relative_chronology_in_early_greek_epic_poetry_-_edd._o._andersen,_d._t._t._haug_-_2012.pdf

RELATIVE CHRONOLOGY INEARLY GREEK EPIC POETRY

This book sets out to disentangle the complex chronology of earlyGreek epic poetry, which includes Homer, Hesiod, hymns and cata-logues. The preserved corpus of these texts is characterized by a ratheruniform language and many recurring themes, thus making the estab-lishment of chronological priorities a difficult task. The editors havebrought together scholars working on these texts from both a linguisticand a literary perspective to address the problem. Some contributionsoffer statistical analysis of the linguistic material or linguistic analysisof subgenres within epic, others use a neoanalytical approach to thehistory of epic themes or otherwise seek to track the development andinterrelationship of epic contents. All the contributors focus on theimplications of their study for the dating of early epic poems relativeto each other. Thus the book offers an overview of the current stateof discussion.

øivind andersen is Professor of Greek at the University of Oslo.He is also currently a fellow of the Institute of Advanced Study at theNorwegian Academy of Science and Letters.

dag t. t. haug is Associate Professor of Latin at the University ofOslo.

Downloaded from Cambridge Books Online by IP 130.60.206.43 on Sat Jul 06 17:18:47 WEST 2013.http://ebooks.cambridge.org/ebook.jsf?bid=CBO9780511921728Cambridge Books Online © Cambridge University Press, 2013

RELATIVE CHRONOLOGY IN

EARLY GREEK EPIC POETRY

edited by

ØIVIND ANDERSEN AND DAG T. T. HAUG


cambridge university press

Cambridge, New York, Melbourne, Madrid, Cape Town,Singapore, Sao Paulo, Delhi, Tokyo, Mexico City

Cambridge University PressThe Edinburgh Building, Cambridge cb ru, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.orgInformation on this title: www.cambridge.org/

c© Cambridge University Press

This publication is in copyright. Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place without the written

permission of Cambridge University Press.

First published

Printed in the United Kingdom at the University Press, Cambridge

A catalogue record for this publication is available from the British Library

Library of Congress Cataloguing in Publication dataRelative chronology in early Greek epic poetry / edited by Øivind Andersen, Dag T. T. Haug.

p. cm.Includes bibliographical references and index.

isbn ---- (hardback). Epic poetry, Greek – History and criticism. . Epic poetry, Greek – Chronology.

I. Andersen, Øivind. II. Haug, Dag. III. Title.pa.r

′. – dc

isbn ---- Hardback

Cambridge University Press has no responsibility for the persistence oraccuracy of URLs for external or third-party internet websites referred to

in this publication, and does not guarantee that any content on suchwebsites is, or will remain, accurate or appropriate.


Contents

Notes on contributors page viiPreface xiAbbreviations xii

Introduction

Øivind Andersen and Dag T. T. Haug

�� : relative chronologyand the literary history of the early Greek epos

Richard Janko

Relative chronology and an ‘Aeolic phase’ of epic

Brandtly Jones

The other view: focus on linguistic innovations in theHomeric epics

Rudolf Wachter

Late features in the speeches of the Iliad

Margalit Finkelberg

Tmesis in the epic tradition

Dag T. T. Haug

The Doloneia revisited

Georg Danek

Odyssean stratigraphy

Stephanie West

Older heroes and earlier poems: the case of Heracles inthe Odyssey

Øivind Andersen

v

Downloaded from Cambridge Books Online by IP 94.189.177.21 on Fri Jul 05 16:37:24 WEST 2013.http://ebooks.cambridge.org/ebook.jsf?bid=CBO9780511921728Cambridge Books Online © Cambridge University Press, 2013

vi Contents

The Catalogue of Women within the Greek epic tradition:allusion, intertextuality and traditional referentiality

Ian C. Rutherford

Intertextuality without text in early Greek epic

Jonathan S. Burgess

Perspectives on neoanalysis from the archaic hymnsto Demeter

Bruno Currie

The relative chronology of the Homeric Catalogue ofShips and of the lists of heroes and cities within theCatalogue

Wolfgang Kullmann

Towards a chronology of early Greek epic

Martin West

Bibliography

General index

Index locorum


Notes on contributors

Øivind Andersen is a professor of Classical philology (Greek) at theUniversity of Oslo; in – he was the Director of the NorwegianInstitute at Athens. He has published Die Diomedesgestalt in der Ilias() and many articles on Homer. His other main research interestsare orality and literacy, and ancient rhetoric.

Jonathan s. Burgess is a Professor in the Classics department at theUniversity of Toronto, Canada. His areas of research include early Greekepic, mythological iconography, and travel literature. His main publica-tions are The Tradition of the Trojan War in Homer and the Epic Cycle() and The Death and Afterlife of Achilles ().

Bruno Currie is Monro Fellow and Tutor in Classics at Oriel College,Oxford, and Lecturer at Oxford University, UK. He is the author ofPindar and the Cult of Heroes (Oxford, ) and co-editor of EpicInteractions: Perspectives on Homer, Virgil and the Epic Tradition pre-sented to Jasper Griffin by Former Pupils (Oxford, ). His mainresearch interests are choral lyric, early hexameter poetry and Greekreligion.

Georg Danek is Associate Professor at the University of Vienna, Austria.He works mostly on Homer, including comparative studies in SouthSlavic Epics. His books are: Studien zur Dolonie (), Epos und Zitat:Studien zu den Quellen der Odyssee () and Bosnische Heldenepen().

Margalit Finkelberg is Professor of Classics at Tel Aviv University,Israel. She has widely published on Homer and epic tradition, Aegeanprehistory, and poetics and literary theory. She is the author of The Birthof Literary Fiction in Ancient Greece () and Greeks and Pre-Greeks:Aegean Prehistory and Greek Heroic Tradition (), and the editor ofthe Homer Encyclopedia I–III ().

vii


viii Notes on contributors

Dag t. t. Haug is an associate professor of Latin at the University of Oslo.His main field of interest is the linguistic study of ancient languages.He has published Les phases de l’evolution de la langue epique (),and many articles on Greek and Latin linguistics, mainly syntax andpragmatics.

Richard Janko is is Gerald F. Else Collegiate Professor of ClassicalStudies at the University of Michigan. He is the author of Homer,Hesiod and the Hymns: Diachronic Development in Epic Diction ()and of The Iliad: A Commentary. Volume IV: Books 13–16 (), whichincludes introductory chapters on the diction and text of Homer. Otherbook publications include Aristotle on Comedy: Towards a Reconstructionof ‘Poetics II’ () and Philodemus Bk. 1: On Poems (Oxford ).

Brandtly Jones is Chair of Foreign Languages and teaches Latin at St.Anne’s-Belfield School in Charlottesville, Virginia, USA. He received hisPh.D. from Cornell University in with a dissertation on ‘RelativeChronology and the Language of Epic’, which he is currently reworkingfor publication. His research interests include Archaic and HellenisticGreek epic poetry, historical linguistics and the poetry of Catullus andVergil.

Wolfgang Kullmann is emeritus professor of Greek at the Universityof Freiburg, Germany. His publications include Das Wirken der Gotter inder Ilias () and Die Quellen der Ilias (Troischer Epenkreis) (). Hehas also published a number of articles on Homer and Greek epic poetry,some of which are collected in Homerische Motive () and in Realitat,Imagination und Theorie: Kleine Schriften zu Epos und Tragodie in derAntike (). His other main field of research is Aristotle’s biologicalwritings.

Ian c. Rutherford is Professor of Greek at the University of Reading.He is the author of Pindar’s Paeans: A Reading of the Fragments with aSurvey of the Genre () and the editor with R. Hunter of WanderingPoets in Ancient Greek Culture (). His chief research interests areGreek poetry, Greek religion, especially pilgrimage and theoria, andcontact between the Aegean and Anatolia in the Bronze Age.

Rudolf Wachter is Professeur associe de linguistique historique indo-europeenne at the University of Lausanne (since ) and extraordi-nary professor of Greek, Latin and Indo-European Linguistics at theUniversity of Basel (since ). He obtained a doctoral degree in Latin


Notes on contributors ix

in Zurich (Altlateinische Inschriften, ) and another in ComparativePhilology in Oxford (Non-Attic Greek Vase Inscriptions, ) and haspublished widely in linguistics, Greek and Roman epigraphy, and on thehistory of the alphabet.

Martin West is an Emeritus Fellow of All Souls College, Oxford. Hehas published many editions of Greek poetic texts, including the Iliad,Hesiod, the Homeric Hymns, the early epic fragments, the elegiac andiambic poets, and Aeschylus, besides such works as Greek Metre (),The Orphic Poems (), Ancient Greek Music (), The East Face ofHelicon () and Indo-European Poetry and Myth ().

Stephanie West is an Emeritus Fellow of Hertford College, Oxford, UK.Her main research interests are in Homer, Herodotus and Lycophron.Her most recent publications on Homer are ‘Phoenix’s antecedents:a note on Iliad ’, Scripta Classica Israelica (), – and ‘DieOdyssee: Inhalt u. Aufbau’, in J. Latacz et al. (eds.), Homer: Der Mythosvon Troia in Dichtung und Kunst (Munich, ), –. She is cur-rently working on a commentary on Herodotus book .


Preface

The present volume on relative chronology in early Greek epic poetry orig-inates from a conference under the same heading which was organized atthe Norwegian Academy of Science and Letters in Oslo in the summerof , with generous financial support from the Academy and from theLudvig Holberg Memorial Fund and with a contribution from the Depart-ment of Philosophy, Classics, History of Art and Ideas at the University ofOslo. At the conference, a dozen of invited and a score of other papers weregiven, on both literary and linguistic aspects of the problem, leading tomuch fruitful discussion. The editors are pleased to present this selectionof essays to a wider audience.

The editors wish to express their sincere thanks to their co-organizer ofthe conference, Anastasia Maravela, who also contributed substantially tothe early stages of the work on the present volume. They also wish to thankPal Rykkja Gilbert for valuable assistance in the final stages of the work.

xi

Downloaded from Cambridge Books Online by IP 130.60.206.43 on Sat Jul 06 17:19:15 WEST 2013.http://dx.doi.org/10.1017/CBO9780511921728.001

Cambridge Books Online © Cambridge University Press, 2013

Abbreviations

AO R. Develin () Athenian Officials 684–321 B.C.Cambridge

Bernabe A. Bernabe (ed.) (–) Poetarum EpicorumGraecorum Testimonia et Fragmenta ( vols.). Berlin,Munich and Stuttgart

Davies M. Davies (ed.) () Epicorum Graecorum Fragmenta.Gottingen

Drachmann A. B. Drachmann (ed.) (–) Scholia Vetera inPindari Carmina ( vols). Leipzig

CEG P. A. Hansen (ed.) (–) Carmina Epigraphica Graeca( vols.). Berlin

FGrH F. Jacoby (ed.) (–) Die Fragmente der griechischenHistoriker. Berlin and Leiden

H M. Hirschberger () Gynaikon Katalogos undMegalai Ehoiai. Munich and Leipzig

IEG M. L. West (ed.) (–) Iambi et Elegi Graeci anteAlexandrum Cantati ( vols.), nd edn. Oxford

LGS D. L. Page (ed.) () Lyrica Graeca Selecta. OxfordLIMC (–) Lexicon Iconographicum Mythologiae Classicae.

ZurichL–P E. Lobel and D. L. Page (eds.) () Poetarum

Lesbiorum Fragmenta. OxfordLSJ H. G. Liddell, R. Scott and H. Stuart Jones ()

Greek–English Lexicon, th edn, Suppl. by E. A. Barberet al. Oxford

M–W R. Merkelbach and M. L. West (eds.) () FragmentaHesiodea. Oxford

PMG D. L. Page (ed.) () Poetae Melici Graeci. OxfordPMGF M. Davies (ed.) () Poetarum Melicorum Graecorum

Fragmenta, vol. I. Oxford

xii


Abbreviations xiii

Rose V. Rose (ed.) () Aristotelis Qui Ferebantur LibrorumFragmenta. Stuttgart

SLG D. L. Page (ed.) () Supplementum Lyricis Graecis.Oxford

TGF B. Snell, R. Kannicht and S. L. Radt (eds.) (–)Tragicorum Graecorum Fragmenta ( vols.). Gottingen

Voigt E.-M. Voigt (ed.) () Sappho et Alcaeus: Fragmenta.Amsterdam

West M. L. West (ed.) () Greek Epic Fragments.Cambridge, Mass. and London


Introduction

Øivind Andersen and Dag T. T. Haug

Early Greek epos is represented for us by several thousands of hexameterverses in essentially uniform dialect, idiom and style. In addition to whathas come down to us more or less intact – the two Homeric poems,Hesiod’s works, a number of hymns (not all of them that early), theAspis – we have chunks of catalogue poetry and fragments and reportsof a number of different works, and we are aware of the existence ofa whole oral epic tradition in which essentially traditional material wastransmitted in conventional forms. How are we to establish a literaryhistory of early Greek epos, to sort out the elements of the medium andthe message chronologically and even genealogically in relation to eachother? In the case of most modern oeuvres, there is no need to ponderthe internal sequence, as dates of composition and publication are known.Indeed, in European literature back until the time of the Greek tragedians,independent information and external evidence often yield absolute datesso that the working out of an internal chronology is not an issue. In the caseof early Greek epos, however, the near total lack of absolute chronologicalpegs and the scarcity of relevant facts and contexts compel us to rely mainlyon internal criteria. That holds true especially for the earlier part of theArchaic Age (the eighth and seventh centuries bc); with time we do gettestimonies that may serve as clues to termini post (or even ante) quemand thus help us establish a relative chronology based on (approximate)absolute dates. For the charting out and pinning down of poets and poems,we are not much helped by authorial self-reference, except, perhaps, in thecase of Hesiod, which does not yield much in the way of chronology, andof the Hymn to Apollo, which may already build on the fiction of a Homerfrom Chios.

Homer and Hesiod obviously could lay claim to pride of place even ata time when much more epic poetry by many more poets was availablethan is the case today. We are not, however, much helped by the ancientbiographical lore about Homer and Hesiod – the vitae, the Certamen and



øivind andersen and dag t. t. haug

scattered evidence – or by what the ancients generally imagined aboutthe age and succession and relationship of those two and of other poets.The tendency of ancient literary biographers to construct neat successionsbetween prominent literary figures and to fashion biographical accountsfrom what is in the poetry is well known. Graziosi () has shown howthe characters and circumstances and dates of Homer, and of Hesiod aswell, have been constructed by posterity and in essence must be understoodin the light of social and political circumstances and within the context ofliterary and ideological axe grinding. Interest in the age and chronologicalrelationship of Homer and Hesiod is not driven by historical curiosity. TheHomeridai, the rhapsodes in general, the Chians, the Athenians, Herodotusand Thucydides, Plato and Aristotle – all conjure up a different Homer.Therefore, although the story of the invention and individualization ofHomer is fascinating and utterly instructive in its own right, the ancientbiographical lore tells us preciously little about the persons it claims to beabout, and is of very little assistance in the quest for absolute and relativedates. The same questions were rehearsed again and again throughoutthe centuries. Thus, for example, while the prominent fourth-century bc

historian Ephorus from Cyme proposed a genealogy according to whichHomer was a younger cousin of Hesiod’s, the somewhat later scholar-historian Philochorus held – with others – that Homer was the older one.

The topic was the theme of a treatise in two books by the fourth-centurybc Platonic philosopher-scholar Heraclides Ponticus, of which nothing isknown other than its title, and the Peripatitic Chamaeleon’s (also fromPontus) allegation that the material in it had been derived from his work.

For the most part, we have only opinions to go by, and very little inthe way of argument. Aristarchus’ collective labelling of the cyclic poets as�� (‘more recent’) in relation to Homer probably does not rely on anymethod of assessing literary development, or any historical investigation; itcannot be separated from Aristarchus’ attitude to the non-Homeric Trojanwar tradition as a threat to Homer’s originality, even as he championedHomer’s unity against the chorizontes who would ascribe the Iliad andthe Odyssey to two different authors. As for the relative chronology ofIliad and Odyssey, ps.-Longinus seems to contribute to an ongoing debate

Graziosi’s comprehensive account also generously refers to other contributions in a similar vein, suchas West (), Burkert ().

FGrH F ad Plut. Vita Hom. .; F. Heraclid. Pont. F .– ap. Diog. Laert. V ; F ap. Diog. Laert. V . Schol. D ad Il. .; schol. A ad Il. .–. On Aristarchus’ attitude, see Severyns (), Ballabriga

(), Burgess ().



Introduction

as he advances the view that the Iliad is the work of a young, vigorouspoet, and the Odyssey that of an aged one. The second-century ad satiristLucian during his sojourn amongst the dead asks Homer whether he wrotethe Odyssey before the Iliad, ‘as most people say’ (Homer answers ‘no’).

Another writer of the same period, the geographer Pausanias – who himselfwas unsure whether the Theogony was a work of Hesiod and believed thatHomer had written the Thebaid – declares that he has undertaken carefulresearch into the question of the age of Hesiod and Homer, ‘but I donot like to write on this matter, as I know the quarrelsome nature of thoseespecially who constitute the modern school of epic criticism’. Sometimes,fortunately, relevant evidence is reported. Thus, in view of the fact thatverses – of the Aspis were transmitted in Book of the Catalogue,Aristophanes of Byzantium, as reported in the hypothesis to the Aspis,suspected that the Aspis was not by Hesiod but by someone else whohad chosen to imitate the Homeric ‘Shield’; this is not only testimony tothe way an Alexandrian scholar would reason, it also contains, if reliable,relevant information on the relationship between two works within theGreek epos.

