 398–400. A witness to the importance of Aristotle’s Organon and physical works can be found in Michael Psellos: ὁ δ’ Ἀριστοτέλης θεολογῶν µὲν ἧττον, φυσιολογῶν δὲ µᾶλλον, καὶ ἀποδεικνὺς ἐπὶ µάλιστα (Theol. min. II 6.16–17). 10 See M. Trizio, “Forging Identities between Heaven and Earth. Com- mentaries on Aristotle and Authorial Practices in Eleventh- and Twelfth- Century Byzantium,” in B. Van der Berg et al. (eds.), Byzantine Commentaries on Ancient Greek Texts, 12th–15th Centuries (Cambridge 2022) 61–99. 11 In the funeral oration dedicated to Anna Komnene, George Tornikes (ed. Darrouzès 283.4–9) refers to a circle of “philosophers” gathered by the princess in order to draft commentaries on those works of Aristotle that had not received attention in the previous scholarly tradition, including Rhetoric. Therefore Robert Browning, “An Unpublished Funeral Oration on Anna Comnena,” PCPhS 8 (1962) 1–12, linked the growth of interest in Aristotelian studies in the 12th century to the “circle of Anna Komnene.” More recent scholarship tends to resize the role of Anna’s circle, considering it as an element of a broader trend to which the extant commentaries on Rhetoric belong as well. For this problem see M. Trizio, Il neoplatonismo di Eustrazio di Nicea (Bari 2016) 22–72; Vogiatzi, ByzZeit 113 (2020) 1069–1088. 158 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 edition of reference is still that published in 1896 by Hugo Rabe in CAG. According to Rabe, all the manuscripts transmitting the three commentaries depend on a single codex, the Vat.gr. 1340.12 This manuscript is composed of two codicological units:13 the first contains the text of the Rhetoric and was copied by four scribes, all to be traced to the Planoudean circle (late 13th/ beginning 14th cent.); the second and older unit, dated to the end of the 12th century or the beginning of the 13th on the basis of palaeography, contains the three commentaries, written by a single anonymous scribe.14 12 Rabe, Anonymi et Stephani in Artem Rhetoricam Commentaria (CAG XXI.2) v– vii. 13 On the concept see J. P. Gumbert, “Codicological Units: Towards a Terminology for the Stratigraphy of the Non-Homogeneous Codex,” in P. Degni et al. (eds.), Il codice miscellaneo. Tipologie e funzioni (Turnhout 2004) 17– 39. 14 The handwriting of the commentaries has been identified also in Marc.gr. Z 257 and Paris.gr. 1925: C. Giacomelli “Aristotele e i suoi commentatori nella biblioteca di Bessarione,” in A. Rigo et al. (eds.), I libri di Bessarione. Studi sui manoscritti del Cardinale a Venezia e in Europa (Turnhout 2021) 219–275, at 234–235, and L. Koch, τὸ τῆς λέξεως συνεχές. Michael von Ephesos und die Rezeption der Aristotelischen Schrift De motu animalium in Byzanz (diss. Hamburg, https://ediss.sub.uni-hamburg.de/handle/ediss/7298). A description can be found in Á. Ibáñez Chacón, “Un escolio en el Vat.gr 1340 y la transmission de la Καινὴ ἱστορία de Tolomeo Queno,” RHT 17 (2022) 145–170, at 151– 154. I. Pérez Martín had identified one of the four scribes responsible for copying the first codicological unit (containing the Rhetoric) as Maximos Planoudes himself: “Un esemplare della Retorica di Aristotele copiato da Massimo Planude e Giovanni Zaride,” in Byzantium. Identity, Image, Influence (abstracts, 19th Intern. Congr.: Copenhagen 1996) no. 8126. However, she has recently retracted this attribution, ascribing this handwriting to the “Ps scribe,” a disciple of Planoudes who imitates his writing almost perfectly: I. Pérez Martín, “La influencia de la escritura de Máximo Planoudes en su entorno,” Scripta 15 (2022) 75–94, at 91. On the manuscript and its complex history in general see P. De Nolhac, La bibliothèque de Fulvio Orsini (Paris 1887) 164, 188, 341; Rabe, Anonymi v–vii; R. Kassel, Der Text der Aristotelischen Rhetorik (Berlin 1971) 52–57; P. Schreiner, Texte zur Spätbyzantinischen Finanz und Wirt- schaftsgeschichte in Handschriften der Biblioteca Vaticana (Vatican City 1991) 65–79; UGO VALORI 159 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 The Vatican manuscript presents the commentaries as self- standing texts, and not in the margins of the relevant Aristotelian passages.15 The first commentary in the manuscript (CAG XXI.2 1–262) is anonymous and is the longest—despite being in- complete due to the loss of eight quires at the beginning of the manuscript.16 It is followed by our anonymous Fragment (XXI.2 323–329) devoted to the last part of the Aristotelian treatise (Rhet. 1416a28–11420b2). The third commentary (XXI.2 263–322) is introduced with the heading ἔτερα σχόλια τοῦ κυροῦ στεφάνου and so is conventionally known as Stephanos’ Commentary. It starts with the analysis of Rhet. 1357a22 and ends abruptly with Rhet. 1410a33 because of the loss of some folios in the Vatican manuscript itself. The Fragment The second text preserved in the manuscript, the Fragment, has an apparently standard structure: it is not a continuous com- mentary but a set of unconnected glosses on specific points in the Rhetoric, though it mainly follows the order of the text com- mented upon. This is a quite common feature of Middle- Byzantine philosophical commentaries.17 What appears unusual is that, after reaching the last section of the Rhetoric and com- P. Canart, “Quelques examples de division du travail chez les copistes byzan- tins,” in P. Hoffman et al. (eds.), Recherches de codicologie comparée. La composition du codex au Moyen Âge, en Orient et en Occident (Paris 1998) 49–67, at 51–52; M. L. Agati, Giovanni Onorio da Maglie, Copista greco (1535–1563) (Pisa 2001) 141. 