Cyril ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 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/ Cyril’s Glossary as a Source for the Lexicon of Ps.-Zonaras Peter Isépy and Paola Carmela La Barbera To the memory of Mark Naoumides (1931–1977) and Klaus Alpers (1935–2022) HE MOST COMPREHENSIVE and widely disseminated Greek-Byzantine lexicon remains in a regrettable state of neglect, and there is still no satisfactory edition of it. The “Lexicon Tittmannianum” or “Lexicon Ps.-Zonarae,” which has preserved for posterity a large amount of late antique and early Byzantine scholarship, was erroneously ascribed by its sole editor to date, Johannes A. H. Tittmann (1808), on the basis of later annotations (in three rather late manuscripts), to the historian John Zonaras, who lived from the end of the 11th century to the mid-12th (after 1159).1 However, it is sufficiently 1 On Zonaras see A. Kazhdan, ODB (1991) III 2229; W. Treadgold, The Middle Byzantine Historians (Basingstoke 2013) 388–399. On the manuscripts reporting the name of a certain “Ioannes” see I. A. H. Tittmann, Iohannis Zonarae Lexicon I–II (Leipzig 1808) I LXIII–LXX. The codices mentioned by Tittmann are Vind.phil.gr. 32, 154 (in both, annotation of Zonaras’ name by a second hand), 332 (annotation by Hofbibliothekar Sebastian Tengnagel, 1563–1636), and Marc.gr. X 2 (13th cent.; olim Monasterii SS. Iohannis et Pauli LV). The annotation in the latter on f. 1r, Ἰωάννοῦ τοῦ µοναχοῦ, ὡς ἐγὼ νοµίζω, is added by an old, but not the first hand (K. Alpers, Das attizistische Lexikon des Oros [Berlin 1981] 27). However, we could not find such a note in the Marc.Z.gr. 492 (see O. Mazzon, in M. Fanelli [ed.], Cipro nella Biblioteca Marciana di Venezia, Manoscritti, testi e carte, Catalogo della mostra [Venice 2002] 142–144), quoted by Alpers, “Zonarae Lexicon,” RE 10A (1972) 737, and Das attizistische Lexikon 27, during our examination of the codex in situ. T PETER ISÉPY AND PAOLA CARMELA LA BARBERA 247 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 clear from the work of Klaus Alpers and Mark Naoumides that the lexicon was not compiled by Zonaras and has its origins in the first half of the 13th century.2 Date The terminus ante quem for the origin of the lexicon is the subscription of its oldest dated witness, Vat.gr. 10, copied by Georgios Timiostaurites in A.D. 1253 (f. 301r).3 A terminus post quem was inferred by Alpers and Naoumides on the basis of an observation made already by Tittmann himself:4 the lexicon mentions in the context of the entry ἤλεκτρον that the altar in Hagia Sophia was made of a certain fine stone (Tittmann 986): ἤλεκτρον· […] ἀλλότυπον χρύσιον µεµιγµένον ὑέλῳ καὶ λιθείᾳ, οἵας ἦν κατασκευῆς ἡ τῆς ἁγίας Σοφίας τράπεζα. Amber: […] Gold of a different form mixed with crystal glass and precious stone, from which the altar of St. Sophia was made. On a literal reading, the imperfect ἦν suggests that the altar, destroyed in the plundering raid of 1204 as witnessed by Niketas 2 On the Lexicon of Ps.-Zonaras see generally Tittmann, Iohannis Zonarae Lexicon; Alpers, RE 10A (1972) 732–763; M. Naoumides, “The Shorter Version of Pseudo-Zonaras, Lexicon,” in John L. Heller (ed.), Serta Turyniana: Studies in Greek Literature and Palaeography in Honor of Alexander Turyn (Urbana 1974) 436–488; Alpers, Das attizistische Lexikon 3–55, and “Zonarae Lexicon,” Neue Pauly 12.2 (2002) 831. Zonaras seems to be excluded as the author of the lexicon already by the fact that its latest source is Zonaras’ commentary on the Canones, written shortly before his death, so not before A.D. 1159. See Alpers, RE 10A (1972) 737, and H.-G. Beck, Kirche und theologische Literatur im Byzantinischen Reich (Munich 1959) 656. It seems that in more recent times, only I. Gregoriades, “Tracing the Hand of Zonaras in the Lexicon Tittman- nianum,” Ελληνικά 46 (1996) 41–42, has attempted to advance the argument in favour of identifying Zonaras as the compiler of the lexicon. 3 Alpers, RE 10A (1972) 737; and on the codex, Giovanni Mercati and Pio Franchi de’ Cavalieri, Codices Vaticani Graeci I (Rome 1923) 6. 4 Tittmann, Iohannis Zonarae Lexicon I LXXII; Alpers, RE 10A (1972) 736– 737, and Naoumides, in Serta Turyniana 486. 248 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 Choniates,5 no longer existed when this lemma was written. Thus, the author of the lexicon would have completed his work between 1204 and 1253. Tittmann himself (LXXII) confesses, “[se] hac dubitatione gravius commotum,” that it would be reasonable to assume that Zonaras would have employed ἐστὶ in lieu of ἦν had he been the author. Nevertheless, he surprisingly considered the arguments in favour of Zonaras’ authorship— basically only the opinions of Joseph J. Scaliger (1540–1609) and Daniel Nessel (1630–1700) as well as the spurious annotations mentioned above—to be so compelling that he regards ἦν rather as a conjecture of Byzantine scribes, “ab illa urbis calamitate non remot[i] […] quasi dolorem expressur[i].” Even if the use of the imperfect in similar cases could well refer only to the experience of the witnessing author himself and not to the existence of the object in question,6 the interpretation of Alpers and Naoumides is confirmed by the combination of two circumstances. First, the reading ἦν is transmitted unanimously by the manuscripts.7 Second, the lemma also occurs in the older 5 Niketae Choniatae Historia 573.13–17 van Dieten: Τά δ’ ἐπὶ Νεὼ τοῦ Μεγίστου ἠσεβηµένα οὐδ’ ἀκοαῖς εἰσιν εὐπαράδεκτα. ἡ µὲν θυωρὸς τράπεζα, τὸ ἐκ πασῶν τιµίων ὑλῶν σύνθεµα συντετηγµένων πυρὶ καὶ περιχωρησασῶν ἀλλήλαις εἰς ἑνὸς ποικιλοχρόου κάλλους ὑπερβολήν, ἐξαισίου τῷ ὄντι καὶ ἀξιαγάστου παρ’ ἔθνε- σιν ἅπασι, κατετεµαχίσθη καὶ διεµερίσθη τοῖς σκυλευταῖς, “The report of the impious acts perpetrated in the Great Church are unwelcome to the ears. The table of sacrifice, fashioned from every kind of precious material and fused by fire into one whole—blended together into a perfection of one multi- coloured thing of beauty, truly extraordinary and admired by all nations— was broken into pieces and divided among the despoilers” (transl. Harry J. Magoulias, O City of Byzantium, Annals of Niketas Choniates [Detroit 1984] 315). 