Both Eratosthenes, Aristophanes’ predecessor as head of the AlexandrianLibrary, and Aristarchus of Samothrace, who succeeded him, attempted tobuttress Homer’s priority over Hesiod by showing that the geographical,ethnographical and socio-cultural information incorporated in the Home-ric epics represents a less advanced and thus earlier stage in relation tocomparable information in Hesiod’s works. According to Strabo, Eratos-thenes pointed out that Hesiod knew many more localities associated withOdysseus’ wanderings than did Homer. The same scholar observed thatHomer was ignorant of the fact that the river Nile had several mouths –something that the younger Hesiod was aware of. Indication of Homer’spriority is his use of early ethnic names, e.g. of ‘Meones’ for the Lydians;

on the other hand, Hesiod, being younger, introduced Hippomenes run-ning naked against Atalante; the terminus ante quem for the innovation ispinned to the th Olympiad, i.e. – bc. Much unsure ground has tobe traversed before the presence of putative items of relevant geographicaland ethnographic knowledge can be securely transformed into evidence for

Ps.-Longinus, Subl. .. Lucian, Ver. hist. .. Paus. ... Eratosth. FGrH I B and Hes. F M–W ad Strabo ... Eratosth. FGrH I B,– and Hes. F ad Strabo ... Hesiod’s more advanced geographical

knowledge is also evidenced by the fact that Homer refers to the Nile as ‘Aigyptos’ while Hesiodknows the river by its later name, cf. schol. HMPQT ad Hom. Od. ..

Schol. A ad Il. .a. Schol. A and T ad Il. .b.




absolute and relative chronology – and then only for that verse or passageof a poem in which it occurs.

In a sense, the Alexandrian philologists’ work on the text of the Homericpoems is all about relative chronology, in so far as they seek to identify accre-tions to what Homer originally wrote. Some statements question wholesections of the text. Most important has been the view of Aristophanes andAristarchus, that the Odyssey reached its �� (‘limit’) or �� (‘end’)at ., which has often been taken to mean that the poem originallyended at that point and that the end of book and all of book areadditions by a later hand. The Alexandrians’ statement is not supportedby argument; we do not even know for sure what they claim; many havetaken their pronouncement as an aesthetic judgement, referring to thereunification of husband and wife as the ‘goal’ of the hero’s journey home.The remark in the scholia that the Doloneia (book of the Iliad) hadbeen drawn up by Homer separately and was included into the poemby Pisistratus can scarcely be relied upon and is not argued for. A feel-ing for the special linguistic and compositional features of the Doloneia,and/or of its untypical content and setting, may have led to this wholesong coming under suspicion at some stage in some circles, prefiguring themodern unease with Book of the Iliad, which was bound to have arisenanyway, but no doubt has been spurred on also by the ancient scholiast’sremark. The fact that the Doloneia’s alleged inclusion in the Iliad underPisistratus would hardly have been politically motivated may point eitherway – the allegation appears the more credible as it is not linked to a polit-ical motive, or less so, because it has no ‘Pisistratean’ motive. Most otherputative Pisistratid additions to the text seem to serve Athenian politicalaspirations, especially in relation to Aegina and Megara. In a number ofancient sources from the fourth-century bc Pseudo-Platonic Hipparchusonwards, and the (lost) Megarian historian Dieuchidas of the same period,the Athenian tyrant Pisistratus or his son Hipparchus are credited (or dis-credited) not only with the odd addition to the Homeric corpus, but witha far-reaching initiative which consisted in bringing ‘Homer’ to Athensand seeing to it that the Homeric epics were recited every year at the Pana-thenaic festival in a fixed order by a series of rhapsodes. While the so-called‘Pisistratean Redaction’ seems to some to be a chimaera, the tradition seemsto us to rely on some real historical initiative during the tyrants’ regime,privileging and canonizing the Homeric epics by means of the ‘Pisistratean

Their collective opinion is somewhat differently reported in schol. MV, Vind. and in HMQ. Schol B to Il. .. On the Doloneia, see Danek (, and in this volume).



Introduction

Rule’. That the Iliad and the Odyssey came into being only at this time, assome have held, is a hypothesis that is not supported by the testimoniesto this process, and flies in the face of the linguistic and literary evidencefor the existence of those poems in a relatively fixed form some to years earlier.

Pictorial representations are another type of ancient evidence that couldhelp us pin down the emergence and existence of epic poems chronolog-ically. If a pictorial representation can be reasonably securely dated, andit can be convincingly shown to represent a scene from a specific poem,then we have at least a terminus ante quem for the existence of that poem.And if we can observe a sudden surge in pictorial representations fromsome part of heroic myth – say, from the subject matter of the Iliad orthe Odyssey – this would consititute a strong indication of the existenceand coming into circulation of a poem. Snodgrass (, ), reviewingrelevant evidence down to late archaic times and refining the methodol-ogy of his predecessors, has shown how tenuous are the links betweenpoetry and pottery in the Geometric and Early Archaic ages. Down untilc. bc only two out of c. sixty pictorial representations of myth maybe confidently judged to have been influenced specifically by ‘Homer’, i.e.the Iliad or the Odyssey as we know them, and during the first half of thesixth century there is only a slight increase. This warrants the postulationof a terminus ante quem for the Iliad and the Odyssey c. – if the poemswere not circulated in a truncated form, an idea that will not appeal tomany today. The very parts of the saga that were developed in the Iliadand the Odyssey appear to have been quite slow to occupy centre stage inthe visual arts. No surge in the popularity of ‘Homeric’ motives indicatesthat new, great poems had suddenly become available and rapidly attainedpopularity. Indeed, ‘Homer’ seems not to have been a source of inspirationand authority for painters even when they depict scenes taken from thesubject matter of the Iliad and the Odyssey. And what is more, the subjectmatter of those two poems is not privileged in comparison with episodesfrom other parts of the Trojan saga (eventually contained in the epic poemsof the kyklos), and even less so in comparison to other heroic myths. Onemight prefer to stress that Snodgrass’s cautious considerations would afterall give us a terminus ante quem for ‘Homer’; we would like to point to thechallenges of charting out the relative chronology that arise from the factthat the existence of our two monumental epics apparently did not have

Notably Friis Johansen () and Fittschen (). The pictorial evidence is put to good use byBurgess ().




momentous consequences and did not monopolize the market. Snodgrass’swork constantly reminds us of the plethora of local oral traditions, vernac-ular oral accounts, and of lost poems that were around for painters, butalso for poets, and for the public at large to refer and relate to.

∗ ∗ ∗

While pictorial representations and the textual testimonia from historians,philologists, etc. must be used for what it is worth, we mainly depend onthe evidence of the corpus itself when we try to work out the chronologicalrelations within the corpus of Greek epic poetry. About the corpus as awhole, we should like to stress the following points, some more pertinent tolinguistic issues, others to literary. First, although the corpus is voluminous,what has been preserved is only a fraction of what was available by, say, bc. In addition, the whole mass of oral popular and local traditionshas vanished, except for scattered remarks. Second, the loss also of allother poetry of the period before c. bc has deprived us of invaluablecomparative material and makes epic poetry stand out in splendid isolation.Third, in Greek epic, tradition and convention possess the poet to such adegree that it is especially difficult to disentangle what is older from whatis not so old. Fourth, and on the other hand, early Greek hexameter poetryis represented for us by several subclasses (heroic, theogonic, didactic,hymnic) and it has various provenances; they are not all grafted similarlyonto the general tradition. Fifth, some of what actually remains of thecorpus is anything but typical in the sense that it must rely on extraordinarypoetic genius; at least the special status of the Iliad and the Odyssey hasbeen communis opinio since the time of Aristotle: how to provide for theindividual talent?

Then there is the question of the nature of those ‘texts’ that we have. Eposof every kind had been written down before the end of the Archaic Age;whether we should think of specimina as ‘oral dictated texts’, or accordingto some other method of textualization, is not our concern here. Oncewritten down, the individual works would to a varying degree be subject tochanges in performance and even in writing, depending, probably, on theindividuality the works exhibited and the status they achieved. Finkelberg() has shown how the ‘multiformity’ that is the hallmark of oraltradition can be plausibly postulated in the case of the Cypria, which seemshardly to have reached a standard, canonical version, while the kinds ofvariations that we find in the case of the Iliad and the Odyssey are notsufficient to meet the definition of multiformity: they remained, both inessence and in detail, the works that had once been written down. That iswhy the Homeric poems, as well as Hesiod’s works and the major hymns,



Introduction

can form the basis for the study of relative chronology. Another approachto Greek epic that has been influential for some time holds that this ideaof the fixed text is misguided: Gregory Nagy’s ‘evolutionary model of thegenesis of epic’ holds that after a fluid period down to c. bc, therefollowed a ‘pan-Hellenic’ period with still no written text down to themiddle of the sixth century, and then a third period with ‘potential textsin the form of transcripts’ down into the fourth century, followed by astandardizing period under Demetrius of Phalerum at the very end ofthe century and further standardizing by the Alexandrian scholars. This‘evolutionary theory of text-fixation through crystallization’ makes awaywith ‘the elusive certainty of finding the original composition of Homer’as it implies that even after the Homeric poems were first fixed in writing,they were characterized by ‘multiformity’, as oral performances continuedto influence the written texts. This approach for all practical purposesignores the difference between aoidoi and rhapsodes. Nagy’s hypothesis ishardly called for in order to explain the kind of differences that we findin Homeric manuscripts. Above all it seems to us not to tally with thelinguistic evidence, which shows that the works of Greek epos are not allof the same fabric, but have ‘crystallized’, if that is the correct word, atdifferent times.

As for the texts that were actually written down, one may ask howfar they represent oral tradition and conform to what is often called oralpoetics, and what, if anything, they owe to the new medium of alphabeticscript being exploited for epic song. Are we dealing with oral poetry, inthe sense that the poems bear witness of composition-in-performance, orhas writing enabled the poet to make things that he otherwise could nothave done? Is Homer the traditional bard, or is he the pan-Hellenic poet?Closely connected with this is the question of what kind of audiences thepoems and texts are meant for, and how high we shall rate the familiarity ofthe audiences with the corpus as a whole, and with the mythical tradition,and to what works or traditions they relate what they actually listen to. Isthe audience the local population and big men of various Aegean localities,as they are occasionally visited by travelling bards or rhapsodes, or perhapsthe multitude of Ionians assembled to celebrate a religious festival? Ormust we think of a generalized, pan-Hellenic audience? Finally, there is theproblem of the authority of the manuscripts that we possess and of othertextual evidence owed to papyri and literary quotations in relation to whatwas once written down. The manuscripts are what we have got. ‘Objects

Nagy (a); cf. Nagy ().




which we can see and touch and smell are the data of history: all else isconstruction.’ (Dover : )

∗ ∗ ∗

In modern times, one may say that the topic of relative chronology in earlyGreek epic has been broached mainly from three angles: (a) on the basis oflinguistic criteria, one has aimed to understand the mechanisms and phasesthrough which the epic idiom was formed and to discern the genuinelyolder from the younger and archaizing elements, thus making possiblea diachronic analysis of the linguistic conglomerate of Greek epic; thisapproach is based on a kind of knowledge of the Greek language and therelationship between Greek dialects that the ancients simply did not possess;(b) on the basis of on literary criteria, one has explored relations of literarydependence and thus chronological posteriority with the aid, if possible, ofexternal information, but above all on the basis of internal considerationsinvolving aesthetic appreciation. Familiar problems in this area are howqualitative evaluations may be transposed into temporal relations (‘betteris older’?) and how the typical and traditional can also be somethingpregnant and individual; this approach also remained largely unexploredin antiquity; (c) on the basis of criteria of material culture and historicalcriteria in general. This last approach is typically oriented towards theabsolute dating of phenomena that occur in the text but belong in theworld. By dating these, one can also establish a relative chronology betweendifferent parts of the corpus. Intimations of the period to which a verse or apassage or a poem belongs can be had, typically, from references to historicalcircumstances (e.g. the reign of King Amphidamas, the destruction ofBabylon, the heyday of Egyptian Thebes, Phoenician seafaring, the politicalgeography of Asia Minor, military tactics, etc.). Apart from the difficultiesinherent in identifying reliable clues in the text as opposed to reading theminto it, such external matter can merely serve as the basis for argumentsfor termini post quem. Arguments in favour of termini ante quem basedon the absence of elements from the text are even more problematic. Thecase is similar in regard to material objects described or mentioned in thetext, on the basis of which Carl Robert () made his ambitious attemptat an archaeological analysis of the Iliad. When securely dated, specificobjects or types of objects may give us relatively objective dating criteria.Objects such as the towering shield of Ajax, the Shield of Achilles, Heracles’shield (the Aspis), the silver-studded sword, Athena’s lamp and many more

See especially Burkert (), Dickie () and West (). Lorimer’s () was another useful survey; the field is now exhaustively treated in the series

Archaeologia Homerica (Buchholz and Matz –).



Introduction

objects, make their entry into the world at some specific time. But again,each gives us only a terminus post quem, and only for the introduction ofthat item into the tradition: the boar-tusk helmet, which doubtless belongsin Mycenaean times, only makes its appearance in the Doloneia, which isby common consent a recent part of the Iliad. The corpus of Greek epos isnot only a linguistic but also a material and socio-political conglomerate.There is no one ‘world of Odysseus’ out there, or an historical ‘Homericsociety’, whose realities are reflected in the Homeric poems. Which is notto say that the poems do not imaginatively conjure up a fairly unitary andconsistent fictitious world that makes sense in its own right.

In this volume, archaeological and historical considerations do not loomlarge, although they play an important role in some of the contributions,e.g. those by Martin West and Wolfgang Kullmann. The contributions tothis volume broadly fall under two headings, linguistic and literary. MartinWest’s contribution is a qualified wholesale attempt at pinning early Greekepic poems down in time, both relative to each other and in absolute terms.It is fitting, we think, to put it at the end, although it is anything but aconclusion to the volume, which opens with Richard Janko’s restatementand refinement of the argument of his path-breaking Homer, Hesiod, andthe Hymns (). Needless to say, no broad scholarly consensus on theissue has yet been reached, and the contributions in this volume point indifferent directions.

1 linguistic perspectives on relative chronology

Linguistic analysis of the corpus of early Greek epic poetry throws lightupon the historical development of the Greek language, of the epic idiomas such and of the oral epic tradition; it may also contribute importantly todetermining the relative chronology between segments and works withinthe epic corpus.

Serious research on the linguistic stratification of the epic languagestarted with Gustav Hinrichs’s dissertation. Like so many other strandsof nineteenth-century Homeric philology, this line of research built uponthe new premises established by F. A. Wolf’s Prolegomena, that Homerbelonged to an oral tradition where everything, including language anddiction, was in a state of flux. With the rapid advancement of Greekdialectology and comparative, historical linguistics, it soon became clear

See Morris (, ); good remarks in Cairns (a: –). Wolf’s work in Latin () is conveniently available in English translation in Wolf ().




that Homer’s Greek was not the kind of ‘original Greek dialect’ that scholarshad thought. Indeed it was not one homogeneous dialect at all, but rathera mixture incorporating elements from various dialects and chronologicallayers. Untangling this mixture is still the task that faces modern linguisticstudies of Homer, including those of Janko, Wachter, Finkelberg, Haugand Jones in this volume. There are two dimensions to the question, thegeographical and the diachronic. Although they are interrelated, we willlook at them one at a time, starting with the geography.

Homer’s basic dialect is Ionic Greek, the dialect spoken in Euboea andAsia Minor. More specifically it has been argued to be Euboean Ionic, sincehe uses forms like �� instead of �� which we find in Herodotus, butthe significance of this has been doubted, since such forms could easilybe changed during the transmission. Moreover, -forms hardly appear inepigraphic attestations of Eastern Ionic.

Importantly, the epic language also contains a number of Aeolic forms.Hinrichs () was able to pinpoint a number of these and developeda phase model to explain them: the Aeolic forms, he argued, reflect theprehistory of the epic diction. At one point, the dialect of epic poetry waspure Aeolic, but as Ionians came to practise it, they gradually replacedAeolic forms with their native Ionic – which is why Aeolic forms are onlyfound when they differ in metrical value from the corresponding Ionicones. Milman Parry further developed this hypothesis by embedding itwithin a general theory of oral composition and formulaic language. Thedemands of composition in performance make the tradition conservative,since useful formulae are preserved even when they contain obsolete lin-guistic material, but gradually poets create new expressions which oust theold ones. This view remains strong today.

But a competing model has been developed, which claims that theAeolic forms are not archaisms of the tradition, but rather borrowings byIonic epic of neighbouring dialectal forms (the diffusionist hypothesis).

This model is defended with extensive argumentation by Jones (this vol.).As these scholars point out, the existence of Aeolic forms is not enoughto prove an Aeolic phase – it is necessary to show that there is a breakin the Ionic tradition, i.e. that there are no Ionic archaisms in the epiclanguage. Since Meister (), the absence of genitives in ∗-��/∗-��(which would be the archaic Ionic form instead of Aeolic -�/-��) has

See Stuber (: –). As shown first by Witte (a). See for instance Janko (), West (), Haug (). Notable supporters of this hypothesis include Strunk () (who in fact denied the very existence

of specifically Aeolic material in the epics), Wyatt () and Horrocks ().



Introduction

been seen as evidence for just such a break. On the other hand, there areIonic sequences of -��-/-��- in other forms, and the argument dependspartly on the linguistic interpretation of the change known as quantitativemetathesis. Another crucial issue is the genitive in -��, which was arguedby Haug () to be an Aeolic archaism that is not directly comparableto the Mycenaean forms which are written -o-jo. If this is right, it wouldstrongly support the hypothesis of an Aeolic phase.

As the momentum of Aeolic influence became clear, interest in theinfluence of other dialects on the epic language also grew. In , Meilletwrote: ‘Just as the Aeolic epic was ionicized, there might have existedan Achaean epic, which influenced the Aeolic one’ (: ). Lexicalstudies supported this, by showing that many Homeric words were foundonly in Arcado-Cypriote. The decipherment of Linear B confirmed thishypothesis. Linear B was shown to be a close relative of Arcado-Cypriote,and at the same time it contained many forms that scholars had previouslythought to be Aeolic.

However, the more we go back in time, the more similar the Greekdialects were to each other. Despite treatments like Ruijgh (), there isno convincing demonstration that there are specifically Mycenaean formsin Homer, as opposed to just archaic forms – and perhaps there cannotbe one, given the state of our knowledge about Greek dialects prior tothe advent of alphabetic writing. We must therefore rather ask ourselveshow old the Homeric forms are, rather than where they come from, andthis brings us to the chronological dimension. Chronology partly relies ondialectology, of course; as noted by Jones, many Aeolic forms will countas archaisms in Homer only on the assumption that Aeolic preceded Ionicin the development of the epic. Similar problems arise with other features,such as tmesis, the separation of preverb and verb (Haug, this vol.). Thishas been argued to be a pre-Mycenaean feature of the epic language, sincethere are no clear cases of tmesis in the Linear B tablets. On the otherhand, the evidence of the tablets is meagre, especially when it comes tocomplex syntactic constructions, so we cannot rule out that tmesis was aliving syntactic feature in Mycenaean times and possibly Homer’s time.The same holds for lexical evidence: Bowra showed that unusual words inHomer are often found also in Arcadian and Cypriote inscriptions – but wecannot know whether the word also existed in pre-historic Aeolic and Ionic.

