Microsoft Word - Rance.doc Greek, Roman, and Byzantine Studies 47 (2007) 193–224 © 2007 GRBS The Etymologicum Magnum and the “Fragment of Urbicius” Philip Rance YZANTINE LEXICA AND ETYMOLOGICA have long been rec- ognised as treasuries rich in citations from lost works of antiquity.1 These great monuments of mediaeval Greek scholarship and encyclopaedism have been subjected to almost two centuries of modern criticism and commentary, of which the highpoint was the enterprise of late nineteenth/early twen- tieth-century German classical philologists. Many obscurities 1 I am grateful to Prof. Klaus Alpers (University of Hamburg) for his kind assistance with this subject. The following abbreviations have been used: Alpers (1990) = K. Alpers, “Griechische Lexicographie in Antike und Mit- telalter,” in H.-A. Koch with A. Krup-Eber (eds.), Welt der Information. Wissen und Wissensvermittlung in Geschichte und Gegenwart (Stuttgart 1990) 14– 38 Cellerini = A. Cellerini, Introduzione all’ Etymologicum Gudianum (Rome 1988) Dain (1946) = A. Dain, L’Histoire du texte d’Elien le Tacticien (Paris 1946) Dain/de Foucault (1967) = A. Dain (texte mis au net et complété par J.-A. de Foucault), “Les stratégistes byzantins,” TravMém 2 (1967) 317–392 Dain/de Foucault (1968) = A. Dain (complété par J.-A. de Foucault), “Urbicius ou Mauricius?” REB 26 (1968) 123–136 Förster = R. Förster, “Studien zu den griechischen Taktikern,” Hermes 12 (1877) 426–471 Greatrex et al. = G. Greatrex, H. Elton, R. Burgess, “Urbicius’ Epitedeuma: an Edition, Translation and Commentary,” Byzantion 98 (2005) 35–74 Lasserre/Livadaras = F. Lasserre and N. Livadaras, Etymologicum Magnum Genuinum, Symeonis Etymologicum una cum Magna Grammatica, Etymologicum Magnum Auctum I α–ἀμωσγέπως (Rome 1976), II ἀνά–βώτορες (Athens 1992) Reitzenstein (1897) = R. Reitzenstein, Geschichte der griechischen Etymologika (Leipzig 1897) Reitzenstein (1907) = “Etymologika,” RE 6 (1907) 807–817 Shuvalov I, II = P. B. Shuvalov, “Урбикий и “Стратегикон” Псевдо- Маврикия,” VizVrem 61 (2002) 71–87; 64 (2005) 34–60 B 194 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” persist in this field, however, in part because some of the most important works lack modern and/or complete critical editions or have never been published.2 The present study directs a relatively narrow focus upon a single entry in the Etymologicum Magnum, the conventional title for an extensive lexical compen- dium produced by an unknown compilator in the mid-twelfth century.3 The article στρατός in the Etymologicum Magnum con- tains what purports to be an extract of an unnamed work by Urbicius, a writer of military treatises in the late fifth/ early sixth century, who in turn drew on a classicizing tradition of tactical writing dating back to the late Hellenistic period. This embedded extraneous item has been variously identified as a heavily modified excerpt from Urbicius’ extant Tacticon or a fragment of a unidentified lost work by Urbicius, though there is no scholarly consensus on this point and none of the argu- ments advanced is wholly satisfying. This paper presents a detailed deconstruction of the text of this lexical article, which 2 For up-to-date introductions to the tradition of Greek lexicography, citing earlier bibliography, see Alpers (1990) and Alpers, “Lexicographie (B.I–III),” in G. Üding with W. Jens (eds.), Historisches Wörterbuch der Rhetorik II (Tübingen 2001) 194–210. For Byzantine etymologica Reitzenstein (1897) remains fundamental, abridged with emendations in Reitzenstein (1907), and summarised with additional bibliography in H. Hunger, Die hochsprach- liche profaner Literatur der Byzantiner II (Munich 1978) 45–48. For studies of classical authors and genres preserved in Byzantine etymologica see e.g. A. Colonna, “Antica esegesi nicandrea negli Etymologica,” Bollettino del Comitato per la preparazione dell’ Edizione Nazionale dei Classici Greci e Latini N.S. 4 (1956) 17–24; C. Calame, Etymologicum Genuinum: Les citations des poètes lyriques (Rome 1970); G. Marcovigi, “Le citazioni dei lirici corali presso l’Etymologicum Genuinum,” Quaderni Triestini per il lessico della lirica corale greca 1 (1970) 11–49; F. Kolb, “Ο ΔΗΜΟΣ ΚΑΘΗΜΕΝΟΣ: Zur Notiz des Etymologicum Magnum über die Diobelie,” Historia 27 (1978) 219–221; G. Massimilla, “Gli ultimi due libri degli Αἴτια di Callimaco nell’ Etymologicum Genuinum,” StIt 16 (1998) 159–170; F. Schironi, I frammenti di Aristarco di Samotracia negli etimologici bizan- tini (Hypomnemata 152 [Göttingen 2004]). 3 The most recent complete edition is T. Gaisford, Etymologicum Magnum (Oxford 1848). Partial edition: specimen glosses in αμ– in Reitzenstein (1897) 223–241. Incomplete critical edition: (as Etymologicum Magnum Auctum) in Lasserre/Livadaras. For studies and earlier bibliography see Reitzenstein (1897) 212–253, (1907) 815–816; Lasserre/Livadaras I xvii–xxii; Cellerini 66–67. PHILIP RANCE 195 will demonstrate that it is a complex amalgam fashioned from several sources that originate in different genres and periods, and locating it within the textual traditions of both Graeco- Roman military literature and Byzantine lexicography. It is hoped in addition that these limited objectives will make a modest but significant contribution to understanding the meth- odology, source-materials, and originality of the compilator of the Etymologicum Magnum and, more broadly, the reception and preservation of classical literature in Byzantine encyclopaedic compilations. The article στρατός in the Etymologicum Magnum incorporates a short self-contained section under the heading Ὀρβικίου, τῶν περὶ τὸ στράτευμα τάξεων, “(from) Orbikios, on the formations of the army.”4 This item, hereafter “the fragment,” comprises a brief explanitory outline of the technical and for the most part classicising terminology for the various sub-divisions of an army and the appropriate commanding officers, followed by a list of terms relating to the deployment of troops on the battlefield. The Ὀρβίκιος in question is doubtless Urbicius, a stratégiste en chambre, about whom almost nothing is known with certainty other than his authorship of military treatises during the reign of Anastasius I (491–518).5 Two (formerly conjoined) works by Urbicius are extant. First, the Tacticon is a brief skeletal epitome of the first part (chs. 1–32) of Arrian’s Ars Tactica, written in A.D. 136 in celebration of Hadrian’s vicennalia. This section of Arrian’s work is a conventional and avowedly antiquarian treatment of the arcane terminology, internal articulation, and tactical evolutions of an idealised infantry phalanx, drawing on an originally late Hellenistic sub-genre of Graeco-Roman mil- itary literature.6 Second, the Ἐπιτήδευμα or “Invention” was 4 Etym.Magn. 728.48–729.18. 5 For the problematic sources for Urbicius’ life and works see esp. Greatrex et al. 40–49; also Förster 449–466; G./J. Gyomlay, Bölcs Leó Taktikája mint magyar történeti kútforrás (Értekezések a nyelv-és széptud. köréből 18.1 [Budapest 1902]) 35–40; Dain (1946) 37–39, 109; Dain/de Foucault (1967) 341–342, 347 (with errors); Dain/de Foucault (1968) 124–130; PLRE II 1190 “Urbicius 2” (with errors); Shuvalov I 83–86, II 35–36 (to be read with caution). 6 Urbicius’ Tacticon is preserved in a single manuscript prototype, the 196 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” originally appended to the Tacticon but separated and trans- mitted as an independent item at some point before the late tenth century.7 In this pamphlet Urbicius recommends, with rhetorical embellishment, his own design for a type of portable chevaux-de-frise called kanones (κανόνες), which he contends would be of great utility in warfare against barbarian horse- archers. This tract falls within the category of amateur com- positions addressed to imperial incumbents and predicting victory through technological innovation, which sought as much to publicize the author’s erudition and curry favour at court as to offer practical counsel. It is of no immediate rel- evance to the present study.8 To these two opuscula modern ___ tenth-century Ambros.gr. 139 (119 B sup.) (93r–95v). The text is edited only by Förster 467–471, who explains (459–461) that this edition is not based on his autopsy of the original MS., but follows the late Friedrich Haase’s col- lated transcription of two seventeenth-century apographs. Förster demon- strated (449–455, 459–466) that Arrian’s Ars Tactica was the sole source for the Tacticon, a view endorsed by all subsequent studies except PLRE II 1190, which erroneously claims Aelian as Urbicius’ model. For Arrian’s Ars Tactica see Flavii Arriani quae exstant omnia II, ed. A. G. Roos, add. et corr. G. Wirth (Leipzig 1968) 129–176; repr. in J. G. DeVoto, Flavius Arrianus, Tactical Handbook and The Expedition against the Alans (Chicago 1993), accompanied by a near-impenetrable English translation. For Arrian’s antiquarian intent: Arr. Tact. 32.2–3. For the work’s date and context: Tact. 44.3 with esp. E. L. Wheeler, “The Occasion of Arrian’s Ars Tactica,” GRBS 19 (1978) 351–365; accepted by M. Devine, “Arrian’s ‘Tactica’,” ANRW II.34.1 (1993) 312– 337, at 315–316, and A. B. Bosworth, “Arrian and Rome: The Minor Works,” ANRW 226–275, at 255, 259–261. 7 Ambros.gr. 139, which contains only the Tacticon, is convincingly dated to 959 by C. M. Mazzucchi, “Dagli anni di Basilio Parakimomenos (cod. Ambr. B 119 sup.),” Aevum 52 (1978) 267–316, esp. 267–282, 292–310. 8 For text, Eng. transl., and comm. of the Epitedeuma see now Greatrex et al. For the characteristics of this class of treatise see also the fourth-century anonymous De rebus bellicis, ed. R. I. Ireland (Leipzig 1982); Eng. transl. and comm. E. A. Thompson, A Roman Reformer and Inventor (Oxford 1952); It. transl. and comm. A. Giardina, Anonimo, Le cose della guerra (Milan 1989); for studies and bibliography: M. W. C. Hassall and R. I. Ireland (eds.), De rebus bellicis (BAR Int. ser. 63 [Oxford 1979]); M. A. Tomei, “Le tecnica nel tardo impero romano: le macchine da guerra,” Dialoghi di Archeologia N.S. IV 1 (1982) 63–88, at 69–84; T. Weidemann, “Petitioning a Fourth-Century Emperor,” Florilegium 1 (1979) 140–150; H. Brandt, Zeitkritik in der Spätantike. Untersuchungen zu den Reformvorschlägen des Anonymus De rebus bellicis (Munich PHILIP RANCE 197 scholarship has conventionally added a third work, the so- called “Cynegeticus of Urbicius,” a short tract concerning large- scale hunting as a method of training cavalry, but both the title and ascription are modern fabrications without manuscript authority that resulted from confused and careless scholarship in the 1930s.9 The erroneous ascription of Maurice’s Strategicon (ca. 590–600) to Urbicius in one tenth-century manuscript (M) is demonstrably spurious and the result of the copyist’s inter- vention.10 ___ 1988); J. H. W. G. Liebeschuetz, “Realism and Fantasy: The Anonymous De Rebus Bellicis and its Afterlife,” in E. Dabrowa (ed.), The Roman and Byzan- tine Army in the East (Cracow 1994) 119–139; B. Meißner, Die technologische Fachliteratur der Antike (Berlin 1999) 277–283. 