Greek, Roman, and Byzantine Studies 46 (2006) 231–248 © 2006 GRBS Regional Texts and the Circulation of Books: the Case of Homer Margalit Finkelberg HIS PAPER is an attempt at contextualization of the transmission of Homer’s text in the Hellenistic, Roman, and Byzantine periods. Its main thesis is that until the emergence of Constantinople as the central authority for the Greek-speaking world, the production, circulation, and trans- mission of books concentrated around self-contained regional centres and that Alexandria was only one such centre among others. Comparison with the patterns of circulation of early Christian texts suggests some useful parallels. 1. Books and central authority Ptolemy II Philadelphus, the second ruler of Hellenistic Egypt, was born on the island of Cos in 308 B.C.E. Some time afterwards, Ptolemy I appointed the poet and scholar Philetas of Cos as his son’s tutor. Ptolemy I’s original intention was probably nothing more than to follow the precedent set by Philip II of Macedon, who invited Aristotle to be tutor to the young Alexander. The king could hardly have been aware that in creating the position of royal tutor he was himself establish- ing a precedent whose historical consequences were to go far beyond the immediate objective of giving a proper education to his heir. Ptolemy II was crowned in 283. Not long after, he founded a library in his capital Alexandria and appointed his second tutor, Philetas’ pupil Zenodotus of Ephesus, as its head (prostatês). Thus a tradition was established according to which the tutor to the children of the royal family was also head of the Library of Alexandria. The great Homeric scholars Zenodotus, Aristophanes of Byzantium, and Aristarchus of Samothrace, each of them responsible for a recension of the Homeric T 232 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS poems,1 take pride of place in this list. The Alexandrian schol- ars were thus members of the Ptolemaic court and an integral part of the Establishment.2 This close connection between scholarship and power is rarely taken into account in current treatments of Hellenistic scholarship.3 We are still invited to envisage the relations be- tween scholars or the texts they produced in the vein of Romantic criticism, that is, as resulting from the interaction of great minds taking place beyond time and space. It seems, however, that approaching Hellenistic scholarship, in Alex- andria and elsewhere, from the perspective of its social context is indispensable for a proper understanding of the history of the texts at our disposal, first and foremost the text of Homer. Since the publication in 1788 of the Venetus A, a magnifi- cently produced manuscript of the Iliad accompanied by copious scholia throwing much light on the Alexandrian philo- logical tradition, it has become increasingly clear that, in spite of the great prestige of the Alexandrian scholars, first and fore- most Aristarchus of Samothrace, the Byzantine text of Homer —and, consequently, our text—contains a negligibly small number of the readings they proposed. Thus, of 385 Zenodo- tean readings attested in our sources, only 4 appear in all the 1 Throughout this paper I translate the Greek term diorthôsis as “recen- sion.” Cf. R. Pfeiffer, History of Classical Scholarship from the Beginning to the End of the Hellenistic Age (Oxford 1968) 94; G. Nagy, Poetry as Performance. Homer and Beyond (Cambridge 1996) 115–116. Another possible rendering of the term, introduced in G. Nagy, Homer’s Text and Language (Urbana/Chicago 2004) 22, 85, is “corrective editing.” 2 The position of royal tutor is explicitly attested for Zenodotus, for his successor Apollonius of Rhodes, and for Aristarchus of Samothrace; in view of this it is difficult to avoid the conclusion that the two other heads of the Library, Eratosthenes and Aristophanes, also held this position. Cf. Pfeiffer, History 154; A. W. Bulloch, “Hellenistic Poetry,” in P. E. Easterling and B. M. W. Knox (eds.), The Cambridge History of Classical Literature I.4 The Hellenistic Period and the Empire (Cambridge 1989) 4. 3 See however Pfeiffer, History 98; S. West, The Ptolemaic Papyri of Homer (Cologne/Opladen 1967) 16; P. E. Easterling, “Books and Readers in the Greek World: The Hellenistic and Imperial Periods,” in Easterling and Knox, Cambridge History 186–187; H. Y. Gamble, Books and Readers in the Early Church. A History of Early Christian Texts (New Haven/London 1995) 177. MARGALIT FINKELBERG 233 medieval manuscripts, 34 in the majority, 12 in about half, 29 in a minority, 64 in a few, and 259 in none; of 874 Aristarchan readings, only 80 appear in all the medieval manuscripts, 160 in the majority, 76 in about half, 181 in a minority, 245 in a few, and 132 in none; and so on.4 As Robert Lamberton puts it, “One might have expected Aristarchus to prevail, but the vulgate text is not his.”5 The other major source demonstrating beyond doubt that the scholars’ texts of Homer exerted no significant influence on those in the hands of the general public is the evidence supplied by the papyri. As is well known, the number of ancient manuscripts containing segments of the Homeric poems, first and foremost the Iliad, unearthed in the sands of Egypt has been especially impressive.6 The evidence of the papyri shows that, whatever the differences between the ancient manuscripts themselves (see below), they were even less influenced by the readings proposed by Alexandrian scholars than was the medieval vulgate.7 4 For the figures see T. W. Allen, Homeri Ilias I Prolegomena (Oxford 1931) 199–204; cf. M. J. Apthorp, The Manuscript Evidence for Interpolation in Homer (Heidelberg 1980) 9; R. Janko, “The Text and Transmission of the Iliad,” in R. Janko (ed.), The Iliad: A Commentary IV (Cambridge 1992) 22 n.9. 