15 The mise en page of the Vatican MS. is typical of many 12th-century MSS. containing commentaries on Aristotle, as noted by A. Németh, “Fragments from the Earliest Parchment Manuscript of Eustratios’ Commentary on Aristotle’s Nicomachean Ethics,” RHT 9 (2014) 51–78; Trizio, in Byzantine Com- mentaries on Ancient Greek Texts 63–67. 16 At f. 43 the signature Θʹ is found: the first quires likely contained the text of Aristotle’s Rhetoric, and certainly contained the opening pages of the Anonymous Commentary. 17 Thus e.g. Micheal of Ephesos’ method, as noted by Trizio, in Byzantine Commentaries on Ancient Greek Texts 77–80. 160 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 menting on it, the commentator goes back to ch. 14 of Book 3 (Rhet. 1415a4) and adds glosses on sections of Aristotle’s text that he has already commented on (Frag. in Rhet. 327.34–329.22). In other words, the author seems to leaf through his book again to comment on specific items he had previously neglected. He also devotes three glosses to the same sentence.18 The glosses in this final section of the commentary are further disordered among themselves, so this anomaly cannot be explained by the mis- placement of a folio. The commentator is very interested in Aristotle’s mentions and quotations of ancient authors: he discusses, completes, and even corrects Aristotle’s references, often making clear that he is consulting a manuscript of the mentioned author. Thus he had access to a large number of manuscripts. For example, he quotes five otherwise unknown verses by the fifth-century B.C. poet Choerilus of Samos,19 two passages from Plato,20 one from Herodotus,21 the extant hypothesis of Isocrates’ Philippus,22 and probably a lost one of the Antidosis.23 Moreover, he corrects two quotations from Sophocles and Euripides using readings that are found in branches of the medieval manuscript tradition of these authors.24 He also compares multiple manuscripts of Aristotle’s Rhetoric, reporting variants.25 Considering the brevity of the ex- tant fragment and, in contrast, the number of manuscripts that its author is seen to have consulted, it is probable that he had 18 Frag. in Rhet. 324.12–15, 328.30–36, 329.13–22. 19 Frag. in Rhet. 328.3–7 = Choeril. fr.2 Bernabé. 20 Frag. in Rhet. 328.31–36 = Pl. Resp. 614B, 325.7–9 = Theag. 128D. 21 Frag. in Rhet. 323.22–324.2 = Hdt. 2.30. 22 Frag. in Rhet. 325.20–25 = P. Mathieu and É. Brémond, Isocrate. Discours IV (Paris 1962) 19. 23 Frag. in Rhet. 325.25–326.3. On this passage see Pinto, BollClass 20 (1999) 11–18. 24 Frag. in Rhet. 323.5–6, 328.12–13. 25 Frag. in Rhet. 329.1–2. UGO VALORI 161 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 access to a library. Finally, interest in the logical aspect of the Rhetoric is shown by the inclusion of five logical diagrams.26 Dating the Fragment The only sure termini for dating the Fragment are a quotation from Proklos (5th cent.) and the oldest part of Vat.gr 1340 (late 12th/beginning 13th).27 However, many features of the text suggest dating the composition of the Fragment to the 11th–12th century. First, as mentioned above, Aristotle’s Rhetoric enjoyed very little circulation and success before this age. Second, our author is particularly interested in the Nicomachean Ethics,28 and this work too garnered renewed interest in the 11th–12th centuries, so that after centuries it now became the subject of commentaries, by 26 Frag. 324.27–29; 325.3–4, 16–18; 326.8, 14–15. For the syllogistic and enthymematic diagrams in the Fragment and Stephanos see C. Rambourg, “Les diagrammes syllogistiques des scholies de Stephanos à la Rhétorique d’Aristote,” ClMed 63 (2013) 279–315. On the logical diagrams in the Byzan- tine MSS. of Aristotle in general see M. Cacouros, “Les schémas dans les manuscrits grecs de contenu logique,” Gazette du livre médiéval 39 (2001) 21–33, and “Mise en page et mise en texte, illustration figurative dans les manuscrits byzantins et post-byzantins d’Aristote à exégèse marginale,” in C. Brock- mann et al. (eds.), Griechisch-byzantinische Handschriftenforschung: Traditionen, Ent- wicklungen, neue Wege (Berlin 2020) 311–324; N. Agiotis, “Remarks on the Greek Tradition of Syllogistic Diagrams on Prior Analytics I 4–6,” in Ph. Hegel et al., Wissen und Buchgestalt (Wiesbaden 2022) 41–85. For an overview of the use of diagrams in rhetorical training see V. Valiavictharska, “Diagrams and the Visual-Oral Nexus: What Visuals Reveal about Oral Pedagogy in Argumentation Training,” and C. D’Agostini, “Visualization Strategies in the Corpus Hermogenianum: Preliminary Remarks on Byzantine Rhetorical Diagrams,” both forthcoming in A. Pizzone et al. (eds.), Reading the Corpus Hermogenianum in the Middle Byzantine Period. 