This book is a companion volume to M.L. West's edition of the Iliad (Stuttgart-Leipzig-Munich, 1998-2000). The first part (3-170, "an amplification of the first twelve pages of the edition", as the author calls it) includes an overview of the pre-Alexandrian transmission of the Homeric text (3-32), W.' views about Zenodotus' text (33-45) and Didymus' sources (46-85), a complete list of the edition papyri (86-138), a brief description of the manuscripts (139-157) and the author's editorial principles (158-170). The second part includes textual discussion of various Iliadic passages (173-285). There is also a bibliography (286-89) and indices (Greek Words, Passages discussed, General Index, 290-304).
W.'s edition has been repeatedly and exhaustively reviewed (Janko 2000, Nagy 2000, Führer-Schmidt 2001, Nardelli 2001a, 2001b). Therefore in this review I will not discuss W.'s editorial principles but his two most important new theories about Alexandrian Homeric scholarship, i.e. the origin of Zenodotus' text and the sources of Didymus.
The main issue in the study of the ancient tradition of the Homeric text has been the nature of Alexandrian readings: are they manuscript variants or mere conjectures? In recent years a fair number of scholars have very enthusiastically espoused a theory (put forth mainly by Martinus van der Valk 1963-64) according to which the Alexandrian scholars offered mere conjectures and, therefore, their arbitrary and worthless readings could be safely brushed aside (main exponents: R. Janko e.g. 1992: 22ff. and H. van Thiel (1991: IXff., 1992; 1996: VI; 1997); meanwhile, some other, methodologically sound, contributions (K. Nickau's  or M. Apthorp's , for instance) have made hardly any impact at all. Two scholars in particular have recently made important contributions toward a fairer assessment of Alexandrian Homeric criticism: G. Nagy (1996a: ch. 5; 1998; 2000; forthcoming) and F. Montanari (1998; 2000; 2002). The latter has brought noteworthy new arguments to bear on the aforementioned main issue as a whole and various important parts of it, e.g. on the form of the Alexandrian ekdoseis or the two Aristarchean editions of Homer. He is in favor of a conciliatory solution to the problem, repeatedly advocated also by the present reviewer (1993; 2001; 2002), namely that Alexandrian Homeric criticism is a mixture of conjectural criticism and selection between textual variants.
In this debate W. cautiously sides with those who believe that Alexandrian Homeric criticism was limited to conjectures. He is convinced that "it is time to challenge this assumption, inherited from Wolf, that collation of different copies was a normal and essential part of what Aristarchus and his predecessors did...there is no evidence that they actively sought out a plurality of different manuscripts for comparative purposes" (36). He also suggests that the Aristarchean readings that differ from both the vulgate known to Didymus and the medieval vulgate are rarely correct (1998: VII). However, in a few instances he claims that some Alexandrian readings arose from diplomatic considerations. Let us examine his arguments more closely, beginning with his theory about the Homeric text of the first diorthotes, Zenodotus.
This text is labelled "eccentric": it includes "inorganic additional lines", it omits lines and passages and ends up being an "abbreviated text"; it features neo-Ionic and other morphologically late forms, at least four unmetrical readings, ca. ten "dual verb forms or participles used as plurals" und shows a tendency to use "irregular forms of pronouns." Conclusion (37): "the text...is so eccentric that it is impossible to regard it as the product of any rational process of selection of readings from alternative sources ... or indeed of any rational thought at all".
The accusations against Zenodotus' edition are part and parcel of the aforementioned view that the three luminaries of Alexandrian Homeric criticism did not consult manuscripts and that Zenodotus was not even a textual critic at all: "the only kind of textual criticism which we know Zenodotus practised was...concerned with...the identification of spurious lines and passages". According to this view Zenodotus did not select among variants, i.e. he neither adopted manuscript readings nor made conjectures of his own. In W.'s opinion this view is further corroborated by the inability of scholars so far (esp. Duentzer 1848 and Nickau 1977) to posit a definitive criterion on the basis of which to distinguish between genuine readings and Zenodotus' conjectures. W. himself (43) identifies only two readings as Zenodotus' conjectures (Il. 6.155, 20.224). This line of argument leads to the formulation of W.' theory about the origin of Zenodotus' text (the first proponent of the theory was Duentzer 1848: 50.; see also Wackernagel 1916: 61 and 73, and van der Valk 1964: 54): Zenodotus' ekdosis was an Ionian rhapsodic text hailing from Ephesus, Zenodotus' hometown; his only intervention on this text was the introduction of the obelos to athetize lines.