9 In its earliest extant form the tract on hunting is appended to the “authentic recension” (MSS. M A ) of Maurice’s Strategicon (12.D), though ec- centricities of style and vocabulary betray its earlier independence. Detailed argument will follow elsewhere; it suffices here to observe that there is no manuscript authority for ascribing this item to Urbicius, nor any reason to connect him with it. The “Cynegeticus d’Urbicius” first surfaced from a be- wildering jumble of errors in A. Dain, La “Tactique” de Nicéphore Ouranos (Paris 1937) 58, and via additional misconceptions became an idée fixe in Dain/de Foucault (1967) 341–342, 352–353, 372, thence passing into the work of many other scholars, most recently Shuvalov I 77–78, 81–83, II 35 with n.4, 40 fig. 2, 42, 46, 48 (with additional misunderstanding). On this tract in general see P. Rance, “Simulacra pugnae: the Literary and Historical Tradition of Mock Battles in the Roman and Early Byzantine Army,” GRBS 41 (2000) 223–275, at 254–258; V. V. Kuchma, “Трактат «Об охоте»,” ADSV 33 (Ekaterinburg 2002) 48–58. 10 Förster 455–459; F. Aussaresses, “L’auteur du Strategicon,” REA 8 (1906) 23–40; Dain/de Foucault (1968) 136; G. T. Dennis (ed.), Das Stra- tegikon des Maurikios, Germ. transl. E. Gamillscheg (CFHB 17 [Vienna 1981]), 15–18 with stemma codicum at 41. Even Shuvalov I 79–81, 84, who wishes to ascribe the Strategicon to Urbicius, has to concede that on codi- cological grounds the ascription to Urbicius (ΟΥΡΒΙΚΙΟΥ) uniquely found in M must have been introduced by the tenth-century copyist and that his exemplar (ε) bore an ascription to Maurice (ΜΑΥΡΙΚΙΟΥ), cf. n.45 below. Shuvalov has attempted to resurrect the long-discredited thesis that Ur- bicius wrote the Strategicon. He presents a highly conjectural case for a two- stage textual evolution of the Strategicon, the earliest version of which was written in the late fifth/early sixth century. This putative “Strategicon of Ur- bicius” was, he claims, revised in the late sixth century during the reign of Maurice (582–602), at which point the many internal references that now date the work to that period were added, and presumably all explicit traces 198 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” The few scholars who have examined the article στρατός have drawn widely differing conclusions concerning the origin and character of the alleged extract of Urbicius. The text has been available to scholarship since the editio princeps by Zach- arias Kallierges in 1499, but until the later nineteenth century it was better known via an appendix to Henri Estienne’s The- saurus Graecae Linguae (1572), where it was originally printed as a discrete fragment without explanation or reference to the Ety- mologicum Magnum, though an editorial note was added in the revised edition of 1865.11 Scholarly interest began with Johann Scheffer, the editor princeps of Maurice’s Strategicon (1664), who conjectured that this “fragmentum Urbicii” had been extracted from a lost work, which, he thought, Maurice had used when compiling his treatise, though Scheffer offered no evidence to substantiate this speculation and his antiquarian guesswork need not detain the modern reader.12 In the earliest critical ___ of Urbicius’ authorship erased. The current text of the Strategicon, he contends, is thus an “Urbician” treatise overlaid by a “Maurician” textual stratum, and this “Maurician recension” was transmitted in the manuscript tradition under Maurice’s name. See Shuvalov I and II, with a slightly different version of the same arguments in idem, “Влияние авар на поздне- римское военное дело,” in V. M. Masson et al. (eds.), Изучение культурного наследия Востока (St. Petersburg 1999) 48–51. Shuvalov’s hypothesis, for the most part a modified rehearsal of arguments originally put forward by R. Vári in the 1890s–1900s (in support of his own case for authorship by a putative eighth-century “Urbicius”), will be rebutted in detail in P. Rance, The Roman Art of War in Late Antiquity: The Strategicon of the Emperor Maurice (Birmingham Byzantine, Ottoman and Modern Greek Monographs, forth- coming). 11 Etymologicum Magnum Graecum, ed. and printed by Z. Kallierges (Venice 1499). H. Stephanus et al., Thesaurus Graecae Linguae³ (Paris 1831–65) VIII 344–345. 12 Arriani Tactica et Mauricii Artis militaris libri duodecim (Uppsala 1664) 383– 384. Scheffer appears to have been ignorant of the context or character of this “fragmentum,” which he knew only as an unreferenced fragment in the appendix to the Thesaurus Graecae Linguae (1572). Scheffer’s seventeenth- century guesswork is based on too many misconceptions and faulty and/or incomplete data to warrant detailed rebuttal. The fragment was also noted without comment by F. Haase, “Über die griechischen und lateinischen Kriegsschriftsteller,” Neue Jahrbücher für Philologie und Pädagogik 5 [14.1] (1835) 88–118, at 108, as “de Ordinibus exercitus.” PHILIP RANCE 199 assessment of the evidence Richard Förster firmly denied a connection between the “Fragment des Urbicius” and the ex- tant Tacticon, observing that their contents exhibit few points of correspondence, though he offered no alternative explanation for the origin of the fragment.13 Karl Krumbacher acknowl- edged the existence of this “lexikalische Artikel” but stopped short of identifying it with Urbicius’ known works, implying that this item is or derives from an additional composition, though the content and context of the fragment make it an unlikely independent work, at least in its current form.14 In contrast, and with greater confidence, the distinguished French codicologist Alphonse Dain identified the fragment as a mod- ified abridgment of Urbicius’ Tacticon, “un dévelopment qui est une adaptation fortemente résumée et stylisée du morceau d’Urbicius, et qui porte encore mentioné le souvenir de son origine.”15 Dain’s status as the most prolific and influential scholar of the Greek tactical tradition writing in the second half of the twentieth century has ensured the endorsement of this view by subsequent scholarship.16 Most recently Geoffrey Greatrex was inclined to accept the essence of this textual relationship, though he rightly observed partial inconsistency in the content of the fragment and the Tacticon.17 There is there- 13 Förster 456 n.2: “Dieses unbedeutende Stück enthalt nur die Namen der Theile des Heeres und deren Führer. Ein Bestandtheil des gleich zu nennenden τακτικόν des Urbicius ist es nicht, kann auch kaum in un- eigentlichem Sinne auf dasselbe zurückgeführt werden, da es nur in einigen Punkten mit demselben inhaltlich stimmt.” 14 K. Krumbacher, Geschichte der Byzantinischen Literatur2 (Munich 1897) 635, 637, “Ausserdem schrieb Orbikios einen Auszug der Taktik des Arrian (τακτικόν) und einen im Etymologicum Magnum erhaltenen Artikel über die Unterabteilungen des Heeres und deren Führer (Ὀρβικίου τῶν περὶ τὸ στράτευμα τάξεων).” 15 Dain (1946) 38–39, reiterated in Dain/de Foucault (1967) 347; (1968) 130, “C’est une adaptation fortement stylisée et un bref résumé du Tacticon d’Urbicius, présenté sous le nom même d’Urbicius.” 16 E.g. PLRE II 1190, “The passages (sic) of Urbicius cited in the EM s.v. στρατός from a later resumé of the Tacticon”; and most recently Shuvalov I 83. 17 Greatrex et al. 42–43 with n.36: “the last section of the entry, however … does not appear to be drawn from either Urbicius or Arrian.” 200 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” fore scope for further investigation to determine the origin and character of the fragment and its relationship to Urbicius and his works, and to conjecture the possible circumstances or in- terest that led to its inclusion in the Etymologicum Magnum. The article στρατός reads as follows: στρατός· τὸ πλῆθος τῶν πολεμούντων. παρὰ τὸ στρέφω, στρέ- ψω, ἔστροφα, ἔστραμμαι, ἔστραπται, στραπτός καί, κατὰ ἀποβο- λὴν τοῦ π, στρατός, ὁ εὐπερίστρεπτος, ἢ ὁ συνεστραμμένος καὶ συνδεδεμένος ὄχλος. ἀλλὰ μὴν οἱ Αἰολεῖς στρόφω λέγουσι τὸ ῥῆμα καὶ στρότον λέγουσι τὸ συνεστραμμένον πλῆθος. ἢ παρὰ τὸ στερρῶς ἵστασθαι στρατός· ἢ παρὰ τὸ στερρῶς ᾄττειν, ἤγουν ἐφάλλεσθαι. Ὀρβικίου, τῶν περὶ τὸ στράτευμα τάξεων. ἰστέον δέ, ὅτι τῶν περὶ τὸ στράτευμα τάξεων καὶ ἡγεμονιῶν, ὁ μὲν πέντε ἀνδρῶν ἀριθμὸς καλεῖται πεμπάς· ὁ δὲ ἄρχων, πεμπάδαρχος· ὁ δὲ τῶν δέκα, δεκάς, καὶ δεκάδαρχος· ὁ δὲ τῶν πέντε καὶ εἴκοσι, λόχος, καὶ λοχαγός· ὁ τῶν πεντήκοντα, πεντηκοστύς, καὶ πεντηκόνταρ- χος· ὁ τῶν ἑκατόν, τάξις, καὶ ταξίαρχος· ὁ τῶν πεντήκοντα καὶ διακοσίων, σύνταγμα, καὶ συνταγματάρχης· ὁ τῶν πεντακοσίων, πεντακοσιοστύς, καὶ πεντακοσίαρχος· ὁ τῶν χιλίων, χιλιοστύς, καὶ χιλίαρχος· ὁ τῶν δισχιλίων μεραρχία καὶ τέλος, μεράρχης καὶ τελάρχης· ὁ τῶν τετρακισχιλίων, φάλαγξ, καὶ φαλαγγάρχης· ὁ τῶν μυρίων, μυριοστύς, καὶ μυρίαρχος. αἰ δύο διφαλαγγαρχίαι, τετραφαλαγγαρχία, ἀνδρῶν μυρίων ἑξακισχιλίων· καὶ ὁ ἄρχων, τετραφαλαγγάρχης. τὸ μέντοι πεζικὸν ἅπαν στράτευμα, ὁμωνύ- μως τοῖς μερικοῖς τάγμασι, λέγεται φάλαγξ· καὶ ὁ ἡγούμενος, στρατηγός· τοῦ δὲ ἱππικοῦ, ἵππαρχος. τὸ δὲ συναμφότερον πεζοί τε καὶ ἱππεῖς, στρατιά. τῆς δὲ στρατιᾶς τὸ μέτωπον λεγόμενον, ὃ καὶ πρῶτον ζυγὸν καλοῦσι, πρωτοστάται· καὶ ὁ ἡγούμενος, πρω- τοστάτης· οἱ δὲ παρ’ ἑκάτερα ταττόμενοι κέρατα, δεξιόν τε καὶ εὐώνυμον, οἱ αὐτοὶ καὶ παραστάται· ὁ δὲ ἀρχὸς ὁμωνύμως· οἱ δὲ ὄπισθεν αὐτῶν ὡς ἐπὶ τὸ βάθος ταττόμενοι, ἐπιστάται· τὸ δὲ ἐπὶ τούτοις πρὸς τὸ βάθος ἔσχατον ζυγόν, οὐρὰ καὶ οὐραγία, καὶ οὐραγός· ὁ αὐτὸς δὲ ὀπισθοφύλαξ· ἡ συνήθεια πεντηκοντοφύ- λακα αὐτὸν καλεῖ. ὁ δέ γε τοῦ παντὸς στρατοῦ ἡγεμονεύων, βασιλεύς. The article comprises two distinct elements: first, the lemma στρατός followed by an etymological gloss that adduces a sequence of typically fanciful derivations; second, a purported excerpt or abstract of an unidentified work or works by Ur- bicius, distinguished by a separate but rather uninformative rubric Ὀρβικίου, τῶν περὶ τὸ στράτευμα τάξεων. This second PHILIP RANCE 201 section has a somewhat anomalous position within the Etymo- logicum Magnum as a whole in that, although an underlying ety- mological interest is implicit in many of the names of military units and officers listed (e.g. a notional five-man unit is intrinsic to the derivation of πεμπάς and πεμπάδαρχος), this section is nevertheless free of explicit etymologising, even where deri- vations are ripe for construal or misconstrual. In substance and genre, therefore, this excerpt takes the form of a specialist word-list, onomasticon, or short self-contained military lexicon, rather than, strictly speaking, an etymologicum. The differing character of the two elements of this lexical article, the gloss and the fragment, and the expressly extraneous origin of the latter, call for separate investigation with regard to their source material. 