5 R. Lamberton, “Homer in Antiquity,” in I. Morris and B. Powell (eds.), A New Companion to Homer (Leiden 1997) 14; cf. Pfeiffer, History 215; West, Ptolemaic Papyri 16. Actually, this was the conclusion arrived at by Wolf im- mediately upon the publication of the Venetus A scholia: F. A. Wolf, Pro- legomena to Homer, transl. with introduction and notes A. Grafton, G. W. Most, and J. E. G. Zetzel (Princeton 1985) 68. In Wolf’s opinion, however, the discrepancy between Aristarchus’ readings and the medieval vulgate was due to the corruption of the “pure form of the Aristarchan text” in the course of time; yet, the evidence of the papyri, some of which go back to the time of Aristarchus himself, unambiguously testifies to the opposite. 6 See further H.-I. Marrou, Histoire de l’éducation dans l’Antiquité6 I (Paris 1964) 244–245; N. G. Wilson, Scholars of Byzantium (London 1983) 18–19; M. Haslam, “Homeric Papyri and Transmission of the Text,” in Morris and Powell, New Companion 60–61; T. Morgan, Literate Education in the Hel- lenistic and Roman Worlds (Cambridge 1998) 69, 105; R. Cribiore, Gymnastics of the Mind. Greek Education in Hellenistic and Roman Egypt (Princeton 2001) 140–142, 194. 7 See Allen, Homeri Ilias 83–85; K. McNamee, “Aristarchus and ‘Every- man’s’ Homer,” GRBS 22 (1981) 247–255, at 247; S. West, “The Trans- 234 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS The only province in which the Alexandrian scholars seem to have had an influence on the texts of Homer that circulated in Hellenistic and Roman Egypt was the numerus versuum, that is, the number of lines constituting the Homeric poems. As the evidence provided by the papyri demonstrates, after ca. 150 B.C.E. variants and especially additions, found in abundance in the earlier papyri (the so-called “wild” or “eccentric” papyri), suddenly disappear. This coincides with the activity of Ari- starchus of Samothrace, who undertook to purge the text of Homer of extra lines, in particular repetitions, which had accumulated in the course of centuries.8 The majority opinion tends to regard this phenomenon as a by-product of the book trade: while customers knew enough about the newly estab- lished length of the poems to be interested in purchasing a copy of the improved text, they did not care much about textual variants. Yet it is hard not to agree with Michael Haslam that this explanation does not really solve the problem: “If we imagine an Aristarchan text available to proprietors of scrip- toria, it has to be explained why they should not simply have reproduced it. We are invited to envisage a public so insistent to have Aristarchan texts that all non-Aristarchan lines were cancelled, yet so ignorant and uncaring that Aristarchan read- ings could be routinely passed over.”9 Probably owing to the above-mentioned tendency to treat scholarship in isolation from its social context, it has been generally overlooked that the standardization of the numerus versuum coincided with major political developments in Ptol- emaic Egypt. In 145/4 B.C.E., after the death of Ptolemy VI, his younger brother usurped the throne of Egypt to become Ptolemy VIII. The new king’s former tutor and head of the Library, Aristarchus of Samothrace, was forced to leave Alex- andria together with his pupils (the so-called secessio doctorum), and a senior military officer named Cydas was appointed chief ___ mission of the Text,” in A. Heubeck, S. West, J. B. Hainsworth, A Commen- tary on Homer’s Odyssey I (Oxford 1988) 44–45; Janko, Iliad 22 n.9. 8 For discussion see West, Ptolemaic Papyri 15–17; Apthorp, Manuscript Evi- dence 1–14. 9 Haslam, in Morris and Powell, New Companion 84–85. MARGALIT FINKELBERG 235 librarian.10 This is exactly when the “eccentric” papyri became obsolete. The coincidence is telling, to say the least, and this is why I find it odd that Homeric scholars as a rule do not try to establish a correlation between the two events. The only excep- tion seems to have been Gregory Nagy, who argued that, just as the first standardization of the text of Homer in the sixth century B.C.E. was due to the regulation of rhapsodic perfor- mances by the Athenian state, so also the second standardiza- tion in the middle of the second century B.C.E. should be explained by similar reasons.11 If we further take into account (a) that Ptolemy VIII was a pupil of Aristarchus and therefore must have been closely familiar with the latter’s text of Homer; (b) that, to judge from the conjecture for Odyssey 5.72 that he proposed,12 he claimed to be a Homeric scholar; and (c) that his new head of the Library was a virtual nonentity as a man of learning, it can be suggested that it was due to the personal initiative of this Ptol- emaic ruler