27 The second codicological unit of the Vatican manuscript has been dated to the 12th century by I. Pérez Martín, in Byzantium. Identity no. 8126. The dating to the 13th century is by Á. Ibáñez Chacón, “Un escolio en el Vat.gr 1340 y la transmission de la Καινὴ ἱστορία de Tolomeo Queno,” RHT 17 (2022) 145–170. 28 Frag. in Rhet. 327.1–15, 329.3–12. 162 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 exegetes such as Eustratios of Nicaea and Michael of Ephesos.29 Third, the influence of Proklos in the interpretation of a passage of Plato’s Republic30 is typical of such a milieu: Michael of Ephesos and Eustratios of Nicaea had a keen interest in Proklos’ philosophy despite his controversial reputation, flagged as a threat to Orthodoxy in the 11th and 12th centuries in particular.31 Also Isaak Komnenos (who has been identified alternatively as Anna Komnene’s uncle or brother) showed a certain interest in Proklos’ philosophy with his Christianised paraphrase of Proklos’ treatises On Providence.32 Finally, the tendency to discuss different readings found in the manuscripts is typical of some 11th–12th century exegetes, who revived a habit of the Neo- platonic commentators.33 29 For an overview of the Byzantine reception of the Aristotelian ethics see L. Benakis, “Aristotelian Ethics in Byzantium,” in Ch. Barber et al. (eds.), Medieval Greek Commentaries on the Nicomachean Ethics (Leiden 2009) 63–69. 30 Frag. 325.12–15, 328.30–36, 329.13–22. The author depends on Procl. In Resp. II 111.23–112.25 Kroll, as he himself says: φησὶν ὁ Πρόκλος (329.20). 31 See M. Trizio, “Eleventh- to Twelfth-Century Byzantium,” in S. Gersh (ed.), Interpreting Proclus. From Antiquity to the Renaissance (Cambridge 2014) 182– 215. Cf. C. Steel, “Neoplatonic Sources in the Commentaries on the Nico- machean Ethics by Eustratius and Michael of Ephesus,” Bulletin de Philosophie Médiévale 44 (2002) 51–57; M. Trizio, “Neoplatonic Source-Material in Eustratios of Nicaea’s Commentary on Book VI of the Nicomachean Ethics,” in C. Barber et al. (eds.), Medieval Greek Commentaries on the Nicomachean Ethics (Leiden 2009) 71–109. 32 Isaak’s metaphrasis is printed in D. Isaac, Proclus. Trois études sur la Providence I–III (Paris 1977–1982), as an appendix to William of Moerbeke’s translation of the same treatise. The MSS. signal a certain Isaak Sebastokrator as author of the work; scholars have identified him as Alexios I Komnenos’ brother or son. Brother: J. Opsomer and C. Steel, On the Existence of Evils (London 2014) 7, 48–49; B. Strobel, Proklos. Tria opuscula (Berlin 2014) 37–42. Son: D. Isaac, Dix problèmes concernant la Providence (Paris 1977) 25–28. In a forthcoming article Aglae Pizzone will argue for dating to the early 12th cent., hence attribution to Alexios’ son. In any case, the author of this text must have belonged to the Komnenian age and family. 33 This practice can be found, for example, in Eustratios of Nicaea (Trizio, UGO VALORI 163 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 Identifying the author In the Vatican manuscript our fragment is followed by an al- most complete commentary, ascribed by the heading to a certain Stephanos. He refers to the Byzantines’ preparations for a pos- sible Venetian attack on the island of Corfu, which took place in the autumn of 1122. This chronological anchor, together with the name of the author, his professional activity as a teacher, and the Constantinopolitan setting of the commentary, enabled Wanda Wolska-Conus to identify him as Stephanos Skylitzes.34 The identification is further strengthened by the fact, cited by Wolska-Conus, that Theodore Prodromos’ monody in honor of Skylitzes refers to his interest in rhetoric and argumentation.35 Skylitzes was a teacher of grammar in the Saint Paul School in Constantinople, where he later became the director, inherit- ing this charge from his elder brother. When he was only thirty (probably in 1126) he was appointed metropolitan of Trebizond, but he could reach that city only fourteen years later because of a revolt in the region. He died of malaria a few years later.36 Stephanos Skylitzes’ Commentary shares many features with our Fragment in terms of structure, content, and style. Like the Fragment, it consists of a set of autonomous and unconnected glosses. Even more importantly, at the end of the commentary the final section of glosses shows a compositional strategy com- parable to that of the Fragment. As the commentary on Book 1 ends, in fact, the commentator comes back to Rhet. 1363a16 and adds glosses on passages found here and there in Book 1 (Steph. In Rhet. 264.34-296.36). As far as I understand from the edited Byzantine commentaries on Aristotle that I was able to consult, in Byzantine Commentaries on Ancient Greek Texts 73) and in the two major commentaries on Rhetoric (Kassel, Der Text 87–88), both datable to the 12th century. 34 Wolska-Conus, in Actes 599–606. 35 Theodoros Prodromos Or. 9.75–80 De Coul. 36 A sketch of his biography is offered by M. De Coul, Theodori Prodromi epistulae et orationes (Turnhout 2023) xxxvi–xxxiix. 