Nevertheless, the decipherment of Linear B allowed for a more correctassessment of the history of Greek dialects and a more correct classification

The most important articles are Bowra (, ).




of Homeric foms as Ionic, Aeolic – or even Mycenaean. With these insightsscholars have been able to proceed from the linguistic stratification of theHomeric language as such to studies of subgenres of the Homeric poems –as in Finkelberg’s contribution to this volume – and to studies of the wholeearly epic tradition, as in Richard Janko’s work.

After Milman Parry’s work it was clear that only dialectal forms thatare metrically irreducible can carry weight. Thus, a formula �� could easily be an Ionicization of Aeolic �� . Itwas therefore an important discovery when Hoekstra could show thatthere are almost no formulae that presuppose Ionic prosody. Instead, Ionicforms cluster in the similes and the speeches, which tend to contain latelinguistic material (Finkelberg, this vol., with references). It is characteristicof much recent work on linguistic stratification in Homer that it tends tofocus on the innovations rather than the archaisms. But this also bringsmethodological problems: we are very poorly informed on what the poet’svernacular, the Ionic of the eighth or seventh centuries, looked like. Unlikesome archaisms, such as specific endings or phonological traits, the recentforms of Homer are often unsystematic and not easily datable. In fact, if arecent-looking Homeric form is not transmitted in inscriptions, we oftendo not know whether it is a recent form at all, or just a nonce-formationmetri gratia. That said, Wachter’s article in this volume clearly demonstratesthe importance of recent forms in Homer: it is not possible to ignore them,or claim that every passage in which they occur is a later addition.

In fact, if there is one thing that the linguistic archaeology of Homerhas shown for certain, it is that archaism of language in a certain passagedoes not guarantee archaism of content or vice versa. The Kunstsprachehas its own, complex stratigraphy, but for the poets who used it, it was asynchronic system: very archaic forms were readily available even as theycomposed new scenes.

Our approach to the evolution of the epic diction and its implicationsfor the evolution of epic as such must therefore be statistical. The first andonly ambitious attempt to apply the knowledge gathered by more thana century’s linguistic analysis of Homer to the whole early epic traditionin Greece was Janko’s Homer, Hesiod and the Hymns (). This bookstudies the frequencies of select features of the epic diction that are knownto be archaisms for Ionic poets. Applying statistical methods, Janko claimsto show that there was indeed a gradual modernization of the diction

Notably Hoekstra () and Shipp (). Conveniently collected in A. M. Parry (b). E.g. Wachter (this vol.) and Hackstein ().



Introduction

proceeding at a more or less regular pace. Moreover, by extrapolatingbackwards from the historical evidence, Janko aimed to show that theAeolic forms of Homer’s language were introduced within a very shortspan of time. Not all scholars agree with his approach and his conclusions,as is apparent in Brandtly Jones’s and Martin West’s contributions to thisvolume.

Linguistic research into the language of Hesiod and the hymns hasfocused on establishing the relationship to Homer’s diction. Althoughthere is some evidence for independent archaisms in these works, the gen-eral conclusion is clearly that Homer’s language is more archaic. However,not all scholars agree that this points to a chronological difference: insteadthey point to the influence of factors such as genre and local traditions.

This is at the core of the problem of correlating linguistic and literarydevelopments in the history of epic. While time is certainly a crucial factorinfluencing the language, there must have been other factors as well. Bardsno doubt attempted to modernize the language they used when the metrepermitted, but this was not their only concern. One can easily imagine thatarchaisms and foreign words lent gravity to the epic style. Haug (this vol.)shows that tmesis was a desired feature of the language, not something thepoets sought to avoid. Janko () himself showed the existence of falsearchaisms (i.e. predilection for old forms in hymns known to be relativelyrecent). This does not falsify the statistical approach to archaisms, but itdoes make it necessary to try to control for other influence than the simplepassage of time.

2 literary perspectives on relative chronology

By means of what criteria can literary, i.e. compositional, priority anddependence be established within the corpus of early Greek epic – andbetween those parts of it which have been preserved and those which havebeen lost? The discussion may be usefully divided into four, viz. regardingrelationships between (a) layers, earlier and later parts, within one and thesame extant poem, (b) individual passages in different extant works, (c)whole existing poems, (d) existing poems and what has been lost and mustbe reconstructed.

To identify layers within one and the same work is the classical field ofHomeric ‘analysis’, the separatist scholarship that came out of the Wolfiantheory that the Homeric epics were products not of an individual poet but

E.g. Edwards () and Hoekstra ().




of an illiterate song culture: the two long poems could not be unitary, andthe analysts proceeded to show that they were not. There is much to belearnt from analytic scholarship of the nineteenth and twentieth centuries(and from the interpretations of the ‘unitarians’ as well). Insight into theways of oral poetry has, however, taught us to judge inconsistencies andrepetitions differently, and not to seek in Homer qualities that are foreignto compositions in an oral tradition, even as that tradition is beginningto exploit writing. In the case of many short passages and episodes oftenregarded as spurious, each passage must be judged individually on thebasis of textual, linguistic and adequate literary criteria. When it comesto longer sections in the Homeric epics, which somehow seem secondary,one approach is to view them as evidence for additions and adaptationsmade by the poet himself in the course of a lengthy, creative labour witha poem monumental in both its conception and its bulk. Granted thatin an illiterate song culture the singers when expanding a song invariablyintegrate the new elements into their common style so that the whole – theadded sections included – emerges as stylistically seamless, the Doloneiaposes a special problem because of its conspicuous divergence from therest of the Iliad, as has been shown by Danek (, this vol.). Apart fromthe Doloneia, some other major problematic segments within poems areparts of Odyssey (catalogue of heroines, encounter with Heracles) andthe end of the Odyssey (from .), the latter parts of the Theogony,the relationship between the two parts of the Homeric Hymn to Apollo,allegedly interpolated passages in the Hymns (e.g. on nymphs in the Hymnto Aphrodite; on Python in the Hymn to Apollo), and the provenance ofAspis – and its relation to the poem’s main narrative.

Comparative literary analysis of individual epic passages aiming to estab-lish the direction of influence and literary borrowing has been an all-timefavourite of Homeric scholars. Arguments will always be informed by aes-thetic judgement; in addition, there are the problems of the fragmentarystate of the corpus, which should always make us ponder the possibilitythat there may be more relevant models in the mass of lost material, and ofshared tradition, which often makes it questionable to privilege a one-to-one correspondence between the verses in question. Some classical itemsare: the relationship between the catalogue of heroines in Odyssey and theCatalogue of Women (see Rutherford in this volume); relationships between

This approach was followed on a large scale in the case of the Iliad by Reinhardt (), whoexplicitly refused to take into account oral theory. Goold () also has in mind a poet who workson a written text over an extended period of time; for the application of this perspective on theOdyssey, see S. West in this volume.



Introduction

individual passages in the Homeric epics and passages in the two Hesiodicpoems (e.g. Theog. – // Od. .– etc.) and the Hymns; betweenpassages in the Theogony and passages in the Works and Days; between theIliad and the Hymn to Aphrodite; between similar passages in the Hymns(e.g. the epiphany scenes in the Hymns to Aphrodite and to Demeter); andbetween specific passages in the Hesiodic epics and passages in the Hymns.

When it comes to individual extant poems, discussion about the rela-tionship between the Iliad and the Odyssey would take pride of place, wereit not for the fact that there is virtual unanimity today about the priority ofthe Iliad. On the literary side, the strongest indication for the posteriorityof the Odyssey is that it consistently steers away from mythological groundcovered by the Iliad (‘Monro’s Law’). The Odyssey thus not only is, butalso views itself as, a sequel to the Iliad. There is such a thing as a ‘senseof text’ in Homer. In Pucci (), the Odyssey’s relationship to the Iliadwas at the heart of one of the earliest ‘intertextual’ readings of Greek epos,demonstrating both the potential gains to be had and the problems ofimporting elements of modern and modish theory into the very differenttextual culture of archaic Greece. In general, there has been an importantshift in perspective in that the younger poem tends to be read more andmore with the older one as a sounding board that contributes to its meaningand message.

While the priority of the Theogony in relation to the Works and Daysseems largely to be taken for granted today, although a good case can bemade for the opposite view, the relationship between Homer and Hesiodis subject to more debate, and must be judged non liquet (see, e.g., thecontributions of Richard Janko and Martin West in this volume). The samemay be said of the relationship of the Catalogue of Women to the Theogony,of which it is presented as a continuation, though its diction, according toJanko in this volume, is slightly more archaic.

Problems of a different order arise when we want to consider extanttexts in relation to what has been lost. How much do extant texts oweto older tradition, or, to put it differently: what can extant texts tell usabout tradition? Some specific questions addressed in the present volumeare: the link between the Iliad, the Odyssey and the Catalogue of Womenand a hypothetical Pylian epos on the family of Neleus (Rutherford); therelationship between the Odyssey and (putative) epic poetry about Heracles(S. West; Andersen); the relationship between the Hymn to Demeter and

See, for this idea, Dowden (). Theiler () is an ambitious attempt to entangle aspects of thegenealogy of the two poems when read against each other.

Schein (, ) makes the most of this. On the Odyssey, see Rengakos ().




epic poetry on the rape and recovery of Persephone, e.g. the ‘Orphic’poem partly transmitted in prose and partly quoted in a Berlin papyrus(Currie).

The main battleground, however, is the relationship between the Home-ric epics and the whole subject matter of the Trojan War such as it was toldand sung before the Iliad and the Odyssey came into being. We may safelyassume that what was told, was for the main part the tale and the episodeswhich were later – and after Homer – to be written down in the poemsof the Epic kyklos – Kypria, Aithiopis, Ilias Mikra, Iliou Persis, Nostoi,Telegony – of which works we only have fragments plus paraphrases oftheir plots with some cataloguing of episodes. Being later than Homer,these works may of course be influenced by the Iliad and the Odyssey.

The critical tools and concepts that are employed for shedding light onthese questions reflect the major twentieth-century scholarly tendencieswithin Homeric studies. In Homeric studies of the latter part of the twen-tieth century, Quellenforschung (source criticism) became, by way of neo-analysis, a kind of Rezeptionsforschung which inquired into the more or lesscreative redeployment of older material and episodes in the extant works.After pioneering work by others, Wolfgang Kullmann’s Die Quellen derIlias () was instrumental in establishing neoanalysis as a new paradigmin Homeric studies. Neoanalysis or, as Kullmann later has preferred it,the study of Motivubertragung (motif transference), was informed by theidea that the poet of the Iliad incorporated into his poem sequences ofnarrative motifs that belonged in other epic poems so that certain scenes inthe Iliad can be read as adaptations of scenes in lost epic poems. Whetherthis ‘technique’ operates subconsciously or consciously, and whether itshould be seen as a compositional resource only or as a way of alludingto, and even citing, other poems, is another matter. Originally the aimof neoanalytical studies was to retrieve one-to-one connections betweenspecific scenes and episodes in the Homeric epics and their assumed modelscenes in specific lost (pre-Homeric) Cyclic epics. Under the influenceof Kullmann, however, the scope of neoanalytical research has broadenedconsiderably to encompass all extraneous mythological material – Cyclic ornon-Cyclic – found in the Homeric epics, such as mythological examples,short or more developed mythological references etc. These are read asallusive references to the contents of an epic tradition embodied in con-crete songs, the contents of which are assumed to be reflected between the

Especially Pestalozzi (), Kakridis (), Schadewaldt (). See the useful survey by Willcock ().



Introduction

lines of the Iliadic and Odyssean narratives. The single most ambitiousrealization of this approach is Georg Danek’s Epos und Zitat: Studien zuden Quellen der Odyssee (b).

The methodological difficulties are considerable. For one thing, wedo not know how closely the post-Homeric Cyclic poems represent pre-Homeric poetry – and mythology as told at large – that would be availableto Homer, nor whether the transference is limited to motifs and narrativesequences only or includes portions of the original wording, if it makessense to talk of fixed wording at all, that is. And most importantly: bywhat criteria do we deem one use of a motif or a narrative incident moresuccessful than another, and what entitles us to think that the better use ofan ingredient is the prior use of it? Dysfunctional, redundant or otherwiseinexplicable elements may be important pointers – the third, loose horse inthe Nestorbedrangnis is a case in point. But it is to a very large extent a matterof judgement. After a period of relative quiet at this point, Adrian Kelly() challenged the rather commonly held view that the Nestorbedrangnisin book of the Iliad is dependent on a similar episode which we know tohave been part of the Aithiopis, where it would seem to sit and fit better,an episode which presumably belongs to pre-Homeric tradition.

Then again there is the larger question of how, if at all, neoanalyticalassumptions go together with the oralist perspective, in which the sharedtraditional material, the typical scenes and the conventional ways of thepoets are constitutive. Neither approach is primarily aimed at establishingchronological sequences; in the case of neoanalysis it is more a way ofexploring how poetry uses (older or contemporary) poetry, while from anoralist perspective the issue of relative chronology cannot come to the fore,because we have the poems in the form which they had at the compara-tively late moment when they were committed to writing. Kullmann ()himself made an attempt at rapprochement between the oralist conceptof ‘composition by theme’ and the neoanalytical one of ‘motif transfer-ence’. The area has later been innovatively explored especially by JonathanBurgess, first in his The Tradition of the Trojan War in Homer and the EpicCycle () and again in Burgess () and in his contribution to thisvolume.

Burgess argues for ‘a textless intertextuality’ positing unlimited interac-tions between the Homeric poems and pre-Homeric oral tradition, poeticand otherwise. In this he is in line with Danek (b, b), whoseresearch also has been conducted at the interface of neoanalysis and orality.

Kelly is answered by Heitsch ().




Both scholars have developed their views by arguing that meanings andstrategies are ‘inherent’ in the typical scene and that motif and expression,and the concept of ‘traditional referentiality’, are instrumental in exploringepic references. According to this principle, an individual oral realizationof an epic element (be it formula, typical motif or plot structure) only yieldsits full meaning within the broader context of the epic tradition. But if theepic – epic language, epic characters, epic episodes – is allusive generallyand always, by definition, so to say, refers to the tradition, to all there is,including even what is suppressed, it becomes difficult to think in terms ofrelative chronology. Poetological theory replaces historical analysis.

Bruno Currie’s studies (, this volume) are more open to the possi-bility of literary ‘interaction’, as he calls it, between specific texts. AlthoughCurrie adheres to the idea that early Greek epic is a product of an oralculture, his view of epic interaction largely rests on a conception of inter-textuality closely associated with authorial intention. And in contrast tooralist analyses that tend to deny that fixed wording was ever quoted inthe context of early Greek epic, Currie is more open to the possibility thatspecific formulations, bound to specific episodes or characters in identifi-able epic poems, were transferred to new contexts in the spirit of allusionto the source poem. In his contribution to this volume he tests his casefor allusive motif transference by examining the direction of interaction intwo archaic hymnic poems that have as their theme the rape and recoveryof Persephone and offers more positive evidence that allusive reference alsoincluded quotation of wording.

∗ ∗ ∗

Tradition, by providing idiom, formular phrasings and typical scenes, andthemes and motifs as well, furnishes the poet with compositional aids.The uses of tradition at all levels result in compositions in which we mayidentify a variety of intricate relationships between what is old and whatis new, in the shape of borrowings, adaptations and transformations. But

The concepts of ‘inherent meaning’ and ‘traditional referentiality’ are developed in Foley (,). Dowden () and Graziosi and Haubold () also aim at showing Homer’s ability toevoke a web of associations and implications by referring to the wider epic tradition. See now for alarge scale application Kelly ().

One may identify two tendencies. E.g. Finkelberg () argues that the ‘meta-epic’ nature ofHomeric poetry is intended to supersede or neutralize other traditions, while Burgess () seesthe ‘meta-cyclic’ nature of Homeric poetry more as parasitic in nature, in the sense that the fullextent of its potential meaning is dependent on the resonance of what is not there. Much depends onhow one envisages the Homeric audience (if there is such a thing) – how mythologically informedit is, and in what way the ‘potential’ context is actually evoked as they listen; on this, see Scodel(, ).



Introduction

the product is not a mechanical melee: the epic idiom ‘functions as alanguage, only more so’ (Foley : ). Synchrony and diachrony areintertwined: the idiom, the themes and the motifs do have their history,they are paradigmatic – but in each work they must be read syntagmati-cally within their individual context. Disentangling this complex of tradi-tional resonance and individual meaning is one of the tasks of the presentvolume.



chapter 1

�� Relative chronology and the literary history of the early

Greek epos

Richard Janko

The relative chronology of early Greek epic poetry is an essential questionfor anyone who wishes to understand the literature, society, history, religionand mythology of early Hellas. Yet the dating, both relative and absolute,of the poems themselves is a field where chaos and confusion reign. Withthe exception of Martin West, who famously stated that Hesiod’s Theogony‘may well be the oldest Greek poem we have’ (: ), until about

most scholars believed that the Homeric poems were composed later inthe eighth century bc, and the Hesiodic poems a generation or two later.The belief that the poems of Homer are older than Hesiod’s concurredwith the views of Eratosthenes and Aristarchus, who marshalled muchevidence, principally from geography, that Hesiod was later than Homer,since Hesiod knew of more remote places in the Mediterranean and namedthem more accurately, e.g. the Nile. Aristarchus argued this in a lost mono-graph entitled �� !" ��#, ‘On the age of Hesiod’. Now,however, many scholars who have made important contributions to ourunderstanding of different aspects of the early Greek epic tradition eitherprofess uncertainty about Homer’s date or argue for dates in the seventhor even the sixth centuries bc.

I am most grateful to Øivind Andersen and Dag Haug for their invitation to contribute a paper onthis topic, and to them and the other members of the audience for lively and stimulating discussion.