that the curious phenomenon of the adoption of the numerus versuum established by Aristarchus and the rejection of the readings he proposed was produced. This conclusion can be further corroborated by comparative evidence. This unequivocally points up that it is through the intervention of the central authority that the standardization of texts circulating in a given community normally takes place, especially when the texts in question are possessed of a high status in the community. As Moshe Greenberg showed in a seminal article, this was certainly the case with the standardi- zation of the text of the Hebrew Bible: The editorial work of the bookmen did not immediately affect the Bibles in the hands of the people. Only after the consol- idation of Rabbinic Judaism between the two revolts (70–132) did a more thorough supervision of the text on the basis of the standard become possible.13 10 On Cydas see Pfeiffer, History 212 and n.2. 11 Nagy, Poetry 144. 12 See further Pfeiffer, History 212. 13 M. Greenberg, “The Stabilization of the Text of the Hebrew Bible, Reviewed in the Light of the Biblical Materials from the Judean Desert,” 236 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS The same with the early Christian tradition. To quote Kurt and Barbara Aland, The more loosely organized a diocese, or the greater the differ- ences between its constituent churches, the more likely different text types would coexist (as in early Egypt). The more uniform its organization, the more likely there would be only a single text type, as exemplified by the Byzantine Imperial text type [of the New Testament] which expanded its influence rapidly from the fourth century to become increasingly the dominant text of the Byzantine church.14 I see no reason why Ptolemaic Egypt should have differed in this respect. It seems indeed reasonable to suppose that in its case too it was first and foremost the intervention of a central authority, the state, that was responsible for the standardization of the Egyptian text of Homer in the middle of the second century B.C.E. Whether this standardization should also be projected onto the rest of the Hellenistic world is however an entirely different question. 2. Regional centres of learning and book production Antiochus I became sole king of the Seleucid realm in 281 B.C.E., and Antigonus II Gonatas became king of Macedonia in 277. Everything points in the direction that these rulers and their successors followed the example of the Ptolemies in turn- ing their capitals into centres of learning. During the reign of Antigonus II, a literary circle flourished at the court at Pella.15 It included the poets Alexander Aetolus, who had also worked in Alexandria, and Aratus of Soli, author of the Phaenomena, ___ JAOS 76 (1956) 157–167, at 161; also 166, “The editing was a continuing process which reached its end by the first Christian century, well before the First Revolt. The standard became all prevalent, however, only after the fall of Jerusalem, when Rabbinic Judaism came into exclusive hegemony. Pre- viously the various stages of the text work coexisted in the Bibles of the people.” Cf. also F. Kenyon, Our Bible and the Ancient Manuscripts5, revised by A. W. Adams (London 1958) 72. 14 K. Aland and B. Aland, The Text of the New Testament. An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism2 (Grand Rapids 1989) 55. 15 See further Pfeiffer, History 107–108. MARGALIT FINKELBERG 237 credited in our sources with a recension of the Odyssey. At some later stage, however, Aratus joined the court of Antioch, where he is said to have been commissioned by Antiochus I to pre- pare a new recension of the Iliad; he apparently returned to Pella after that king’s death.16 Moreover, the evidence at our disposal leaves no room for doubt that by the end of the third century B.C.E. both Antioch and Pella had royal libraries of their own: sometime after 223 the poet Euphorion of Chalcis was appointed by Antiochus III as head of the library at Antioch,17 whereas the very fact that in 168 B.C.E. the royal Macedonian library (which probably con- tained the copy of the Homeric poems made for Cassander of Macedonia long before the foundation of the library at Alex- andria) was seized by Aemilius Paulus and transferred from Pella to Rome shows clearly enough that such a library had been in existence.18 All this allows us to conclude that, even if on a lesser scale than Alexandria, the capitals of the two other 16 See Vita I: “and he also prepared a recension of the Odyssey, a recension that is called ‘Aratean’, in the manner of ‘Aristarchean’ and ‘Aristo- phanean’. Some say that he went to Syria and was introduced to Antiochus and was requested by him to produce a recension of the Iliad, because (so he said) its text had been spoilt by many” (ka‹ tØn ÉOdÊsseian d¢ di≈ryvse, ka‹ kale›ta¤ tiw diÒryvsiw oÏtvw ÉArãteiow …w ÉAristãrxeiow ka‹ ÉAristofãneiow. tin¢w d¢ aÈtÚn efiw Sur¤an §lhluy°nai fas‹ ka‹ gegon°nai par' ÉAntiÒxƒ ka‹ ±ji«syai Íp' aÈtoË Àste tØn ÉIl¤ada diory≈sasyai, diå tÚ ÍpÚ poll«n lelumãnyai); Suda A3744 “Aratus … A recension of the Odyssey” (ÖAratow … diÒryvsin ÉOdusse¤aw), in J. Martin (ed.), Scholia in Aratum vetera (Stuttgart 1974) 8.18–24; 22.18. On Aratus see Pfeiffer, History 120–122; A. Rengakos, Der Homertext und die hellenistischen Dichter (Stuttgart 1993) 10–11, 165–166. Wolf, Prolegomena 166, suggests that Aratus’ work on the Iliad was completed by Rhianus, whose edition is often mentioned in the scholia. 