1. The gloss s.v. στρατός The compilator of the Etymologicum Magnum most frequently cites among his sources two other etymologica, which he styles τὸ μέγα Ἐτυμολογικόν and τὸ ἄλλο Ἐτυμολογικόν. The first of these is the Etymologicum Genuinum, which was compiled at Constantinople around the mid-ninth century and drew on the writings of numerous earlier lexicographers and scholiasts. This work was discovered during the nineteenth century in two tenth-century manuscripts, Vat.gr. 1818 (A) and Laurent.S.Marci 304 (B), which do not preserve the earliest recension but two differing abridgements; the difficulties inherent in reconstruct- ing the original text from these manuscripts, and from the com- parative testimony of derivative lexica, are largely responsible for the Etymologicum Genuinum remaining for the most part un- published.18 The second or “other etymologicum” is the so-called 18 The critical edition initiated by Ada Adler continues in preparation under the direction of Klaus Alpers. Another edition projected by N. A. Livadaras, “Ce qu’apportera l’édition de l’Etymologicum Genuinum,” Ἐθνικὸν καὶ Καποδιστριακὸν Πανεπιστήμιον Ἐπιστημονικὴ Ἐπετηρὶς τῆς Φιλοσο- φικῆς Σχολῆς 24 (1973–74) 331–336, is in progress (as Etymologicum Magnum Genuinum) in Lasserre/Livadaras (α–βώτορες only). Individual sections: specimen glosses in αμ- in Reitzenstein (1897) 11–44; A. Colonna (ed.), Etymologicum genuinum, littera Λ (Quaderni Athena 4 [Rome 1967]); N. A. Livadaras, “Etymologicum Genuinum. Μεταγραφὴ καὶ ἔκδοσις τοῦ ἐξι- τήλου κειμένου τοῦ κώδικος Β,” Ἀθηνᾶ 70 (1968) 37–82; K. Alpers, Bericht 202 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” Etymologicum Gudianum, the archetype for which has been iden- tified in the Vat.Barberin.gr. 70. While the southern Italian pro- venance of this Urhandschrift is generally acknowledged, its date remains disputed, with arguments for the late tenth century accepted here.19 The relationship between the Etymologicum Genuinum and the Etymologicum Gudianum is complex; the compi- lator of the Gudianum certainly had the Genuinum at his disposal as an exemplar, but appears also to have utilised some of the ___ über Stand und Methode der Ausgabe des Etymologicum Genuinum (mit einer Ausgabe des Buchstaben Λ) (Copenhagen 1969); G. Berger, Etymologicum genuinum et Etymologicum Symeonis (Β) (Beitr.klass.Philol. 45 [Meisenheim am Glan 1972]). E. Miller, Mélanges de littérature grecque (Paris 1868) 1–318, published from codex B the lemmata in Etymologicum Genuinum (as “Etymologicum Florentinum”), collated with Gaisford’s 1848 edition of the Etymologicum Mag- num, but his accuracy cannot be relied on. Studies: Reitzenstein (1897) 1– 69, and (1907) 812–814; A. Colonna, “Un antico esemplare dell’ Etymo- logicum Genuinum,” Bollettino del Comitato per la Preparazione dell’ Edizione Nazionale dei Classici Greci e Latini N.S. 13 (1965) 9–13; R. Pintaudi, “Etymo- logica,” RendIstLomb 107 (1973) 10–24; Lasserre/Livadaras I v–xi; Cellerini 60–62; Alpers, Bericht 3–24, and “Eine byzantinische Enzyklopädie des 9. Jahrhunderts. Zu Hintergrund, Entstehung und Geschichte des griechi- schen Etymologikons in Konstantinopel und im italogriechischen Bereich,” in G. Cavallo et al. (eds.), Scritture, libri e testi nelle aree provinciali di Bisanzio (Spoleto 1989) I 235–69; Alpers (1990) 28–31; I. C. Cunningham, Synagoge. Συναγωγὴ λέξεων χρησίμων (SGLG 10 [Berlin/New York 2003]) 13–14, 22, 35–38. 19 The only complete edition is F. W. Sturz, Etymologicum Graecae Linguae Gudianum (Leipzig 1818), using a single corrupt and heavily interpolated (class IV) manuscript (Guelpherb.Gud.gr. 29/30), see remarks of Cellerini 12– 13; Alpers (1990) 29. Partial editions: specimen glosses in αμ– in Reitzen- stein (1897) 109–136; E. L. de Stefani, Etymologicum Gudianum quod vocantur (Leipzig 1909–20) fasc. 1 (α–βωμολοχία), 2 (βωμολόχοι–ζειαί). For manu- scripts see Reitzenstein (1897) 70–109; Cellerini 12–13, 21–29; S. Maleci, Il Codice Barberinianus Graecus 70 dell’ Etymologicum Gudianum (BollClassici Suppl. 15 [Rome 1995]). Date: Barberin.gr. 70 has been convincingly redated to the late tenth-century by K. Alpers, “Marginalia zur Überlieferung der grie- chischen Etymologika,” in D. Halfinger and G. Prato, Paleografia e codicologia greca I (Alessandria 1991) 523–541, at 531–540. Cellerini in his stemma at 69 appears also to favour the late tenth century, though he does not argue for this in his discussion at 21–24. Maleci (6) dates this MS. to the eleventh century, without argumentation, and Schironi, I frammenti 23–24, to the twelfth. PHILIP RANCE 203 same sources directly.20 Aside from these two earlier etymologica, the sources of the Etymologicum Magnum include an abridgement of Stephanus of Byzantium, Diogenianus’ epitome, Eulogius’ Ἀπορίαι καὶ λύσεις, the Lexicon Αἱμωδεῖν, George Choero- boscus’ Epimerismi ad Psalmos, scholia on Pindar, and other grammatical, rhetorical, and lexicographical works.21 The compilator of the Etymologicum Magnum freely modified the ma- terial he drew from his two principal sources—abbreviating and expanding explanations, and/or transposing text; adding glosses and imposing a stricter alphabetical sequence; altering or suppressing citations, rewording lemmata and interpolating new references. In short he was not an unoriginal copyist but sought to craft a novel and individual work according to his own design.22 Given our knowledge of the sources of the Etymologicum Magnum and of the working practices of its compilator, recon- struction of the textual history of the gloss στρατός presents relatively few problems. The compilator combined the corres- ponding articles in the Etymologicum Genuinum and Etymologicum Gudianum, and by amalgamating and transposing clauses ac- cording to a fastidious cut-and-paste methodology, and var- iously reducing and expanding the text of his two exemplars, he critically refashioned these sources into a new text. The process is tabulated below, with borrowings from the Genuinum underlined and those from the Gudianum in bold type. 20 Reitzenstein (1897) 98–155, (1907) 814–815; E. L. de Stefani, “Per le fonti dell’ Etimologico Gudiano,” BZ 16 (1907) 52–68; Cellerini 30–63. 21 O. Carnuth, De Etymologici Magni fontibus (Berlin 1873); A. Kopp, “Zur Quellenkunde des Etymologicum Magnum,” RhM 40 (1885) 371–376; R. Reitzenstein, “Zu den Quellen des sogenannten Etymologicum magnum,” Philologus 48 (1889) 450–455; 49 (1890) 400–420; Reitzenstein (1897) 248– 253, 351–352; (1907) 816. 22 Reitzenstein (1897) 241–248 demonstrates the compilator’s methodol- ogy using specimen glosses in αμ-. 204 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” Etym. Genuinum στρατός· τὸ πλῆθος τῶν πολεμούντων, ὁ δὲ τόπος στρατόπεδον· λέ- γεται γὰρ τοῦ στρατοῦ τὸ πέδον. εἴρηται δὲ παρὰ τὸ στρέφω, ἔνθα ἔστραπται στραπτὸς καὶ στρατός, τὸ συνε- στραμμένον πλῆθος. ἀλλὰ μὴν οἱ Αἰολεῖς στρόφω λέγοντες τὸ ῥῆμα καὶ στρότον λέ- γουσιν. ἐνήλλακται τὸ ε εἰς τὸ α, ὡς Ἄρτεμις Ἄρταμις.23 Etym. Magnum στρατός· τὸ πλῆθος τῶν πολεμούντων. παρὰ τὸ στρέφω, στρέψω, ἔστρο- φα, ἔστραμμαι, ἔστραπται, στραπτὸς καί, κατὰ ἀπο- βολὴν τοῦ π, στρατός, ὁ εὐπερίστρεπτος, ἢ ὁ συν- εστραμμένος καὶ συνδε- δεμένος ὄχλος. ἀλλὰ μὴν οἱ Αἰολεῖς στρόφω λέ- γουσι τὸ ῥῆμα καὶ στρότον λέγουσι τὸ συνεστραμμέ- νον πλῆθος. ἢ παρὰ τὸ στερρῶς ἵστασθαι στρα- τός· ἢ παρὰ τὸ στερρῶς ᾄττειν, ἤγουν ἐφάλλε- σθαι.24 Etym. Gudianum στρατός· γίνεται πα- ρὰ τὸ στερρῶς ᾄτ- τειν, ἤγουν ἐφάλλε- σθαι, ἢ παρὰ τὸ στρέφω, ἔστραμμαι, στραπτός, ὡς γέγρα- πται γραπτός, καὶ στρατός, ὁ εὐπερί- τρεπτος, ἢ παρὰ τὸ στερρῶς ἵστασθαι στρατός· ὁ συν- εστραμμένος καὶ συνδεδεμένος ὄχ- λος.25 Furthermore, the sources for the glosses in the two older etymo- logica can in turn be established. The source for the Etymologicum Gudianum is an anonymous alphabetised collection of Homeric epimerisms compiled between the mid-eighth and mid-ninth centuries: 23 Text: edited from MSS. A and B by A. R. Dyck, Epimerismi Homerici (SGLG 5 [Berlin/New York 1983–95]) II 674. Variants: λέγεται–πέδον A : om. B; εἴρηται A : γίνεται B; ἔστραπται A : om. B; στραπτὸς A : ἐστραπτὸς B; στρότον A : στρωτον B; τὸ ε εἰς τὸ (B : om. A) α, ὡς (hab. B τὸ) Ἄρτεμις Ἄρταμις. 24 Text: Gaisford 728.40–47 with minor repunctuation and the following emendations: ἄλλα Gaisford : ἀλλὰ; στροφῶ Gaisford : στρόφω; στροτὸν Gaisford : στρότον; ἄττειν Gaisford : ᾄττειν. For στρόφω and στρότον see Dyck, Epim.Hom. σ 9 and σ 52, app. crit. (II 659, 674), citing R. Meister, Die griechischen Dialekte I (Göttingen 1882) 48, 52. It is not possible to determine whether εὐπερίστρεπτος in Etym.Magn. is due to an error or editorial choice. Dyck edits Epim.Hom. σ 9 (II 659.50) εὐπερίτρεπτος (citing MSS. G O), and this reading is corroborated in the derivative Et.Gud. Cramer’s editio princeps of MS. O (Oxon.bibl.Nov.Colleg. 298) of Epim.Hom., however, reports εὐπερί- στρεπτος at 236r: Anecdota Graeca I (Oxford 1839) 389.23–27. 