W.'s hypothesis sounds plausible but his view of Zenodotus' performance as a textual critic is not borne out by the available evidence. I will begin with W.'s two assumptions concerning Zenodotus' practice, the alleged failure of the three great Alexandrian philologists to collate manuscripts and Zenodotus' failure to expand his editorial interventions beyond the introduction of the obelos to mark spurious lines. Montanari has very recently addressed and correctly rejected the latter assumption: "it is inconceivable for there to be a theoretical and essential separation which discriminates between verse athetesis and single word alteration...by addressing the issue of the authentic text and how to devise the critical-methodological tools to obtain it, Zenodotus achieved a major breakthrough" (2002: 129-30).
Much more weighty is W.'s other assumption, that neither Zenodotus nor Aristophanes or Aristarchus relied on manuscript authority but that "the first scholars known to have cited manuscript authority for variant readings are Aristarchus' contemporaries Callistratus and Crates" (36). Years ago I turned to the Hellenistic poet-scholars in my attempt to determine whether the Alexandrians relied on manuscript authority (Rengakos 1993; cf. also 2001; 2002). As is well known, the poetic works of Hellenistic scholar-poets make up a remarkably rich collection of memoria Homerica, and on those grounds the study of those works as witnesses to that period's Homeric text is extremely interesting. On the testimony of Hellenistic poets a considerable number of Alexandrian readings can be shown to be of a documentary character. Conjunction errors pointing to the older Homeric tradition, use of a Homeric variant common to a Hellenistic poet, an early Ptolemaic papyrus and an Alexandrian edition, simultaneous occurrence of a vulgate reading and of a variant departing from the vulgate, clear anticipation of readings which had hitherto been known under the name of later Homeric critics -- all this cumulative evidence points to the conclusion that the Alexandrians must have compared different copies of the Homeric text available to them and must have chosen among variae lectiones.
Apart from the indirect clues provided by the Hellenistic poets, there is also direct evidence that the great Alexandrian philologists relied on manuscript authority. First, the scholia: the famous Didymus scholion on Il. 9.222 states clearly that Aristarchus "encountered a certain reading in several manuscripts" (more on that below). The situation is similar in the early Ptolemaic papyri: one may cite as examples four instances from S. West's collection that testify to the use of several manuscripts, a fact often neglected in discussions of the issue in question. On pap. Il. 12 (280-240 BC) S. West (1967:137) notes: "a second, rather cursive, hand...has in many places corrected mistakes and inserted variants. It is not clear whether the latter all come from a single text, or are a selection from various texts, a kind of primitive apparatus criticus". Similarly, see S. West (1967: 223) on pap. Od. 31 (250-200 BC): "the text has undergone a double process of correction and collation. The original scribe appears to have had two MSS. at his disposal...". Also, on pap. Od. 126 (2nd c. BC) see S. West (1967: 263): "a second hand...has inserted variants and corrections". Finally, on pap. Il. 51 (1st c. BC) see S. West (1967: 133): "it looks as if the signs were originally inserted, whether in this papyrus or in an ancestor, by someone who collated his copy with a Vulgate text provided with marginal signs". Is it plausible, then, that the three most significant Alexandrian Homeric critics did not collate manuscripts and that the practice was initiated by a minor Alexandrian Homeric critic such as Callistratus? Or, to quote Montanari again (2002: 134): "should we conceive of a paradoxical Zenodotus who, despite his taste and his concern for the Homeric text, made every effort not to look at other copies he may have come across, not to note the points at which they departed from his own copy and not to ask himself any questions about those differences?" and with regard to the practice of Ptolemaic papyri (2002: 135): "is it believable that, if the diorthotes of a scriptorium compared copies and corrected them as part of his work as a craftsman, such a procedure was not adopted by this new type of diorthotes who, in his capacity as an erudite scholar and an intellectual, was working on the Homeric text?"