6 On this phenomenon see R. Kühner and B. Gerth, Ausführliche Grammatik der griechischen Sprache II.1 (Hannover 1904) 145–146; F. Pfister, Die Reisebilder des Herakleides (Vienna 1951) 105. Thanks are due to Aldo Corcella (Basilicata) for drawing our attention to Pfister’s work. 7 Tittmann I LXXII: “Codices, quos inspexi, omnes ita habent.” An examination of the text of several other manuscripts from the 13th–14th centuries available to us and not mentioned by Tittmann has led to the same result: Bonon. 3559 f. 213r, Monac.gr. 308 f. 112r, Ottob.gr. 252 f. 120v, Pal.gr. 46 PETER ISÉPY AND PAOLA CARMELA LA BARBERA 249 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 Suda (composed probably ca. A.D. 1000),8 in the following form (η 200 Adler): ἤλεκτρον· ἀλλότυπον χρύσιον µεµιγµένον ὑέλῳ καὶ λιθίᾳ. οἵας ἐστὶ κατασκευῆς ἡ τῆς ἁγίας Σοφίας τράπεζα. In other words, Tittmann was correct in his assertion that the original reading of this passage is ἐστὶ, and ἦν represents a chronologically conditioned expression of the pain experienced as a result of the destruction of the altar in 1204.9 However, it was not the conjecture of certain scribes, as he assumed, but rather an intervention of the compiler of the lexicon himself. However, one manuscript, Vat.gr. 11 (ff. 1r–104v), yet seems to suggest that the lexicon originated in the 12th century—at least if we follow Giovanni Mercati and Pio Franchi de’ Cavalieri’s dating to the full 12th century, which was accepted by Ada Adler, Gyula Moravcsik, and Julien Leroy.10 It was especially on the basis of the chronological implications of the ‘amber lemma’ that Alpers and Naoumides ruled out the possibility that the Vaticanus was written already in the 12th century.11 This can undoubtedly be confirmed by a palaeographic com- f. 233r, Paris.Coisl. gr. 346 f. 141r, Paris.gr. 2408 f. 111r, Philadelphia MS 2 f. 54r, Urb.gr. 158 f. 152r, and Vat.gr. 875 f. 132v. 8 Kazhdan, ODB III 1930. 9 It is worthy of note that the core of the Suda’s information is derived from the Synagoge η 57 (drawing from Cyril): ἤλεκτρον· ἀλλότυπον χρυσίον µεµιγµέ- νον ὑέλῳ καὶ λιθίᾳ. Therefore, the section οἵας ἐστὶ κατασκευῆς ἡ τῆς ἁγίας Σοφίας τράπεζα appears to be a later addition, not attested before the Suda. This indicates that this addition is likely to be original to the compiler of the Suda or, at the very least, to have been created during the same period. It also supports the hypothesis that the information regarding the altar was estab- lished long before the plundering of 1204. 10 Codices Vaticani 7; cf. A. Adler, “Suidas,” RE 4A (1931) 714 (“Perhaps the oldest manuscript, Vatic. 11 from the 12th century, offers a shorter redaction, where S. is almost not used”); G. Moravcsik, Byzantinoturcica2 I (Berlin 1958) 346; J. Leroy, “L’or dans les manuscrits grecs d’Italie,” StBiz 14–16 (1977) 115–123, at 121 n.4. 11 Alpers, RE 10A (1972) 737; Naoumides, in Serta Turyniana 440. 250 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 parison.12 A confrontation of the writing of Vat.gr. 11 with dated manuscripts suggests that this codex was copied in the second half of the 13th century or even between the end of the century and the beginning of the 14th. The script is reminiscent of the ‘scritture arcaizzanti’ of the early Palaeologan era, as evidenced by examples such as Vat.gr. 2153 (A.D. 1273), Ottob.gr. 381 (1281/ 2), Vat.gr. 191 (1296), etc.13 This means that there is no longer any evidence to suggest that the Lexicon Ps.-Zonarae was com- posed before the first half of the 13th century. Author In addition to the spurious attribution to John Zonaras, other potential author names are proposed in the manuscripts of the Zonaras lexicon.14 A certain (κύρ[ι]ος) Ἀντώνιος (µοναχός or φιλόσοφος) is men- tioned at least in the codices Athos Docheiariou 260 (A.D. 1420), Athos Laurae K 71 (14th cent.), Laur. 5.7 (A.D. 1474),15 Bodl. Barocci 95 (15th cent.),16 Bonon. 3559 (A.D. 1291/2),17 Const. Seragl. G. I.4 (A.D. 1464),18 Dresden Da. 41 (16th cent.),19 Mosc. 12 The authors thank Domenico Surace for sending them a sample specimen from the Vaticanus (f. 59r). 13 See e.g. G. Prato, “Scritture librarie arcaizzanti della prima metà dei Paleologi e i loro modelli,” S&C 3 (1979) 151–193; F. Acerbi and A. Gioffreda, “Manoscritti scientifici della prima età paleologa in scrittura arcaizzante,” Scripta 12 (2019) 9–52. 14 Alpers, Das attizistische Lexikon 22–35, and our research into the manu- scripts; on the attribution to Zonaras see n.1 above. 15 On f. 404r: Τὸν αὐθέντην καὶ συντακτὴν τοῦ ἔργου οὐκ ἐξεγένετο µαθεῖν. Ἡ γὰρ ἐπιγραφὴ τοῦ βιβλίου οὐδενὸς ἐµέµνητο. ἐν δὲ τῶ άω ὑποβλεφάρω ἐνέγραπτο ἐρυτροῖς καὶ τοῦ µέλανος µείζοσι γράµµασιν Ἀντωνίου. Τίνος δὲ καὶ πόθεν καὶ κατὰ τίνα χρόνον, οὐδαµῶς. 16 On f. 1r: λεξικὸν τοῦ ἁγίου κυρίλλου ἀλεξανδρείας‧ συντεθὲν παρὰ τοῦ κυροῦ ἀντωνίου τοῦ φιλοσόφου‧ ἔρανος λέξεων. 17 On f. 1r in marg. manus prima: ἀντωνίου (µον)αχοῦ. 18 On f. 9r: ἀντωνίου µοναχοῦ. 19 Antonii Monachi lexicon: F. Schnorr von Carolsfeld, Katalog der Handschriften der Sächsischen Landesbibliothek zu Dresden (Dresden 1979) 239. PETER ISÉPY AND PAOLA CARMELA LA BARBERA 251 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 Sinod. 