17 Suda E3801. Cf. Pfeiffer, History 122. 18 On Cassander see Ath. 620B: “Cassander who was king of Macedonia, of whom Carystius says in his Historical Commentaries that he admired Homer to such a degree that he knew most of his verses by heart and even had in his possession the Iliad and the Odyssey written down especially for him” (Kãsandron tÚn Makedon¤aw basileÊsanta, per‹ o fhsi KarÊstiow §n ÑIstori- ko›w ÑUpomnÆmasin ˜ti oÏtvw ∑n filÒmhrow …w diå stÒmatow ¶xein t«n §p«n tå pollã: ka‹ ÉIliåw ∑n aÈt“ ka‹ ÉOdusse¤a fid¤vw gegramm°nai). On the expres- sion fid¤vw gegramm°nai see Wolf, Prolegomena 157 n.37; Nagy, Poetry 158 and n.20. 238 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS major Hellenistic kingdoms were centres of learning in their own right. It was, however, the fourth Hellenistic kingdom that pro- duced a library destined to become the full-scale rival of the Library of Alexandria. In 197 B.C.E. Eumenes II succeeded Attalus I on the throne of Pergamum. There is good reason to suppose that scholarly research and book collection at the Attalid court predate this event.19 Yet it was Eumenes II who founded a library, and it soon became the second major in- stitution of this kind in the Hellenistic world. After several un- successful attempts to bring to Pergamum some of the leading scholars of the time, among them Aristophanes of Byzantium, Eumenes II succeeded in appointing the Stoic philosopher Crates of Mallos as head of the library. Crates’ rivalry with Aristarchus of Samothrace, the chief librarian at Alexandria, is too well known to be dwelt upon here: suffice it to say that Crates’ interpretative strategies were very different from those of Aristarchus and that he too prepared a recension of Homer which, naturally enough, was heavily influenced by the strat- egies in question.20 Note that not only Zenodotus, Aristophanes, and Aristarchus but also other scholars who occupied the position of chief librarian at Alexandria and other Hellenistic courts either produced their own texts of Homer or wrote extensive com- mentaries on the Homeric poems. The first Ptolemaic tutor, Philetas of Cos, collected rare Homeric words and suggested 19 See esp. E. Kosmetatou, “The Attalids in the Troad. An Addendum: an Episode of the Perils of the Aristotelian Corpus,” AncSoc 33 (2003) 53–60, at 56–58; cf. Pfeiffer, History 246–247. 20 On Aristarchus and Crates see J. I. Porter, “Hermeneutic Lines and Circles: Aristarchus and Crates on the Exegesis of Homer,” in R. Lamber- ton and J. J. Keaney (eds.), Homer’s Ancient Readers. The Hermeneutics of Greek Epic’s Earliest Exegetes (Princeton 1989) 67–114; M. Finkelberg, “‘She Turns about in the Same Spot and Watches for Orion’: Ancient Criticism and Exegesis of Od. 5.274 = Il. 18.488,” GRBS 44 (2004) 231–244. On the re- cension of Crates see Suda K2342; G. Nagy, “The Library of Pergamon as a Classical Model,” in H. Koester (ed.), Pergamon: Citadel of the Gods (Harrisburg 1998) 215. On the rivalry between Pergamon and Alexandria see J. Wyrick, The Ascension of Authorship. Attribution and Canon Formation in Jewish, Hellenistic, and Christian Traditions (Cambridge [Mass.] 2004) 214–220. MARGALIT FINKELBERG 239 interpretations which were still influential at the time of Aristarchus of Samothrace; Zenodotus’ successor Apollonius Rhodius wrote a book Against Zenodotus and suggested his own readings of the Homeric text; Apollonius’ successor Eratos- thenes dealt extensively with Homeric problems in his chrono- logical and especially geographical work; Aratus and Crates prepared their own recensions of the Homeric poems; and so on.21 It seems to follow from this that the credentials of an expert Homerist were a major requirement for the position of chief librarian, both in Alexandria and elsewhere. Given the fact that the superior status of Alexandrian scholarship was generally recognized in antiquity, it may be asked why all these alternative texts of Homer continued to be produced. At the end of the preceding section, we came to the con- clusion that the standardization of a text can only be achieved through the intervention of a central authority. No overarching central authority, however, was in existence in Hellenistic Greece. Or, to put it the other way round, there were no fewer than four such authorities, one for each of the major Hellenistic kingdoms. It follows from this that, when speaking of the authoritative Hellenistic text of Homer, we should abandon the usual practice of privileging Alexandria, which after all was only one Hellenistic capital among others, and envisage at least four such texts—Egyptian, Syrian, Macedonian, and Perga- mene. Indeed, it would be hard to imagine circumstances under which, say, a Seleucid king might choose to educate his heir by using an edition of Homer imported from the court of his arch-enemy (six wars were waged between Ptolemaic Egypt and Seleucid Syria between 274 and 168 B.C.E.): it is much more likely that he would wish to emphasize his state’s cultural self-sufficiency by adopting an edition produced by a Homeric scholar working at his own court or at that of one of his predecessors. In so far indeed as the library was an institution of the state and Homer the most esteemed text in the Greek world, the production of the state text of Homer must have 21 On Philetas of Cos see Pfeiffer, History 90–91; on Apollonius Rhodius 146–147; on Eratosthenes 163–168; on Aratus see above, n.16; on Crates, n.20. 