Concerning the aforementioned features of Zenodotus' Homeric text that led to W.'s assumption that Zenodotus used an Ionic rhapsodic text to which he only introduced the obelus:
1) Was the Homeric text of the first diorthotes an "abbreviated text"? Nickau (1977: 20) has successfully refuted this very widespread misconception: according to his calculations Zenodotus' text was at most 138 lines shorter than the "Aristarchean vulgate". As Nickau correctly points out, this figure amounts to very little in a poem of approximately 15,700 lines and certainly does not make the Ephesian's text "substantially shorter". Nickau (1977: 62ff.) has also correctly identified Zenodotus' criteria for athetesis or omission of lines (W. seems not to distinguish between these two essentially different practices): they involve "genaue Beobachtung homerischer Kompositionsweise", testify to an ability to distinguish between "formelhafte Wiederholung des Typischen und untypische Verwendung des Formelhaften sowie verflachender Wiederholung des individuell gepraegten" and do not have anything to do with Zenodotus' purported tendency to athetize lines on the basis of aprepeia. This is a fundamentally different portrait of Zenodotus from that sketched by W. who suggests that Zenodotus' text was not the product of "rational thought" (37).
2) Do irregular forms of pronouns occur often in Zenodotus' text? Years ago (Rengakos 1993: 110ff.) I discussed the so-called "free use" of reflexive pronouns, i.e. the use of singular forms for the plural and vice versa and of the third person for the second and first. This use is amply manifested in Zenodotus' Homeric text as well as in the works of Apollonius of Rhodes and Callimachus. I reached the following conclusion: "It is a known fact, and has not been challenged since K. Brugmann (1876) established it in his pioneering study, that Aristarchus was the first to purge the Homeric text from this widespread free use, which may reach back to the Indo-European past and to which Zenodotus did not object. In other words, the forms in question appeared in the contemporary, early Alexandrian vulgate and the Hellenistic poets as well as Zenodotus simply followed the vulgate". The free use of pronouns cannot be considered a special feature of Zenodotus' Homeric text.
3) Do the neo- or hyper-ionic forms that occurred in Zenodotus' text indicate that he used an Ionic rhapsodic text? Wilamowitz, Wackernagel and Pasquali put forth views similar to this at first sight plausible hypothesis. Van der Valk (1964: 54-55) discussed them in detail and reached the following conclusions: Zenodotus' text contains "uncertain and false Ionisms"; Didymus does not mention city editions from Miletus or Smyrna while the relatively numerous readings that the scholia trace to the edition from Ionian Chios do not include a single Ionic form. "Accordingly, the theory of the Ionian texts may be considered to be a false one". Nagy (1998) even suggests that hyper-Ionisms occurred in "performances in an Attic-speaking context".
4) Even the few cases of the use of a dual for plural attested for Zenodotus' text do not prove that his original text was Ionic. According to Aristarchus, Eratosthenes and Crates of Mallos made the same mistake (sch. Ariston. Il. 24. 282; see recently Matthaios 1999, 378ff.).
Thus not much is left of the theory of the "Ionian rhapsodic text". The consensus today (Montanari 2002: 121ff.) is that Zenodotus made his textual interventions on a base-text to which he introduced obeloi and other signs to mark omissions or to introduce selected readings. This base-text may very well have been of Ionian extraction, which may account best for some Ionic variants in Zenodotus' Homeric text. Zenodotus, though, revised the text making "qualitative" as well as "quantitative" alterations (in Nagy's terminology  bringing in "horizontal" and "vertical" variants): he both collated manuscripts, a practice already common in his time (cf. the Hellenistic poets and the Ptolemaic papyri), and introduced his own conjectures.
W.'s theory about the sources of Didymus takes up the views of K. Lehrs (1882), which Ludwig energetically refuted at the time (1884: 43-44). To an extent departing from Lehrs W. analyzes the wording of several Didymean scholia and concludes that Didymus made direct use of the editions of Antimachus, Rhianus (either directly or through Callistratus), Callistratus (through which he also had knowledge of the Argolic edition), Aristarchus, Seleucus (where he found readings from the Cypriot, Cretan and the polystichos), as well as the editions from Massalia, Chios and perhaps Sinope. From Aristarchus' ekdoseis and hypomnemata he took only the readings of Zenodotus and Aristophanes. As mentioned above, W. concludes that Aristarchus did not collate manuscripts.