503 (Vlad. 485) (A.D. 1537),20 Ottob.gr. 252 (A.D. 1291/2),21 and Urb.gr. 158 (2nd half of 13th cent.).22 The assertion of Richard Reitzenstein that “no author name has any legitimacy”23 was based on the belief that “Antonios Monachos” was written in the codices by second hands—which is certainly not the case. Nevertheless, Heinrich Stein’s24 and Curt Wachsmuth’s25 attempt to identify the mentioned Antonios with the author of the so-called “Melissa”26 does not seem to be a plausible solution: not so much because the only argument would be to agree on a very common name—this Florilegium, at least in the oldest manuscripts, is also attributed to a certain “Antonios Monachos.”27 What weighs much more heavily is that the “Melissa”—regardless of authorship—existed at the latest by the 11th century, at least a century before the com- 20 On f. 439v. Τὸ παρὸν θαυµάσιον βιβλίον τοῦ σοφωτάτου κυροῦ ἀντωνίου: Alpers, Das attizistische Lexikon 28, no. 61; Vladimir (archim.), Систематическое описание рукописей Московской Синодальной (Патриаршей) библиотеки I (Mos- cow 1894) 715. 21 On f. 1r: Συναγωγὴ λέξεων ἀπό τε θεοφίλων ἁγίων π(ατέ)ρων βίβλων καὶ αὐτῆς δήπου τῆς θύραθεν (spatio circiter 6 litterarum interiecto) Ἀντ(ω)ν(ίου) (µον)αχ(οῦ): M. Agati and P. Canart, I manoscritti grammaticali greci del card. Guglielmo Sirleto (Vatican City 2022) 128–131. 22 On f. 1r: (συ)ναγωγὴ (sic) λέξεων … πόνηµα τοῦ κυ(ρίου) Ἀντωνίου: R. Stefec, “Die griechische Bibliothek des Angelo Vadio da Rimini,” Römische historische Mitteilungen 54 (2012) 95–184, at 143–144 n.189. 23 R. Reitzenstein, Geschichte der griechischen Etymologika (Leipzig 1897) 280 n.1. 24 Herodoti Historiae II (Berlin 1871) 479–482. 25 “Ueber das byzantinische Florilegium ‘Parallela’ und seine Quellen,” Studien zu den griechischen Florilegien (Berlin 1882) 90–161, at 109–110. 26 On the “Melissa” see M. Richard, “Florilèges spirituels grecs,” Diction- naire de Spiritualité 5 (1964) 475–512, at 492–494 (“Pseudo-Antoine, Loci com- munes”). 27 Alpers, RE 10A (1972) 737. The title τοῦ ὁσίου πατρὸς ἡµῶν καὶ µακαρίου Ἀντονίου µοναχοῦ is in Patmos 189 (11th cent.), Bodl. Barocci 128 f. 219r (end of 11th cent.), Paris.gr. 914 f. 266r (12th cent.). See also Richard, Dictionnaire de Spiritualité 5 (1964) 494. 252 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 position of the Lexicon Tittmannianum. Nevertheless, it is quite possible that the name Antonios was taken from the “Melissa” or another lexicographical tradition.28 However, in the absence of a sound stemmatic basis, it cannot be completely ruled out, particularly in light of the age of some of the manuscripts, that a certain Antonios was the author of our lexicon. In the case of Laur. 9.27, which erroneously ascribes the lexi- con to a Saint Maximos,29 it seems reasonable to posit influence from the so-called Florilegium of Ps.-Maximos (10th–11th cent.). As Marcel Richard was able to point out, that Florilegium is the basis of the later pseudo-Antonian “Melissa.”30 Dulcis in fundo, the name of Saint Cyril is also incorrectly given as the lexicon’s author in Athos Iviron 76 (13th cent.) and 77 (14th), Bodl. Barocci 95 (15th),31 and Scoraliensis y III 8 (15th).32 There is no doubt that the author of our lexicon must have been a cleric or monk who, in compiling it, took into account the desideratum of theological studies, which did not exist in Byzan- tium stricto sensu.33 Although the name of this clergyman cannot be determined with certainty—Antonios may be a possibility— 28 The so-called Lexicon Ambrosianum (cf. Alpers, RE 10A [1972] 737, 746) is also referred to as ἐπιµερισµοὶ σὺν Θεῷ κατὰ στοιχεῖον: Ἀντωνίου µοναχοῦ in Zabord. 95 f. 1r (end of 13th cent.): Λ. Πολίτης and Μ. Πολίτης, Κατάλογος χειρογράφων Ἱερᾶς Μονῆς Ζάβορδας (Thessaloniki 2012) 63. 29 On f. 1r: ἐκλογὴ διαφόρων λέξεων συντεθεῖσαι κατὰ στοιχεῖον παρὰ τοῦ ἁγίου µαξίµου. 30 On the widespread Florilegium of Pseudo-Maximos see Richard, Diction- naire de Spiritualité 5 (1964) 488–492, and S. Ihm, Ps.-Maximus Confessor. Erste kritische Edition einer Redaktion des sacro-profanen Florilegiums Loci communes (Stutt- gart 2001) IV. 31 See n.16 above. 32 On f. 1r: Λεξικὸν τοῦ ἁγίου Kυρίλλου Ἀλεξανδρείας συντεθὲν παρὰ τοῦ κυροῦ Ἀντωνίου τοῦ φιλοσόφου. Ἔρανος λέξεων: G. De Andrés, Catálogo de los Códices Griegos de la Real Biblioteca de El Escorial II (Madrid 1965) 226–227, no. 331. 33 Alpers, Das attizistische Lexikon 15–16, with reference to H.-G. Beck, “Bildung und Theologie im frühmittelalterlichen Byzanz,” in P. Wirth (ed.), Polychronion: Festschrift Franz Dölger (Heidelberg 1966) 69–81, at 77. PETER ISÉPY AND PAOLA CARMELA LA BARBERA 253 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 the name Nikephoros also seems a quite plausible suggestion: Alpers34 has pointed to the lemma ἀντωνυµία, i.e. on pronouns (Tittmann 185–186), which the author of the lexicon had drawn from Georgios Choiroboskos. In this passage, Choiroboskos introduced his own name as an example: ἀντὶ τοῦ εἰπεῖν Γεώργιος, λέγω ἐγώ.35 However, in the Lexicon Ps.-Zonarae, we no longer find Γεώργιος, but—unanimously in the manuscripts—Νικηφόρος, i.e. ἀντὶ τοῦ εἰπεῖν Νικηφόρος, λέγω ἐγώ. It thus appears reasonable to consider Alpers’ Nikephoros as a valid variant for the name of the author of the lexicon.36 It remains to be seen whether his plausible suggestion that the man in question could have been Nikephoros Blemmydes (1197–1272) will ever be confirmed. Our subsequent results will show at least that the author of the lexicon, like Nikephoros Blemmydes,37 was living during the Latin occupation of the capital, and more precisely, when he compiled his lexicon, in eastern Greece. Manuscripts and recensions Once again, it is Alpers and Naoumides who have done the most work in gathering the extant manuscripts of the Lexicon Ps.-Zonarae. Combining the relevant publications and adding today’s research resources (especially the Pinakes database), the number of manuscripts that contain all or part of the lexicon is an impressive 143. The codices in question date from the 13th century to the 18th and are found in a wide range of collections, from Alexandria, Sinai, the Seraglio, and Mount Athos to Western libraries, in- 34 RE 10A (1972) 738. 