25 Text: Sturz 513.18–22 with minor repunctuation; emend ἄττειν Sturz 513.18 : ᾄττειν; ἐσφάλλεσθαι Sturz 513.19 : ἐφάλλεσθαι Etym.Magn. 728.47, Epim.Hom. σ 9 (Dyck II 659.48). PHILIP RANCE 205 στρατός· γίνεται παρὰ τὸ στερρῶς ᾄττειν, ἤγουν ἐφάλλεσθαι. ἢ παρὰ τὸ στρέφω ἔστραμμαι στραπτός, ὡς γέγραπται γραπτός, καὶ στρατός, ὁ εὐπερίτρεπτος. ἢ παρὰ τὸ στερρῶς ἵστασθαι στρατός· ὁ συνεστραμμένος καὶ συνδεδεμένος ὄχλος.26 The gloss in the Etymologicum Genuinum was drawn from two sources. (a) The majority of the gloss derives from a Homeric epimerism, possibly known in the form of a scholion rather than an alphabetised lexicon. A substantially similar text is pre- served in the aforementioned collection of Epimerismi Homerici, which must derive from the same tradition but is not the direct source for the gloss in the Etymologicum Genuinuim: στρατός· ὄνομα ῥηματικὸν παρὰ τὸ στρέφω, ἔνθεν τὸ ἔστραπται στραπτὸς καὶ ἐνδείᾳ τοῦ π, ὡς ἐν τῷ θάλπω θάλπαμος καὶ θάλαμος, γίνεται στρατός, οἱονεὶ τὸ συνεστραμμένον πλῆθος. καὶ τὸ ε τρέπεται εἰς α, ὡς Ἄρτεμις Ἄρταμις· οἱ δὲ Αἰολεῖς στρόφω λέγοντες τὸ ῥῆμα καὶ στρότον λέγουσιν.27 (b) The first line, στρατός· τὸ πλῆθος τῶν πολεμούντων, ὁ δὲ τόπος στρατόπεδον· λέγεται γὰρ τοῦ στρατοῦ τὸ πέδον, is derived from the so-called ῥητορικόν, a designation that the compilator of the Etymologicum Genuinum elsewhere applies to an earlier lexical compilation, now identified as a late recension of the anonymous Συναγωγὴ λέξεων χρησίμων, an originally later eighth-/early ninth-century alphabetised adaptation of the (fifth-century?) glossarium of rare words falsely ascribed to Cyril of Alexandria. Different versions of the Synagoge were in- dependently the sources for the identical glosses s.v. στρατός in both the Lexicon of Photius (ca. 840) and the Suda (ca. 1000).28 26 Epim.Hom. σ 9 (Dyck II 659.48–660.51). For date see Dyck I 6–7, with 9–10, 14–16, 23–27 for Et.Gud.’s use of Epim.Hom. 27 Epim.Hom. σ 52 (Dyck II 674.26–31). Cf. Cramer I 394.9–14 (237r). For this form used by οἱ Αἰολεῖς cf. Sappho fr.16.1; Alcaeus fr.382.2, and pos- sibly fr.300.1 (ed. E.-M. Voigt, Sappho et Alcaeus [Amsterdam 1971]). For general discussion of the ambiguous evidence for Et.Gen.’s use of Epim.Hom. see Dyck I 36–38. 28 Συναγωγὴ λέξεων χρησίμων σ 250: στρατός· τὸ πλῆθος τῶν πολε- μούντων, ὁ δὲ τόπος στρατόπεδον· λέγεται γὰρ τοῦ στρατοῦ τὸ πέδον. Cf. identically Photius Lexicon (ed. R. Porson [Cambridge 1822] 542.21–22); Suda σ 1183 (ed. A. Adler IV 442). For demonstration of the sources of and relationships between these and other lexical works see Cunningham 20–42 206 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” This gloss already had a long pedigree: it was earlier partially registered in the fifth-century Lexicon of Hesychius, and in ex- tant lexica it may be traced back to the surviving abridgement of the Lexicon Homericum of Apollonius Sophista (originally ca. A.D. 100), and given the derivative character of the latter work in all likelihood derives from an earlier Homeric glossary or scholion, though which of Apollonius’ many sources supplied this ma- terial is not known.29 2. The “Fragment of Urbicius” Reitzenstein has already observed that the fragment of Ur- bicius does not occur in either the Etymologicum Genuinum or Etymologicum Gudianum, nor in any of the other known sources of the Etymologicum Magnum.30 The compilator seems therefore to have drawn the fragment from an alternative and unknown source, and it remains to be established whether he knew an original text by Urbicius directly or via another lexical com- pendium, and to what degree he modified its contents. A translation of the fragment follows: (From) Orbicius, on the formations of the army. One should know, with regard to the formations and commands of the army, that a force of five men is called a pempas, of which the commander is a pempadarchos; a force of ten men is a dekas under ___ with summary at 13–14; note that in the stemma at 14 the second Σ should read Σ΄. For the thorny question of the identity of the λεξικὸν ῥητορικόν see K. Alpers, “Das Lexikon des Photius und das Lexicon Rhetoricum des Etymologium Genuinum,” JÖByz 38 (1988) 171–191, whose conclusions are corroborated by Cunningham 20–42, but acrimoniously contested by C. Theodoridis, “Das Lexicon des Patriarchen Photius und das Lexicon Rhetorikon des Etymologicum Genuinum,” JÖByz 42 (1992) 95–141. 29 Hesychius σ 1972 (ed. P. A. Hansen [Berlin 2005] 355): στρατός· τὸ πλῆθος τῶν πολεμούντων στρατιωτῶν; Apollonius Sophista Lexicon Homeri- cum (ed. I. Bekker [Berlin 1833]) 145.17: στρατός· τὸ πλῆθος τῶν πολεμούν- των, στρατόπεδον δὲ ὁ τόπος. For Apollonius’ sources see H. Schenck, Die Quellen des Homerlexikons des Apollonius Sophistes (Hamburger philol. Stud. 34 [Hamburg 1974]). 30 Reitzenstein (1897) 250: “ebenso aus Orbikios entlehnte längere Stück … Die Zahl der benutzen Nebenquellen ist bei diesem vielleicht erst durch die allmählige Arbeit verschiedener Grammatiker enstandenen Werk kei- nesfalls gering.” PHILIP RANCE 207 a dekarchos; twenty-five men a lochos under a lochagos; fifty men a pentêkostus under a pentêkontarchos; a hundred men a taxis under a taxiarchos, two hundred and fifty men a suntagma under a suntag- matarchos; five hundred men a pentakosiotus under a pentakosiarchos; a thousand men a chiliostus under a chiliarchos; two thousand men a merarchia or telos under a merarchês or telarchês; four thousand men a phalanx under a phalangarchês; ten thousand men a muri- ostus under a muriarchos. Two diphalangarchiai are a tetraphalangar- chia, of one thousand six hundred men, of which the commander is a tetraphalangarchês. An army which is wholly infantry is called a phalanx, having the same name as its constituent units, of which the leader is a stratêgos, while a hipparchos leads the cavalry; but an army that is both infantry and cavalry is a stratia. The so-called “front” (metôpon) of the army is what they also call the first rank, the prôtostatai, of which the leader is a prôtostatês. The men de- ployed on either wing, the right and the left, these are both parastatai, and the commander is similarly named. The men de- ployed in depth behind them are epistatai, while their last rank is called the “tail” (oura) or ouragia, under an ouragos; this man is also an opisthophulax, who is customarily called the pentêkontophulax. The man who heads the whole army, however, is the emperor. The fragment divides broadly into two parts: an initial outline of an organisational structure and hierarchy for an army, fol- lowed by an explanation of terminology concerned with tac- tical deployment and the stationing of personnel. It is first necessary to determine whether and to what extent either part corresponds to the contents of Urbicius’ Tacticon. Urbicius begins his Tacticon with a similar summary of the subdivisions of an army, but it is markedly different in content and origin. Here Urbicius reiterates a conventional and largely artificial numerical schema for the organisation of an idealised phalanx, as delineated by his model Arrian (Tact. 10) and the other authors who drew on the same late Hellenistic tradition of tactical writing (principally Asclepiodotus and Aelian). The base unit for this system was a squad or file (lochos) of 16 men, which is successively doubled via a series of increasingly larger units to arrive ultimately at a hypothetical army of 16,384 heavy infantry, supported by half as many light infantry (8,192) and half that again in cavalry (4,096). Within such abstract dis- cussions of the tactical manoeuvres of an infantry phalanx the value of the magic number 16,384 was its sequential divisibility 208 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” by two down to, in theory, a two-man file (16,384 = 214). This numerical sequence, which was probably formulated, or at least canonised, in a lost tactica by the Stoic philosopher Posei- donius of Apamea (ca. 135–51 B.C.), is in some measure reflec- tive of Hellenistic philosophical and arithmetical idealism, but it served military theorists as a convenient illustrative model when explaining military formations and tactical evolutions. To the extent that this sub-genre portrays a “real” army, it is most probably the Seleucid army of the late second century B.C.31 31 For this numerical schema cf. Asclep. 2.7–10, 6.1–3, 7.11 (with com- ments of L. Poznanski, Asclépiodote. Traité de tactique [Paris 1992] 41, 44); Ael. 8.3–9.10, 15.2–16.3, 20.2; Arr. Tact. 9–10, 14, 18; Syrianus De re strat. 15.62–76 (ed. G. T. Dennis, Three Byzantine Military Treatises [CFHB 25 (Washington 1985)] 1–135, as “Sixth-Century Anonymus, On Strategy”). The figure 16,384 also recurs in Mamluk military treatises via a late mediaeval Arabic translation of Aelian’s Tactica, see G. Tantum, “Muslim Warfare: A Study of a Medieval Muslim Treatise on the Art of War,” in R. Elgood (ed.), Islamic Arms and Armour (London 1979) 187–201, at 190, 194–195. For the textual relationship between the tactica of Asclepiodotus, Aelian, and Arrian and cognate opuscula see Dain (1946) 26–40; A. B. Bosworth, “Arrian and Rome: The Minor Works,” ANRW II.34.1 (1993) 226–275, at 253–255, 258–259, 262–264; M. Devine, “Aelian’s Manual of Hellenistic Military Tactics,” AncW 19 (1989) 31–64, at 32–33, and “Arrian’s ‘Tac- tica’,” ANRW II.34.1, 312–337, at 316–330, contra Förster 426–449; P. A. Stadter, “The Ars Tactica of Arrian: Tradition and Originality,” CP 73 (1978) 117–128, at 117–118. The hypothesis that most satisfactorily ac- counts for the points of similarity and divergence requires that Aelian and Arrian (chs. 1–31) drew independently on a common lost source (Dain’s “Techne Perdue”), to which Arrian introduced minor changes, interpola- tions, and glosses. The so-called Lexicon Militare, ed. H. Köchly and W. Rüstow, Griechische Kriegsschriftsteller II.2 (Leipzig 1855) 217–233 (hereafter Lex.Mil.), also derives independently from this source. The unknown author of this source and Asclepiodotus both had direct access to the lost treatise of Poseidonius of Apamea, of which Asclepiodotus’ work is an abridged sum- mary. It has been suggested that a lost tactical memorandum known to have been written by Polybius underlies this tradition of tactical writing: K. K. Müller, “Asklepiodotos 10,” RE 2 (1896) 1637–41, at 1640–41; A. M. De- vine, “Polybius’ Lost Tactica: The Ultimate Source for the Tactical Manuals of Asclepiodotus, Aelian, and Arrian?” AHB 9 (1995) 40–44; partially en- dorsed by B. Campbell, OCD3 s.v. “Asclepiodotus”; F. Walbank, “Polybius