Reported by Strabo, .., ... Cf. e.g. schol. A (Aristonicus) ad Il. .a Erbse, $�� %&'�� ( �)�� #'��#� *#��+�,��, -��. ��.� �, ‘�� !" ��#’. The existence of this title was first recognizedby Schroeder (); for the form of the cross-reference cf. schol. A (Aristonicus) ad Il. ., /�� #�� 0� �#��0�. �� 1��, ��.� �, ‘�� ’2 �3�� 4,� 5��.In his lost work On the Homeland (sc. of Homer) Aristarchus argued that Homer was an Athenian(Janko : , ). It is symptomatic of the present confusion that the new Loeb edition of Hesiod(Most ) omits from the Testimonia the opinions on Hesiod’s date of two of the best scholarsin antiquity. M. L. West mentioned neither these scholia nor Aristarchus when he so influentiallyargued that Hesiod antedates Homer; his silence is curious, since he refers to Aristarchus’ argumentsand conclusion elsewhere (: –, cf. –).

Downloaded from Cambridge Books Online by IP 142.150.190.39 on Fri Jul 05 16:34:44 WEST 2013.http://dx.doi.org/10.1017/CBO9780511921728.003


��

Thus I broach this topic with some trepidation. This is not because Ido not know what I believe about it; indeed, that belief has been furtherstrengthened over time. In I published my doctoral thesis underthe title Homer, Hesiod and the Hymns: Diachronic Development in EpicDiction. This title reflects what I believe that I proved in that book, namelythat the poems of Hesiod post-date the Homeric epics, and that the longHomeric Hymns are mostly later still. The sole exception to this is theHymn to Aphrodite, where I then accepted arguments from the detectionof exemplum and imitatio, which I felt dated the poem after the Theogony(Janko : , –, fig. ), rather than my linguistic methodology,which strongly suggested a date soon after that of the Iliad (Janko :–, fig. ). I now think I ought to have had more faith in my ownmethod, and prefer to date the Hymn to Aphrodite to Homer’s time, withits oddities explained by the influence of the original Aeolic tradition onwhich it also drew. My primary focus throughout was on the relativechronology of the poems in the corpus, and not on their absolute date,although I did offer some very tentative suggestions about absolute dates.

At the time, my conclusion that Homer’s poems are earlier than Hesiod’sdid not seem much of a revelation, since, with the notable exception ofWest’s view, it matched the scholarly communis opinio. What was innovativewas that I had argued for this chronology from statistics, rather than fromthe relatively impressionistic methods that had been used hitherto (thoughthe conclusions of Eratosthenes and Aristarchus were better founded thanthat). I based my conclusions on a study of eleven linguistic changes thatare very common in the corpus of early hexameter poetry.

This was not a completely new approach. In the nineteenth century sev-eral German doctoral theses had tried to quantify one linguistic criterion,the observance or neglect of the effect of digamma. In the s Hoekstra(, ) had revealed how the modification of traditional formulaewas instrumental in introducing recent and indeed post-Homeric formsinto the Homeric Hymns, although he did not quantify his results. Garbrah

My reason for believing that the Hymn postdates the Theogony was that Hym. Ven. , [�+'61�] �.� ��.� 7�� , seemed to imitate the proem to the Theogony, [-�8 �] ��9�7��: �� | ��+�, . . . 7�� (Th. –). However, the case of the Hymn toHermes, where the opening verses survive in two versions with oral variants (Janko : ), showsthat short passages comprising the beginnings of Homeric Hymns were so traditional that they couldhave a relatively fixed form over extended periods of time: cf. Il. .–, where Homer drawson an earlier version of the proem to the Hymn to Hermes (Janko : ). Once the Hymn toAphrodite is correctly placed just after the Iliad, my diagram of the relations between the poems(Janko : fig. ) concurs to a remarkable degree with the results of the fine cluster-analysiswhich Brandtly Jones presented in Oslo and includes in his article ‘Relative Chronology within (an)Oral Tradition’, Classical Journal () – at –; his other arguments appeared too latefor me to address them here.



richard janko

() had undertaken a statistical analysis of the language of portionsof the Odyssey. Still more significantly, Edwards () had quantified fre-quently occurring aspects of Hesiodic diction, above all the usage of o-and a-stem accusative plurals, in order to explain the peculiar a-stem short-vowel accusative plurals that sometimes occur in his poems. But nobodyhad tried to correlate the behaviour of a number of different linguisticindices in order to see whether any consistent pattern emerged.

The work was at first undertaken by hand, with months of mind-numbing counting. I was encouraged to persevere by the late Sir DenysPage, to whom I had sent my early ideas (Professor Page did not live to seethat my results disprove his analytic theories, which I had set out to try tosupport). However, before long, with the help of Dr John Chadwick, theReverend A. Q. Morton and Dr John Dawson of the Literary and LinguisticComputing Centre at Cambridge, I was able to use computerized textsand key-word-in-context concordances of the early epos, at a time whencomputer-files still consisted of stacks of punch-cards and reels of papertape.

I well recall how, after all the counting had been done, in I stayedup one March night until three o’clock in the morning, while the windhowled outside my garret, in order to see whether any pattern emergedfrom all the results. To my excited eyes, as I plotted the first simple graphby hand, it seemed obvious that there was indeed such a pattern, whichmade me feel that all the effort had been worthwhile.

Statistical analysis is a method which, if properly applied and understood,affords a basis for a kind of objectivity that seems to belong more to thenatural and social sciences than to the humanities. Accordingly, I felt thatthe reliability of my conclusions went beyond the intuitions of previousresearchers. The conclusions to which statistical analysis leads are simplynumerical measurements of probability. At first sight, these differ in naturefrom the results of philological arguments. However, experience leads meto conclude that this is a false dichotomy, because the conclusions to whichwe come in philological or historical inquiry are very often probabilisticalso. Even in philology, our conclusions are only provisional; we mustalways be prepared to abandon them if a new hypothesis better explainsthe evidence. A statistical conclusion too is probabilistic, but with thedifference that we can state with some degree of precision how likely it isthat the alternative interpretation of the same data (what statisticians callthe ‘null hypothesis’) is in fact correct. It seemed then, and still seems tome now, that the statistical odds that these phenomena in the language ofthe early Greek epic are random are almost infinitesimally small.



��

Against this background, for the benefit of those who have not absorbedthe argument of Homer, Hesiod and the Hymns, I shall try to summarizeit in a way which I hope will easily be understood; I shall also integrateinto the argument some fresh observations that I have made during theintervening decades. Let me first set out the four presuppositions of mymethod.

First, the language of Greek epic is a Kunstsprache made up of elements ofdifferent dates, some of them very old; indeed, a few demonstrably go backto the middle of the second millennium bc. This Kunstsprache is morecomplex than that of any other oral poetry that we know, probably becauseof the demands of the hexameter, which is more fixed in its form than are theverse-patterns of other traditions, like the South Slavic deseterac. However,even South Slavic epic has metrically convenient oddities of Kunstsprachelike the interchange of the vocative and the nominative. Just as the Greekbards use ��1��4�� ;�+� or <�� =� �� as nominatives, whenthese adjectives were originally vocatives, so too the singers readily replacethe nominative form of a name with a vocative when they need to fill anextra syllable; thus when Avdo Mededovic wanted to say ‘Vuk asks’ in foursyllables, he would sing pita Vuce (Mededovic b: l. ). Onceoral poets master their special epic language, it becomes a language likeany other. The fact that it is a Kunstsprache does not imply that it willnot change; on the contrary, so long as it is part of an oral tradition, itsmorphology changes over time, as older forms are replaced by youngerones. This process of morphological replacement occurs more rapidlyoutside fixed formulae and more slowly within them, since formulae playan essential role in composition and may be retained for a long time evenafter they become unmetrical. While the Kunstprache is being used by oralbards for composition-in-performance, it will change along with changes intheir own spoken language. To quote Milman Parry, to whom this crucialpoint is owed:

The language of oral poetry changes as a whole neither faster nor slower than thespoken language, but in its parts it changes readily where no loss of formulas is

Several linguists have recently denied this (Hackstein : –; Haug : –; Hajnal :–). However, those who hold this view seem to neglect the archaeological evidence for objectsthat go back to the Early Mycenaean period, like the body-shields that inspired some of the phraseswhich once had syllabic r like �� '1�>��. Of course, such phrases could have been developedin Central Greece (where a mixture of Mycenaean and West Greek was probably spoken) rather thanSouthern Greece, so the ‘Mycenaean’ phase of the epic tradition may be more chronological thangeographical in significance.

As Haug (this vol.) shows, the rate of replacement in syntactical features like tmesis is likely to havediffered from that of morphological features; but this does not affect my argument.



richard janko

called for, belatedly when there must be such a loss, so that the traditional dictionhas in it words and forms of everyday use side by side with others that belong toearlier stages of the language. (in A. M. Parry b: )

This admirable insight was long overshadowed by Parry’s other claim thatall or most epic verse consists of formulae, which would of course implythat their language hardly changes over time. We owe the rebuttal of thislatter assumption to Hoekstra () and Hainsworth () in particular.

My second presupposition was that the decipherment of Linear B byVentris and Chadwick and the reconstructions of historical linguistics,many of them decisively confirmed by the decipherment, are in theiroutlines largely accurate and reliable, so that we can say that, for instance,phrases with the digamma preserved are older than those which presupposeits loss. Indeed, the method can sometimes be used to confirm those results.For example, when statistics show that the contraction of 7+ to �? spreadsthrough the epic as the poems become later, this helps to confirm that 7+�‘good’ is derived from Indo-European ∗

ə1sus and is cognate with the Hittiteadjective assus of the same meaning.

My third presupposition was that, in order to obtain statistically depend-able results, it is in most cases useless to study elements of the language thatare rare, like infrequent linguistic forms or isolated items of vocabulary;the occurrence of the latter depends too much on the vagaries of content,and one cannot usually know against which other words they must be mea-sured. Instead, one must study common features, like phonetic changes andmorphological endings. Fortunately I was able to identify more than ten ofthese (and a few more since) that are so frequent that they yield statisticallysound samples in all but the shortest poems in the corpus. Each statistic isbased on counting dozens or even hundreds of occurrences (‘populations’,in statisticians’ terminology) of tiny things, about many of which the poetscould hardly have been aware. One can usually obtain some usable resultsfrom poems as short as to lines, but everything depends on thefrequency of the particular phenomenon that is being studied.

My fourth presupposition, related to the third, was that this methodapplies only to large bodies of text. It cannot date brief passages or particularlines, although one can certainly observe more recent or more archaic formswithin them, and doing so afforded me many insights into the creativity ofHomer when I was commenting on the Iliad. (It is important to emphasizethat the minimum size of the body of text cannot be precisely defined, since

See Janko (: n. ), with further examples. This is the weakness of Hackstein ().



��

it is contingent upon the number of instances of the phenomenon thatis under observation and the level of probability that is desired from theresult.) Because of this restriction, the method presupposes the generalintegrity of the transmitted texts of the early epic. But so do most scholars,except for a few diehard analysts and the surprising number of thosewho think there never was a fixed text. The work of the unitarians andneoanalysts seems to me decisively to support the literary integrity of thebulk of the Homeric epics. Indeed, if we were to seek to date Homer’spoems on the basis of the date of their latest lines, we ought to put theminto the Roman or Byzantine eras, when concordance-interpolations werestill entering the papyri and codices in significant numbers. Thus I am notcomfortable with efforts to date whole poems solely on the basis of singlelines or brief passages.

Subsequent study of the textual transmission of the Homeric poemsconvinced me that they may contain minor interpolations that arrivedthere at a very early date, before many manuscripts of the poems existed. Iconcluded that ‘a pessimistic editor is entitled to suspect that some spuriouslines permeated the whole paradosis so early that the Alexandrians couldnot detect them’ (Janko : ). These would be far harder to discoverthan are the later concordance-interpolations, which are betrayed by theirabsence from earlier papyri and/or quotations, but are not all expelled evenin West’s new text of the Iliad. In that text West also identifies a numberof early short interpolations. While I rarely share his confidence that theycan be recognized, I agree that in principle they are possible and perhapseven probable; but they are unlikely to affect the statistics. On the otherhand, the years that I spent in writing a commentary on the longest booksof the Iliad strongly reinforced my confidence that so skilfully structured apoem could only have been composed by a single poet at a single time, eventhough he was using the techniques of oral composition-in-performance,and that we have his poem much as he dictated it. Thus a position that isboth oralist and unitarian is fully compatible with this statistical method.

My original aim had in fact been to uncover consistent patterns ofvariation in the separate books of the Iliad or Odyssey, where clustersof more archaic or innovative diction would indicate the existence ofpassages of divergent date and origin. However, I never found any, despitemonths of arduous labour; indeed my results converted me to a unitarian

This is effectively the error that I made myself when I dated the Hymn to Aphrodite after theTheogony (above, n. ).

West (–), with West (a: –). See Janko (, ). See Janko ().



richard janko

0

20

40

60

80

100

1 2 3 4 5 6 7 8 9 10 11 12

Book-number

Percentage

Fig. . Linguistic variation in the Iliad, books –.

position. Since a book that managed to prove nothing in particular wouldnot have been interesting to read or to write, and I had made otherdiscoveries that seemed important, I presented those instead. I suppliedin an appendix (Janko : –) the raw data for the individual booksof Homer so that others could test for themselves my denial that suchvariations exist. Few have done so, except for Danek (; this volume:–), who discerns only qualitative differences between the Doloneiaand the rest of the Iliad, in aspects like vocabulary and formular usage.Qualitative divergences such as those to which he points will never beable to be detected by a quantitative method such as mine, and theirexistence and significance will always be a matter of judgement ratherthan of statistical evidence. As he acknowledges, there are no decisivequantitative differences between the diction of the Doloneia and that ofthe rest of the Iliad; I failed to detect any elsewhere either. Figure .

Janko (: ). Also, the format of my publication limited the number of illustrations and ofwords. My omission was regretted by one reviewer (Fowler ).



��

0

10

20

30

40

50

60

70

80

90

100

Iliad Odyssey Theogony Erga

percen

tag

es

Fig. . Diminishing archaisms from Homer to Hesiod.

presents as a graph the variations in eleven very common features of epicdiction (which I shall explain below) in the first twelve books of the poem,based on the data in my appendix A. One can see even by eye that thedivergences between books exhibit no consistent pattern. The rest of theIliad and Odyssey presents a similar picture, indicative apparently of randomvariation.

When, on the other hand, I plotted the results for the whole epic corpus,I at once found an almost consistent decline in these eleven very commonfeatures, as they evolve from the Iliad to the Odyssey to the Theogonyand Works and Days. Figure . shows the different choices made in thefour poems in fourteen features of epic diction, including three additionalones – the incidence of 40 as against contracted 4�, the failure tocontract 7+ and 7#- to �? and �(-, and the use of �� rather than ��.According to the evidence, each feature comprises the retention of an

Janko (: –). I have revised the calculations of the results for -� and -�� in accord withthe suggestions of Jones (this vol.).



richard janko

0

10

20

30

40

50

60

70

80

90

100

Iliad Odyssey Catalogue Theogony Erga Hy. Dem.

* false archaism † regional variation

percen

tag

es

Fig. . Diminishing archaisms in epic diction.

archaism that could have been replaced by a more recent form. The onlyfeature among the fourteen that does not belong to this pattern is anAttic-Ionic innovation, the use of n-mobile (�8 71��# ��) before aconsonant; I include this feature to show that it disappears in works ofBoeotian origin like those of Hesiod.

Classicists are notoriously averse to numbers, but figure . should makeit clear to anyone that there is a coherent pattern in these results. It emergesthat the Odyssey is nearly always more advanced than the Iliad, the Theogonyis always more advanced than the Odyssey, and the Works and Days is nearlyalways more advanced than the Theogony. Yet the Odyssey is extremelyclose to the Iliad, and the Works and Days stands in a similar relationto the Theogony. On the other hand, the frequency of n-mobile before aconsonant increases over time in Ionic and Attic poems and diminishesin Boeotian ones, so that in this Hesiod seems more archaic than Homer.In figure . I have added the results for the Catalogue of Women and the



��

Table . The statistical basis of figure 1.3 (%)

Iliad Odyssey Catalogue Theogony Erga Hy Dem

digamma (observed) . . . . .masc. gen. in -� . . . .

gen. plur. in -�� . . . . . .gen. sing. in -�� . . . . . .gen. sing. in ∗-�� not

–�# . . . . .

dat. plur. in–�� /-�� /-� �

. . . . . .

dat. plur. in -�� @/-� @not -��/-��

. . . . .

o-stem acc. plur.before vowel

. . . . .

a-stem acc. plur. beforevowel

. . . . .

A�- not ;��- . . . . . .40 not 4� . . . .�� not ��

7# not contracted . . . . . .n-mobile before cons.

per vv.. . . . . .

total no. of instances (sample size)digamma (7�+�

excluded)

masc. a-stem gen. sing.

a-stem gen. plur.

o-stem gen. sing.

gen. sing. in ∗-�� or-�#

o- and a-stem dat.plur.

o- and a-stem shortdat. plur.

o-stem acc. plur.

a-stem acc. plur.

oblique cases of ;�+�

40 and 4�

�� and ��

7# and �(

n-mobile beforeconsonant

no. of verses c.



richard janko

Homeric Hymn to Demeter. The data on which this graph is based are givenin table ., to show that the numbers of instances involved are high; weare not counting just a few occurrences.

As one expects with statistics, figure . reveals two outlying results in theHesiodic Catalogue of Women. However, both are based on small samples,and most of its other results fall between those for the Odyssey and thosefor the Theogony. There is, however, a tremendous scatter in the HomericHymn to Demeter, a phenomenon to which I shall return. In the first fivepoems, a general pattern is evident. The Iliad is the most archaic poem interms of its diction, the Odyssey is next, and Hesiod comes last, generallyin the sequence Catalogue of Women, Theogony and Works and Days. Thisis of course only a relative sequence; it is not translated into centuries orgenerations, let alone years. But it is a sequence, and unless the statisticsare somehow wrong, it is therefore a fact. As a fact, it is a phenomenon inneed of explanation.