35 Th. Gaisford, Georgii Choerobosci Epimerismi in psalmos III (Oxford 1862) 35.2. 36 See also Alpers, Das attizistische Lexikon 11 with n.3. 37 PLP 2897 s.v. Βλεµµύδης Νικηφόρος; J. A. Munitiz, Nikephoros Blemmydes: A Partial Account (Leuven 1988), esp. 14–28. See also C. N. Constantinides, Higher Education in Byzantium in the Thirteenth and Early Fourteenth Centuries (Nicosia 1982) 7–9. 254 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 cluding small ones such as the Pontificio Collegio Greco in Rome (no. 3).38 38 See the detailed list in Alpers, Das attizistische Lexikon 22–35 (129 mss.); Neue Pauly 12.2 (2002) (“ca. 130”); and “Ein Handschriftenfund zum Cyrill- Glossar in der Staats und Universitätsbibliothek Bremen,” in W. Hörandner et al. (eds.), Lexicographica Byzantina (Vienna 1991) 11–52, at 11 with correction in n.1 (“knapp 130”); and the database Pinakes (08/29/2024: 145 mss.). On the witnesses to the shorter version of the lexicon see Naoumides, in Serta Turyniana 437–448, and below. To be excluded or deleted from the list of mss.: Paris.gr. 2655 (A), Atheniensis 1082 (Aa), Bucharest Acad.Rom. 608 (Ab), Meteora M. Metamorph. 498 (Am), Patmos M. Ioannou t. Th. 320 (Ap), Vind. suppl.gr. 147 (Av), and Alexandr. 123 (Naoumides 437–438), which transmit a Cyril text interpolated by our lexicon (Alpers, RE 10A [1972] 758–760, and Das attizistische Lexikon 36), Konstant. Metoch. P. T. 43, Athos M. Pant. 824, Dresden Da. 39, and Marc. X.23, mentioned by Naoumides 487–488 (Alpers, Das attizistische Lexikon 36–37). To be added to Alpers’ list: Uppsal. 16 (Naou- mides 487); Vat.gr. 874 (Alpers, in Lexicographica 11 n.1), and the fragmentary mss. Leiden BPL 484 ff. 19–24, his apographon Leiden Periz. Q 59 ff. 98–100v, and Voss. Misc. 1. In addition, the following codices included in Pinakes have withstood pre- liminary examination via illustrations or the catalogues in question: Athen. Mous. Benaki 115 (3) f. 194r (88.26–89.18 Tittmann); Cantabr. Univ. Add. 1880.6 ff. 1r–9v (M. Di Franco, “Description of Cambridge, University Library, MS Add. 1880.6,” https://cudl.lib.cam.ac.uk/view/MS-ADD- 01880-00006/1; 08/29/2024); Cantabr. Kk. 5.25 ff. 1–262v (Ch. Wright, “Description of Cambridge, University Library, MS Kk.5.25,” https:// cudl.lib.cam.ac.uk/view/MS-KK-00005-00025/12; 08/29/2024); Halle, Hauptbibl. Waisenhauses, W. J. 20 (copy of Monac. 263 with annotated readings from Monac. 308, sent by Ignaz Hardt in March 1808 to Tittmann, who was unfortunately unable to use it in time for his edition; we thank Stefano Valente for his kind advice), Monac. 66 f. 34v.18–30 (731.1–732.13 Tittmann; M. Pradel, Katalog der griechischen Handschriften der Bayerischen Staats- bibliothek München II [Wiesbaden 2013] 18); Paris.gr. 1000 f. 185v (88.26–89.19 Tittmann); Vat.gr. 2222 ff. 318r–348v (S. Lilla, Codices Vaticani Graeci, Codices 2162–2254 [Vatican City 1985] 266–267); Göttingen UB 12 (18th cent.; W. Meyer, Die Handschriften in Göttingen [Berlin 1893] 6); Neap. II D 36 ff. 1–108 (cf. Vat.gr. 1753; Maria R. Formentin, Catalogus Codicum graecorum Bibliothecae nationalis Neapolitanae II [Rome 1995] 41), Marc.gr. VII.38 ff. 156v–157r (133/ 1879 Tittmann; E. Mioni, Bibliothecae Divi Marci Venetiarum codices graeci manu- scripti II [Rome 1960] 71). To be neglected are the codices Paris.gr. 2662 f. PETER ISÉPY AND PAOLA CARMELA LA BARBERA 255 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 From Tittmann’s work onwards, a fundamental division of the manuscripts into a longer and a shorter version of the glossary emerged.39 The long version usually bears the title Συναγωγὴ λέξεων ἐκ διαφόρων βιβλίων συλλεγεῖσα παλαιᾶς τέ φηµι γραφῆς καὶ τῆς νέας καὶ αὐτῆς δήπου θύραθεν vel. sim., which seems to be the original title of the lexicon.40 The short version, at least in some manuscripts, has the heading ἔρανος λέξεων ἐκ τῶν ἔξωθεν φιλοσόφων καὶ τῶν καθ’ ἡµᾶς διδασκάλων,41 hence the term “ἔρανος-version.” Naoumides identified some 57 manuscripts as witnesses to the shorter version, clarified their filiation, and presented the edition of a few selected passages.42 The majority of the manuscripts, however, represent the long version, or vulgata, which was edited by Tittmann on the basis of a mere three manuscripts (A, D, K).43 In 1981 Alpers adopted an exemplary approach, employ- ing both the long and the short version in editing fragments of 17v–32v and Ambr. C 178 inf. 282r–v (Cyril). The following mss. could not be verified: Romanus Vallic. F 14 f. 109–110(?); Marc.gr. XI.24 (collocaz. 1293) f. 88v–95v, and Utrecht UB 10 (mentioned by Naoumides 488). 39 Tittmann I XXX–XXXI; Alpers, RE 10A (1972) 733–734. 40 In Ath. M. Doch. 260, Ath. M. Dion. 224, Ath. Vatop. 416 and 514, Ath. Laur. I 85 and K 71, Bodl. Can. 65, Bonon. 3559, Seragl. G. I. 4, Harl. 5572, Laur. 57.50, 59.38, Laur.Conv.Soppr. 146, Marc.gr. Z. 140 and 492, Monac. 263 and 510, Mutin.gr. 165, Ottob.gr. 252, Paris.Coisl.gr. 178, 902, 1142, 2408, 2619, 2633, 2634, 2664, Suppl.gr. 461, Rom.Coll.gr. 3, Scorial. Ψ.III.16, Sinait. 2103 and 2128, Upps. 16 (<…>? λέξεων), Urb. 158, Vat.gr. 10, 868, 872, 873, 875, and 876, Vind.phil.gr. 77 and 322. 41 E.g. in Ambr. B 81 Suppl., Haun. Add. 280 4°, Vat.gr. 11, Laur. S. Marci 301, Monac. 308, Paris.gr. 1123, Vind. Phil. 77 and 233. 42 Serta Turyniana 437–448; at 482–488, the edition of the letters θ and ξ. 43 Tittmann I XIX–XX. A = Monac. 510 (Augustanus 119; 14th cent.); D = Dresden Da. 37 (today: Moscow Rossijskij gosudarstvennyj arhiv drevnih aktov (РГАДА), Ф. 1607 21; 14th/15th cent.); K = apographon Kulenkampianum (from an unknown manuscript of the 13th cent.; K. Alpers, Theognostos ΠΕΡΙ ΟΡΘΟΓΡΑΦΙΑΣ, Überlieferung, Quellen und Text der Kanones 1–84 [diss. Hamburg 1964] 13–16). We would like to thank Alessandro Musino for sending us copies of Alpers’ dissertation. 256 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 the Atticistic lexicon of Oros (5th cent.) preserved solely in the Lexicon Ps.