164 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 this feature can be found only in these two texts. Like the Fragment, Skylitzes’ Commentary is rich in quota- tions from ancient authors, and sometimes comments on the quotations included in Aristotle’s Rhetoric, supplementing them with another portion of the text, or reconstructing the context from which the quotation is drawn, e.g. a passage from Isocrates (Pan. 1) and four verses of Euripides (Hipp. 986-989).37 Like the author of the Fragment, Skylitzes quotes some material gen- erally used for the interpretation of ancient literary works, such as scholia and lexica.38 In sum, as argued by Wolska-Conus,39 Stephanos Skylitzes is overall a rhetorician (or a grammarian, I would say), and this attitude is particularly consistent with that of the Fragment. Nevertheless, alongside his attention to these aspects, there is also a certain interest in logical argumentation, as can be seen, for example, in the use of the syllogistic diagrams, which are present in both Stephanos and the Fragment.40 Moreover, the author of the Fragment shows a profound knowl- edge of Aristotle’s Nicomachean Ethics,41 and Stephanos Skylitzes explicitly states that he has written some scholia on this work (τὰ ἐν τοῖς ἠθικοῖς µου σχόλια, Steph. In Rhet. 277.28). The two commentaries show analogies as regards language and style as well. At least three expressions found in Stephanos and the Fragment are almost exclusive to these two texts, among all the edited 11th- and 12th-century commentaries on Aristotle:42 37 At Steph. In Rhet. 319.22–26 and 301.11–16. 38 He quotes a scholion to Pindar’s Pythian Odes (269.7), the Suda lexicon (285.18), and a lexicon of Homer (306.29). 39 Wolska-Conus, in Actes 605: “Il est aussi, et avant tout, un rhéteur.” 40 On the centrality of logical theory in Stephanos' Commentary see Vogiatzi, Byzantine Commentaries on Aristotle’s Rhetoric 4–12. It should be noted that these two aspects of Skylitzes’ personality are not so far apart: in Pro- dromos’ monody (Or. 9), he is portrayed both as a rhetorician skilled in the use of enthymemes and as a grammarian dedicated to schedography. 41 Frag. in Rhet. 327.1–15; 329.3–12. 42 I consider all the works attributed in the TLG to Michael of Ephesos, UGO VALORI 165 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 ὡς … ἔµαθες, appears twice by Stephanos, once in the Fragment, and once in Leo Magentenos;43 ἀνώτερον εἴρηται, three times in Stephanos’ Commentary and once in the Fragment;44 πλέκ(ειν) συλλογισµόν, twice in the Fragment, once in Stephanos, and once in Eustratios of Nicaea’s commentary on Eth.Nic. 1.45 Lexi- calizations in metacommunicative strategies and specialized exegetical vocabulary like these often characterize a commenta- tor’s style. Of this kind (though more numerous owing to the greater length of the extant text) are, for instance, the charac- teristic expressions of Michael of Ephesos’ style, as listed by Concetta Luna.46 Finally, it is remarkable that the two commentaries are com- plementary: Stephanos Skylitzes stops with the analysis of Rhet. 1410a33, whereas the Fragment starts with Rhet. 1416a28, up to the very end of the treatise. Common features, same manuscript tradition, and almost perfect complementarity: these things point to the conclusion that Stephanos’ Commentary and the Fragment were not only written by the same author but are also part of the very same work. In other words, the Fragment was originally designed to be the last section of Stephanos’ Commentary. The current set- up of the text is probably due to displacement of a quire in a manuscript. Displacements of quires were quite common in medieval manuscripts because of dire repairs, damage, or the absence of binding. In fact, being unbound was nothing remarkable for Eustratios of Nicaea, Stephanos Skylitzes, and Leo Magentenos, as well as the anonymous commentaries on Nicomachean Ethics VII and Rhetoric. 43 Steph. In Rhet. 282.9, 298.8, 318.22; Frag. in Rhet. 325.13; Magent. In Top. 1.99 Kotzabassi. 44 Steph. Ιn Rhet. 274.2, 315.24, 315.29; Frag. in Rhet. 325.23. 45 Steph. Ιn Rhet. 312.2; Frag. in Rhet. 325.2, 326.7; Eustrat. In Eth.Nic., CAG XX 43.2. 46 C. Luna, Trois études sur la tradition des commentaires anciens à la Mètaphysique d’Aristote (Leiden 2001) 197–212. 166 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 Byzantine books: quires could be left in their loose form even for many years, especially in secular and scholarly books47. First and last quires, moreover, were more fragile and particularly prone to this kind of accident. Something similar may have happened to Stephanos’ Commentary: the Fragment occupies about seven pages of the Vatican manuscript (ff. 163–166) and could easily have been the last quire of a codex containing the whole text. As mentioned above, all the extant manuscript tradition of Byzantine commentaries on the Rhetoric comes from the Vat.gr. 1340. But in this manuscript there is no trace of displacements of quires and, what is more, there is no codicological caesura between the commentaries: Stephanos’ Commentary begins on the same page on which the Fragment ends (f. 166). The two texts, moreover, are clearly divided by a decorated band and a heading. Therefore, displacement of a quire in the