Even before I wrote, great effort had gone into arguing that linguisticdifferences between the poems depend on genre alone; the claim was thatsome genres are more conservative in their use of old formulae than areothers. Much of my book was devoted to examining this claim. It has someplausibility only in the case of genealogy, where the Catalogue of Womenis consistently slightly more archaic than the Theogony, an acknowledgedpoem of Hesiod which contains rather less genealogy as a proportion ofits total number of lines. Yet the Catalogue must be either contemporarywith Hesiod or later than he, since it is attached in content, structure,and the manuscripts to the end of the Theogony (Janko : , –).Almost everyone ‘knows’ (or thinks they know) first, that the Catalogueis spurious, and secondly that it was composed after the foundation ofCyrene in bc; indeed, many hold that it was written in the sixthcentury. I shall argue below that neither of these views need be correct.

When poems such as the Iliad and Odyssey display a greater usage ofarchaic diction than do the Theogony or Works and Days, and when theydo so consistently, with ‘clusters’ of early results (e.g. the Odyssey is nearlyalways slightly more ‘recent’ than the Iliad), it is counter-intuitive to

Notably Pavese (, ). Drager (), who ascribes to Hesiod both the Catalogue and the end of the Theogony, is a recent

exception; likewise Arrighetti (: –). So, notably, West (: –), Fowler (). West’s arguments are successfully rebutted by Drager

(: –). Brantly Jones (this vol.) has proved that the Odyssey is more archaic than the Iliad in one result, the

genitive in -�, once ambiguous cases before vowels are discounted. I accept his correction, whichhardly affects the overall picture. My tables reflect this correction.



��

suppose that this could be explained by anything other than linguisticchange through time. I shall look at other possible explanations shortly;but we should only embrace a different kind of explanation if we haveother, compelling types of evidence that point us in that direction. It isboth simpler and more logical to accept that the preponderance of earlydiction in Homer as compared with Hesiod is because the texts of Homerwere somehow fixed at an earlier date than were those of Hesiod.

There is one major complexity to be addressed before we turn to theimplications of this relative chronology of Homer and Hesiod. Figure .reveals that the Homeric Hymn to Demeter does not behave like the textsof Homer and Hesiod. The Hymn has more recent diction than Hesiod insix features, but more archaic diction in seven others (I exclude n-mobile,as this phenomenon varies by region). The more recent features are (a) therarity of 40 rather than contracted 4�; (b) the infrequency of elided o-and a-stem dative plurals in -�� ’ and -�� ’ before vowels; (c) the rarity of a-stem accusative plurals before vowels; (d ) the rarity of long dative plurals in-�� and -�� ; (e) the rarity with which the uncontracted o-stem genitivesingular in ∗-�� could be restored; and (f ) the rarity with which the effectof digamma is observed. Of these six criteria, all are based on statisticallyreliable samples. The reason why the poet does not archaize in observing thedigamma, or in observing the earlier frequency of uncontracted genitivessingular in ∗-��, is that he saw no reason to do either thing; he probablydid not know that there were archaisms at stake in these features of hislanguage, since the existence of digamma and of the uncontracted genitivein ∗-�� was long in the past of the Attic dialect when he composed hisHymn.

The pattern is different in other linguistic features, some of whichpoets could have consciously manipulated in order to achieve particularliterary effects. The seven more archaic features of the Hymn to Demeterare (a) the frequent use of the stem A�- instead of ;��- in the name ofZeus; (b) the frequent retention of the archaic genitive plural in -��; (c)the retention of the archaic genitive singular in -�; (d ) the retention ofuncontracted 7# both on its own and in compounds; (e) the retention ofthe archaic genitive singular in -��; (f ) the frequent use of the accusativeplural in -�#� before a vowel; and ( g) the frequent use of archaic ��rather than ��. However, no fewer than four of these phenomenarest on unreliably small samples of eight occurrences or fewer: these are(c) the retention of the archaic genitive singular in -�, (d) the retentionof uncontracted 7# instead of �(, (e) the frequent use of the accusativeplural in -�#� before a vowel, and (g) the frequent use of �� rather



richard janko

than ��. Accordingly, no weight can be placed upon these criteria. Butthe other three are statistically significant, and are distinctive because theyare features which the composing bard certainly could have manipulatedconsciously. First, the retention of the Attic stem A�- instead of the EastIonic ;��- in the name of Zeus is a feature which one can attribute to thevernacular Attic dialect of the poet. The second and third features arethe retention of the archaic genitive plural in -�� and the retention ofthe archaic genitive singular in -��. Both features are sonorous disyllabicendings, and I believe that the poet consciously favoured them in order toenhance the solemnity and archaic flavour of his verse.

Examination of most of the other long Homeric Hymns and of theHesiodic Shield of Heracles reveals even more numerous departures fromthe pattern, so much so that they have no ‘cluster’ of results at all (Janko: –, with table and fig. ); I forbear to demonstrate this here,and refer you to my book and to a related article (Janko b), where Iargued from historical evidence that the Shield of Heracles and the Hymnto Pythian Apollo were both composed in central Greece in the first twodecades of the sixth century bc, with the Shield being prior. I explained thepresence of anomalous results like those in the Hymn to Demeter, the Shieldof Heracles and the Hymn to Pythian Apollo as follows. Once we are dealingwith poems from a very late stage in the tradition, the poets are no longerlearning their diction exclusively from their older living contemporaries.They are also hearing renditions of texts fixed at much earlier stages ofthe tradition, from which they are learning to adopt certain archaisms,i.e. those which are immediately noticeable to a listener. These poets didnot care whether they put a formerly uncontracted disyllable in heavy orlight position in the hexameter, but they felt passionately about producinga grand-sounding poem, packed with euphonious polysyllables like -��and -��. The same can be demonstrated for all subsequent composers ofGreek hexameters, from Theocritus and Apollonius of Rhodes to Quintusof Smyrna and Nonnus (Janko : ).

I called this phenomenon ‘false archaism’, which still seems the bestname for it. Poets who practise ‘false archaism’ are imitating older modelsselectively. They could not imitate them perfectly, even if they wished todo so. However, if poets are free to imitate aspects of older models from thesame tradition, one may well ask what value an analysis of their languagecan possibly have for chronological purposes. In effect, this phenomenonof ‘false archaism’ might seem completely to demolish the edifice of graduallinguistic innovation that I had carefully constructed.



��

Yet I do not believe that it does. Just because some epic poets practised‘false archaism’, it does not logically follow that they all did so. The consis-tency of the results that I obtained for Homer and Hesiod shows that theirpoems are not subject to ‘false archaism’ but are genuinely archaic com-positions, whose the bards learnt their diction from their contemporariesand not from some far older, fixed text. If such texts were available in anyform to Homer and Hesiod, they have had no discernible effect on thetexture of the language of those poets. The later poets, on the other hand,like those of the Hellenistic and Roman periods, offer no such consistentpatterns.

So what is the significance of the fact that the poets of the long Hymnsto Hermes, Demeter and Pythian Apollo practise ‘false archaism’, whereasthe texts of Homer and Hesiod, including the Catalogue, show no sign ofit? The only explanation is that those poets who had access to texts fixed atmuch earlier stages of the tradition were themselves later than Homer andHesiod, who had no such access to such ‘fossilized’ material. Later poetsknew the work of much older predecessors, as did their Hellenistic andRoman successors. This accords with the growing diffusion of literacy inthe Greek world, and with the gradual increase in the number of writtentexts and of performances based on them, such as are documented forAthens in the time of the Pisistratids. The picture of a coherent evolutionof the epic diction remains consistent.

Hence one must ask what it means that the texts of the two poems ofHomer and the three Hesiodic poems were ‘fixed’ at different stages in theevolution of the diction. I have already suggested that this reflects a relativechronological order. But the fact that the poems were fixed at differenttimes has other important implications. If we believe, as I do, that theTheogony and the Works and Days are by a single poet, the evolution inHesiod’s diction from the Theogony to the Works and Days shows that theepic Kunstsprache could evolve even during the lifetime of a single poet;the same is true, mutatis mutandis, for Homer, where ‘Longinus’ (Subl..) plausibly suggested that the Odyssey is the work of the poet’s old age.Others are willing to accept Kirk’s hypothesis (: –) that the epicswere accurately transmitted by memory over a considerable period beforethey were written down; but I have never seen any decisive argument infavour of this view, since these are not sacred texts like the Rig Veda. It is

Hence the account of my method in Martin (: ) is gravely mistaken when it is said thatthere is no criterion for distinguishing false archaism, when the criterion is lack of consistency inthe diction, i.e. the lack of ‘clusters’.



richard janko

far easier to accept that the epics acquired a fixed form at different times,that we have them in much the form in which they acquired this fixity,and that their language differs in these subtle ways precisely because theywere fixed at different times. Others, again, may prefer to believe that theauthors of all these poems wrote them down in this form. However, likeKirk’s theory, this too is an unnecessary hypothesis, because it assumes thatthese poems could not have been composed without their authors usingwriting. This belief, in my view, reflects our own lack of imagination andof experience of an oral culture rather than the realities of skilful oral epiccomposition in a largely illiterate society.

Since the evidence shows that the poems were fixed at different stagesin the ongoing evolution of the tradition, those who wish to ascribe theirpresent form to editors must accept that they have been uniformly and con-tinuously updated by experts in historical linguistics to keep them distinctfrom each other in terms of language. This is not credible. Accordingly,I supported very tentatively in my book, and more definitely since, thehypothesis of Milman Parry and Albert Lord that these epics are oral dic-tated texts, and that their linguistic form accurately reflects the differentstages during the tradition when they were taken down. I believe, followingLord, that the slower pace of dictation permitted the best poets to producebetter poems than they could have sung. The simplicity, coherence andcomprehensiveness of this explanation of how the texts were fixed havenever remotely been matched by any other theory.

Although some reviewers of my book wrote that they could not followall the statistics, on the whole they accepted my results, at least as regardsrelative chronology. However, as time went on it began to appear thatthe book was being more cited than read. Unfortunately, it went out ofprint within ten years. Also, it presupposes much expertise in historicallinguistics and the Greek dialects, as well as an intimate knowledge of earlyGreek hexameter poetry. Although I kept the statistics as elementary aspossible, the fifty-two tables of numbers are dreary stuff in themselves. Inan account of the language of early Greek epic I supplied further statisticswhich support my case (Janko : –, esp. –), but none of myrecent critics takes account of these additions to my theory.

During the s I began to notice some really strange pronouncementsabout the date of Homer. These often began by setting aside the linguisticevidence, allowing scholars to adopt other approaches, above all the notion

Lord (, : –, : –, –, –, –, : –); cf. Parry in A. M. Parryb: .



��

that the texts were never fixed. This was in my view already refuted bythe statistical results presented in my book. Such critics usually cited,sometimes apparently at second hand, my appendix E. There I offered,among a number of tables giving different scenarios for the absolute datingof the epics, provided that their relative dates were upheld, the observationthat, if the date of the composition of the Iliad were set at around –

bc, that of the Odyssey would fall at around –, the Theogony in –, and so on. The citation of these exact dates became a sort of talismanfor these scholars, allowing them to overlook the mountain of statisticalevidence, because it seems absurd to give an exact date in a case where therewill never be documentary evidence for the precise years in which thesetexts were fixed. I quote an example of such a critique, written by a scholarwhose work I admire in other respects:

[T]here is still the internal evidence of linguistic forms to consider. On the basis ofsuch evidence, Janko () has insisted that the Homeric poems must date fromthe eighth century, and he came to conclusions about the relative dating of manyother poems as well. Janko’s study displays an unsettling confidence that epic verseeverywhere was developing in the same way in a robotically steady manner. Thestatistical quantification of forms here does not make enough allowance for thevarying length of the different poems that are being compared, and indeed the verymeagerness of what survives in some cases can only lead to misleading percentages.Variance in diction and form between poems was undoubtedly [my emphasis]caused by subject matter, poetic function, local dialects, and the preferences orability of composers. The epic tradition on the whole was a swirling flux ofcrosscurrents, which a rigid statistical analysis cannot hope to measure. And evenif we accept conclusions about chronological relativity, that means nothing aboutthe historical time of the poems. An unverifiable argument about the LelantineWar is the peg on which Janko hangs the whole frame of his relative dating toa historical timeline. Finally, the very desirability of assigning specific dates toindividual poems is open to question. Assuming a moment in time for the fixationof early epic does not change the fact that a lengthy process of oral compositionlay behind it. (Burgess : –)

A footnote adds ‘[t]o be fair, Janko is certainly not unaware of these issues’,and gives page-references that amount to per cent of my book.

E.g. Taplin (: n. ). Others make a similar error over the level of exactness in dating, whenthey hold that I believe that, because the Theogony is linguistically slightly more advanced than itssequel the Catalogue, the latter was necessarily composed before the former (e.g. Haubold :

n. ). As I had noted (Janko : –), ‘linguistic tools are inevitably somewhat blunt’: greaterdivergence than this is found within the attested work of Hesiod himself, i.e. between the Theogonyand Works and Days. It is of course possible that Hesiod dictated the Catalogue before he dictatedTheogony –.



richard janko

As for ‘epic verse developing everywhere in a robotically steady manner’,I wrote many pages on the influence of local dialects, since I wanted toestablish the origin of each poem. I found that one can easily distinguishIonian from mainland compositions, which have Boeotian influence inHesiod and Attic features in the Hymn to Demeter. As for ‘not mak[ing]enough allowance for the varying length of the different poems that arebeing compared’, what matters is the number of occurrences, not the lengthof the poem, although the two things are related. Part of the purpose ofstatistics is that it enables one to compare populations of different sizes.Every result was subjected to tests such as t-tests and � tests to make surethat it was statistically reliable, as my footnotes reveal; results based oninadequate sample-sizes were set aside. Again, when it is suggested thatthe diction was affected by ‘the preferences . . . of composers’, this can onlybe partly true; in many cases, such as how many uses of the contractedgenitive singular in -�# can be replaced by the uncontracted form in ∗-��,the bards cannot have been aware of this difference and had no experts inhistorical linguistics to advise them.

West (: –) suggests that ‘the major determinant of the quantityof younger forms in a given poet is the extent to which his language divergesfrom the formulaic, and this depends on many other factors apart from hisdate’. However, it is not clear how one could be certain that this is the majordeterminant rather than the date: in the Homeric poems the proportionof speeches, which constitute the least formulaic part of the diction, isthe highest in the corpus, yet these are the poems with the most archaicdiction. Moreover, I did examine, at some length, the question of ‘variancein diction and form between poems’ being caused by ‘subject matter, poeticfunction, local dialects, and the preferences or ability of composers’, andwrote as follows:

It might be alleged, for example, that battle-poetry has more ancient antecedentsthan stories of nautical adventures, genealogies or gnomic and rustic verse. Suchan approach is almost impossible to test on present evidence, at least within theearly Greek tradition, with one important exception, and that is in Hesiod. It issometimes assumed that gnomai and accounts of rural activities are likely to bethe least traditional of all the types of hexameter poetry of the archaic period.Certainly it is true that the diction of the Works and Days is very advanced: but itis nonetheless not as advanced, by a good margin, as the narrative in the Hymn toDemeter, which is in terms of content reminiscent of scenes in Phaeacia and Ithacain the Odyssey. Moreover, Works and Days – consists of mythic narrative, and

Griffin (). It does not of course follow from this fact that the Homeric poems cannot be orallycomposed, as Griffin argued.



��

not of moral and agricultural precepts . . . And yet, when the language of thesetwo hundred lines is studied scientifically, it is established that, although it issomewhat less advanced in the linguistic criteria than the remainder of the poem,it is on the whole more advanced than the diction of the Theogony. Hesiod’sdiction advanced, as we know, but it advanced despite the content. (Janko :)

However, the critic whom I quoted is partly correct that ‘if we acceptconclusions about chronological relativity, that means nothing about thehistorical time of the poems’; it is true that relative and absolute chronologyare different questions. I would only say that, if the artistic or historicalevidence for the date of Homer did point decisively to the seventh or sixthcentury, then Hesiod would have to be later still by some margin or other.If people wish to accept this, they will need to explain how it is that Hesiodis imitated not only by Alcaeus but also by Semonides of Amorgos, since inanother appendix in my book (–) I surveyed all the apparent cases, inthe epos and other early poetry, of exemplum and imitatio. There I showedthat there are a number of ways in which this can reliably be recognizedfrom the adaptation of formulaic diction. As for the eighth-century dateof Homer and the Lelantine War, what I actually wrote was ‘if . . . theLelantine War can indeed be dated to Archilochus’ time, and if Hesiod wasconnected with it, and if Homer was not prior to the mid-eighth centuryand not later than its end, then the relative datings will be corroborated,but no universal agreement on these points exists at present’ (–). Therewere a lot of ‘ifs’ in this sentence, and deliberately so, since the historicalevidence is hard to confirm; but the arguments about the Lelantine Warhave not changed, even if West is no longer so confident about them.

Finally, if critics question the desirability of assigning dates to individualpoems, I do not see how one can understand the literary history of earlyHellas without knowing at least the relative dates of the major poems. Onemight as well say that ignorance is better than knowledge. Other critiquesof my general approach have been similar to those discussed above.

I gave the figures in appendix B (–). West (: –): contrast West (: –). Here is another: ‘[Janko] uses various linguistic criteria to establish that the language of the Iliad

and Odyssey is less “modern” than that of other hexameter poems. On the basis of this observationhe establishes a relative chronology of Homer, Hesiod and the Homeric Hymns; he then assumesan eighth-century date for the Iliad and uses it to calculate the absolute dates of the other poems.There are various possible objections . . . : in the first place, the eighth-century date is, by hisown admission, simply assumed. Secondly, Janko’s own linguistic criteria measure choices betweenalternative epic forms which sound more or less archaic, but which were all current within theIonian poetic language, given that they feature, with greater or lesser frequency, in all or most early



richard janko

When one critic suggested that ‘assuming a moment in time for thefixation of early epic does not change the fact that a lengthy process of oralcomposition lay behind it’, nobody denies this; in fact, projecting back fromHomer’s usage to the beginnings of various linguistic phenomena confirmsthe view of Meister, Parry, West, Ruijgh and now Haug (: –) thatthe epic tradition resided in Asiatic Aeolis until shortly before Homer’stime (Janko : –, with fig. on p. ). Here I must address thediscussion (Chapter below) by Brandtly Jones, who earns my thanks forbeing the first to acknowledge the statistical dimension of the prehistory ofthe epic diction. His account of my sequence of changes in the prehistoryof epic diction seems to misunderstand my argument in several ways.