240 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS been considered not only a scholarly enterprise but also a mat- ter of national prestige. The Antigonid kingdom of Macedonia was dissolved in 168 B.C.E., and in 149/8 Macedonia became a Roman province; it was followed by Pergamum in 133. Seventy years later, Pom- pey made Syria a Roman province, and in 31 B.C.E. Egypt, the last independent Hellenistic kingdom, fell to Octavian. The centuries-old bond between Hellenistic scholarship and the State was broken. The Roman takeover in the East also had far-reaching repercussions in terms of the physical existence of the Hellenistic libraries. As already mentioned, the Pella library was transferred to Rome as early as 168 B.C.E. Somewhere around 41 the Pergamum library was either entirely or in part transferred to Alexandria by Marcus Antonius, to recoup the damage caused by fire in the siege of the city by Julius Caesar. Consequently, Alexandria and Antioch, now the capitals of Roman provinces, remained the only two former Hellenistic centres where the tradition of a library as the focus of learning and book production continued unbroken into the Roman era. Aristonicus and Didymus, who worked in Alexandria in the time of Augustus, were still able to consult the manuscripts by Zenodotus, Aristophanes, and Aristarchus, and the same was true of Nicanor and Herodian in the second century C.E.: the work of these four scholars forms the main corpus of the scholia preserved in Venetus A.22 Strabo testifies to the fact that Tiberius (14–37 C.E.) supported the Museum, and there is no reason to think that the other emperors behaved differently.23 Our evidence concerning Antioch relates to a later period (see be- low), but the fact that, excluding the short interval between 194 and 201, the city retained its status as capital of the province of Syria allows hardly any doubt that its position as a centre of 22 See further M. West, Studies in the Text and Transmission of the Iliad (Munich/Leipzig 2001) 46–50. 23 Strab. 17.1.8 (794): “The Museum is also a part of the royal palaces; it has a public walk … and a large house, in which is the common mess-hall of the men of learning who share the Museum (t«n metexÒntvn toË Mouse¤ou filolÒgvn éndr«n). This group of men not only hold property in common, but also have a priest in charge of the Museum, who formerly was ap- pointed by the kings, but is now appointed by Caesar” (transl. H. L. Jones). MARGALIT FINKELBERG 241 learning was sustained during this period as well.24 Meanwhile, significant developments were taking place in the production and circulation of books. Classical literature was read in Origen’s school at Caesarea, a city that gradually emerged as a major centre of learning and book production, and Origen adapted to the Old Testament the system of mar- ginal signs used by the Alexandrians (233–238 C.E.). A greater need of educated men in the civil service is clearly discernible from the time of Diocletian at the end of the third century, and Greek was gradually replacing Latin as “the language of pow- er” of the Eastern Empire.25 The two latter developments must have further secured the central position of Homer, the corner- stone of Greek paideia, in the production and circulation of books. Whose text of Homer was regarded as authoritative during this period is not clear, but the location of the major scriptoria, alongside the evidence concerning the transmission of the text of the New Testament, strongly suggests that there were at least two such texts—the Egyptian (Alexandria) and the Syrian (Antioch). Toward the end of the third/beginning of the fourth century, the New Testament Koine text, which would eventually take over from the Alexandrian Koine to become the Byzantine Im- perial Text, was consolidated in Antioch and rapidly expanded its influence. I believe that this should be regarded as a fact of major importance in any reconstruction of the history of the Byzantine text of Homer, and I find it surprising that, to the best of my knowledge, T. W. Allen has been the only scholar to suggest that the history of the New Testament text may throw 24 Cf. Strab. 16.2.5 (750): “Antioch is the metropolis of Syria; and here was established the royal residence for the rulers of the country. And it does not fall much short, either in power or in size, of Seleuceia on the Tigris or Alexandria in Egypt.” Three hundred years later Libanius, himself an Antiochene, refers to the city as “the metropolis of Asia” (Or. 11.130, 187). See also below, with n.33. 25 K. Butcher, Roman Syria and the Near East (London 2003) 332, and 270– 277. Cf. Marrou, Histoire II 51–52, 122–123; M. Sartre, The Middle East under Rome, transl. C. Porter and E. Rawlings (Cambridge [Mass.] 2005) 275–291. 