as Military Expert,” in P. R. Hill (ed.), Polybius to Vegetius. Essays on the Roman Army and Hadrian’s Wall presented to Brian Dobson (Hadrianic Society 2002) 19– 31, at 21. However, N. Sekunda, “The Taktika of Poseidonius of Apameia,” PHILIP RANCE 209 Urbicius presents this contrived organisational hierarchy as follows (Tact. praef.): τὸ πλῆθος τοῦ παντὸς πεζικοῦ στρατεύματος ὀφείλει ἔχειν ἄν- δρας μυρίους ἑξακισχιλίους τριακοσίους ὀγδοήκοντα τέσσαρας. αἱ δὲ ὀνομασίαι τῶν ἡγεμόνων τοῦ τοιούτου στρατεύματός εἰσιν αὗται· λοχαγὸς ὁ δεκαὲξ ἀνδρῶν ἡγούμενος. διλοχίτης ὁ δύο λόχων ἡγούμενος, ὅ ἐστιν ἀνδρῶν λβʹ. τετράρχης ὁ τεσσάρων λόχων ἡγούμενος, ὅ ἐστιν ἀνδρῶν ξδʹ. ταξιάρχης ὁ λόχων ὀκτὼ ἄρχων, τουτέστιν ἀνδρῶν ρκηʹ. ἑκατοντάρχης ὁ ἑκατὸν μόνον ἀνδρῶν ἡγούμενος. συνταγματάρχης ὁ λόχων ιϛʹ ἡγούμενος, ὅ ἐστιν ἀνδρῶν σνϛʹ· οὗτος δὲ καὶ ξεναγὸς καλεῖται. δεῖ δὲ τοὺς σνϛʹ ἔχειν ἐπιλέκτους πέντε, σημειοφόρον, οὐραγόν, στρατοκή- ρυκα, σαλπιγκτὴν καὶ ὑπηρέτην. πεντακοσιάρχης ὁ λόχων λβʹ ἡγούμενος, ὅ ἐστιν ἀνδρῶν φιβʹ. χιλιάρχης ὁ λόχων ξδʹ ἄρχων, τουτέστιν ἀνδρῶν ͵ακδʹ. μεράρχης ὁ λόχων ρκηʹ ἡγούμενος, τουτέστιν ἀνδρῶν ͵βμηʹ. οὗτος δὲ καὶ τελάρχης ὀνομάζεται. φαλαγγάρχης ὁ λόχων διακοσίων πεντηκονταὲξ ἄρχων, ἤγουν ἀνδρῶν ͵δϟϛʹ. οὗτος δὲ καὶ στρατηγὸς καλεῖται. διφαλαγγάρχης ὁ λόχων φιβʹ ἡγούμενος, τουτέστιν ἀνδρῶν ͵ηρϟβʹ. τὸ δὲ τάγμα τοῦτο καλεῖται κέρας. τετραφαλαγγάρχης ὁ λόχων ͵ακδʹ ἄρχων, ὅ ἐστιν ἀνδρῶν μυρίων ἑξακισχιλίων τπδʹ. τοῦτον γὰρ τὸν ἀριθμὸν κάλλιστον καὶ ἁρμόζοντα τῇ ὁπλιτικῇ τῶν πεζῶν τάξει ἐκρίναμεν. The mass of the whole infantry army ought to comprise 16,384 men. The terms for the commanders of such an army are as follows: a lochagos is the commander of 16 men. A dilochitês is the commander of two lochoi, which is 32 men. A tetrarchês is the commander of four lochoi, which is 64 men. A taxiarchês is the officer of eight lochoi, which is 128 men. A hekatontarchês is the commander of only a hundred men. A suntagmatarchês is the com- mander of 16 lochoi, which is 256 men; this man is also called a xenagos. The 256 men should include five select men: a standard- ___ in Sekunda, Hellenistic Infantry Reform in the 160’s BC (Łodz 2001) 125–134, at 128–129, argues persuasively that Poseidonius originated the tradition. For what it is worth, the speculative attempt of L. Poznanski to reconstruct what Polybius’ lost tactica might have looked like envisages a work quite different in content from the Poseidonian tradition: “A propos du ‘Traité de Tac- tique’ de Polybe,” Athenaeum 58 (1980) 340–352, and “Essai de reconstitu- tion du Traité de Polybe d’après le livre III des Histoires,” AntCl 49 (1980) 161–172. 210 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” bearer, a rearguard (ouragos), a herald, a bugler, and an adjutant. A pentakosiarchês is the commander of 32 lochoi, which is 512 men. A chiliarchês is the commander of 64 lochoi, which is 1,024 men. A merarchês is the commander of 128 lochoi, which is 2,048 men; this man is also termed a telarchês. A phalangarchês is the officer of 256 lochoi, which is 4,096 men; this man is also called a stratêgos. A diphilangarchês is the commander of 512 lochoi, which is 8,192 men. This unit is called a wing. A tetraphalangarchês is the officer of 1,024 lochoi, which is 16,384 men. For we have judged this number to be the best and most convenient for a combat for- mation of infantry. It is difficult to concur with Dain’s conclusion that this pas- sage was the source for the article in the Etymologicum Magnum. The binary sequence of the late Hellenistic tradition differs from the decimal system outlined in the fragment. There are points of conceptual and terminological correspondence: both texts include a suntagmatarchês (commanding 250 or 256 men), a pentakosiarchês (500 or 512), a chiliarchês (1,000 or 1,024), a merarchês or telarchês (2,000 or 2,048), a phalangarchês (4,000 or 4,096), and a tetraphalangarchês (16,000 or 16,384). But the author of the fragment has clearly attempted to reconcile two conflicting organisational systems, in that for the most part he presents a decimal sequence (5, 10, 25, 50, 100, 250, 500, 1,000, 10,000) which at first stands in sharp contrast to the unit-strengths Urbicius describes in the Tacticon (16, 32, 64, 128). For some of the larger formations the two systems, in rounded figures, coincide (250/256, 500/512, 1,000/1,024), but the author of the fragment struggles to accommodate other elements of the Hellenistic tradition, which more crudely in- trude into his preferred decimal model; thus he incongruously includes formations of 2,000 (rounded 2,048) and 4,000 (4,096) men, and having completed his decimal sequence at 10,000 he abruptly adds that a tetraphalangarchia comprising 16,000 men is composed of two diphilangarchiai, though he does not otherwise mention the latter formation or name its officer. There are also contradictions in the use of terminology: for Urbicius a lochos under a lochagos is a 16-man unit, but in the fragment contains 25 men; while according to Urbicius the term stratêgos is syn- onymous with phalangarchês, the commander of the 4,096-strong sub-unit called a phalanx, but in the fragment the stratêgos is the PHILIP RANCE 211 commander of the entire army, which may, apparently, also be called a phalanx. Other officer-titles listed by Urbicius are ab- sent from the fragment (διλοχίτης, τετράρχης, ἑκατοντάρχης, ξεναγός). Furthermore, a large number of organisational and hierar- chical terms which appear in the fragment are nowhere used by Urbicius. In the Tacticon Urbicius supplies only the titles of officers but not the names of the units they command, as found in the fragment. In some cases it is perfectly conceivable that the author of the fragment could have construed the names of units from their officers’ titles (λόχος < λοχαγός, τάξις < ταξι- άρχης, σύνταγμα < συνταγματάρχης, μεραρχία < μεράρχης, τέλος < τελάρχης, διφαλαγγαρχία < διφαλαγγάρχης, τετρα- φαλαγγαρχία < τετραφαλαγγάρχης), but in other instances the terminology of the fragment is wholly without parallel in Urbicius’ work (πεμπάς, πεμπάδαρχος, δεκάς, δεκάδαρχος, πεντηκοστύς, πεντηκόνταρχος, χιλιοστύς, πεντακοσιοστύς, μυριοστύς, μυρίαρχος). Above all, none of the information in the second part of the fragment concerning tactical deployment occurs in the eleven chapters of Urbicius’ Tacticon, and the author could not therefore have drawn on this text for his definitions of the terms—μέτωπον, πρωτοστάται, παραστάται, ἐπιστάται, οὐρά, οὐραγία, οὐραγός, ὀπισθοφύλαξ, πεντη- κοντοφύλαξ, or ἵππαρχος.32 The concluding reference to the 32 Urbicius later mentions a ἱππάρχης (Tact. 5), but this officer is the com- mander of a specific 512-strong cavalry unit according to the conventions of the late Hellenistic organisational schema, and thus quite distinct from the ἵππαρχος in the fragment, who commands all the cavalry in the army. Ur- bicius also mentions an οὐραγός as one of the five select men (ἐπίλεκτοι) of a 256-man suntagma, but the completely different context and the additional terminological synonyms in the fragment (οὐρὰ καὶ οὐραγία, καὶ οὐραγός· ὁ αὐτὸς δὲ ὀπισθοφύλαξ … πεντηκοντοφύλακα) indicates that Urbicius’ Tacticon was not the source. In ancient tactical writers the term οὐραγός was used with two distinct meanings. First, the last man in every file, i.e. the final rank of a formation, were collectively the οὐραγοί or “file-closers,” whose important role in maintaining cohesion from the rear during combat is frequently acknowledged (Xen. Mem. 3.1.8, Eq.Mag. 2.3, Cyr. 3.3.41–42; Asclep. 2.2, 3.6; Ael. 5.1; Arr. Tact. 6.6; Maurice Strat. 12.B.16.27–29). Second, in the Hellenistic tradition οὐραγός, as the “rearguard,” also desig- nated a single supernumerary officer attached to a larger field unit, whose 212 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” emperor’s overall command of the armed forces is also without parallel in the Tacticon. To summarise: the fragment contains a different numerical system from that outlined in Urbicius’ Tac- ticon, which for the most part uses different terminology for units and officers, and contains none of the information about deployment found in the fragment. In short, less than a third of the content of the fragment can have originated in the Tacticon. If one insists upon the authenticity of the heading Οὐρβικίου, it might initially be tempting to speculate—as previously did Scheffer and (by implication) Krumbacher and Förster—that the Etymologicon Magnum preserves a fragment of a lost work by Urbicius, and it is not impossible that Urbicius wrote other treatises.33 But this hypothesis would require Urbicius to have ___ duty was to supervise drills and manoeuvres from the rear (Asclep. 2.9, 3.6, 6.3; Ael. 9.4, 14.8; Arr. Tact. 10.4, 12.11). The potential confusion between the two types of οὐραγοί is illustrated at Asclep. 3.6. The latter sense is meant in both the Etym.Magn. and Urbicius’ Tacticon. 33 Insofar as the tenth-century topographical compendium known as the Patria Constantinopoleos may be trusted, Urbicius was credited with “writing military works” (τοῦ ἱστορήσαντος [τὰ] στρατηγικά) of unspecified number: Patr.Const. 