However, he is right that my method of calculation of the introductionof quantitative metathesis in the a-stem genitives was flawed. I shouldhave excluded from the calculation the cases of -�� before vowels, becausethese could, as he suggests, be instances of elided -�. When this is done,the relative date at which a monosyllabic scansion of the a-stem masculinegenitive singulars enters the epic diction turns out to be −. units (ratherthan −. as I had thought) on the Common Scale, where the Iliad is set at and the Theogony is set at +., because the result for the Iliad changes.Similarly, I should not have omitted the forms in -�� and -�� from thetotals of a-stem genitive plurals that admit quantitative metathesis, since

hexameter texts. Choices and preferences between available words and expressions may thus beinfluenced by the subject-matter of the poem, the specific area in which it was composed, and thetastes of particular bards and audiences, and not simply by the date of composition’ (Graziosi :–). These are very similar arguments.

A different critique is that of Sauge (). (I am grateful to M. Sauge for a copy of his book.)In order to set aside my results and argue that the epics were recorded in sixth-century Athens,Sauge has to reject Hoekstra’s overwhelming proof that epic poets modified their formulae ( n.). Instead, he suggests that the epic diction (but not the texts) was fixed in the early eighthcentury(!), perhaps by Cynaethus on Chios (, ); but Cynaethus was alive in c. – bc

(Burkert ). He holds that the poets learned and employed two different pronunciations at once.Thus, he believes (–), in phrases involving initial digamma they pronounced the phoneme orignored it according to whether it is ‘observed’ or ‘neglected’ (and in order to furnish an objectionto one of my results he rejects the consensus that ‘initial’ digamma in enclitics is in fact in a medialenvironment, as the usage of Sappho and Alcaeus proves). I have not seen Blumer (: I. –);like Sauge, he does not address my argument from ‘clusters’ (Fowler : ), which is the crucialone.

Regarding -��, -�� and -�, my point is indeed that the metrical and not the phonological shapeof the termination changed, but I do not see how this affects the argument: if -�� was ever presentin the epic diction, why was it changed back to -�? Regarding the change in the admission ofelision in the dative plurals, Greek has or has not admitted elision in different words at differenttimes, and it is surprising that Jones misconstrues me as thinking that such a change could havereflected a general change in the practice of elision in all words. Nor is it ‘bizarre’ to suggest thatthe datives of �- and a-stems were treated differently, since these represent syncretisms of differentendings: in the o-stems Mycenaean has distinct endings in -o-i and -o (from the instrumental in∗-ois), whereas in the a-stems it has only -a-i.



��

in the case of -�� these contracted forms go a step beyond quantitativemetathesis and in that of -�� they are an equivalent response in a dialectother than Ionic. When this is done, the relative date at which a-stemgenitive plurals with a monosyllabic scansion enter the tradition turns outto be −. rather than −.. The sequence of changes that results fromhis corrections is as follows:

. -�� ceases to be sole o-stem genitive singular ending: −... Contraction in ∗-�� first appears: −... o- and a-stem dative plurals begin to admit elision: −... Initial prevocalic digamma begins to be neglected: −... Monosyllabic scansions begin to replace -��: −... The Ionic declension ;�� etc. first appears: −... ‘Short’ o- and a-stem dative plurals which cannot be elided first appear:

−... Monosyllabic scansions begin to replace -�: −.., . o-stem and a-stem accusative plurals begin to be biassed towards

appearing before consonants: not before −. and −. respectively.Jones notes that the small sample of a-stem genitive singulars in theTheogony (only in total) is a ‘narrow and insecure footing on whichto base the date of the Aeolic phase of the epic tradition’ (this vol.: n. ).The sample is indeed too small, but of course we cannot obtain anothersample of Hesiod’s diction as it was at that moment – this is all we have;and it is not the sole basis for the chronology, since there are also theresults for the other works of the corpus and the behaviour of the genitiveplurals. Thus the Works and Days has a small population of a-stem genitivesingulars (only instances); again this is a low number, but per cent ofthem are in -�, as opposed to . per cent in the Theogony. Hence theirincidence in the Theogony may be abnormally low; if the true percentageof forms in -� there were higher, it would yield an earlier date for theappearance of the monosyllabic singular endings.

Jones’s corrections, while valid, do not in my view disprove the existenceof the Aeolic phase. The innovations that I associate with the start of theIonic phase all appear between −. and −.. The introduction of thedeclension ;��, which is unquestionably Ionic, falls in the middle of thisrange at −.. If the arrival of monosyllabic genitives did predate the Ionicphase, a question where the data point in different directions, they couldat first have been in the contracted forms - and -�� known in Lesbian

Jones’s calculations of this (below, pp. –) are correct in both cases. For the original list see Janko (: –).



richard janko

Aeolic. It is unlikely, given their phonetic similarity to each other, that theforms -� and -�� entered the Ionic tradition gradually, by borrowingfrom an adjacent tradition; the discrepancy in the dates of their appearanceis likely to be a result of the vagaries of statistics based on small populations,especially in the case of the genitive singular.

The opponents of the Aeolic phase have still not explained why thetradition did not retain uncontracted -�� and -:�� if it formerly usedthem (not to mention why it first used Aeolic��, rather than Mycenaeanpo-si, and then replaced��with��). Jones (this vol.) is right that theBosnian singers put together their Kunstsprache from different dialects ofSerbo-Croatian where these supply metrically convenient alternatives; butthis analogy will not explain why the Ionian bards used -� and -��, sincetheir own archaisms -�� and -:�� would have been metrically equivalent.The words cited by Jones (this vol.) with different vocalisms like ��'��and #�:�� are all rarer morphophonemically than the a-stem genitives.Accordingly, they are more likely to have been created by bards by analogywith the vocalism of their current speech than to have been retained froman older phase of the tradition.

Finally, Ionic and Attic had two successive phases of quantitativemetathesis: the first occurred where there had been an intervocalic ∗-s-, thesecond where there had been an intervocalic ∗-w-. Hence Ionic experiencedquantitative metathesis in -��, deriving from ∗-asom, before it lost inter-vocalic digamma. However, the sequence of changes in the traditional dic-tion, even as corrected by Jones, still places the introduction of both of themonosyllabic a-stem genitives (nos. and in the list above) after, indeedwell after, the loss of initial prevocalic digamma (no. ). Initial prevocalicdigamma and medial intervocalic (and therefore, regarding the second syl-lable, prevocalic) digamma constitute very similar phonetic environments;in those dialects which we can observe in inscriptions, the loss of digammain both environments takes place in the same manner and at the sametime. Hence this sequence of changes supports the existence of the Aeolicphase.

Thus there can be no doubt that the epos acquired its East Ionic veneernot long before the time of the Iliad – whenever that was – and that ithad already had these features for some considerable time when it wastaken back across the Aegean to appear in the poems of Hesiod. Recent(non-statistical) linguistic study by the late C. J. Ruijgh suggests that in

So already Janko (: n. ). See my statistical article ‘The use of ��, �� and �� in Homer’ (Janko ), to which only

D. Haug (this vol.) refers.



��

fact the linguistic gap between Homer and Archilochus is actually greaterthan I and others had argued, and that the Greek alphabet was borrowedfrom the Phoenicians in about bc, far earlier than has been thought;Ruijgh (, ) dates the Homeric texts to the ninth century bc. Hisarguments, which adversaries of the Aeolic phase have not yet addressed,merit serious consideration.

I wish to conclude by considering the date of one particular poem – theCatalogue of Women. Hirschberger’s new commentary (: –) offersan exhaustive discussion of the scholarly views about its date. As we haveseen, its language is almost identical with that of the Theogony, of which itis formally the continuation. Its terminus ante quem is the date when theShield of Heracles was composed, since the Shield is a continuation by adifferent poet of the ehoie of Alcmena in book of the Catalogue; anyonewho doubts this must deny that the Catalogue is a unitary composition,a denial to which Hirschberger raises cogent structural objections (:). I argued on historical grounds (Janko b) that the Shield of Heracleswas composed by a pro-Theban poet in c. – bc, just before the FirstSacred War; the Shield was itself cited by Stesichorus, who said it was byHesiod. Nearly everyone’s terminus post quem for the composition of theCatalogue is the foundation of Cyrene in Libya by colonists from Therain about bc, since the poem referred to the nymph Cyrene. Accordingto a scholiast on Pythian (Pyth. .a Drachmann), Pindar took the storyfrom an ehoie of Hesiod. By this argument, anyone who wishes to date theCatalogue to before c. must deny the unity of the poem, and assumethat the passage about the nymph Cyrene was added later.

However, matters are not so simple. The fragments of the Cataloguethat are quoted tell us that the nymph Cyrene lived by the river Peneiosin Thessaly (fr. M–W) and gave birth to Aristaeus, presumably byApollo (fr. M–W). Pindar locates her wrestling match with a lion onMount Pelion, although Callimachus (Hymn .–) and the Hellenistichistorian Acesander (FGrH F ) put it in Libya. Hesiod mentioneda lion somewhere in his poetry (fr. M–W). A Thessalian locationfor this combat is perfectly possible, since there were lions in northernGreece until much later times. On the other hand, there is simply noevidence that the Catalogue made Apollo take his bride to Libya, eventhough most scholars assert this. Pindar may have invented this episode.

In the hypothesis to the Ps.-Hesiodic Scutum. Cf. Hdt. .–, Xen. Cyr. . and Paus. .., with Alden (: –). E.g. West (: ). Contrast Janko (: , ), Kohnken (: , ), Dougherty (:

), Drager (: –).



richard janko

It is also disputed whether Chiron’s dialogue with Apollo on Mount Peliongoes back to Hesiod, or is another invention of Pindar’s; this conver-sation does not match the style of the Catalogue. Cyrene’s name, bythe way, originated in Greece, since its form follows familiar principlesof pre-Hellenic onomastics: for the very common suffix -ana cf. Athana(Athene), Mycene, Messene, Peirene and so on, and for the root compare(Anti)cyra. It is simply bizarre that the city of Cyrene be linked with Thes-saly, unless that was where the nymph was originally imagined to havelived before the city was founded. D’Alessio has now proposed (: –) that, although the Catalogue said that Cyrene was beloved by Apolloand the father of Actaeon (fr. A), the ehoie of Cyrene on which Pindardrew was in the Megalai Ehoiai, which was a different and probably laterpoem. His suggestion rests on the fact that the scholia to Pindar consis-tently cite the Megalai Ehoiai rather than the Catalogue, and may well becorrect.

Another Hesiodic passage brings Chiron into contact with Aristaeus’son Actaeon, again on Mount Pelion. They meet in Hirschberger’s F , afragment which Merkelbach and West mistakenly chose to omit. Actaeonwas Aristaeus’ son by Autonoe in Thebes. According to this genealogy, ifApollo took Cyrene to Libya for her honeymoon, he brought her backto Greece (unless Aristaeus flew back once he had become a god). I amjoking, of course, but the need for such fanciful hypotheses shows howimprobable it is that Cyrene went to Libya in the Catalogue. Hence thefoundation of Cyrene in Libya cannot be used as a terminus post quem forthe Catalogue. Hirschberger can find no other solid evidence for dating thepoem so late. She places the poem between the foundation of Cyrene andthe composition of the Shield, i.e. between and c. , and certainlynot later in the sixth century (: ). It would be more prudent to relyon the language of the poem, and on its degree of geographical knowledge(which is comparable to that of the Theogony), than on the unproven andunlikely hypothesis that the Catalogue made Apollo whisk Cyrene fromThessaly to Libya and back.

Despite the fact that the diction of the Catalogue is in some ways frac-tionally more archaic than that of the Theogony, I see no reason why itshould not be by the same poet as the Theogony, of which it presents itselfas the continuation. That poet calls himself Hesiod. He moves from listingthe gods’ unions with goddesses, starting with Zeus and Metis (Theog.

Kohnken (: –); cf. Drager (: ). West (: ); contrast Janko (). It is included by Most (). The piece could however be

from the Megalai Ehoiai.



��

–), to their unions with nymphs or mortals (–); then, with anappeal to the Muses, he lists the unions of goddesses with mortals (–). This is followed by a summation, and then a renewed appeal to theMuses leads into the Catalogue. Even Merkelbach and West think that theend of the Theogony was composed by the same poet as was the Catalogue.My results showed that the end of the Theogony, the authenticity of whichWest had questioned, belongs statistically with the Catalogue of Women,and so I believe that the whole composition is likely to be by Hesiod. ThusI hope that this linguistic method can contribute, among other insights, toa re-evaluation of the Catalogue; for we apparently have more poetry byHesiod than we thought. Recent studies show that such a re-evaluationis already under way (Drager ; Arrighetti ; Hirschberger ;Hunter ).

West (: –). It is interesting that Haug (this vol.) has found a discrepancy in the use oftmesis in Theog. –; again I wonder whether Hesiod dictated the Catalogue first.



chapter 2


Brandtly Jones

Numerous attempts to establish the relative chronology of the songs ofthe Greek epos, or of particular passages therein, have depended on thelinguistic evidence in our received texts. These studies, while producingsome suggestive results, have not solved the problem of chronology for earlyGreek poetry. The language of epic developed in parallel with the colloquialspeech of the singers in many ways, but the Kunstsprache has rules and ten-dencies which work counter to the diachronic development of the singers’dialect. Also, assumptions about the history of the development of the epicsheavily influence the definition of linguistic material as ‘archaic’. As we shallsee below, one’s model for the development of the songs of the Greek epos,whether through a series of dialectal ‘phases’ or through the diffusion ofshared poetic material among various local traditions, has a profound effecton attempts to elicit chronology from the language of the epics.

1 the mixed dialect of epic

The dialect of the epics mixes archaic forms with innovations; it likewisemixes Ionic forms with forms showing Aeolic phonology and morphology.Linguistic archaisms by definition precede innovations, and concentrationsof innovative forms are generally taken to indicate a later work or passage.

Still, the presence of any linguistic innovation indicates that a passage couldonly have been composed after that innovation had occurred. For example,

It is important to note, however, that specific examples of apparent archaisms in the epics do notindicate that the passages containing them entered the tradition at an early date. Epic recycledformulaic language, and it created new diction on intra-epic analogy. Thus, we often cannot insiston a correlation between (apparent) archaic language and the age of a passage or poem. This, ofcourse, is the fundamental problem of any dating based on archaisms.

The epic language was capable of producing forms and paradigms which never existed outsideof epic as well. Such well-documented phenomena as metrical lengthening provided for a certainamount of flexibility and artificiality, but the tendency was toward using actual vernacular formson the whole. Cf. Hackstein (: esp. –), whose treatment of epic employs the language of




a (metrically guaranteed) ‘neglect’ of digamma could not reasonably haveentered epic language until initial [w-] had lost consonantal force in somerelevant dialect. Clearly, however, the effects of digamma persisted in epicsinging well after the sound itself had disappeared, and thus, any apparent‘observance’ of digamma, say, need not have entered the text before theloss of the sound in colloquial speech.

In addition to the diachronic depth of the hexameter tradition suggestedby the deployment of certain archaisms in the epics, the presence ofAeolisms in the dialect mixture of so-called ‘epic Ionic’ raises interestingquestions about the origins and development of the Greek epic tradition.While some features once considered Aeolic have proved to be merelyarchaisms, epic language features a number of elements generally agreedto be specifically Aeolic on the basis of inscriptional and literary evidence.

One Aeolic feature in particular, the -� � dative plural, is clearlyproductive beyond its original distribution, as witnessed in the frequentdative plural 7�� . Indeed, this epic creation is the dominant form inHomer. Certain other features, which may in fact be archaisms, or at leastreflect archaisms, include the genitive singular of the masculine a-stems,e.g. ��B��, etc., and the genitive plural of the a-stems of both genders,e.g. 3��. These are excluded from some lists since the endings in longalpha are not specific to the Aeolic dialects. I will suggest below, however,that their presence in the epics is owed to an Aeolic epic tradition, thoughnot a diachronically anterior phase of development as has been suggested.

2 competing explanations for the mixed dialect

Aeolic forms embedded within epic language pose a distinct problem forthe definition of ‘archaism’ for the epic language. Aeolisms are effectively

Homer to make inferences regarding the development of Ionic. The interaction of the vernacularand the Kunstsprache will be a primary concern of this discussion as it relates to arguments for relativechronology.

The digamma in Hesiod presents problems if we take seriously the biographical indications thathe was a Boeotian; his poetry neglects initial digamma more frequently than Homer, though theBoeotian dialect retained the sound well into the classical period and beyond.

That is to say, digamma is both a productive and moribund feature of epic. On the one hand, itis an important part of the Kunstsprache, licensing hiatus at certain metrical positions even beyondlegitimate historically w-initial words, but it has been lost from the Ionian vernacular by the timeof the fixation of the epics, so that the poets increasingly create new phraseology which neglects theeffects of digamma.

For reliable assessments of Aeolic features in the epic language, see Horrocks (: ), Haug(: ), Wathelet (: ).

7�� (× in Homer) versus C�� (�) (× ) and C�� (�) (× ). Cf. Wathelet (: ), Haug(: ).



brandtly jones

Aeolic

“Achaean”

Stacked/ “Phase” model “Diffusion” model

Aeolic “Achaean”Ionic

IonicNon-extant Aeolic

epicNon-extant Arcado-

Cyprian epic

Homer, Hesiod,etc.

Homer, Hesiod,etc.

Fig. . The stacked/‘phase’ model versus the ‘diffusion’ model.

archaisms within epic diction if the epic language passed through an Aeolicphase. Specifically Ionic phraseology, on the other hand, is tantamount to akind of innovation if the tradition as a whole only came to Ionian bards ata late date. A given morphological category might show metrically distinctvariants in Ionic and Aeolic which were each current in their respectivedialects, and thus not diachronically ordered in absolute terms. Under aschema of dialectal phases, however, an Aeolism in the Ionic hexametertradition would count as an archaism, since it would have been retainedfrom a time preceding the transfer of the tradition to Ionian singers.