242 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS light on the transmission of the text of Homer.26 As to the more general question of the degree to which such extrapolation from the early Christian to the Hellenic manuscript tradition would be methodologically valid, I suggest, together with Harry Gamble, that in cases such as this the lack of informa- tion should be compensated for by the evidence supplied by the larger social context: Yet no differentiating features of early Christianity require us to think that the publication and circulation of early Christian texts proceeded along unique or idiosyncratic lines. Without evidence to the contrary, it ought to be supposed that Christian writings were produced and disseminated in much the same way as other literature within the larger environment.27 I suggest that Homerists should in turn watch closely the early Christian manuscript tradition, if only for the reason that eventually both the text of the New Testament and that of Homer emerge as Byzantine Imperial Texts, produced in the same scriptoria and treasured in the same libraries. This draws our attention again to the two major centres of learning that sustained their position from the Hellenistic period—Alexan- dria and Antioch. 26 Allen, Homeri Ilias 196 n.1: “A certain parallel may be found in the medieval text of the Greek Scriptures. This, known as the Byzantine text, survived, to the almost complete extinction of the other versions, owing to political and geographical circumstances, viz. the Saracenic conquest of Egypt, and the decay of Greek in the West.” As we shall see, although the Arab conquest of Egypt undoubtedly contributed to the isolation of Alex- andria from the rest of the Greek-speaking world, it was far from being the decisive factor that marginalized the Alexandrian recension of the New Testament. 27 Gamble, Books 93–94; cf. 82–83: “Any item of early Christian literature is available to us only in consequence of a long process of transmission. Given our indebtedness to that process, it is easy to forget that the trans- mission of a text over time is but an accidental function of its currency in its own time—of the extent to which it was duplicated, distributed, and used soon after it was composed … Since early Christian sources shed little direct light on these issues, we must again attend to the larger context and ask first how non-Christian Greek and Latin literature was published and circulated during the early centuries of the church.” MARGALIT FINKELBERG 243 3. Regional into dominant The fourth century C.E. was one of the most significant periods in the history of books in the Greek-speaking world. To begin with, it saw some major shifts in the technology of book production. Parchment took over from papyrus, the papyrus scrolls were transferred to codices, and the copies came to be made by state-supported scribes. Changes in the political sphere were no less impressive. In 330 the seat of government was transferred to Constantinople. The importance of this fact for the history of books is hard to overestimate, for a single city was to become, for the first time in history, the central religious, cultural, and educational authority for the entire Greek-speaking world. The foundation of Constantinople was accompanied by a considerable increase in the civil service: government required greater numbers of administrators of liberal education and a good prose style. In 357 Constantius II issued an edict encouraging liberal education (Cod.Theod. 14.1.1); he also founded in Constantinople a library of secular books transcribed by professional scribes working at imperial expense (Them. Or. 4.59–61). That this library was the object of considerable care on the part of the Christian rulers of the Eastern Empire follows from the fact that in 372 Valens or- dered the employment of seven copyists to maintain and repair the books (Cod.Theod. 14.9.2). Since, as we have seen, there had been several centres of learning and book production in the Greek-speaking world, each issuing its own texts, it is evident that the choice of one recension of a given text over another for the imperial library and scriptorium was crucial for its survival. Again, this is best illustrated by the history of the Christian texts. As the Church Father Jerome makes clear, towards the end of the fourth century the circulation areas of the Greek Old Testament texts closely corresponded to the geographical distribution of the major scriptoria. Hesychius was known as the editor of the text used in Alexandria and Egypt, while “in Constantinople and as far as Antioch copies made by the mar- tyr Lucian are regarded as authoritative; the provinces between these two read the Palestinian manuscripts prepared by Origen 244 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS and widely promoted by Eusebius and Pamphylius.”28 Kurt and Barbara Aland remark on this: Thus Jerome mentions three major types of text for the Greek Old Testament, and his statement confirms the view expressed above that these types … survived because they were the texts of the great scriptoria (of Alexandria, Caesarea, and the episcopal sees from Antioch to Constantinople, i.e., the Antiochene school), and consequently circulated almost exclusively.29 As this evidence demonstrates, it is the Antiochene text of the Greek Old Testament that was regarded as authoritative in Constantinople; as we saw, this was true of the text of the New Testament as well.30 When in the seventh century the Arabs conquered both Antioch and Alexandria, this affected only the Alexandrian recension of the New Testament, which became much less widespread.31 The reason is clear: by the very fact of choosing the Antiochene text several centuries earlier, the Con- stantinople authorities had guaranteed its survival even when Antioch itself ceased to be part of the empire.32 28 Jerome Praef. in Paralip.; transl. Aland and Aland. 