Vatican man- uscript itself is impossible: one must postulate that the displace- ment happened in an antigraph of the Vaticanus that preserved the commentary in its complete form. The last quire of the man- uscript, which would have held Stephanos’ Commentary on Rhet. 1416a 28–1420b 2, could have fallen out (in the case of a bound manuscript) and been relocated by someone at the begin- ning of the codex. This procedure was not uncommon: some- thing similar can be observed, for example, in Paris.gr. 1897 (Dik- tyon 51523), whose last quire fell out and a copy of it was replaced at the beginning of the codex by Nicephorus Gregoras.48 Thus, the scribe of the Vat.gr. 1340 copied the text in the altered order that he found in his model. He separated the two commentaries and created what we now refer to as “Fragment of Commentary on Aristotle’s Rhetoric.” Finally, the last folios of the Vatican manuscript itself fell out. These last folios contained 47 D. Bianconi, Cura et studio. Il restauro del libro a Bisanzio (Rome 2018) 93-111. 48 D. Bianconi, “Le pietre e il ponte ovvero identificazione di mani e storia della cultura,” Bizantinistica 8 (2006) 135–181, at 174–181, and Cura et studio 88–89. UGO VALORI 167 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 the section of Stephanos’ Commentary dealing with Rhet. 1410a 33–1416a 28, which is in fact lost. Curiously, the opening pages of Stephanos’ Commentary are lost as well,49 and this loss cannot have occurred in the Vatican manuscript itself, again for codicological reasons. Therefore, an antigraph of the Vaticanus should also be responsible for the loss of a quire or some folios at the beginning of the commentary.50 Stephanos’ authorial practices51 The textual shift involving the Fragment is not the only structural anomaly present in Stephanos’ Commentary: upon close reading, the reader is struck by a general impression of disorder, which points to a specific typology of work. Given the general outlook and layout of the Commentary, I would argue that this exegetical work has been transmitted in the form of a provisional and informal draft, probably intended to circulate only in the author’s milieu. Stephanos’ Commentary shows many alterations in the order of the glosses, the final sections on Rhetoric Books 1 and 3 being the most telling clue,52 but there are also smaller regressions back to sections of the treatise that Stephanos has already commented 49 Stephanos’ Commentary starts with a note on Rhet. 1.2, 1357a22. 50 Even more curiously, it is worth noting that both the Anonymous and Stephanos are missing their initial sections (albeit for different reasons). Consequently, we do not possess any Byzantine commentary on the opening pages of the Rhetoric. 51 Henceforth I will consider Stephanos’ Commentary in its complete form, therefore including the Fragment. 52 Steph. In Rhet. 294.34–296.36; Frag. in Rhet. 327.34–329.22. As noted above, here the commentator, after reaching the end of the book, returns to sections of the Aristotelian treatise he has already commented on. He sometimes signals this with the expressions τὸ δ’ ἀληθέστατον τῆς ἱστορίας γραφήσεται κάτω ἐν τῷ τέλει τοῦ λόγου (In Rhet. 285.32), περὶ οὗ ὁδῷ προϊόντες ἐροῦµεν (Frag. in Rhet. 324.14–15), or ὄπισθεν πρὸ δύο ἴσως τετραδίων (Frag. in Rhet. 327.34). However, these expressions are likely to have been added at a later stage. In some MSS. the relevant sections are also signaled with small blank spaces and division signs. 168 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 on.53 The impression of a disorderly work is further reinforced by some of glosses going back to the same passage for a second time, adding information drawn from further bibliographical material.54 A paradigmatic example is provided by the two glosses on Rhet. 1365a30. Here Aristotle quotes a hemistich from the Odyssey (22.347): καὶ ὁ ποιητής φησιν “αὐτοδίδακτος δ’ εἰµί.” At first (276.23–24) Stephanos erroneously ascribes the quotation to Pindar.55 But only two pages later (278.22–32) he returns to the same passage, and here Stephanos (1) quotes the lines from Olympian Odes (2.86–88) to which he was referring before, to- gether with an exegesis drawn almost verbatim from the scholia to Pindar;56 (2) gives the reference to the Odyssey, thus correcting his previous attribution;57 (3) quotes another passage from the scholia to the Olympian Odes, without any apparent reason.58 Clearly, there is here a chaotic accumulation of conflicting material that would be impossible to find in a ‘fair copy’ of a commentary. Scholars have advanced a few hypotheses on the reason be- hind this textual dynamic. Otmar Schissel suggests that these 53 Steph. In Rhet. 269.7–12, 278.22–32, 306.16–36, 315.28–316.19, 319.1– 4; Frag. in Rhet. 325.19–326.3. 54 In addition to the passage analyzed below, one can consider the glosses on Rhet. 1373a23 (Steph. In Rhet. 285.18–34, 295.27–296.21); Rhet. 1409a17 (In Rhet. 318.19–20, 319.1–4); Rhet. 1417a13 (Frag. in Rhet. 324.12–15, 328.30–36, 329.13–22). 55 Ὁ ποιητὴς Πίνδαρος ἦν αὐτοδίδακτος, ἤτοι “ἐκ γενετῆς εἰµι σοφός,” φησί, “καὶ οὐκ ἀπὸ διδασκάλου τοιοῦτος γέγονα.” 