-Zonarae.44 The genesis of the two versions was seen differently by Naoumides and Alpers: while the former attributed the two versions to the expansion of a smaller lexicon into a larger one,45 the latter has plausibly demonstrated that the existence of these two ‘recensions’ must be explained by two copies of one and the same “Urschrift, [...] ein[es] Text[es] mit zahlreichen Zusätzen an den Rändern.”46 For the additional material included in the vulgata, as opposed to the short version, cannot be explained by the incorporation of new sources alone. Instead, the ‘new ma- terial’ represents an expansion of the existing glosses and parts of glosses drawn from all sources, including those that are distinctive of the Lexicon Ps.-Zonarae and are not found in other lexicons. Moreover, Alpers identified numerous instances where it is evident that the short version represents a text that has been abridged in comparison with the vulgate.47 Glosses from ‘Cyril’s’ lexicon in Ps.-Zonaras In the course of his work, Ps.-Zonaras consulted a con- siderable number of late ancient and Byzantine lexicographical materials and glossaries with the objective of producing a comprehensive reference work. Among these was the lexicon attributed to the Church Father Cyril (375/80–444), patriarch of Alexandria. This extensive alphabetical lexicon (over 20,000 entries), which Paul Maas labelled as “the most gruesome of all dic- tionaries,”48 is the first extant lexicon transmitting pagan, 44 Alpers, Das attizistische Lexikon 149–260. 45 Serta Turyniana 483–486. 46 Das attizistische Lexikon 19. 47 Das attizistische Lexikon 19–20, 36–42. 48 Review of A. B. Drachmann, Die Überlieferung des Cyrillglossars (Copen- hagen 1936): ByzZeit 37 (1937) 379–381, at 380: “Daß die heutige Forschung an diesem greulichsten aller Wörterbücher nicht abgewandten Blickes vor- übergehen darf…” PETER ISÉPY AND PAOLA CARMELA LA BARBERA 257 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 Christian, and Jewish glosses at the same time.49 Due to its broad and complex tradition—it was extremely popular over the centuries and widely copied—the lexicon of ‘Cyril’ remains unpublished to this day. Moreover, although in the 19th and 20th centuries scholars made efforts to trace the history of this lexicon,50 much work remains unprinted. At present, the most updated sources of information on ‘Cyril’ are Anders Bjørn Drachmann’s study from 1936, in which he identifies the re- censions of the lexicon, especially the two major ones, labelled g and v, and Mark Naoumides’ study from 1979, consecrated in particular to the v-recension. Drachmann’s edition of the entire lexicon (based on fewer than 20 manuscripts) is yet to be pub- lished; it is preserved in the archives of the Corpus Lexicographorum Graecorum at the Danish Academy in Copenhagen. Also unpub- lished are important studies by Naoumides on the g-version (with a partial edition of a few letters), preserved in the archives of the University of Illinois at Urbana-Champaign. 49 D. Papanikolaou, “Sacred, Profane, Troublesome, Adventurous: The Lexicon Cyrilli across Ages and Manuscripts,” BRL 96 (2020) 1–18. 50 Drachmann, Die Überlieferung; Naoumides, in Serta Turyniana, and “The v-Recension of St. Cyril’s Lexicon,” ICS 4 (1979) 94–135. See further R. Reitzenstein, “Die Ueberarbeitung des Lexicons des Hesychios,” RhM N.F. 43 (1888) 443–460; G. Wentzel, “Beiträge zur Geschichte der griechischen Lexikographen,” in SBBerl 26 (1895) 477–487; K. Latte, Hesychii Alexandrini Lexicon I (Haunia 1953) XLIV–LI; P. Burguière, “Cyrilliana. Observations sur deux manuscrits parisiens du Lexique de Cyrille,” REA 63 (1961) 345–361, “Cyrilliana (II). Observations sur deux manuscrits parisiens du Lexique de Cyrille,” REA 64 (1962) 95–108, and “Cyrilliana (III). Remarques sur la com- position du Lexique de Cyrille,” REA 72 (1970) 364–384; Alpers, in Lexico- graphica 11–52, and “Zum Bremer ‘Cyrill-Glossar’ (Bremen, SuUB msc 0011): eine philologisch-historische Nachlese,” in Ch. Brockmann et al. (eds.), Griechisch-byzantinische Handschriftenforschung (Berlin 2020) 351–362; S. Lucà, “Il lessico della Ps.-Cirillo (Redazione V1): da Rossano a Messina,” RSBN N.S. 31 (1994) 45–80; A. Corcella, “Le trappole di Cirillo: Hesych. ε 4016 L., π 12 H.,” Eikasmos 28 (2017) 225–234; Papanikolaou, BRL 96 (2020) 1–18; U. Hagedorn, Das sogenannte “Kyrill”-Lexikon in der Fassung der Handschrift E I (2005 (https://kups.ub.uni-koeln.de/1813/1/Hagedorn_Kyrillos_Hauptdatei.pdf; 11/12/2024). 258 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 Rebus sic stantibus, taking full advantage of the unpublished works of Naoumides that we have been able to study,51 our aim here is to determine more precisely, if possible, the Cyrillean sources used by Ps.-Zonaras. In this respect, Naoumides and Alpers have once more made the most significant steps. In his study on the short version of Ps.-Zonaras, Naoumides suggested that the Cyrillean entries in this version of the lexicon “agree […] more particularly with g, v.”52 A few years before, relying on Drachmann’s unpublished edition of the entire Ps.-Cyril Lexicon, Alpers posited that a manuscript belonging to the g-recension constituted the direct model for the Cyrillean glosses in the Lexicon Ps.-Zonarae.53 As evidence for this claim he cited a sample of eleven lemmata according to the edition of Tittmann, nine of which are found in the g-recension only, one in g and Hauniensis 1968 (h),54 and one in both the g- and v-recensions. These glosses are: 1145 καυλῷ. τῷ ἄκρῷ τοῦ δόρατος (Cyr. καυ 3, g) 1152 κάρδοπος. ἔνθα τὰ ἄλευρα µαλάσσονται (Cyr. καρ 30 κ. : καµπτρα, εν ᾑ τα αλεθρα µαλασσουσιν [sic], g) 1327 µαλθακός. ὁ τρυφερός. (Cyr. µαλ 4, g) 1327 µαλακός. ὁ ἀναπεπτωκὼς, ὁ νοσερὸς, ὁ ἀσελγής. (Cyr. µαλ 9, g) 1328 † µαγέβ. νότου ἢ µεσηµβρίας † (Cyr. µαγ 1, g) 1328 µαζουρώθ. ἀστὴρ οὕτω καλούµενος (Cyr. µαζ 3, g) 1330 µαγάς. σανὶς τετράγωνος, ὑπόκυφος, δεχοµένη ἐφ’ ἑαυτῆς τὰς τῆς κιθάρας νευρὰς καὶ ἀποτελοῦσα τὸν φθόγγον. (Cyr. µαγ 2, g) 1335 † µάψαρ. ὀχύρωµα. † (Cyr. µαψ 3, g) 1336 µανείην. χολωθείην. (Cyr. µαν 11, g) 51 This has been made possible in the framework of the PatriarX project. We are especially grateful to Paolo Scattolin, whose foundational contri- butions enabled the achievement of this research. 