29 Aland and Aland, Text 66. Cf. T. W. Allen, Homer. The Origins and the Transmission (Oxford 1924) 314–320. 30 Cf. L. Vaganay, An Introduction to New Testament Textual Criticism2 (Cam- bridge 1991) 126: “This period saw the triumph of the Byzantine text, otherwise known as the ecclesiastical or imperial text. The name is of little importance but the fact of the predominance of this form of the text is strik- ing. It developed from the Antiochene recension and dominated the whole of the Greek-speaking East.” Cf. Kenyon, Our Bible 169–178; Aland and Aland, Text 55 (quoted above, 236). According to Aland and Aland (66), in spite of the fact that in 332 Constantine ordered from Eusebius of Caesarea fifty parchment copies of the New Testament for use in the churches of Constantinople, it is far from certain that Caesarea developed its own distinctive New Testament text alongside those of Alexandria and Antioch. For a different opinion see Vaganay (below, 247). 31 Antioch was reconquered by the emperor Nicephorus II Phocas in 969, only to be lost again in 1085, this time to the Seljuk Turks. Thirteen years later, it was taken by the Crusaders, and remained in their hands until its capture by the Mamluks in 1268. 32 Cf. Kenyon, Our Bible 174: “The subsequent acceptance of the An- tiochian or Syrian type as the received text of the Greek New Testament must have been due to the predominant influence of Constantinople.” MARGALIT FINKELBERG 245 What can all this tell us of the history of the text of Homer? We saw that Antioch had a royal library since the Hellenistic period, and that this library, like those of the other Hellenistic kingdoms, must have possessed its own text of Homer. The recension of the Iliad commissioned by Antiochus I from Aratus comes to mind in this connection. It is immaterial for our pur- poses whether or not it was Aratus’s recension that eventually became the Seleucid state text of Homer: what matters is the king’s intention to produce such a text. We also saw that, as far as the Greek Old Testament and New Testament texts are concerned, the Constantinople authorities consistently pre- ferred Antioch (or, sometimes, Caesarea, see n.30) over Alex- andria, and that in terms of the regional circulation of books Constantinople and Antioch belonged to the same area. Finally, as Libanius amply demonstrates, in the fourth cen- tury—that is, in the time of the consolidation of the Antiochene recension of the New Testament—the citizens of Antioch continued to found schools of secular learning and the city sustained its position as the centre of “Hellenic paideia.”33 This circumstantial evidence is of course insufficient in itself to allow the unequivocal conclusion that it was the Antiochene rather than some other regional text of Homer that made its way to Constantinople, eventually to become the Byzantine Imperial Text. Yet, given the scarcity of Alexandrian readings in the medieval vulgate, the available evidence seems to explain satis- factorily enough why it was some other recension of Homer 33 See esp. Or. 11.188.8–10: “You [the citizens of Antioch] … have been lavish in building temples of the Muses [i.e. schoolrooms] for students to attend, and as a reward to the goddesses, and you both employ citizens and hold no grudge against foreigners as teachers” (MoÊsaiw te flerå polutel«w ofikodome›sye n°oiw m¢n §ndiatr¤bein, ta›w d¢ yea›w g°raw, ka‹ didaskãloiw pol¤taiw te xr∞sye ka‹ oÈ fyone›te j°noiw); 270.6–9: “Moreover, if she [An- tioch] be inferior to any in respect of her walls, she yet surpasses that town [sc. Constantinople] … both in the wit of her inhabitants and in the pursuit of philosophy, and she rises superior to the city still greater [sc. Rome] in the most noble feature of all, in Greek education and oratory” (ka‹ mØn ∏w m¢n ≤ttçtai katå toÁw to¤xouw, taÊthw kre¤ttvn g¤netai … ka‹ tª t«n §noikoÊn- tvn ésteiÒthti ka‹ tª t∞w sof¤aw éskÆsei, t∞w d¢ ¶ti me¤zonow t“ kall¤stƒ kall¤vn §st¤n, ÑEllhnikª paide¤& ka‹ lÒgoiw). Transl. A. F. Norman, slightly changed. See also Wilson, Scholars 28–30. 