3.22 (ed. T. Preger, II 220.6–11). This may refer to the Tacticon and Epitedeuma only, or to these opuscula and/or additional works, though it is unlikely that the topographer was in possession of detailed information. On the doubtful reliability of this work, and specifically in relation to the “two Urbicii,” see A. Berger, Untersuchungen zu den Patria Konstantinupoleos (Poikila Byzantina 8 [Bonn 1988]) 211–212, 223–224, 228–229, 404–406, 586–587, 607; Greatrex et al. 40–41. All previous scholars have read the word στρατηγικά (some MSS. τὰ στρατηγικά) in this passage as generic “military works.” Recently Shuvalov I 83, 86, has attempted to argue that here is a reference to the actual title of a major treatise, i.e. Shuvalov’s putative “Strategicon of Urbicius” (cf. n.10 above), but this is no more than his wishful thinking and it is highly implausable that the tenth-century topog- rapher accurately preserved, knew, or meant a specific title. For other examples of this middle-Byzantine usage of (τὰ) στρατηγικά cf. Nicephorus Phocas De velitatione 21.1, ed. G. Dagron and H. Mihăescu, Le Traité sur la guérilla (De velitatione) de l’empereur Nicéphore Phocas (Paris 1986) 119.11, οἱ τὰ τακτικὰ καὶ στρατηγικὰ ἀναγραψάμενοι. See also the extensive source- notice at the beginning of the Tactica of Nicephorus Uranus (Constantinop.gr. 36) listing the various τακτικὰ ἤγουν στρατιγικά (sic) he has consulted: F. Blass, “Die griechischen und lateinischen Handschriften im alten Serail zu Konstantinopel,” Hermes 23 (1888) 219–233, at 225; Dain, La “Tactique” de Nicéphore Ouranos 13, 89–90, 93–95; Dain (1946) 150–151; Dain/de Foucault PHILIP RANCE 213 written another work, similar to the Tacticon, but substantially at odds with its content, which was somehow known uniquely to the twelfth-century compilator of the Etymologicum Magnum, but otherwise undocumented in the rich tradition of Greek, Roman, and Byzantine tactical writing. Furthermore, this proposition does not in any case acknowledge and account for the distinctive character of the fragment. We have already noted indications of the author’s attempt to amalgamate and reconcile conflicting sources descending from different literary traditions, which marks out the fragment as the construction of a compilator or antiquarian rather than an informed and coherent composition on military matters. The most striking characteristic of the fragment, however, is its lexical interest. The organisational section contains obvious genre terms or “tactica-speak” used by the writers of tactical treatises (σύν- ταγμα, συνταγματάρχης, πεντακοσίαρχος, μεραρχία, τέλος, τελάρχης, μεράρχης, φαλαγγάρχης, διφαλαγγαρχίαι, τετραφα- λαγγαρχία, τετραφαλαγγάρχης, μέτωπον, πρωτοστάτης, ἐπι- στάται) or words too commonplace to establish a connection with a particular source or genre (λόχος, λοχαγός, δεκάδαρχος, τάξις, ταξίαρχος, φάλαγξ, χιλίαρχος, στρατηγός, ἵππαρχος, στρατιά, στράτευμα, τάγμα, ζυγόν, κέρατα, βάθος, οὐρά, οὐ- ραγία, οὐραγός, ὀπισθοφύλαξ). But this lexical article is not merely a rehearsal of the standard technical vocabulary of Greek tactical writing; on the contrary, the fragment is con- spicuous for its assemblage of rare words seldom attested in antiquity outside specialist lexica and in some cases otherwise unknown. These include poetic or archaicising forms (ἀρχός); terminology not conventionally found in the late Hellenistic tradition followed by Urbicius (παραστάται, πεντηκοντοφύ- λαξ);34 and in particular a significant number of Atticising ___ (1967) 371–372. Cf. also Constantine VII Praecepta in J. F. Haldon, Three Treatises on Imperial Military Expeditions (CFHB 28 [Vienna 1990]) Text C 106.196–199, βιβλία στρατηγικά; Psellus Chron. 7.16 (ed. Renauld II 181.12–13), ἀπὸ τῶν τακτικῶν βιβλίων καὶ στρατηγικῶν καὶ πολιορκητικῶν. 34 πεντηκοντοφύλαξ is otherwise unattested. παραστάται has a rather ten- uous place in the technical vocabulary of the late Hellenistic tradition from which Urbicius’ Tacticon descends. Of the three chief representatives of this sub-genre, only Asclep. 2.4 defines παραστάται, i.e. comrades deployed in 214 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” usages (πεμπάς, πεμπάδαρχος, δεκάς, πεντηκοστύς, πεντακο- σιοστύς, πεντηκόνταρχος, χιλιοστύς, μυριοστύς, μυρίαρχος). This terminological eccentricity both underlines the lexical (rather than military) purpose of the author and points to a source within the genre of lexica or etymologica rather than tactica. The only previous lexical compilation to contain these words is the Onomasticon of Julius Pollux. This work, produced in the later second century, is topical rather than alphabetic in ar- rangement, and covers a wider range of subjects, including warfare. It functions primarily as a thesaurus rather than a lexi- con or etymologicum, compiling synonyms and specialist vocab- ularies, and serving principally as a handbook for Atticising rhetorical composition. It has not survived in its original form; all manuscripts derive from four incomplete and interpolated copies that in turn descend from a common hyparchetype, an epitome possessed and interpolated by Arethas of Caesarea (ca. 900–932). Examination of the vocabula militaria of the Onomasti- con reveals the same body of vocabulary as found in the frag- ment (1.127–128): καὶ ὁ μὲν ἐκ δεξιᾶς τοῦ πρώτου ζυγοῦ πρωτοστάτης, καὶ πᾶν τὸ μέτωπον, πρωτοστάται. ὁ δὲ παρ’ ἕκαστον ταττόμενος, παρα- στάτης, ὁ δὲ ὑπ’ αὐτὸν35 ἐξόπισθεν, ἐπιστάτης. τὸ δὲ ἐπὶ πᾶσιν ἐν τῷ βάθει ζυγόν, οὐρὰ καὶ οὐραγοί· καλοῖτοι δ’ ἂν ὁμωνύμως καὶ ___ the same rank: ὅταν δὲ λόχῳ λόχος παρατεθῇ, ὥστε λοχαγὸν λοχαγῷ καὶ οὐραγὸν οὐραγῷ καὶ τοὺς μεταξὺ τοῖς ὁμοζύγοις παρίστασθαι, συλλοχισμὸς ἔσται τὸ τοιοῦτον, οἱ δὲ ὁμόζυγοι τῶν λόχων πρωτοστάται ἢ ἐπιστάται διὰ τὸ παρ’ ἀλλήλοις ἵστασθαι παραστάται κεκλήσονται. Ael. 29.3 uses the word once but does not supply a definition (παραγγελοῦμεν ἐξελίσσειν τοὺς μετα- τεταγμένους παραστάτας εἰς οὓς προεῖχον τόπους); while it does not occur at all in Arrian’s Tactica. The fourth witness to the tradition, the Lexicon Militare, corroborates the definition given by Asclepiodotus but is clearly not derived from it, and this coincidence renders more likely the presence of παραστάται in the Urtext of this tradition (Lex.Mil. 8): παραστάται· οἱ ὁμό- ζυγοι τῶν λόχων πρωτοστάται καὶ ἐπιστάται διὰ τὸ παρ’ ἀλλήλους ἵστασθαι (= Suda π 444). For πεντηκοντοφύλαξ and παραστάτης see nn.36 and 38 be- low. 35 E. Bethe, Pollucis Onomasticon (Leipzig 1900–37) 41.16, prints παρ’ αὐτὸν but the reading ὑπ’ αὐτὸν in MSS. A V is undoubtedly correct—an ἐπιστάτης stands behind (ἐξόπισθεν) a πρωτοστάτης, while it is παραστάται who stand alongside him (παρ’ ἕσκατον). PHILIP RANCE 215 ὁ ἄρχων αὐτῶν οὐραγός. καλοῦνται δὲ καὶ ὀπισθοφύλακες, καὶ τὸ ἔργον οὐραγεῖν, καὶ ὀπισθοφυλακεῖν. τὸ δὲ σύμπαν στρατιά, στρατός, στράτευμα, στρατιωτικόν, φάλαγξ, τάγμα, σύνταγμα. μέρη δ’ αὐτοῦ μυριοστύς, χιλιοστύς, πεντηκοστύς, λόχος, δεκάς, πεμπάς. καὶ οἱ ἄρχοντες οἱ μὲν τοῦ παντὸς στρατηγοὶ καὶ συ- στράτηγοι καὶ ὑποστράτηγοι, ὥσπερ οἱ ἀποχειροτονηθέντες ἀπο- στράτηγοι. ταξίαρχοι καὶ οὐραγοὶ καὶ μυρίαρχοι καὶ χιλίαρχοι καὶ λοχαγοὶ καὶ ἑκατατόνταρχοι καὶ δεκάδαρχοι καὶ πεμπά- δαρχοι, καὶ τῶν ἱππέων ἵππαρχοι καὶ φύλαρχοι. Θηβαίων δὲ ἴδιον Βοιωτάρχης, καὶ Λακεδαιμονίων βασιλεύς. Juxtaposition of the texts of the fragment and the Onomasticon confirms that Pollux was the source for much of the vocabulary concerning deployment: Poll. Onom. 1.127 (Bethe 41.14–19) καὶ ὁ μὲν ἐκ δεξιᾶς τοῦ πρώτου ζυγοῦ πρωτοστάτης, καὶ πᾶν τὸ μέτωπον, πρωτοστάται. ὁ δὲ παρ’ ἕκαστον ταττόμενος, παραστάτης, ὁ δὲ ὑπ’αὐτὸν ἐξόπισθεν, ἐπιστάτης. τὸ δὲ ἐπὶ πᾶσιν ἐν τῷ βάθει ζυγόν, οὐρὰ καὶ οὐραγοί· καλοῖτοι δ’ ἂν ὁμωνύ- μως καὶ ὁ ἄρχων αὐτῶν οὐραγός. καλοῦνται δὲ καὶ ὀπισθοφύλακες, καὶ τὸ ἔργον οὐραγεῖν, καὶ ὀπισθο- φυλακεῖν. Etym. Magn. (Gaisford 729.9–17) τῆς δὲ στρατιᾶς τὸ μέτωπον λεγόμε- νον, ὃ καὶ πρῶτον ζυγὸν καλοῦσι, πρωτοστάται· καὶ ὁ ἡγούμενος, πρω- τοστάτης· οἱ δὲ παρ’ ἑκάτερα τατ- τόμενοι κέρατα, δεξιόν τε καὶ εὐώ- νυμον, οἱ αὐτοὶ καὶ παραστάται· ὁ δὲ ἀρχὸς ὁμωνύμως· οἱ δὲ ὄπισθεν αὐ- τῶν ὡς ἐπὶ τὸ βάθος ταττόμενοι, ἐπι- στάται· τὸ δὲ ἐπὶ τούτοις πρὸς τὸ βάθος ἔσχατον ζυγόν, οὐρὰ καὶ οὐ- ραγία καὶ οὐραγός· ὁ αὐτὸς δὲ ὀπι- σθοφύλαξ· ἡ συνήθεια πεντηκοντο- φύλακα αὐτὸν καλεῖ. The author of the fragment slightly modified the wording of his exemplar and introduced three minor expansions, presumably with the aim of clarification, though certainly in two instances, and possibly all three, these interventions led him to err: he misunderstands the definition of παραστάτης36 and perhaps 36 As noted above, παραστάται denotes the comrades stationed either side of a man in his rank, so e.g. Hdt. 6.117.3; Xen. Cyr. 3.3.59, 8.1.10, Hell. 6.5.43; Polyaen. 2.10.4; Asclep. 2.4; Lex.Mil. 8 (= Suda π 444). The author of the fragment appears to have misunderstood the definition in Pollux 1.127 ὁ δὲ παρ’ ἕκαστον ταττόμενος, παραστάτης, “the men deployed to each side,” and conceived instead οἱ δὲ παρ’ ἑκάτερα ταττόμενοι κέρατα, δεξιόν τε καὶ εὐώνυμον, οἱ αὐτοὶ καὶ παραστάται, i.e. apparently the men deployed on each flank of the formation. 216 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” also ἐπιστάτης,37 while the hapax πεντηκοντοφύλαξ adduced as a synonym for ὀπισθοφύλαξ appears to be the author’s own formulation, the sense of which, both etymological and mil- itary, defies explanation.38 With regard to the Atticising terminology for the units and unit-commanders of a decimal-based army outlined in the frag- ment, it is necessary to emphasise how rare these words are. Although a small number are individually registered in other lexica, Pollux’s Onomasticon and the Etymologicum Magnum are the only works in which all these terms are assembled, other than Pollux’s own source for this recherché vocabulary, Xenophon, chiefly the Cyropaedia.39 In particular, in the Cyropaedia Xen- 37 It is not clear whether he has fully understood the meaning of ἐπι- στάτης. Tactical manoeuvres that sought to increase or reduce the depth of a formation relied on a fundamental organisational arrangement in which each man in a file was alternately designated πρωτοστάτης or ἐπιστάτης; thus in an eight-man file positions 1, 3, 5, 7 were πρωτοστάται, while 2, 4, 6, 8 were ἐπιστάται: e.g. Asclep. 2.3, 5.2; Onas. 20; Ael. 5.1–4; Arr. Ect. 5, Tact. 6.4–6, 12.4–10; Syrianus De re strat. 15.56–61. Thus correctly Pollux 1.127: ὁ δὲ ὑπ’ αὐτὸν (πρωτοστάτην) ἐξόπισθεν, ἐπιστάτης. The author of the fragment, however, writes οἱ δὲ ὄπισθεν αὐτῶν (πρωτοστάτων) ὡς ἐπὶ τὸ βάθος ταττόμενοι, ἐπιστάται, importing the phrase ὡς ἐπὶ τὸ βάθος from Pollux’s subsequent definition of οὐραγοί, and wrongly implying that be- hind the front rank of πρωτοστάται all the men “deployed in the depth of the formation” (i.e. “through the ranks”) are classed as ἐπιστάται. 38 I can offer no explanation for πεντηκοντοφύλαξ other than the ob- servation that this word is attested only in the fragment and appears to be without historical foundations. As a synonym for ὀπισθοφύλαξ a “fifty- guard” makes no intrinsic or etymological sense, and why a rearguard might be so styled remains elusive. The author’s claim that “custom (ἡ συνήθεια) calls him (the rearguard) a pentêkontophulax” implies access to an alternative tradition, but this word is possibly his own fabrication, perhaps inspired by a corrupt text of his model: cf. Poll. 1.127: οὐραγός. καλοῦνται δὲ καὶ ὀπισθοφύλακες, καὶ τὸ ἔργον οὐραγεῖν, καὶ ὀπισθοφυλακεῖν; Etym. Magn. 729.16–17: οὐραγός· ὁ αὐτὸς δὲ ὀπισθοφύλαξ· ἡ συνήθεια πεντηκον- τοφύλακα αὐτὸν καλεῖ. 