An unbiased analysis of the wording of the scholia in question would again lead to a substantially different assessment of various aspects of the issue at hand and to a different general conclusion. Nagy (forthcoming) thoroughly studies the material and I will limit myself to the essential points here. The basic question is: do the Didymus and Aristonicus scholia cite Aristarchus as a source for other editions and, if yes, which ones? The identification of Aristarchean fragments in the scholia is a notorious problem (a collection of Aristarchus' fragments is still a great desideratum) and I will therefore discuss only the most unambiguous cases. As mentioned above and as W. himself admits, Aristarchus relied on the editions of Zenodotus and Aristophanes and so I will not cite evidence for that. On Il. 6.6. Aristonicus notes that Aristarchus initially adopted the reading of the ἀρχαῖα ἀντίγραφα (μεσσηγὺς ποταμοῖο Σκαμάνδρου καὶ Στομαλίμνης) and introduced it in his hypomnemata. Then he "ran across another one" (ὕστερον δὲ περιπεσών; Didymus notes in agreement: ὕστερον δὲ εὑρών, μεσσηγὺς Σιμ́οεντος ἰδὲ Χάνθοιο ῥοάων, which he adopted. This clearly refutes the theory that Aristarchus did not consult different manuscripts: the critic consulted first the ἀρχαῖα ἀντίγραφα and adopted their reading and then discovered in the course of his further research, i.e. in another edition, a different reading he deemed preferable. The editions used by Aristarchus himself (not Didymus!) are not named in the Didymus scholia on Il. 9.401: "some" (ἔνιοι; in some cases this may indicate Zenodotus, cf. e.g. on Il. 9.159) or "several" in the famous scholion on Il. 9.222: "Aristarchus respectfully refrained from emending the reading he found in several editions" (ὅμως ὑπὸ περιτῆς εὐλαβείας οὐδὲν μετέθεηκεν, ἐν πολλαῖς οὕτως εὑρὼν φερομένην τὴν γραφήν). According to the Didymus double scholion on Il. 19.386 (see Erbse ad loc.), which West does not discuss, Aristarchus even cited a reading from the city editions: first he read τὸ δ' εὖτε (πρότερον δὲ γράφων) and then τὸ δ' αὖτε (ὕστερον δὲ μετὲγραφεν): Didymus himself states that the second reading came from the city editions.
It is clear from the Didymus scholion on Il. 18.10-11 that, apart from the editions of Zenodotus and Aristophanes, Aristarchus used directly that of Rhianus too (see also Nagy [forthcoming] with different arguments). The scholion reports that the two lines were not found in Rhianus' edition "possibly because Patroclus was not a Myrmidon but a Locrian. Aristarchus suggests that the phrase (sc. Μυρμιδόνων τὸν ἄριστον) should be interpreted as 'the best in the Myrmidon contingent after him [sc. Achilles]". This is clearly Aristarchus' retort to Rhianus' assumed objection. I will not discuss other, better known testimonia and will only add a word about the most important piece of evidence, the debated (Führer-Schmidt: 2000: 5f.) Didymus scholion on Il. 1.423-24: in view of the above discussion the scholion reports the views of Aristarchus and not those of Didymus, as Nagy (forthcoming), following Ludwich 1884, has convincingly shown anew.
Apart from the ἀρχαῖα ἀντίγραφα, the ἔνιοι, the πολλαί, the city editions and those of Zenodotus, Aristophanes and Rhianus, i.e. all the editions that Aristarchus himself cites, he also relied on the indirect Homeric tradition, a fact also neglected by W. Aristarchus (but also his rival Crates of Mallos: cf. Rengakos 1993: 28ff.) repeatedly makes use of Homeric quotations in post-Homeric poets in order to establish the correct wording of a Homeric verse: on Il. 11.101 he cited Poseidippus, on Il. 14.499 Callimachus (along with Antimachus), on Il. 21.575 Stesichorus, on Il. 23.88 Anacreon, on Od. 4.1 Callimachus again (along with Zenodotus) and on Od. 6.244 Alcman. Is it plausible to assume that a textual critic consistently took into account the indirect tradition of his text and at the same time completely disregarded the direct one?
In conclusion, one cannot agree with W.'s views about Zenodotus' Homeric text and Didymus' sources. The Alexandrians both collated manuscripts and introduced their own conjectures into the Homeric text. The work of the pioneers of our discipline deserves a less dismissive assessment.
[[For a response to this review by M.L. West, please see BMCR 2004.04.17.]]
Apthorp, M. J. 1980. The Manuscript Evidence for Interpolation in Homer. Heidelberg. Brugmann, K. 1876. Ein Problem der homerischen Textkritik und der vergleichenden Sprachwissenschaft, Leipzig.
Duentzer, H. 1848. De Zenodoti studiis Homericis, Göttingen.
Fuehrer, R.-Schmidt, M. 2001. "Homerus redivivus" (rev. of Homerus, Ilias, rec. M. L. West, vol. I, Stuttgart-Leipzig 1998), GGA 253: 1-32.