246 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS than an Alexandrian one that eventually became the standard text of the Byzantine Empire.34 This conclusion, however, demands certain qualifications. First, in both the Hellenistic and the Roman periods there was much interchange between scholars working at different cen- tres of learning. As we saw, Aristarchus reacted to the critical methods of Crates and the readings he introduced, which clearly indicates that these two scholars were aware of each other’s work (above, 238). It is true that Aristophanes of Byzan- tium is said to have been imprisoned in Alexandria and thus prevented from accepting the invitation from Pergamum, but we saw that Aratus, for one, moved freely between Pella and Antioch, and the same was probably true of other scholars.35 The secessio doctorum, Ptolemy VIII’s exiling of Aristarchus and his pupils from Alexandria, seems to have especially con- tributed to the spread of Alexandrian scholarship over the Hel- lenistic world.36 In view of this, it is reasonable to suppose that Alexandrian texts of Homer, and especially the one produced by Aristarchus, continued to be consulted in Egypt and else- where during the Roman period as well: this was probably one of the ways by which Alexandrian readings entered the medie- val vulgate.37 Second, it should be kept in mind that the general as- sumption concerning the transliteration of books set off by the transition from uncial to minuscule writing in the ninth cen- tury, namely, that one minuscule copy was made from one uncial copy and became the source of all subsequent copies, 34 Cf. Allen, Homer 324: “Alexandria naturally continued to produce books … but it had never a monopoly for which nothing but the indigenous papyrus fitted it.” With Arthur Ludwich, Allen was among the few scholars who paid attention to the historical context of the consolidation of the vul- gate text of Homer. Both emphasized that, rather than to the activity of the philologists, the uniformity of the medieval vulgate was due to the condi- tions of publication, which eventually led to the survival of one recension of Homer over the others. 35 On Aristophanes see Pfeiffer, History 172; on Aratus see above, n.16. 36 Cf. Pfeiffer, History 253. 37 See Allen, Homeri Ilias 85, 209–216; Apthorp, Manuscript Evidence 9; Janko, Iliad 22 n.9; Haslam, in Morris and Powell, New Companion 94–95. MARGALIT FINKELBERG 247 does not apply to texts that were in especially great demand: as far as the texts that enjoyed a firm position in the Byzantine curriculum are concerned, it was often more than one uncial manuscript that came to be transliterated.38 This would of course be especially true in the case of Homer.39 Since it is thus highly likely that more than a single manuscript of the Homeric poems was transcribed into minuscule at the end of the first millennium,40 we cannot exclude the possibility that an Alex- andrian text of Homer was one of them. Yet, as the example of the New Testament manuscript tradition demonstrates, al- though copies of other recensions of the New Testament con- tinued to be produced here and there, they were as a rule made to conform to the dominant Syro-Byzantine text and were contaminated by its influence. Leon Vaganay’s assessment of the relevant evidence appears to be in place here: Generally speaking, the Alexandrian recension, which had not been very widespread since the Arab conquest (seventh century), is not common … The Caesarean recension has more witnesses but they are almost all contaminated by a Byzantine influence … Some other groups are so disfigured that they seem to repre- sent the Syro-Byzantine text slightly corrected with Caesarean variants.41 If we take into account that the medieval tradition of the Homeric poems has also proved to be heavily contaminated,42 it would be reasonable to suppose that in their case, too, the situation could not be much different. Again, this does not mean that the Alexandrian recension disappeared without trace. Yet, although it is likely that it influenced the dominant 38 See further L. D. Reynolds and N. G. Wilson, Scribes and Scholars. A Guide to the Transmission of Greek and Latin Literature (Oxford 1968) 52–53; Has- lam, in Morris and Powell, New Companion 89. 39 On the privileged position of Homer in the Byzantine curriculum see e.g. Reynolds and Wilson, Scribes 55; R. Browning, “The Byzantines and Homer,” in Lamberton and Keaney, Homer’s Ancient Readers 136. 40 See Haslam, in Morris and Powell, New Companion 92–93. 41 Vaganay, Introduction 127. 42 See Haslam, in Morris and Powell, New Companion 89–95, for a bal- anced assessment. 248 REGIONAL TEXTS AND THE CIRCULATION OF BOOKS text in many and various ways, the state of the latter repeatedly demonstrates that it was not an Alexandrian text of Homer that formed the basis of the one we have.43 I hope that the above appraisal of the social and historical context to which it originally belonged shows clearly enough that we should hardly be surprised at this outcome.44 May, 2006 Department of Classics Tel Aviv University P.O.B. 39040, Ramat Aviv Tel Aviv 69978, Israel finkelbe@post.tau.ac.il 43 Cf. Kenyon, Our Bible 171, on the text of the New Testament: “The Alexandrian text is represented, not so much by any individual MS. or version, as by certain readings found scattered about in manuscripts which elsewhere belong to one of the other groups.” 44 An earlier version of this paper was read at the international workshop “The Dominion of Letters: the Roles of Books in Ancient Societies” held at the Institute for Advanced Studies, Jerusalem, in June 2005. My thanks are due to the participants in the workshop for their comments and discussion. I am also grateful to the anonymous referee and the editor of this journal for their helpful remarks.