Scholars have also argued for a continuous Ionic tradition spanning fromthe Bronze Age to the archaic period directly, without a discontinuity ofan Aeolic phase. The epics were ‘Ionian’ as far back as it makes sense touse such a term, and they admitted phraseology through language contactwith neighbouring Aeolians over centuries, perhaps through an Aeolic epictradition or simply through colloquial speech. Under this ‘diffusion’ model,Aeolisms could have entered the Ionic tradition at virtually any time duringthe development of the tradition. Thus, Aeolisms per se are not indicative ofchronology. Figure . gives a simplified graphic depiction of the competingmodels. The diachronic dimension is oriented from earlier at the top tolater at the bottom.




Each model has its vocal and its tacit supporters; the epics containsome specifically Aeolic forms which cannot, pace Strunk, be pushedback into Mycenaean times or otherwise explained away. Naturally, giventhe contiguity of Aeolis and Ionia, and especially with the northwardexpansion of the Ionians, linguistic contact is assured. Aeolisms in the epic,though, do not necessarily point to a previous phase. Horrocks offers thefollowing assessment:

There is no reason to suppose that all the Aeolic forms in the epic dialect camefrom the Aeolian tradition during the period in which it was first Ionicised, evenif one accepts that there was an Aeolic phase. The real issue is whether or notthere is evidence of a break in the Ionian tradition, a point at which, going back inhistory and reconstructing where necessary, characteristic Ionic phenomena peterout leaving a gap between the earliest recoverable Ionic forms and their knownor reconstructed Bronze age antecedents. If there is no evidence for such a gap,there is no room for an Aeolic phase, and the diffusion theory must be adopted toexplain the full set of Homeric Aeolisms. (: )

Indeed, a direct line of transmission from the Bronze Age to generations ofIonian bards, and ultimately to a ‘Homer’, is the most economical modelof development.

A pattern of diffusion obtains in the South Slavic epics, which remainthe closest comparandum for the Greek tradition. About the mixed dialectof the epic register, John Miles Foley notes:

South Slavic singers use both ijekavski (chiefly Bosnian and Croatian) and ekavski(chiefly Serbian) forms, not seldom in the very same line . . . [A]lthough nativespeakers normally use only the forms appropriate to their particular geographicalcontext in most registers, South Slavic singers customarily and systematically have

Some of the most prominent and explicit advocates for an Aeolic phase in the history of epicdiction include R. Janko, M. L. West, C. Ruijgh, P. Wathelet, K. Meister, M. Parry, L. R. Palmer,A. Hoekstra and D. Haug. Many others implicitly adhere to the model, including Chantraine, whofrequently explains forms from the standpoint of the ‘earlier’ phase of the epic. The most even-handedtreatment, in my view, is that of Horrocks (, ) who ultimately sides with the diffusionists,though not without reservations; Wyatt’s view () is more extreme, denying the existence of anyAeolic epic poetry and attributing the Aeolisms to colloquial language contact. Others who haveexplicitly argued against the phase model include D. G. Miller, J. Hooker, K. Strunk, T. Webster,J. Mendez Dosuna and M. Peters.

Strunk (). For recent critique of Strunk’s thesis, cf. Haug (: –). Defending the Aeolic phase, and rebuffing critics, such as Wyatt and Webster, who attempt to

downplay or refute Aeolisms in Homer, Janko notes that the ‘hostility to Aeolisms arises from apreference for the neater theory that the epos was brought straight to Ionia from Attica and thePeloponnese, thus descending directly from major southern centres of Mycenaean civilisation: theAeolic forms are attributed to passing exchanges with a parallel epic tradition’ (: ).



brandtly jones

recourse to forms and syntactic features from both dialects when they code-switchto the traditional performance idiom.

Though we can never superimpose the habits of one oral tradition ontoanother, the parallel with the Greek tradition is very suggestive: we see con-current local traditions which share traditional phraseology from differentregions. The exchange is ongoing, rather than punctual.

3 a gap in a continuous ionic tradition?

Aeolisms alone cannot substantiate a model of discrete dialectal phasesin the development of the epic corpus, as their mere presence could bethe result of ongoing contact, i.e. diffusion. Without some demonstrablebreak in the development of Ionic forms attested in the epics, the proposedphase theory is a solution in search of a problem. Proponents of dialectalphases see just such a gap in the masculine a-stem genitive singulars andthe a-stem genitive plurals.

The phonological history of the a-stem genitives is straightforward.Mycenaean shows evidence for a genitive in -a(h)o, and this seems tobe the common Greek ending. Most dialects simply preserved the

Foley (: –). While epic language is on the whole coincident with colloquial speech, theguslar, like the aoidos, depends on both words and grammatical forms that are no longer a part ofthe singers’ conversational language, but remain an integral feature of their traditional register. Forthe guslars, many of these archaisms are Turkicisms which have obsolesced. Morphologically, theaorist is preserved for both metrical and phonological/phraseological reasons like in-line rhyme.Regarding a mixture of features from distinct dialects, Foley notes that deployment of the variantsdijete/dete ‘child’ depends strictly on the metrical and phraseological environment, e.g. dijete Halile‘child Halil’, syl. colon versus A sede mu dete besjediti ‘But the child began to address him’; ‘Thesinger speaks the poetic language fluently by speaking it multidialectally, and according to the rulesof the register for fashioning verbal signs’ (). Cf. the line (.) Ovcijem te zapojila mlekom‘She began to nurse you with sheep’s milk’, showing both dialectal variants deployed in a single linein a concatenation unlikely to exist in conversational speech or any other register of South Slavic.This deployment is a function of competence. ‘While ekavski or ijekavski speakers naturally favortheir “home” dialect to a large degree, traditional epic phraseology always and everywhere entails autilitarian mixture of forms, sorted not according to the singer’s individual speech habits but rathermetri causa’ (). Witte (b) noted that the Aeolic forms in Homer occurred almost exclusivelywhere they provide a metrical alternative to Ionic, an idea developed further by Parry and others.

A colleague of mine who has done field research in the Balkans has told me, only half jokingly giventhat region’s turbulent recent history, that the singers seem capable of borrowing each others’ songs‘even while they’re shooting at one another!’.

Cf. Myc. su-qo-ta-o /sugw otao/ ‘(of the) swineherd’. The nominative singular of the ∗-eh2 stems wasoriginally endingless but acquired -s after the analogy of the o-stems in most Greek dialects, e.g.Att. ��, Dor. ��D�. Some dialects preserved, or recreated, forms without -s in the masc.a-stem nom. sg., e.g. Boet. �E3��D, -�4�D, also in North-West Greek. Thus, the original masc.gen sg. ∗-eh2es >

∗-as seems to have been replaced by analogy with the thematic stems: -os: ∗-osyo:: -a : ∗-ayo. ∗-ayo subsequently developed to ∗-Dh�, then -D�, with further contractions, QM orparadigm remodelling according to dialect.




ending, often with subsequent contraction, but in Attic–Ionic, long alphaunderwent fronting/raising to long ∗-æ:- which in turn merged with low/e/, eventually spelled with <�>, in Ionic. The ending -D� thus devel-oped to ∗-��, and this sequence in Ionic underwent so-called ‘quantitativemetathesis’, (QM) whereby descriptively an input of certain long vowelsbefore certain short vowels produced an output with quantities reversed.

There was a related sound change whereby a long vowel shortened beforea subsequent long vowel, with restrictions similar to those for QM. Thusfor the genitives which were subject to quantitative metathesis we have thefollowing development:

gen. sg. m. a-stem -D� >∗-�� > -�� (nearly always with synizesis)

gen. pl. m./f. a-stem -�� >∗-:�� > -�� /-�� (nearly always with

synizesis)The epics do not show the expected mid-stages in eta <�>. This wouldseem to be exactly the break in a continuous Ionic tradition that Horrocksdescribed as the necessary evidence for an Aeolic phase. We shall see,however, that the diffusion model can account for the forms from thesecategories and that other evidence makes the phase model unlikely.

4 janko’s treatment of the a-stem genitives

If indeed there were any Aeolic phase, it had given way to an Ionic phasebefore any of the songs of the epos became fixed. A slow transfer ofthe tradition, perhaps with continued sharing of epic material after ‘firstcontact’, would be tantamount to diffusion and thus of limited value forarguments regarding relative chronology of archaic hexameter poetry. Onlyunder a ‘stacked’ model do Aeolisms count as archaisms. Janko argues forjust such a punctual transfer of the tradition from Aeolian to Ionian bards.Indeed, his definition of ‘archaism’, and thus his dating method as a whole,depends crucially on the priority of an Aeolic phase of the epic tradition.In Homer, Hesiod, and the Hymns (), Janko claims that his statisticalanalysis of the songs of the Greek epos provides new evidence for the Aeolicphase of the epic tradition, and even indication for a relative date for thetransfer of the tradition to Ionian bards.

Janko’s dating of the transfer of the tradition to Ionian singers dependson an extension of his common scale. This scale, based on the aggregate

Attic, of course, underwent ‘Attic Reversion’ at the ∗-æ: stage, whereby ∗æ: merged with <�> inmost environments but ‘reverted’ to /a/ after �, �, and �.

The sound change is restricted to certain combinations of vowels. Cf. Peters (: n. ),Mendez Dosuna (: n. ).



brandtly jones

percentages of a number of linguistic features of the epic language, is meantto reflect the relative date of composition for each poem of the epos. Eachfeature measured by Janko shows some degree of innovation in the Iliad,and more in the Theogony, which are set at an arbitrary distance of threeunits on the common scale. If we assume a constant rate of change, thisdistance of three units corresponds to the length of time separating the Iliadand Theogony, and we can theoretically project back to the point on thecommon scale at which a particular innovation entered the epic language.The proposed relative moments of development in the epic language areas follows:

. -�� ceases to be the sole o-stem genitive singular ending: −... Contraction in ∗-oo first appears: −... o- and a-stem dative plurals begin to admit elision: −... Digamma is first neglected: −... -� ceases to be the sole a-stem masculine genitive singular ending:

−... Quantitative metathesis first appears in -��: −... The Ionic declension ;�� etc. first appears: −... ‘Short’ o- and a-stem dative plurals which cannot be elided: −... o- and a-stem accusative plurals begin to be biased towards appearing

before consonants: not before −. and −. respectively.Using this method, Janko finds a very interesting result for the formsshowing QM, numbers and above. About this finding Janko notes:

The point at which quantitative metathesis appeared in the traditional diction isestablished at −. units on the common scale, according to the a-stem genitives.This should represent the stage at which the Ionians took over the Aeolic tradition.It does not give the date of quantitative metathesis in Ionia, which must by Meister’s

This is of course a flawed assumption. While the whole principle of Janko’s method is based onglottochronological presuppositions, the fallacy of these presuppositions is cast in starkest reliefwith these ‘extended results’ of the statistics.

Quoted from Janko (: –). This list is compressed and elliptical. The development of -��is tied to the development of Common Greek ∗-ohyo, and thus the surface representation -��may merely reflect the metrical effects of the pre-form rather than the specifically Aeolic morph.Likewise, the a-stem genitives may have a similar story. Quantitative metathesis, of course, appearedin ∗-:��, not -�� as suggested in number ; also, the import of number would seem also to bethe introduction of QM, as -� and ∗-�� were metrically indistinguishable. Finally, the assertion ofnumber , the admission of elision in dative plurals, is opaque. It is unclear whether Janko wouldsuggest that these forms differed from the rest of epic language in their deployment of elision, or ifthe assertion is that elision first entered for all relevant words at this point on the common scale.While it strikes me as unthinkable that elision would be a late entry into epic language, it is equallybizarre to suggest that the o- and a-stem dative plurals should have been treated differently for somereason. Janko does not, to my knowledge, clarify this point.




argument have happened somewhat earlier, at least in the area which moulded theepic diction to its final shape. (: –)

The agreement of the masculine a-stem genitive singulars and the a-stemgenitive plurals would seem to buttress this claim; both are ostensibly linkedto the introduction of QM into epic diction, but a detailed investigationreveals that the metrics are actually different. Number gives the point onthe common scale at which the figures for Iliad and Theogony project %archaism in -�. That is to say, at this relative date, all Greek hexametersinging would have used -� exclusively in this category. Number , on theother hand, measures % innovation in the genitive plurals by comparingthe number of forms showing QM versus the number of archaic a-stemgenitive plurals in -��. This excludes forms transmitted as -�� (× Il.,× Theog.) and -�� (× Theog.).

While we would expect % archaism to be tantamount to % inno-vation, the presentation of the facts from the epics merely creates theillusion of agreement between these two features. Because of the treatmentof ambiguous sequences, these figures as tallied are not the same. Themasculine a-stem genitive singulars with apparent QM occur relativelyfrequently before a vowel in Homer, though never in Hesiod. In such acase, we can, and perhaps must, restore the archaic ending with elision. Asequence of -�� + V-, scanned with synizesis, may simply be a graphicrepresentation of an earlier ∗-D’ + V- with elision of -�. The Iliad andOdyssey each show over twenty such examples, but again, Hesiod none.QM is only guaranteed at line end or before a consonant. Thus, if we setaside ambiguous examples and calculate instead the point at which ‘Quan-titative metathesis first appears in -�’, i.e. % innovation for this feature,using the method parallel to that used for the genitive plurals, we get −..If we calculate the point at which -�� ‘ceases to be the sole genitive pluralending of a-stems’, we get −.. These results are clearer in table ..

We may take as an example the ending of the first line of the Iliad: �� . We find examples of -�� + V- in the Iliad and in the Odyssey. Nowhere else in the epos do we findthis type of ambiguous example, save one example in Hymn. Hom. Ven.

Or, just as likely to my mind, elision of ∗-��. To give a concrete example, the sequence at Il. .,�� | may thus represent ∗��@ �� |. Cf. Chantraine (: ).

The mathematics involved is a series of proportions based on the percentages obtained from theIliad and Theogony, set three units apart on the common scale. Those percentages are presented inTables . and . Janko calculates that ‘-� ceases to be the sole a-stem masculine genitive singularending: −.’ by treating the difference between the Iliad and Theogony as three units and thencalculating the proportional distance on the common scale between the value for the Iliad and %.Thus, the difference between . and ., or .%, is equivalent to units on the common scalefor this feature. The difference between % archaism and the Iliad, in this case .%, is .%.To determine the distance on the common scale between the Iliad and % archaism, we use the



brandtly jones

Table . Common scale results withambiguous examples excluded

‘% archaism’ ‘% innovation’

-�/-�� −. −.-��/-�� −. −.

Table . Distribution of masculine a-stem genitivesingulars in the Iliad and Theogony

-D� %-D� -�� /� +V ∼+C Grand Total

Iliad 77.9

Theogony 41.7

Table . Distribution of a-stem genitive plurals in the Iliad andTheogony

-D�� %-D�� -�� -�� (syniz.) -��/-�� Grand Total

Iliad 83.9 /

Theogony 61.5 /

Contrary to Janko’s presentation, when we compare ‘apples to apples’,the forms showing quantitative metathesis do not agree on this point,and do not in fact suggest a specific point of transfer from Aeolic toIonic speakers, even assuming a constant rate of language change. This,of course, does not prove that an Aeolic phase never existed, nor does itdisprove the basic notion that a statistical increase in the frequency of the

proportion .:=.:x, x=.. Since the Iliad is set at zero on the common scale, ‘-� ceasesto be the sole a-stem masculine genitive singular ending’ at −. units. Rather than following thissame procedure for the genitive plurals (i.e. (.−.):=(−.):x, or .:=.:x, x=.),Janko instead calculates that ‘[q]uantitative metathesis first appears in -��: −.’ by setting asideforms in -�� or -�� and comparing the percentage of forms in -�� per the total of these formsand forms in -D��, .% Il., .% Theog.; and so, (.−.):=.:x, or .:=.:x, x=..The same method for the genitive singulars produces % innovation at −.; guaranteed -�� .%Il., .% Theog. It must be noted that while the Iliad shows over examples of the masculinea-stem genitive singular, the Theogony shows only twelve total. This is narrow and insecure footingon which to base the date of the Aeolic phase of the epic tradition.




Table . Distribution of masculine a-stem genitivesingulars in the Iliad and Odyssey

-D� %-D� -�� /� +V ∼+C Grand Total

Iliad 77.9

Odyssey 72.7

innovative forms showing QM could reflect a diachronically later stage ofthe tradition. One of Janko’s primary claims is that the Odyssey shows ahigher frequency of innovations than the Iliad for all the features he tests,including those showing QM.

The distribution in Homer of endings with and without QM for themasculine a-stem genitive singulars and the a-stem genitive plurals is pre-sented in tables . and .. For the a-stem masculine singulars Jankoseparates out those examples of -�� which occur before a consonant andat line end, i.e. secure examples of QM, from those which occur before avowel; these, as we have just seen, are in fact ambiguous as they may repre-sent the archaic ending with elision. The manner of counting which Jankoemploys, however, effectively classes these ambiguous sequences with thesecure innovative forms. That is, for the masculine a-stem genitive singu-lars, Janko counts the number of forms showing the archaic/‘Aeolic’ -D�ending and renders this as a percentage of the total number of masculinea-stem genitive singulars. Consequently, the ambiguous sequences before avowel are treated as ‘non-archaic’ even though there is strong reason to seethem as mere graphic representations of an archaism. In fact, these formsare simply ambiguous and must be entirely eliminated from the calculus.

When we omit these forms from consideration we find that the trend forthis feature reverses (table .). Under this improved calculus the Odysseyshows a higher preference for the archaic morph than does the Iliad.

The numbers are taken directly from Janko’s count (: Table , Table ), which I haveindependently confirmed. Janko’s numbers differ slightly from those of Chantraine (which Edwardsfollows). Chantraine’s numbers (: ) are slightly lower: for -D� Il., Od.; for irresolvable-�� before a consonant for Il., for Od. As neither Janko nor Chantraine give a full account ofthe raw data, I cannot account for the discrepancy.