39 πεμπάς: Xen. Cyr. 2.1.22, 24, 26, 30, 3.22; 4.5.5; Hell. 7.2.6. πεμπάδαρ- χος: Xen. Cyr. 2.1.22, 23, 26, 30, 3.21; Eq.Mag. 4.9–10; cf. Hesych. δ 2703 πεμπαδάρχης. δεκάς: Xen. Cyr. 2.1.22, 24, 26, 30, 3.22; 4.2.27, 5.5; cf. Hesych. δ 2703. πεντηκοστύς: Thuc. 5.68.3; Xen. An. 3.4.22; cf. Harp. I 208.3 Dind., 177 Keaney (= Phot. Lex. μ 653; Suda λ 65, μ 1259). In support of πεντηκοστύας Harpocration cites Xenophon on the Spartans, although PHILIP RANCE 217 ophon has Cyrus the Great create and train an idealised Per- sian army organised along decimal lines according to the fol- lowing sequence: Constituent units Unit-title Officer-title Total πεμπάς πεμπάδαρχος 5 2 πεμπάδες δεκάς δεκάδαρχος 10 5 δεκάδες λόχος λοχαγός 50 2 λόχοι τάξις ταξίαρχος 100 10 τάξεις χιλιοστύς χιλίαρχος 1,000 10 χιλιοστύες μυριοστύς μυρίαρχος 10,000 In the context of sixth-century Persian history the detailed structure and terminology of this army is pseudo-historical, though it may partly reflect Xenophon’s knowledge of authen- tic Achaemenid military organisation of his own day, perhaps modified by contemporary Greek practices.40 In this organisational section of the fragment there is again evidence that the author has attempted to elaborate and amplify his source material without access to additional or alternative information. Two examples suffice. First, the only “non-Urbician” officer-title in the fragment which cannot be accounted for as a borrowing from Pollux’s Onomasticon is πεν- τηκόνταρχος, and in this instance it is telling that the author ___ Xenophon never uses this word of a Spartan unit but does refer to Spartan officers styled πεντηκοντῆρες (codd. πεντηκοστῆρες), Lac.Pol. 11.4, 13.4; Hell. 3.5.22, 4.5.7. Harpocration must have assumed, probably correctly, a Spartan institution of πεντηκοστύς on the grounds that at An. 3.4.21–2 (cf. Thuc. 5.66), in a non-Spartan context, Xenophon refers to πεντηκοντῆρες commanding πεντηκοστύες. The notoriously conflicting evidence for Spar- tan military organisation, including πεντηκοστύς and πεντηκοντῆρες, is sur- veyed by J. K. Anderson, Military Theory and Practice in the Age of Xenophon (Berkeley/Los Angeles 1970) 225–236, and more successfully by J. F. Lazenby, The Spartan Army (Warminster 1985) 5–10, 52–53. χιλιοστύς: Xen. Cyr. 2.4.3; 6.3.13, 31, 32; 7.1.22, 5.17; cf. differently Hesych. ε 1285, Suda κ 274, χ 306. μυριοστύς: Xen. Cyr. 6.3.20. μυρίαρχος: Xen. Cyr. 3.3.11; 6.3.20, 21, 22; 8.1.14, 15, 4.29; Polyaen. 1 praef. 2; cf. μυριάχης in Hdt. 7.81 (twice). 40 On the decimal system of the Achaemenid army see P. J. Junge, “Hazarapatis,” Klio 33 (1940) 13–39; G. Widengren, “Recherches sur le féodalisme iranien,” Orientalia Suecana 5 (1956) 79–182, at 160–166; E. Ben- veniste, Titres et noms propres en Iranien ancien (Paris 1966) 67–70. 218 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” has strayed from the tradition: Pollux did not provide him with a classicizing term for a commander of a 50-strong πεντηκο- στύς and so the compilator imported or guessed at a πεντη- κόνταρχος. While this formulation for a “commander of fifty” makes intrinsic sense and is consistent with patristic and Byz- antine usage, it does not in fact occur in the Greco-Roman tactical genre; indeed in classical literature πεντηκόνταρχος is restricted to a small number of Attic authors, for whom it meant exclusively a petty officer on a trireme, a historical con- text that baffled later lexicographers.41 Second, and similarly, 41 Administrative assistant of a τριήραρχος: Dem. 50.18, 19, 24, 25; Xen. Ath.Resp. 1.2; Pl. Leg. 707A, with J. S. Morrison and J. F. Coats, The Athenian Trireme (Cambridge 1986) 111. Atticist lexicographers in the Roman Empire knew that πεντηκόνταρχος correctly applied to a naval officer, but it is doubtful whether any understood the historical context of the classical Athenian trireme, and certainly some rationalised this grade as an un- historical “commander of a penteconter”: Poll. 1.96 τριήραρχος, πεντηκόνταρ- χος, ναύαρχος; 1.119 πλοῖα, οἱ δὲ ἄρχοντες τριήραρχοι καὶ πεντηκόνταρχοι καὶ ναύαρχοι; Harp. I 245.1–3 Dind. (210–211 Keaney) πεντηκόνταρχος· ὁ τῆς πεντηκοντόρου ἄρχων, ὡς δηλοῖ Δημοσθένης ἐν τῷ περὶ τοῦ ἐπι- τριηραρχήματος. ὅτι δὲ πεντηκόντορος ἐκαλεῖτο ἡ ναῦς ὑπὸ νʹ ἐρεσσομένη πρόδηλον (= partially Phot. Lex. 411.7–9; Suda σ 981). πεντηκόνταρχος is first used as a generic “leader of fifty” in LXX Ex. 18:21, 25, Deut. 1:15, 4 Reg. 1:9–11, 13–14, Is. 3:3, 1 Macc. 3:55, and this usage thence passed into biblical commentaries, patristic writings, and derivative chronicles. In the tactical genre πεντηκόνταρχος is unattested, though a πεντηκονταρχία is documented both as a 64-man unit of light infantry in the artificial organi- sational schema of the late Hellenistic tactical tradition (Asclep. 6.3; Ael. 16.1; Arr. Tact. 14.3; Lex.Mil. 30 [= Suda τ 96]) or a generic “50-man” unit (Onas. 34.2). Only in middle Byzantine sources does πεντηκόνταρχος, -ης occur as a specific officer-grade: Niceph. Phocas Praec.Milit., ed. E. McGeer, Sowing the Dragon’s Teeth: Byzantine Warfare in the Tenth Century (Washington 1995) 1.1 (p.12.8–10), 1.4 (14.39), 10 (18.109–110); Anon. De re militari 1.175, ed. Dennis, Three Byzantine Military Treatises 246–326, at 254; Sylloge Tacticorum, ed. A. Dain (Paris 1938) 1.26, 20.2, 20.4, 35.6, 45.12, 46.5; Niceph. Uranus Tactica 56.1 (ed. McGeer 88.8–13), 64.4 (148.43); cf. Const. Porph. De thematibus 1.28, ed. A. Pertusi (Rome 1952). πεντηκόνταρχος also occurs in chronicle sources, though it is unclear whether this is a generic usage or corresponds to an actual rank or title: V.Pachomii 75.9 (BHG 1396a), ed. F. Halkin, Le Corpus athénien de Saint Pachome (Geneva 1982) 11– 72, at 38, δέκαρχοι, πεντηκόνταρχοι, ἑκατόνταρχοι, χιλίαρχοι; Mirac. S. Demetrii, ed. P. Lemerle, Les plus anciens recueils des miracles de Saint Démétrius I (Paris 1979) 230.20. Later Byzantine lexica register both the spurious “com- PHILIP RANCE 219 the fragment lists a 500-strong unit called a πεντακοσιοστύς commanded by a πεντακοσίαρχος, both terms absent from Pollux’s Onomasticon. In the idealised army of the Hellenistic tactical tradition a πεντακοσίαρχης commands a body of 512 men, but this corps is correctly termed a πεντακοσιαρχία.42 Urbicius’ Tacticon rehearses this organisational schema, but supplies only the officer-titles and not the names of the units they command. Faced with this terminological gap, and with no additional corroborative sources at his disposal, the author of the fragment construed the term for the 500-man command of a πεντακοσίαρχος as a πεντακοσιοστύς, a unique usage he appears to have coined himself in accordance with the Atticis- ing conventions for unit-names set out in Pollux’s word-list (e.g. πεντηκοστύς, χιλιοστύς, μυριοστύς).43 Conclusions The following conclusions may be drawn from the foregoing analysis. The article στρατός in the Etymologicum Magnum does not contain a “fragment” (in the conventional sense) of the Tacticon of Urbicius, still less of a putative lost work by that author. Rather, this short self-contained military glossary labelled Ὀρβικίου, τῶν περὶ τὸ στράτευμα τάξεων is a com- ___ mander of a penteconter” and the biblical/patristic “commander of fifty,” e.g. Phot. Lex. 411.4–5 πεντηκόνταρχος· ὁ τῆς πεντηκοστῆς τοῦ τέλους ἄρχων; 7–9 πεντηκόνταρχος· ὁ τῆς πεντηκοντόρου ἄρχων· ἐκαλεῖται δὲ πεντηκόν- τορος ἡ ὑπὸ πεντήκοντα ἐρεσσομένη ναῦς (= Suda σ 981); Lex.Seg., ed. I. Bekker, Anecdota Graeca I (Berlin 1814) 195–318, at 297.11–12, πεντηκόνταρ- χος· ὁ ἄρχων τῆς πεντηκοστῆς τοῦ τέλους καὶ τῶν πεντηκοστῶν. 42 Asclep. 2.10, 3.3; Ael. 9.6. 10; Arr. Tact. 10.9; Lex.Mil. 13; Syrianus De re strat. 15.68–69. 43 πεντακοσιοστύς occurs otherwise only in the sixth-century anonymous De scientia politica dialogus 4.15, ed. C. M. Mazzucchi, Menae patricii cum Thoma referendario, de Scientia politica dialogus (Milan 1982) p.3.15, though here the author does not in any case mean the title of a unit, but merely the elaborately Atticising numeral 500: ἀνὰ πεντακοσιοστὺν ἀνδρῶν ἱππέων, “up to five hundred mounted men.” It is highly improbable that the author of the fragment knew this work, and πεντακοσιοστύς is most likely to be his own formulation. The only other occurrence of πεντακοσιοστύς is in Eu- stathius Ad Il. Π 173 (1052.46: III 827.27 van der Valk), but since Etym. Magn. was one of Eustathius’ principle lexical sources, a direct transmission can be assumed. 220 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” posite of two sources descending from two different literary traditions: A, a representative of the late Hellenistic tradition of tactical writing initiated by Poseidonius of Apamea, which concerned the internal structure, deployment, and manoeuvres of an idealised and hypothetical phalanx of 16,384 men; B, a witness to an older tradition of Atticising vocabulary for a pseudo-historical army organised along decimal lines, originat- ing in Xenophon’s Cyropaedia, but in this instance mediated via Pollux’s Onomasticon. In terms of substance, nothing precludes identifying source A as any of the direct or indirect descendants of Poseidonius’ lost work (Asclepiodotus, Aelian, Arrian, Ur- bicius, Syrianus Magister), such is the uniformity of content among these treatises, but given the citation of Ὀρβικίος in the heading it is reasonable to accept that source A was Urbicius’ Tacticon, which provided the author of the article with an initial structural framework. The author probably had recourse to Urbicius’ spare and jejeune opusculum, rather than the more expansive representatives of this Hellenistic sub-genre, precisely because it offered a succinct and convenient resumé of termin- ology; indeed the principal reason for the preservation of the Tacticon, especially in such a high-quality production as Ambros. gr. 139 (B 119 sup.), was its utility as a glossary of the arcane military vocabulary of the ancients rather than its relevance to contemporary practice, and one might conjecture that its later, and perhaps even original, function was as a guide to reading classical military literature.44 It also appears that in the minds of some Byzantine scholars of the tenth to thirteenth centuries the