Haslam, M. 1997. "Homeric Papyri and Transmission of the Text", in: I. Morris-B. Powell (eds), A New Companion to Homer, Leiden: 55-100.
Janko, R. 1992. The Iliad: A Commentary. Vol. IV: books 13-16, Cambridge.
---. 2000. "West's Iliad" (rev. of Homerus, Ilias, rec. M. L. West, vol. I, Stuttgart-Leipzig 1998), CR 50: 1-4.
Lehrs, K. 1882. De Aristarchi studiis Homericis, Leipzig3.
Ludwich, A. 1884-85. Aristarchs Homerische Textkritik nach den Fragmenten des Didymos, I-II, Leipzig.
Matthaios, S. 1999. Untersuchungen zur Grammatik Aristarchs: Texte und Interpretation zur Wortartenlehre, Göttingen.
Montanari, F. 1979. Studi di filologia omerica antica I. Pisa.
---. 1998. "Zenodotus, Aristarchus and the Ekdosis of Homer", in G. W. Most (ed.), Editing Texts / Texte edieren. Aporemata II. Göttingen 1998: 1-21.
---. 2000. "Ripensamenti di Aristarco sul testo omerico e il problema della seconda ekdosis", in: Poesia e religione in Grecia. Studi in onore di G. Aurelio Privitera, Napoli: 479-486.
---. 2002. "Alexandrian Homeric Philology. The Form of the Ekdosis and the Variae Lectiones", in: Epea Pteroenta. Festschrift für Wolfgang Kullmann zum 75. Geburtstag (edd. M. Reichel, A. Rengakos) 119-140 Stuttgart.
Nagy, G. 1996a. Poetry as Performance: Homer and Beyond. Cambridge.
---. 1996b. Homeric Questions. Austin.
---. 1998. "Aristarchean Questions." BMCR 98.7.14.
---. 2000. Review of Martin L. West, ed., Homeri Ilias. Recensuit / testimonia congessit. Volumen prius, rhapsodias I-XII continens. Stuttgart and Leipzig: Bibliotheca Teubneriana, 1998. BMCR 00.09.12.
---. (forthcoming). Review of Martin L. West, Studies in the Text and Transmission of the Iliad, Munich-Leipzig 2001, Gnomon.
Nardelli, J.-F. 2001a. Review of Martin L. West, ed., Homeri Ilias. Recensuit / testimonia congessit. Volumen alterum rhapsodias XIII-XXIV et indicem nominum continens. Munich and Leipzig: Bibliotheca Teubneriana, 2000. BMCR 01.06.21.
---. 2001b. "Éditer l'Iliade I. La transmission et ses débats: perspectives critiques", Gaia 5, 2001: 41-118.
Nickau, K. 1977. Untersuchungen zur textkritischen Methode des Zenodotos von Ephesos. Berlin and New York.
Pasquali, G. 1952. Storia della tradizione e critica del testo, Firenze2.
Rengakos, A. 1993. Der Homertext und die hellenistischen Dichter. Hermes Einzelschriften 64. Stuttgart.
---. 2001. "Apollonius Rhodius as a Homeric Scholar", in: Th. D. Papanghelis-A. Rengakos (eds), A Companion to Apollonius Rhodius, Leiden: 193-216.
---. 2002. "The Hellenistic Poets as Homeric Critics", in: F. Montanari (ed.), Omero tremila anni dopo. Atti del Congresso Genova 6-8 luglio 2000, Roma: 143-157.
Schmidt, M. 1997. "Variae lectiones oder Parallelstellen: Was notierten Zenodot und Aristarch zu Homer?" ZPE 115:1-12.
Thiel, H. van. 1991. Homeri Odyssea, Hildesheim.
---. 1992. "Zenodot, Aristarch und andere", ZPE 90: 1-32.
---. 1996. Homeri Ilias, Hildesheim.
---. 1997. "Der Homertext in Alexandria", ZPE 115: 13-36.
Valk, M. van der Valk. 1963-64. Researches on the Text and Scholia of the Iliad, I-II, Leiden.
West, M. L., ed. 1998 / 2000. Homeri Ilias I / II. Stuttgart, Munich, Leipzig.
---. 1999. "The Invention of Homer." Classical Quarterly 49:364-382.
---. 2001. Response to BMCR 00.09.12 and BMCR 01.06.21. BMCR 01.09.6.
West, S. 1967. The Ptolemaic Papyri of Homer, Köln-Opladen.