52 Serta Turyniana 484. 53 RE 10A (1972) 745. Naoumides completed his study “before K. Alpers’ extensive article ‘Zonarae lexicon’ appeared in Pauly Wissowa” (Serta Turyni- ana 486 n.131). 54 Drachmann, Die Überlieferung 15: “der vg-Text lässt sich mitunter aus h verbessern.“ PETER ISÉPY AND PAOLA CARMELA LA BARBERA 259 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 1340 µεµψίµοιρος. ὁ µεµφόµενος τὸ ἀγαθὸν, ὁ φιλεγκλήµων. (Cyr. µεµ 1, gh) 1354 µεθορµίζεται. µετάγεται. (Cyr. µεθ 6, vg) According to Tittmann’s edition, these entries are transmitted by the three witnesses he used (A, D, K).55 An examination of other witnesses to the long version identified by Alpers confirms that the eleven Cyrillean glosses are found throughout the tra- dition of this version.56 Moreover, further sample collations of the most representative witnesses of the redactio brevior57 show that they were also handed down in this tradition. The fact that these Cyrillean glosses are in both the short and long versions of Ps.-Zonaras’ lexicon leads to the important con- clusion that they must have been added by the author himself. Before looking closer at their text, an inquiry into the tradition of the g-recension of Cyril’s lexicon appears necessary. To this end, we rely on the unpublished studies of Naoumides.58 While Drachmann located three manuscripts of this recension, Γ (Crypt. Z.α.V), Λ (Laur.Plut. 59.16) and Δ (Laur.Conv.Soppr. 74), Naou- mides identified ca. 40 witnesses and attempted to clarify their interrelations. In his view, the g-tradition splits into two 55 See n.43 above. 56 Based on Alpers, Das attizistische Lexikon 16–22 (stemma at 21), the tra- dition of the long version splits into two branches: we collated manuscripts from both Vat.gr. 11 on one side and Laur.Plut. 5.7 and 9.27 on the other. 57 According to Naoumides (Serta Turyniana) the tradition of the short ver- sion (which he labels p) splits into two branches a and b (his stemma at 460). We collated manuscripts of both branches: from a, Ω (Rom.Coll.gr. 3) and P (Paris.gr. 2665, the archetype of the p3-group); from b, Α (Paris.gr. 2655, the archetype of the p1-group) and Σ (Monac. 308). 58 In the following, we provide some preliminary information obtained by studying the Naoumides Archives. The folder containing his studies on the g-recension has been brought to light by Paolo Scattolin, who presented his findings at the 24th International Congress of Byzantine Studies (2022: https://byzcongress2022.org/index.html; 12/10/2024). The comprehensive analysis of the g-recension based on Naoumides’ unreleased paper will be published in a forthcoming study of this matter in the frame of the PatriarX Project. 260 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 branches. On the one hand, we find the groups g1 (represented, among others, by Drachmann’s Λ Γ), g2, and Drachmann’s MS. Δ. As a whole, g1, g2, and Δ represent an ‘expanded’ branch, augmented by additions and interpolations, which are absent from the rest of the tradition. On the other hand, we find the groups g3, g4, g5, g6, g7, and g8, conversely transmitting a version smaller in size and closer to the g-archetype. On this basis, we have collated the text of the eleven Cyrillean glosses in the lexicon of Ps.-Zonaras with the most representative witnesses of ‘Cyril’s’ g-recension according to Naoumides’ re- construction. The objective was to potentially identify the spe- cific branch of the ‘Cyril’ tradition to which Ps.-Zonaras had access. We list here those manuscripts identified by Naoumides as belonging to the various g-families and to which we had access.59 A synoptic table shows the occurrence of the eleven glosses in these witnesses. Δ = Laur.Conv.Soppr. 74 ‘Expanded’ version g1 Crypt. Z.α.V (Drachmann’s Γ) Laur.Plut. 59.16 (Drachmann’s Λ) Marc.gr. X.30 Vat.gr. 874 g2 Vat.gr. 867 Laur.Plut. 57.48 – – – g3 Cantabr. O.I. 66 Shorter version g4 Paris.gr. 2617 Paris.gr. 2632 Paris.gr. 2672 59 For each family, we have endeavoured to collate the most representative witnesses according to Naoumides’ reconstruction. The manuscripts that we were unable to examine are: For g1, Vind.Phil.gr. 212, 244, and Patm. 601; for g2, Laur. S. Marco 320 and Lond. Add. 14773; for g3, Leovaerd. 38, Athos Μονή Διονισίου 355, Kozani (Macedonia) Κοβεντάρειος Δηµοτική Βιβλιοθήκη 80; for g5, Haun. GKS 1970; for g7, Haun. GKS 1968; for g8, Paris.gr. 2664A. The Messan. S. Salv. 167 transmits a mixed text: first n, then (from the gloss µήδεα) g7. Cf. Latte, Hesychii I XLVIII. For this reason we have not taken it into account. PETER ISÉPY AND PAOLA CARMELA LA BARBERA 261 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 Paris.gr. 2673 Ambr. G 15 sup. Oxon.Bodl.Clark. 3 g5 Laur.Plut. 58.30 Vind.phil.gr. 319 g6 Vat.Ott.gr. 259 Vat.gr. 344 Ambr. F 1 sup. g7 Matr.Complut. 166 [Z 22] g8 Laur.Plut. 59.49 x = the entry is present in all witnesses. --- = the entry is absent in all witnesses. ~ = the entry is present in certain witnesses only. Ps.-Zonaras (Tittmann) g1 g2 Δ g3 g4 g5 g6 g7 g8 1 1145 καυλῷ· τῷ ἄκρῷ τοῦ δόρατος. (Cyr. καυ 3, g) ~ (om. Marc. gr. X.30) --- x --- x --- --- x x 2 1152 κάρδοπος· ἔνθα τὰ ἄλευρα µαλάσσονται. (Cyr. καρ 30 κ. : καµπτρα, εν ᾑ τα αλεθρα µαλασσουσιν [sic], g) ~ (om. Crypt. Z.