56 Steph. In Rhet. 278.27–28: λέγει οὖν “ὥσπερ κόρακες πρὸς ἀετὸν φιλονει- κοῦσιν οἱ διδακτοὶ πρὸς τοὺς φύσει αὐτοδιδάκτους.” αἰνίττεται δὲ εἴς τε ἑαυτὸν τοῦτο καὶ εἰς Σιµωνίδην = schol. Pind. Ol. 2.158c Drachmann. 57 In Rhet. 278–29–30: τὸ δὲ “αὐτοδίδακτος δ’ εἰµι” ἐν τῷ χʹ Ὀδυσσείας Ὅµη- ρος λέγει περὶ τοῦ κιθαρῳδοῦ Φηµίου παρακαλοῦντος Ὀδυσσέα µὴ κτανθῆναι. 58 In Rhet. 278.31–32: τέως δ’ οὖν τὰ εἰρηµένα εἶπεν ὁ Πίνδαρος πρὸς Βακχυλίδην τάχα ἀποτεινόµενος τρόπον τινὰ ἀνταγωνιστὴν αὐτοῦ γεγονότα = schol. Pind. Ol. 2.154c. UGO VALORI 169 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 glosses were added by another exegete, probably a pupil of the primary author.59 On the other hand, Wanda Wolska-Conus interprets these sections as later additions redacted by Stephanos himself, and resulting from further research.60 More recently, Camille Rambourg describes Stephanos’ Commentary as a compilation of older exegetical texts, arguing that the ‘turning back’ in the order of the glosses is due to a change of model.61 In my opinion, a close reading of the content of the displaced glosses supports Wolska-Conus’ conclusion. Both the displaced glosses in the commentary and those in the final sections on Books 1 and 3 are all drawn from other sources, namely ancient texts, other copies of the Rhetoric, scholia, lexica etc., corrob- orating the idea that they derive from further research.62 59 Schissel, RE 3A (1929) 2364–2369. He hypothesizes Stephanos to be the pupil and Michael Italikos the primary author. Schissel’s article preceded Wolska-Conus’ identification of the Στέφανος of the heading as Stephanos Skylitzes. 60 W. Wolska-Conus, in Actes 605 n.55. 61 C. Rambourg, in Commenting on Aristotle’s Rhetoric 15 n.1. 62 Steph. In Rhet. 278.22–32 < Pind. Ol. 2.86–88; schol. Pind. Ol. 2.154c, 2.158c. In Rhet. 294.34–295.20 < another exegete (µάταιος δὲ ἧν ὁ οὔτως αὐτὸ ἐξηγησάµενος, ὄτι […]); schol. Pind. Ol. 78c, 82a; Pind. Ol. 13.55–56. In Rhet. 295.21–26 < Apollod. Bibl. 1.7.10; schol. D Il. 9.563. In Rhet. 295.27–296.21 < probably some lexica of Attic Greek; schol. Pind. Ol. 29d, 15a–d, 16a–c. In Rhet. 296.22–30 < Plut. Sol. 10.1–6. In Rhet. 296.31–36 < probably a juridical lexicon. In Rhet. 306.16–23 < schol. A Il. 3.144. In Rhet. 306.24–30 < a lexicon of Homer (οὕτως ἐν τῷ γάµµα τὸ λεξικὸν τοῦ Ὀµήρου λέγει) (= schol. D Il. 3– 243). In Rhet. 306.31–36 < Od. 5.334; Pind. Ol. 2.30. In Rhet. 315.28–316.13 < Hermog. De id. 1.3. Frag. in Rhet. 325.19–326.3 < hyp. in Isoc. Phil. 230 Mandilaras and maybe a lost hypothesis of Isoc. Antid. (cf. P. M. Pinto, “I ‘parenti’ di Isocrate. Sullo scambio, 4 e un commento anonimo alla Retorica di Aristotele,” BollClass 20 [1999] 11–18) or Isoc. Antid. 1–14. Frag. in Rhet. 327.26–328.11 < Choeril. fr.2 Bernabé (= 1 Naeke). Frag. in Rhet. 328.12–29 < Soph. Ant. 223–236. Frag. in Rhet. 328.30–36 < Pl. Resp. 614B. Frag. 329.1– 2 < some MSS. of the Rhetoric presenting a different reading (ἔν τισιν τῶν ἀντιγράφων οὐ κεῖται). Frag. in Rhet. 329.3–12 < Arist. Eth.Nic. 1112a18–b13, 1113a9–11, 1112b21–24. Frag. in Rhet. 329.13–22 < Procl. In Resp. II 111.6– 112.25 Kroll. 170 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 Something similar can be seen in the autograph exemplars of Eustathios of Thessalonike’s commentaries on Homer. These manuscripts show several marginal notes and little sheets of paper added by Eustathios himself that contain additional material drawn from new sources.63 In our case too, the Frag- ment provides evidence supporting the hypothesis that the displaced glosses were written by Stephanos himself, and not by another compiler as argued by Schissel. As said above, a single passage of Aristotle’s Rhetoric is com- mented on by three glosses. At Rhet. 1417a13–14 Aristotle mentions “Alcinous’ tale made in sixty verses for Penelope” (ὁ Ἀλκίνου ἀπόλογος, ὃς πρὸς τὴν Πηνελόπην ἐν ἑξήκοντα ἔπεσιν πε- ποίηται). Aristotle refers to Od. 264–285, 310–343, where Odys- seus recounts his adventures (the same stories that he had previously narrated to Alcinous) to Penelope, in a few lines. In commenting on Aristotle’s reference for the first time, Stephanos notes that Alcinous’ tale is mentioned in the tenth book of Plato’s Republic. Clearly, Stephanos has misread Aristotle’s mention and, what is more, is reminding readers of a page of the Republic (10.614B) where the tale told to Alcinous is mentioned very cursorily: Ἀλλ’ οὐ µέντοι σοι, ἦν δ’ ἐγώ, Ἀλκίνου γε ἀπόλογον ἐρῶ, ἀλλ’ ἀλκίµου µὲν ἀνδρός, Ἠρὸς τοῦ Ἀρµενίου. This detail can be understood only by reading the other glosses that are devoted to the same passage. The second gloss (328.30– 36) is a quotation of the relevant passage of Plato’s Republic, whereas in the third (329.13–22) Stephanos reveals his actual source: he explicitly mentions Proklos’ Commentary on the Republic, where we find nearly two pages in the modern edition (II 111.6–2.112.25 Kroll) devoted to the Platonic pun Ἀλκίνου/ ἀλκίµου. Clearly, the reference to the Republic in relation to the story of Alcinous was due to a memory of Proklos’ Commentary as 63 M. van der Valk, Eustathii archiepiscopi Thessalonicensis Commentarii I (Leiden 1971) xii–xviii; E. Cullhed, “The Autograph