historically-distant Ur- bicius had acquired a wholly unmerited reputation as a great general and influential military theorist, whose name might thus serve as a mark of authenticity and antique authority in 44 The character of Urbicius’ Tacticon is aptly summarised by Mazzucchi, Aevum 52 (1978) 282, “si riduce a un dizionario della falange.” Note that this codex (349r–352v) also contains the archetype of Ad Basilium patricium Nau- machica, a broadly analogous dictionary of nautical technicalia compiled by an unknown contemporary from classical sources, including Homer and Pollux’s Onomasticon. This work was presumably produced to assist Byzan- tine readers in their comprehension of the so-called Corpus Nauticum. See A. Dain, Naumachica (Paris 1943) 57–68; Dain/de Foucault (1967) 363. PHILIP RANCE 221 discussions of military matters. The citation of Urbicius in the twelfth-century Etymologicum Magnum, in an article that derives only to a small degree from his writings, may therefore reflect a similar misconception of his significance.45 In reality, in the 45 Urbicius’ modest extant oeuvre of two short and formerly conjoined opuscula, the Tacticon and Epitedeuma, appears to have had no impact upon contemporaries nor any discernible Nachleben. He and his work pass un- noticed in late fifth-/sixth-century sources and he is never referred to, cited, nor quoted in the rich corpus of Byzantine military literature, which terminates ca. 1010. This 500-year obscurity stands in contrast to the re- emergence of Urbicius in the second half of the tenth century, though in sources and contexts of dubious historical credibility. The scribe of Mediceo- Laurent.gr. 55.4, the important collection of Greek, Roman, and Byzantine tactica compiled ca. 940–950, uniquely ascribed Maurice’s Strategicon to Urbicius, an historical impossibility according to internal indications of date, and demonstrably not the ascription the scribe found in his exemplar ε. This error appears in essence to be the product of a misreading of ΜΑΥΡΙΚΙΟΥ as ΟΥΡΒΙΚΙΟΥ, and certainly the copyist was capable of blunders of this magnitude (e.g. at fol. 159v he wrongly assigns the poliorcetic treatise of Aeneas to Aelian via a similarly careless reading of the superscription ΑΙΝΕΙΟΥ as ΑΙΛΙΑΝΟΥ), but his ascription of the Strategicon to Urbicius implies that the tenth-century scribe had at least heard of an Urbicius and knew him to be an author of military treatises. Cf. bibliog- raphy n.10 above. Slightly later, as previously noted (n.33), the frequently- inventive topographer who compiled the Patria Constantinopoleos reports (3.22) that an Urbicius who lived during the reign of Anastasius was known for “writing military works,” and describes him as patricius and magister mili- tum per Orientem, founder of an important church of the Theotokos, and epo- nym of a district of Constantinople, all details unreported in any other source. On this basis PLRE II 1291 locates “Urbicius 2 Barbatus” (the epithet is modern and erroneous) in the fasti of magistri militum per Orientum within the lacuna of 506–516/8, though this apparently prominent dig- nitary does not exist outside the Patria, and Greatrex et al. 41 have recently expressed doubts concerning his historical reality, tentatively identifying a muddled doublet of the renowned praepositus sacri cubiculi of the same name (PLRE II 1188–90, Urbicius 1), who was an influential courtier and pious benefactor in the reign of Anastasius, but who in the Patria (1.58, 70; 3.6) is transformed into a semi-legendary figure misdated to the reign of Constan- tine. By the mid-thirteenth century the reputation of Urbicius the tactician had become embellished to such a degree that his name could be listed in a peculiar catalogue of the most illustrious generals of all time: Theodorus II Ducas Lascaris In laudem Iohannis Ducae Imperatoris 14, ed. L. Tartaglia, Teodoro II Duca Lascari, Encomio dell’ Imperatore Giovanni Duca (Naples 1990) 73.711–715, δεῦρο δὴ πᾶς βασιλέων ἀρίστων ἄλλος κατάλογος τῶν στρα- 222 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” lexical notice that bears his name, Urbicius’ Tacticon has been ___ τηγίᾳ μὲν ἐκλαμψάντων, ἀρετῇ δὲ καὶ φρονήσει βιωσάντων βασιλικῶς, ἤτοι Βρούτοι καὶ Κάτωνες, Ἀντώνιοί τε καὶ Ἀννίβαι, Ἀδριανοὶ καὶ Τραϊανοί, Πομπήϊοί τε καὶ Οὐρβίκιοι καὶ πᾶς ἄλλος βασιλικὸς χορὸς ἀνδραγαθίᾳ καὶ χάριτι κλεϊζόμενος (“Here indeed is another full list of the best emperors who were distinguished for their generalship, or those who with virtue and prudence have lived in the manner of emperors, all the Brutuses and Catos, Antonys and Hannibals, Hadrians and Trajans, Pompeys and Urbiciuses, and the all the rest of the imperial chorus celebrated for their courage and generosity”). The imperial or quasi-imperial status accorded to Urbicius here presumably reflects the now-complete fusion of Urbicius the stratégiste en chambre and Mauricius the emperor. It is not clear what precisely inspired this tenth-century re-invention of Urbicius, though one likely source of con- fusion was the existence of a panegyrical epigram which extols the merits of a military work written by an Urbicius during the reign of Anastasius. In these twelve lines of verse, without doubt composed by Urbicius himself or a close associate, the treatise in question introduces itself to the reader as a re- vival or reworking of an ancient text “which once the Emperor Hadrian had beside him in his wars, / which for ages lay disused and nearly for- gotten, / but in the reign of the firm-handed Emperor Anastasius / I was released into the light again” (ἣν πάρος Ἀδριανὸς μὲν ἄναξ ἔχεν ἐν πολέ- μοισι, / κρύψε δ’ ἀεργίη χρόνον ἄσπετον ἐγγύθι λήθης, / ἀλλ’ ὑπὸ καρτερό- χειρος Ἀναστασίου βασιλῆος / ἤλυθον ἐς φάος αὖθις). The object of praise is unquestionably Urbicius’ Tacticon, an epitome of Arrian’s Ars Tactica, the latter written expressly to celebrate Hadrian’s vicennalia and fashioned with elements of panegyric to gain imperial interest and endorsement (cf. bib- liography n.6 above). Furthermore, this epigram immediately precedes Urbicius’ Tacticon in Ambros.gr. 139 (92v), the unique manuscript prototype, where it serves as a verse preface. It also found its way into the Palatine Anthology 9.210, in which the lemma similarly links it to “a tactical book of Urbicius” (εἰς βίβλιον τακτικῶν Οὐρβικίου). No other Greek, Roman, or Byzantine military treatise is the subject of panegyric, and this unusual lit- erary adjunct to what was, in Byzantium, a traditionally low-brow technical genre, certainly attracted scholarly interest and even prompted the compo- sition of supplementary hexametric verses during the reign of Leo VI (886– 912). See Förster 462–463; Dain/de Foucault (1968) 124–127; A. Cameron, The Greek Anthology: from Meleager to Planudes (Oxford 1993) 149–150, 333, 336; Greatrex et al. 40; contra Shuvalov I 83–85, II 41 with 40 fig. 2 (with errors and to be read with great caution). I plan to treat this complex topic in a separate study, but it suffices here to suggest that the rhetorical hyper- bole of these antique laudatory verses may have led unwary scholars of the tenth-century Macedonian Renaissance to invest Urbicius with an unde- served fame quite disproportionate to his minor historical and literary significance. PHILIP RANCE 223 so radically modified and augmented that the contribution of his work is all but obliterated and reduced to a deeply-buried and barely-discernible textual substratum, in effect an endo- skeleton fleshed out using material drawn from Pollux’s Ono- masticon. The relative debt owed by the “fragment” to these two sources is illustrated below, with borrowings from Pollux sig- nified in bold and those from Urbicius underlined: Etym. Magn. 728.48–729.18 Ὀρβικίου, τῶν περὶ τὸ στράτευμα τάξεων. Ἰστέον δὲ, ὅτι τῶν περὶ τὸ στράτευμα τάξεων καὶ ἡγεμονιῶν, ὁ μὲν πέντε ἀνδρῶν ἀριθμὸς καλεῖται πεμπάς· ὁ δὲ ἄρχων, πεμπάδαρχος· ὁ δὲ τῶν δέκα, δε- κὰς, καὶ δεκάδαρχος· ὁ δὲ τῶν πέντε καὶ εἴκοσι, λόχος, καὶ λοχα- γός· ὁ τῶν πεντήκοντα, πεντηκοστύς, καὶ πεντηκόνταρχος· ὁ τῶν ἑκατόν, τάξις, καὶ ταξίαρχος· ὁ τῶν πεντήκοντα καὶ διακοσίων, σύνταγμα, καὶ συνταγματάρχης· ὁ τῶν πεντακοσίων, πεντα- κοσιοστύς, καὶ πεντακοσίαρχος· ὁ τῶν χιλίων, χιλιοστύς, καὶ χιλί- αρχος· ὁ τῶν δισχιλίων μεραρχία καὶ τέλος, μεράρχης καὶ τε- λάρχης· ὁ τῶν τετρακισχιλίων, φάλαγξ, καὶ φαλαγγάρχης· ὁ τῶν μυρίων, μυριοστύς, καὶ μυρίαρχος. αἰ δύο διφαλαγγαρχίαι, τετραφαλαγγαρχία, ἀνδρῶν μυρίων ἑξακισχιλίων· καὶ ὁ ἄρχων, τετραφαλαγγάρχης. τὸ μέντοι πεζικὸν ἅπαν στράτευμα, ὁμωνύ- μως τοῖς μερικοῖς τάγμασι, λέγεται φάλαγξ· καὶ ὁ ἡγούμενος, στρατηγός· τοῦ δὲ ἱππικοῦ, ἵππαρχος. τὸ δὲ συναμφότερον πεζοί τε καὶ ἱππεῖς, στρατιά. τῆς δὲ στρατιᾶς τὸ μέτωπον λεγόμενον, ὃ καὶ πρῶτον ζυγὸν καλοῦσι, πρωτοστάται· καὶ ὁ ἡγούμενος, πρω- τοστάτης· οἱ δὲ παρ’ ἑκάτερα ταττόμενοι κέρατα, δεξιόν τε καὶ εὐώνυμον, οἱ αὐτοὶ καὶ παραστάται· ὁ δὲ ἀρχὸς ὁμωνύμως· οἱ δὲ ὄπισθεν αὐτῶν ὡς ἐπὶ τὸ βάθος ταττόμενοι, ἐπιστάται· τὸ δὲ ἐπὶ τούτοις πρὸς τὸ βάθος ἔσχατον ζυγόν, οὐρὰ καὶ οὐραγία, καὶ οὐραγός· ὁ αὐτὸς δὲ ὀπισθοφύλαξ· ἡ συνήθεια πεντηκοντοφύ- λακα αὐτὸν καλεῖ. ὁ δέ γε τοῦ παντὸς στρατοῦ ἡγεμονεύων, βασιλεύς. In this process of amalgamation the author of the article not only struggled to reconcile the two conflicting traditions, but also sought to amplify his source material on the basis of his own often-misconceived deductions, which suggest that he was both devoid of expertise in the practice and literature of war- fare and writing at a significant chronological remove from his sources. While it is not possible to divine when and by whom this mélange was executed, the multi-source, “cut-and-paste” method of its author resembles that of the compilator of the 224 THE ETYM.MAGN. AND THE “FRAGMENT OF URBICIUS” Etymologicum Magnum as a whole, as previously exemplified by his blending of the glosses s.v. στρατός from the Etymologicum Genuinum and Etymologicum Gudianum. Given this similarity of technique, combined with the primarily lexical interest of the item, it is legitimate to speculate that the “fragment” was not a pre-existing article, but that its author and the compilator of the Etymologicum Magnum were one and the same. October, 2006 Seminar für Alte Geschichte und Epigraphik University of Heidelberg Marstallhof 4 69117 Heidelberg, Germany prr@fastnet.co.uk