α.V & Laur.Plut. 59.16) x x --- x x x --- x 3 1327 µαλθακός· ὁ τρυφερός. (Cyr. µαλ 4, g) x x x x x x x x x 4 1327 µαλακός· ὁ ἀναπεπτωκὼς, ὁ νοσερὸς, ὁ ἀσελγής. (Cyr. µαλ 9, g) x ~ (om. Laur. 57.48) x x x x x x x 5 1328 † µαγέβ· νότου ἢ µεσηµβρίας † (Cyr. µαγ 1, g) x --- x --- x --- --- --- --- 6 1328 µαζουρώθ· ἀστὴρ οὕτω καλούµενος. (Cyr. µαζ 3, g) x --- x --- x --- --- x x 7 1330 µαγάς· σανὶς τετράγωνος, ὑπόκυφος, δεχοµένη ἐφ' ἑαυτῆς τὰς τῆς κιθάρας νευρὰς καὶ ἀποτελοῦσα τὸν φθόγγον. (Cyr. µαγ 2, g) x x x x x x x x x 8 1335 † µάψαρ· ὀχύρωµα.† (Cyr. µαψ 3, g) ~ (om. Marc. gr. X.30) --- x --- x --- --- --- x 262 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 9 1336 µανείην· χολωθείην. (Cyr. µαν 11, g) x --- x x x x x x x 10 1340 µεµψίµοιρος· ὁ µεµφόµενος τὸ ἀγαθὸν, ὁ φιλεγκλήµων. (Cyr. µεµ 1, gh) ~ (om. Crypt. Z.α.V & Laur.Plut. 59.16) x x x x x x x x 11 1354 µεθορµίζεται· µετάγεται. (Cyr. µεθ 6, vg) x x x x x x x x x Our analysis demonstrates that these glosses manifest in a systematic manner exclusively within the entire g4-family, codex Δ, and one witness of g1, Vat.gr. 874. Conversely, only some of them are present in the other families of g and, often, only in part of the witnesses of each group. However, it is possible to take this a step further. Two cases indicate a closer affiliation between Ps.-Zonaras and g4 in oppo- sition to Δ and the Vaticanus. First, in the interpretamentum of entry no. 4 (1327 µαλακός) Ps.-Zonaras follows g4 and Vat.gr. 874 which transmit (λέγεται δὲ καὶ) ὁ ἀσελγής, while Δ shows (λέγεται δὲ καὶ) ἐπὶ τοῦ ἀσελγοῦς. The second case concerns the interpretamentum of no. 7 (1330 µαγάς = ‘bridge of the cithara’), in which Ps.-Zonaras offers correct ὑπόκυφος (‘somewhat humped’) with g4 (three witnesses in particular) against the corruption ὑπόκουφος (‘somewhat light’) found in Δ, the Vaticanus, and the entire rest of the g-tradition. In other words, even if the adjective κοῦφος was well known in com- parison with the rarer κῦφος and might also have spread through multiple independent errors,60 the reading of the ‘g archetype’ must be stemmatically reconstructed as erroneous ὑπόκουφος. Now, without speculating on how g4 arrived at the correct reading ὑπόκυφος, it seems legitimate in this particular situation to regard the agreement in the lectio genuina as conjunctive: in 60 Even three witnesses of the g4 family (Q, S, O: see below) offer the wrong ὑπόκουφος. On the other hand, the rarity of ὑπόκυφος is also shown by the fact that some mss. give other erroneous readings. Among the witnesses of the short version, for instance, Paris.gr. 2665 and Monac.gr. 308 exhibit ὑπό- κρυφος, while Rom.Coll.gr. 3 shows ὑπόκ[.]υφος, with a letter erased before υ. PETER ISÉPY AND PAOLA CARMELA LA BARBERA 263 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 light of the established proximity between g4 and Ps.-Zonaras, the fact that only g4 and the Lexicon Ps.-Zonarae provide the rare correct reading ὑπόκυφος in a sea of erroneous ὑπόκουφος permits us to conclude, with a reasonable degree of certainty, that the ‘Cyril’ glosses entered the lexicon in this specific form from a g4 text. According to Naoumides’ research, the family g4 is repre- sented by six codices:61 Pa: Paris.gr. 2617 (ca. A.D. 1321, Cyprus)62 Pb: Paris.gr. 2632 (copied by Θεόδωρος θύτης, A.D. 1380, Cyprus)63 Q: Ambr. G 15 sup. (14th cent., Greek-oriental)64 R: Paris.gr. 2672 (14th cent., Cyprus)65 S: Paris.gr. 2673 (1301/2, Cyprus)66 O: Bodl. Clark. 3 (copied by the scribe Λέων, 2nd half of 15th cent., Athos)67 A more detailed examination of the relationship between Ps.-Zonaras and the text of the g4 witnesses, with an aim of possibly identifying whether the Lexicon Ps.-Zonarae has closer 61 We adopt the sigla used in Naoumides’ unpublished material. 62 R. Browning and C. Constantinides, Dated Greek Manuscripts from Cyprus to the Year 1570 (Washington 1993) 183–185, pl. 60. 63 RGK II 174. 64 S. Lucà, “Testi medici e tecnico-scientifici del Mezzogiorno Greco,” in G. De Gregorio et al. (eds.), La produzione scritta tecnica e scientifica nel Medioevo: Libro e documento tra scuole e professioni (Spoleto 2012) 551–605, at 576 n.57 (“greco-orientale”). In the 16th century the codex most likely belonged to the Venetian mathematician Matteo Macigni (ca. 1510–1582): C. Giacomelli, “I libri greci di Matteo Macigni. Contributo allo studio di una biblioteca uma- nistica,” PP 74 (2019) 361–420. 65 J. Darrouzès, “Autres manuscrits originaires de Chypre,” REByz 15 (1957) 131–168, at 153; M. Couroupou and P. Géhin, “Nouveaux documents chypriotes,” REByz 59 (2001) 147–164, at 160 n.36. 66 Browning and Constantinides, Dated Greek Manuscripts 132–134, pl. 34 and 182b. 67 RGK I 239. The codex was purchased by Clarke in October 1801 on Patmos: J. Monk and C. Blomfield, “Gaisford’s Account of Clarke’s MSS,” Museum Criticum or Cambridge Classical Researches 1 (1814) 128–132, at 130. 264 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 stemmatic affiliations with one of the g4 members, reveals that there is no further distinctive connection in textu with any of the g4 manuscripts: Ps.-Zonaras (ed. Tittmann): 1145 καυλῷ. τῷ ἄκρῷ τοῦ δόρατος. καυλὸς Q | τοῦ ἄκρου Pa Pb R S : τὸ ἄκρον Q R 1152 κάρδοπος. ἔνθα τὰ ἄλευρα µαλάσσονται κάρποδος Pa Pb Q | ἔνθα τὰ ἄλευρα] κάµπτρα‧ οὗ τὰ ἄλευρα Pa Pb (κάµπρα) S : κάµπτρα‧ ἡ τοῦ ἀλεύρου R | µαλάσσονται] µαλάσσουσιν codd. 1327 µαλθακός. ὁ τρυφερός. µαλφακός Q | ὁ om. codd. | τρυφερός] τρυφερός‧ εὐµάλακτος‧ ἐκλελυµένος Pa Pb : τρυφερός‧ εὐµάλακτος‧ ἀπαλός O R S 1327 µαλακός. ὁ ἀναπεπτωκὼς, ὁ νοσερὸς, ὁ ἀσελγής. ὁ1+2 om. codd. | ἀναπεπτωκὼς] ἀναπεπτωκός Pa Pb R S | νοσερός] νοσσερός R | ὁ ἀσελγής : λέγεται δὲ καὶ ὁ (ὁ om. Pb) ἀσελγής(-ῆς R, -είς Pa[s.l. -η-] Pb) O Pa Pb Q R S 1328 † µαγέβ· νότου ἢ µεσηµβρίας. † µαγέβ] Rp.c., non leg. Ra.c. : µαγεοῦς Pb | ἢ] ἠ S | µεσηµβρίας Pa : µεσηµβρίαν Pb S : µεσηµβρία Rp.c., non leg. Ra.c. 1328 µαζουρώθ· ἀστὴρ οὕτω καλοῦµενος. οὕτω] ὁέτο Pa : ὁεύτω Pbut vid. 1330 µαγάς. σανὶς τετράγωνος, ὑπόκυφος, δεχοµένη ἐφ’ ἑαυτῆς τὰς τῆς κιθάρας νευρὰς καὶ ἀποτελοῦσα τὸν φθόγγον. σανὶς] ῥανὶς Pb | (τετράγονος Pa Pb) | ὑπόκoυφος Q S O | τὰς om. R S | τῆς om. R | νευρᾶς Pb : om. R 1335 † µάψαρ. ὀχύρωµα. † ὀχύρωµα] ὀχύροµα Pb 1336 µανείην. χολωθείην. µανείειν Ra.c. | χολοθείην Pb R S O Q 1340 µεµψίµοιρος. ὁ µεµφόµενος τὸ ἀγαθὸν, ὁ φιλεγκλήµων. µεµψίµοιρος. φιλεγκλήµων, µεµφόµενος τὸ ἀγαθὸν codd. 1354 µεθορµίζεται. µετάγεται. µετάγεται‧ µεταβάλλεται‧ codd. In light of this evidence and the findings of Naoumides con- cerning the tradition of Ps.