Manuscripts Containing Eu- stathius’ Commentary on the Odyssey,” Mnemosyne 65 (2012) 445–461. UGO VALORI 171 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 quoted in the third gloss, and this reminiscence operates from the very first gloss. However, the first gloss has a connection to Proklos that is too implicit and cursory, such that it could not have been grasped by any compiler. The most reasonable ex- planation is that Stephanos himself is the author of all three glosses. He simply persisted in his mistake, revealing his source at the end—probably when he had the opportunity to obtain a copy of Proklos’ Commentary on the Republic. Since Stephanos himself is the author of the displaced glosses, he probably put them in a different place in the commentary out of practical reasons (such as lack of space) without producing a fair copy afterwards. In other words, Stephanos’ Commentary was not a published work but an informal draft—containing material useful for a course64 or a later publication—which the author has not yet reorganized. Other structural anomalies cor- roborate this conclusion: for instance, a sort of recapitulation of Book 1 is placed at the beginning of the commentary on Book 2 of the Rhetoric (Steph. In Rhet. 297.1–298.5), whereas the same does not apply to the beginning of Book 3. Many studies in Greek palaeography and Byzantine philology, in fact, have shown that informal drafts of Byzantine works (especially scholarly and exegetical works) circulated within an inner circle of fellows and pupils.65 At that stage, they were not complete and could still be susceptible to further corrections and improvements. This precariousness was often testified to by the supports used: generally margins of manuscripts, loose quires, single sheets, or scrolls. We have mentioned Eusthathios of Thessalonike’s compo- sitional practices, but many other Byzantine authors exhibit 64 Both Kennedy, Greek Rhetoric 320, and Vogiatzi, Byzantine Commentaries on Aristotle’s Rhetoric 8, believe Stephanos' Commentary to have a teaching aim because of its “personal tone.” 65 In addition to the studies cited below, see J. Irigoin, “Éditions et ré- éditions d’auteur au début du XIIIe siècle. À propos de l’Histoire de Nicétas Choniatès, REG 91 (1978) 571–574; L. Canfora, in N. Bianchi et al. (eds.), Fozio. Bibliotheca (Pisa 2019) xiv–xx. 172 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 similar strategies. Limiting ourselves to the 11th–12th centuries, Michael Psellos (PG 122.1029A) talks about his habit of writing drafts (σχέδια) on rolled sheets of parchment (εἰλητάρια) that he used to distribute to his fellows and pupils.66 Such practices were common also for Michael of Ephesos, a very prolific Aristotelian commentator67 whose teaching or per- sonal material—probably written in the margins of his own manuscripts—has been copied and transmitted in the form of marginal notes. That is the case for: the exegesis of a passage of Nicomachus of Gerasa’s Introductio arithmetica;68 four notes on Asklepios’ commentary on Aristotle’s Metaphysics Z;69 a huge group of notes on Aristotle’s De interpretatione.70 Another important 12th-century writer shows the same habit. In some notes on his own work, John Tzetzes not only mentions some texts τὰς ἔτι ἀµυδρὰς οὔσας ἐν σχεδαρίοις (schol. Ar. Ran. 843b) but also seems to be very concerned with the possible theft of such a preparatory material that was to circulate only among 66 For an analysis of this passage see B. Atsalos, La terminologie du livre- manuscrit à l’époque byzantine (Thessalonike 1971) 168–170. It has also been argued that some material in Oxford, Barocci 131, ff. 397v–446v (Diktyon 47418) is actually a collection of “reading notes made by Psellos from Lydos and other sources when preparing his lessons on natural philosophy”: I. Pérez Martín, “The Transmission of Michael Psellos’ Writings on Meteorologica,” in J. S. Codoñer et al. (eds.), Textual Transmission in Byzantium: Between Textual Criticism and Quellenforschung (Leuven 2014) 291–307, esp. 300. Text: I. N. Pontikos, Anonymou philosophika symmeikta = Anonymi miscellanea philosophica: A Miscellany in the Tradition of Michael Psellos (Athens 1992). 67 See P. Golitsis, “Michel d’Éphèse,” Dictionnaire de philosophes antiques VII (1989) 609–616. 68 F. Acerbi and B. Vitrac, “Les mathématique de Michel d’Éphèse,” REByz 80 (2022) 229–255, at 242–253. 69 P. Golitsis, “Trois annotations de manuscrits aristotéliciens au XIIᵉ siècle,” in D. Bianconi, Storia della scrittura e altre storie (Rome 2014) 33–52, at 34–37. 70 The notes are in Paris.gr. 1917 (Diktyon 51544). Michele Trizio is pre- paring an article on them; I thank him for sharing his research. UGO VALORI 173 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 a small group of trusted people.71 Moreover, a few passages of Tzetzes’ scholia on ancient works testify to the interaction between notes made in the margins of manuscripts, Tzetzes’ teaching, and his ‘published’ works.72 To sum up, there was a phase in the transmission of (at least some) Byzantine literary works73 that included circulation among a small group and quite fluid form, from the point of view of both composition and material arrangement. Sometimes the material was reused as a basis for finalized works designed to have a larger circulation, sometimes it kept its informal set-up. Given the structure of Stephanos’ Commentary, it can be ar- gued that this work, as we read it today, never moved beyond that stage. Some conclusions To review the hypothesis presented here within an organic, even if still provisional, framework: What we now know under the labels ‘Stephanos’ Commentary’ and ‘Fragment of Com- mentary on Aristotle’s Rhetoric’ were originally a single text, con- sisting of personal or teaching notes by Stephanos Skylitzes, a 12th-century teacher in the Saint Paul School in Constantinople. This ephemeral and unrefined text was designed for a limited circulation, maybe among Stephanos’ pupils, and was probably written on a working copy composed of single sheets or quires, 71 Many instances in this sense have been gathered by A. Pizzone, “Bureaucratic Discourse, Signature and Authorship in John Tzetzes: A Com- parative Perspective,” Acme 73 (2021) 43–66. 72 A. Pizzone, “Cultural Appropriation and the Performance of Exegesis in John Tzetzes’ Scholia on Aristophanes,” in Byzantine Commentaries on Ancient Greek Texts 100–129. 73 This was not unique to the Byzantine era: it can be traced, in instance, in Antique literary composition, as shown in T. Dorandi, L’officina dei classici. Come lavoravano gli autori antichi (Rome 2007) 29–46 and 65–82. Such a com- positional method could cause textual displacements also in Antiquity, as argued by L. Canfora, “Traslocazioni testuali in testi greci e latini,” in E. Flores (ed.), La critica testuale greco-latina oggi. Metodi e problemi (Rome 1982) 299– 315. 174 UNIFYING FRAGMENTS OF BYZANTINE SCHOLARSHIP ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 as well as in the margins of a copy of the Rhetoric. Such a mode of composition was common to many authors of the same period. Stephanos’ Commentary and the Fragment bear the marks of some revisions and second thoughts, namely displaced and sometimes contradictory glosses scattered in the text or gathered in the last section of commentary on Rhetoric Books 1 and 3. Stephanos himself produced these revisions using bibliographi- cal material that he accessed subsequently. Someone, before or after Stephanos’ death, had access to his (or his circle’s) private notes and copied them in the form of a book. The textual displacement involving the Fragment may have occurred before or after this moment: in Stephanos’ working copy itself, or in a copy of that. The loss of the first part of Stephanos’ Commentary may have occurred in one of these moments as well: the Commentary as we read it today is incomplete at the beginning too. Finally, a copyist—probably the scribe of Vat.gr. 1340 himself —wrote out the commentary in the altered order that he found in his model, separating the two commentaries definitively. In the Vatican manuscript itself, as proved by codicological evidence, the loss of the last folios occurred, thus the loss of Stephanos’ Commentary on Rhet. 1410a33–1416a28, namely until the start of what we have known as ‘Fragment of Com- mentary’. These conclusions regarding the origin, nature, and form of the commentary should, in my view, make possible a com- prehensive reconsideration of this text, including a new edition based on modern criteria,74 to replace that of Rabe (1869), so 74 I am preparing such an edition in the framework of my Ph.D. project. My intention is to produce an edition of Stephanos’ Commentary based on a complete collation of the available manuscripts, the identification of the author, and his compositional strategies. UGO VALORI 175 ————— Greek, Roman, and Byzantine Studies 65 (2025) 154–175 that the text will be as complete and clear as possible for the reader.75 December, 2024 Università degli Studi di Milano Dipartimento di studi letterari, filologici e linguistici ugo.valori@unimi.it Syddansk Universitet Institut for Kultur og Sprogvidenskaber valori@sdu.dk 75 This article benefited from my participation in the project ‘A Rhetoric for the Empire: Education, Politics, and Speech-making in the Byzantine Millennium’ (Carlsberg Foundation), whose research environment was par- ticularly inspiring for me. I would first like to thank Pasquale Massimo Pinto, who suggested that I work on the Anonymous Fragment for my Master’s Thesis and guided me in various stages of my research. I am also grateful to Frédérique Woerther for guiding me during the early phases of this work and encouraging me to propose an attribution for the Fragment. Aglae Pizzone, Stefano Martinelli Tempesta, and Michele Trizio read earlier versions of this article and offered valuable and irreplaceable suggestions. Also, the criticisms raised by Inmacu- lada Pérez Martín regarding a very preliminary draft of the present paper allowed me to improve my arguments. Helpful insights came from conver- sations with Nunzio Bianchi, Stratis Papaioannou, and Baukje van den Berg, and from the suggestions of the anonymous reviewer. To all of them, I extend my gratitude. The responsibility for any errors remains mine alone.