-Cyril, the Cyrillean tradition can be represented as follows: PETER ISÉPY AND PAOLA CARMELA LA BARBERA 265 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 If this is indeed the case, a final point to consider is the histori- cal and geographical context in which Ps.-Zonaras accessed his g4 source of Ps.-Cyril’s lexicon. First, the fact that all these witnesses were produced outside of Constantinople provides a basis for the inference that the archetype of g4 either was produced outside of Constantinople or was transferred outside of the capital before giving rise to its descendants. Moreover, if we consider the date of origin of the extant manuscripts of g4, all of which are later than our lexicon —the earliest, S, dating to 1301/2—we come to the important conclusion that Ps.-Zonaras, who worked in the first half of the 13th century, used as his source a Ps.-Cyril manuscript of the g4 text (now lost) that was older than all the extant witnesses of this family. This manuscript, from a stemmatic point of view, was either the deperditus g4 itself, a lost descendant of it, or its brother. These observations prompt a reconsideration of Alpers’ per- spective on the production milieu of the lexicon. On the basis of Ps.-Zonaras’ comprehensive utilisation of sources, he postulated that the author could only have accessed this material in Constantinople. Nevertheless, he put forth the contradictory suggestion that Nikephoros Blemmydes, who was absent from the capital after 1205 for the rest of his life,68 could be considered 68 He spent his life mostly on the Greek coast of Asia Minor (Smyrna, 266 CYRIL’S GLOSSARY AS A SOURCE FOR PS.-ZONARAS ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 a prospective candidate for the authorship of the lexicon.69 It is not possible to accept both of the hypotheses; only one of them can be sustained. In light of our results, it seems highly doubtful that Constan- tinople was the place of origin of the lexicon. Ps.-Zonaras, at least as far as the lexicon of ‘Cyril’ is concerned, used a manu- script belonging to a family (g4) which, as far as we know today, had no descendants within Constantinople, but several outside of it, particularly in Cyprus and on Mount Athos, in the early 14th century. In considering this evidence, it is important to re- member that during the Latin occupation, the centralised Byzantine Empire was fragmented into several smaller—once peripheral—states, such as Nicaea, Epirus, or Trebizond.70 This political fragmentation had a significant impact on cultural production: the intellectual elite and Byzantine scholarship no longer had their focus in Constantinople, and in regions pre- viously considered peripheral, vibrant scriptoria emerged.71 With that in mind, we can assume that the Lexicon Ps.-Zonarae was compiled very probably in the territory controlled by the Empire of Nicaea. The location of the g4 manuscripts of ‘Cyril’ in Cyprus and Mount Athos is very revealing in this respect: the Empire of Nicaea was closer to the French-ruled Royaume de Chypre than to any other Greek-speaking state that emerged after 1204, and it also managed to occupy Mount Athos in the course of its expansion against the Crusader states, especially Latmos, Ephesos, Rhodes) and in Palestine, then in Thrace, Macedonia, and Thessaly: PLP 2897 s.v. Βλεµµύδης Νικηφόρος, and Munitiz, Nikephoros Blem- mydes 14–28. 69 Alpers, RE 10A (1972) 739. 70 An overview in D. Stathakopoulos, A Short History of the Byzantine Empire (London 2014) 150–154; see also A. Laiou (ed.), Urbs Capta: The Fourth Crusade and its Consequences (Paris 2005). 71 E. Crisci and P. Degni, La scrittura greca dall’antichità all’epoca della stampa (Rome 2011) 179–182. PETER ISÉPY AND PAOLA CARMELA LA BARBERA 267 ————— Greek, Roman, and Byzantine Studies 65 (2025) 246–267 after the Battle of Pelagonia (1259).72 This situation would be well suited to the biography of a person such as Nikephoros Blemmydes, who, during the Latin occupation, was compelled to seek his numerous sources outside of the capital, on Mount Athos and in the east. In any case, it would seem that he has been quite successful in spite of these circumstances.73 February, 2025 Department of Greek and Latin Philology Ludwig-Maximilians-Universität, Munich peter.isepy@klassphil.uni-muenchen.de Department of Cultures and Civilization Università degli Studi di Verona paolacarmela.labarbera@univr.it 72 M. Angold, A Byzantine Government in Exile. Government and Society under the Laskarids of Nicea. 1204–1261 (Oxford 1975) 279–296, and “Byzantium in Exile,” in David Abulafia (ed.), The New Cambridge Medieval History V (Cam- bridge 2008) 543–568, esp. 563; also D. Jacoby, “The Latin Empire of Constantinople and the Frankish States in Greece,” V 525–542. 73 The elaboration of this article was made possible by the synthesis of multiple research findings presented at the conference on the Cyril lexicon at Villa Vigoni, Loveno, in October 2024. The authors would like to express their gratitude to the conference organisers, Giuseppe Ucciardello and Stefano Valente, as well as all the participants, for their valuable suggestions. The contribution of Paola Carmela La Barbera is part of the PRIN 2022 project PatriarX: Paths of the Transmission of Saint Cyril’s Lexicon CUP B53D23022190006, id. 202279SXRM). Thanks are due to Edward Mezger for thoughtful correction of the English manuscript.