If we treat the -�� + V- sequences as unambiguously archaic, the Odyssey, not the Iliad, still shows ahigher preference for the archaic ending: Il. .% -�, Od. .% -�. The proportion of archaismto innovation is not, of course, the only metric we could apply. Table . provides the numberof occurrences per , lines of the a-stem genitives. By this reckoning, again, the Iliad shows agreater frequency of (guaranteed) use of the innovative -��, this time per line. It is interesting tonote that the Iliad has occasion to employ a masc. a-stem gen. sg. more often than the other poems



brandtly jones

Table . Distribution of masculine a-stem genitivesingulars in the Iliad and Odyssey (corrected)

-D� %-D� (-�� /� +V) ∼+C Grand Total

Iliad 89.2 ()

Odyssey 94.1 ()

Table . Distribution of a-stem genitive plurals in the Iliadand Odyssey

-D�� %-D�� -�� -�� (syniz.) -�� Grand Total

Iliad 83.9

Odyssey 80.2

The genitive plural a-stems show no such ambiguity, only a (perhapsmerely graphic) distinction between forms in -�� (nearly always withsynizesis) and forms in contracted -�� or -��. For this category, the Iliaddoes in fact show the archaic morph in -�� somewhat more frequently(.%) than the Odyssey does (.%), as shown in table .. Hesiod,naturally, shows far fewer examples (table .). The percentages, in com-parison with the Homeric figures, seem to indicate a marked increase inthe frequency of forms with QM, though the numbers are small. Edwardsdissects the numbers into traditional versus apparently innovative phrase-ology. Of the occurrences of -�� in the Theogony, for example, arefound in Homer in the same metrical position, eight of those ‘in phraseswhich occur in Homer and which almost certainly did not originate inHesiod’ (: ), that is, traditional phrases. The six examples in Works

in question, but the Theogony is extremely different from the other poems in terms of its per linedeployment of the gen. pl. of the a-stems (though the total number of examples is only ). Hesiod’susage in both of these categories is again demonstrably different from that of Homer.

The non-Ionic contracted genitive in -�� occurs once in the Theogony and once in Works and Days.Homer shows a total of three disyllabic examples of -��, elsewhere with synizesis. The formsare �#�� (Il. ., .) and 3#�� (Od. .). The semantic connection between the onlytwo forms showing disyllabic scansion of the genitive plural a-stem ending is striking. Janko citesWyatt’s attempt (: n. ) ‘to get rid of �#�� etc. . . . by alleging that the analogy of s-stemsinfluenced these forms, which is unparalleled and incredible’ (Janko : n. ). Given theextreme isolation of these forms within epic diction, however, and the likely phonological historyof the forms in question, an analogical explanation is not unlikely.




Table . Frequency of a-stem genitives

Total masc.a-stem gen. sg.

Total a-stemgen. pl. -� -�� -�� -�� /��/��

Iliad . . . . . .pt .Odyssey . . . . . .pt .Theogony . . . . . .pt.Works and Days . . . . . .pt .

and Days are not in obviously traditional phrases, according to Edwards,though ‘traditional’ here means essentially ‘repeated (in Homer)’.

The genitive singulars in Hesiod show an even more striking pattern:‘the -D� genitives in the Theogony and the Works and Days belong withoutexception to forms which occur with an -D� genitive in Homer, often inthe same or a similar context’ (Edwards : ). This points clearly totraditional diction in the repertoire of the poet of Theogony and Works andDays which was shared with the poet of the Iliad and Odyssey. Edwardsgoes on to demonstrate that traditional influence ‘is much less apparent inHesiod’s use of genitives in -��, -�’ (: ).

The numbers for this morph in particular are especially small, so smallthat Janko declines to give percentages for works other than Homer andHesiod (including the Catalogue, though not the Aspis). The total numbersdecline precipitously outside of Homer, with just twelve total examples inthe Theogony (.% -�) and nine in the Works and Days (.% -�).Janko’s pronouncement that ‘[c]learly little can be learnt about the hymnsfrom a morpheme as rare as this’ (: ) could be extended to includeHesiod (and perhaps even Homer) as well.

5 evidence for the full development of qm

in epic language

The a-stem genitives are not of course the only environment where we canlook for the effects of QM in Homeric diction. To assess the appearance ofa gap, we must search for Ionic archaisms alongside their Ionic innovativeforms; we would thus have evidence of continuity of the tradition. Such

It would seem very likely, of course, that a great deal more of the work of Homer and Hesiod wouldfit the usual definition of ‘traditional’ if only we had access to the great mass of lost archaic hexameterpoetry. The comparative evidence for a didactic poetic tradition in hexameters, in which traditionthe Works and Days fits, is even more sorely lacking than that for narrative/epic or theogonic verse.



brandtly jones

evidence does exist; for example, the subjunctive of athematic F ��'� showsthe following development and distribution of forms:

st pl. aor. subj. ∗ ��'�� > ��-�-'�� (Il. .) > ��-'�� (Il. . =.)

The epics (the Iliad at least), show what appears to be the archaic Ionicphonology with fronting of ∗D to /e/ extant in the form ��'�� as well asthe later form showing QM, ��'�� scanned with synizesis. Additionally,the forms for ‘ship’ present the full development of the Ionic forms:

∗�DG-�� > �� (Il. .) > �� (Il. .), �� (Od. .)

The declension of ‘ship’ shows a very mixed picture, and one that shouldbe treated as a separate phenomenon from the a-stems. The forms aresystematically disyllabic in the simplex, with apparent archaisms in ��,�� and ��, and innovative forms in ��, �� and ��. Except incompounds, we never find the ‘Attic declension’ form �� in the epics.Thus, with the exception of ��D × at Od. ., the forms show shorteningrather than QM. We thus have some evidence for Ionian phonology in theappropriate sequences, but not for the a-stem genitives.

6 further evidence against an aeolic phase

The proponents of an Aeolic phase assert that the Ionian bards took overthe Aeolic tradition and modernized where they could, retaining thoseforms for which they had no replacement. This theory implicitly makescertain predictions, however, which are clearly wrong. Every sequence of<��> or <��> or <�> which appears in epic is, if we accept an Aeolicphase, predicted to have a non-Ionic origin, at least among polysyllables.Homer shows a good number, e.g. �H�� ‘vigorous, stout’ (<∗aygyawo-),C� ‘I burned’ (< ∗ekahwa <

∗ekawsa), etc. The existence, indeed

The spelling <��> for historical <�> is a non-issue. ‘The verb-stems in e normally have -��- beforeo-vowels but -�- before e-vowels (�'��, �':��; >��, ��'��, ��3��). This looks like abardic convention’ (Janko : ). Cf. Chantraine (: ff.).

Forms like �� and �D�� are systematically disyllabic in the simplex in Homer, but show QMin compounds (cf. Chantraine : ): �4�� (Od. ., ), �� (Il. ., .,.), �� (Il. .) �� (Il. .); �� (Od. .), ��>� �� (Od..). Likewise, the long vowel of I��, 7�, I�8� contrasts with the short vowel of the compoundJ� 1�� (Il. .). An analogous distribution is found in Hesiod, archaic lyric, Herodotus, andin inscriptions from the Ionic speech area. Hdt. I�� ∼ J�3��; �� ∼ -��, ��1��;�� ∼ �� #�. Mendez Dosuna (: –).

This excludes disyllabic forms such as I��, �� for reasons put forth by Mendez Dosuna, cf.n. .

Meister (: ) quotes numerous other forms for which we find spellings of <��>, <��>,<��> and <��> in words once containing PGk. /a/. My thanks to Martin Peters (p.c.) forbringing these forms to my attention.




frequency, of such sequences presents real difficulties for an Aeolic phase,per se. Meister himself, in the face of these problems, asks, ‘weshalb lie>Homer nicht ∗�� ∗'�� K�� stehen, da er doch �� '�� L��sprach?’ The duals present similar problems.

The athematic dual forms #��:�� (Od. .), �� #�:��(Il. ., .), #�:�� (Il. .), and 1��:�� (Il. .) are allformed from verbs in -��. This inflection for contract verbs, the so-called‘Aeolic’ inflection, was foreign to Ionic; thus, if these forms were a partof an Aeolic tradition which was taken over into Ionic, the Ionian bardshad no metrically equivalent forms to replace them. Even if we imaginethat Ionic had developed and retained the dual ending in -�� from PGk.∗-�D�, there would be no metrical equivalent form with a long final stemvowel. The mechanical approach would yield the hybrid ∗ #��.

Mais les formes de duel en -�� presentent pourtant l’� ionien attique a la placede l’D du grec commun. C’est de meme un �, non un D qui est note a la fin duradical des duels de verbes en -� comme �� #�:��, ou dans les infinitifscomme 4�:'�� de 4��. Ces formes athematiques de verbes contractes sontcomposites et ont ete accommodees au vocalisme de l’ionien.

This Ionicization stands in contrast to the process envisioned for a transferfrom Aeolic to Ionic. Against this tendency to Ionicize the phonology takenover from Aeolic sources stands the case of the -� and -�� genitives,which are anomalous whether Ionian singers acquired these formationspunctually (under the ‘stacked’ model) or through ongoing exchange anddiffusion. QM is metrically guaranteed by the time of the textualizations ofthe epics. The sound change is assured. Assuming an Aeolic epic transferredto Ionians after QM, we should expect a model ∗��1��D4��D� to producean Ionicized ∗��1��4��, to judge from the situation of the duals

Meister (: ). The so-called ‘Aeolic’ inflection designates a characteristic athematic inflection in Aeolic for contract

verbs in other dialects; e.g. Ionic and elsewhere ��, 1��, ��'�� etc.∼ Lesb./Thess. ��'�,1��'�, ��'D'�, etc. Cf. Blumel (: –), Hock (: ff., ff.). Despite being a part ofMycenaean and Aeolic, the ‘Aeolic inflection’ plays a very tiny role in the language of the epics. TheIonic contract verb treatment pervades the language of the epos; Wyatt notes that such lines as

M '�8�� 1��# @ ��#� '�� 3�� (Od. .)N� O��3�� # � 3��, P�� Q�'�� (Od. .)

have no metrical equivalents in Aeolic. Such examples can be multiplied. Proponents of an Aeolicphase must explain why the nominal system of epic shows a good deal of Aeolic influence, but theverbal system very little.

Chantraine (: ). The duals for contract verbs in -�� only rarely show -:�� (��:��from ��), showing instead -�� (��, �'��, etc.). The poets clearly employeda certain amount of modification and analogy to these forms which were not a part of colloquialspeech.



brandtly jones

above. Instead, we find the hybrid ��1��4��. Nothing preventedthe poets from Ionicizing the ending, though, of course, this was not thepractice. Mendez Dosuna argues, rightly, that the process for these formsin fact worked in the opposite direction:

The forms in -D�, -��, which appear systematically in place of -��, -:��, morein agreement with Ionic phonetics (cf. also �D��, �� , etc.), should not beinterpreted as the residue of an Aeolic phase or, on the other hand, Proto-Ionic.Such Aeolism (or archaism) would be inorganic since -��, -:�� would not havealtered the metrical value of -�, -��. Pace Ruijgh . . . it seems more likely thatthe process operated in the opposite direction . . . : the forms in -�, -��, takenfrom a distinct poetic tradition, supplanted those in -��, -:�� because, at the timewhen the homeric text was being established, these endings were no longer usedin colloquial Ionic. It is as if the singers preferred the artificial forms in -D� and-��, which would seem authentic because they had a correlate in other (literary)dialects, to authentic forms in -��, -:��, which would seem artificial because theywere not in current use. (: – n. )

It is not clear when the Ionic forms in eta would have given way to formsin alpha. We have seen that the a-stem genitives do not provide us withthe full range of the stages of QM in Ionic, but the verb ��'�� ∼

�� '�� does. Why should ��'�� show archaic Ionic phonology butthe genitives not? The singers showed a preference for living forms wherepossible, even those living in neighbouring dialect regions which practisedhexameter singing.

7 an alternative explanation for the distribution of qm

I suggest then that we have a special application of Parry’s principlethat:

the language of oral poetry changes as a whole neither faster nor slower than thespoken language, but in its parts it changes readily where no loss of formulas iscalled for, belatedly when there must be such a loss, so that the traditional dictionhas in it words and forms of everyday use side by side with others that belong toearlier stages of the language. (in A. M. Parry b: )

The form occurs x in the Iliad, always in the phrase A�.� ��1��4�� |. This is an obviousdeclension of the common nominative formula ��1��4�� ;�+� | found in the same lineposition. The phonology of this nominative formula, i.e. in <>, could even have influenced orreinforced the practice of singing the genitive ending in -�.

The dialect mixture does strongly imply sharing among singing traditions, not simply borrowingfrom neighbouring colloquial forms. Phraseology would have been borrowed and adapted becauseof its metrical utility, and that utility was only apparent in versified speech rather than simpleconversation.




When we allow that the language of oral poetry was a reflex of many localsingers using and adapting material shared among them, and that there wereclearly traditions in Aeolic and Ionic in contact, the preservation of obsolete∗ �:�'��, for which there was no living replacement, and the loss of ∗-��,∗-:��, for which there was a replacement, becomes comprehensible. Theapparent gap in the development of the a-stem genitives does not providediagnostic evidence for an Aeolic phase preceding the Ionic tradition.

If Ionians received the tradition from Aeolians at a time when QMwas an accomplished fact in Ionic, and from that point Ionic developedthe tradition briefly before the texts of Homer were established, all ofthe examples of innovative QM entered the tradition recently, i.e. after thetransfer, and all of the -� and -�� forms are de facto older in the tradition.Ionian bards could create some archaic-looking forms by analogy, but giventhe brief window between the transfer and textualization, the effects of anysuch propagation would likely be minimal. Under such a scenario, examplesof -� versus -�� are archaic versus innovative both in absolute linguisticterms (i.e. the archaic variant is a pre-form of the innovative variant) andin terms of the tradition (i.e. Aeolic precedes Ionic).

If we accept the diffusion hypothesis, however, the picture becomescloudier. Not only do the archaic forms have the potential to be recycled,as always, but also, rather than entering the Ionian tradition at its birth,Aeolic forms, including those in -� and -��, could enter the Ionictradition over generations. Even after QM occurred in the Ionic vernacular,and innovative forms began to enter the tradition, the Ionian traditioncould still borrow forms in -� and -�� from Aeolic sources, and/orcreate them on the model of -� and -�� forms already there. Theselate imports would have an archaic shape but could have entered thetradition after forms showing QM entered. Forms showing the effects ofQM would still be linguistic innovations, but without the Aeolic phase wecan no longer identify specific examples of the archaic variant as ‘older’in terms of the tradition itself. Given this situation, the number of formsin -� or -�� can tell us very little since we can never be sure whethera particular example entered the tradition before QM in Ionic or later bydiffusion.

Despite all of this, we must face the fact that the genitives with QMare rare in the epics. The categories as a whole are not very large, especiallythe singulars, but even more important is the observation by Hoekstra that‘the evidence for the existence of formulae originally built upon quantitativemetathesis is extremely slight’ (: ). Attempting to date the soundchange provisionally, Hoekstra proposes the following:



brandtly jones

[W]e may reasonably suppose that in the second half of the seventh century, towhich the Nikandre-inscription probably belongs, the metathesis itself was still arecent development in Central Ionic. So if we assume that in the regions of Chiosand Smyrna the change in quantity was completed no more than a century earlier,that is to say about half of that time (i.e. one or at most two generations of singers)before the composition of the Iliad, we shall not be very far from the truth. Thestate of affairs we find in Homer is in accordance with this approximate dating ofthe phenomenon.

Hoekstra’s attitude is that the epics were in a process of decay from for-mulaic originals to less formulaic modifications. The bardic technique wasone of near constant reworking (which is in fact the foundation of Janko’sanalysis), but to call this process ‘decomposition of the formulaic tradition’(: ) casts the process as one of degeneration, rather than constantregeneration. He is right to note that disyllabic examples of -�� and -��are vanishingly rare, not showing any evidence for formulaic involvement,and perhaps even the ‘personal creations of Homer himself or of his imme-diate predecessors’. Forms of -��+ V- which could be restored to ∗-D’(�)+ V- outnumber secure cases of monosyllabic -�� in both the Iliad andOdyssey, but are unknown in Hesiod and nearly all the hymns. This replace-able prevocalic -�� also shows some involvement in formulaic language,e.g. �� , *�� @&�# ��, �� 4'�'6��. Guaranteed examples of QM are somewhat more rare in the epics,and this scarcity has led to a supposition that the sound change itselfwas recent. The extra-Homeric evidence, however, does not support thissupposition.

Hoekstra (: ). The famous Nikandre inscription of Naxos (CEG .), found at Delos anddated to the seventh or sixth century, shows distinct signs for etymological /e/, spelled <�> (e.g.��3 R��), and the sound resulting from the raising of /a/, spelled <�> (e.g. =��); thedifferent spellings reflect a time when the sounds had yet to merge in Central Ionic. The inscriptionis composed in hexameters and shows two guaranteed examples of QM each scanned with synizesisand spelled with the sign for <�>: A��|�� and ��:� (= ��:�� ). These forms are not,despite the spelling, the missing ∗-�� and ∗-�� of the continuous Ionic tradition, but ratherequivalents to the epic forms in -�� , -�� . The masculine P�� is frequent at line end in theepics, which also show �� but never �� . Mendez Dosuna (: ) notes that metricalinscriptions systematically show synizesis, rather than disyllabic QM.

Hoekstra (: ). The forms are extremely limited: �#��, 3#��, L��, ��, and perhaps'�4�� . Mendez Dosuna (: s.v.), developing an idea of Schwyzer, seeks to demonstrate thatso-called QM in Greek was more likely a process of desyllabification of the first element, creatinga diphthong and triggering compensatory lengthening of the second element. The notion, then,that these disyllabic forms would have been the primary output of QM followed by contraction orsynizesis seems wrong and not in accord with the practice of the epic or, indeed, the tragedians.The monosyllabic scansion seems the primary outcome of the sound change. That said, however,solutions for the distraction of the monosyllable to yield the ultimate disyllabic output are difficultto justify.




Table . Distribution of a-stem genitive plurals in Hesiod

-�� %-�� -�� -�� (syniz.) -�� -�� Grand Total

Theogony 61.5

Works and Days 54.5

Catalogue 66.7

Table . Order of changes in ∗hedrahon and ∗hippewon

∗hedrahon ∗hippewon

i. loss of h (<∗s, yod) hedraon ∗hippewonii. fronting of a: to æ: ∗hedræ:on ∗hippewon

iii. QM/shortening- hedreon ∗hippewoniv. Contraction hedron ∗hippewonv. loss of intervocallic -w- hedron hippeon

vi. QM/shortening- hedron hippeon

The evidence from Attic and Ionic points strongly to an early date forQM over a hiatus left after the loss of s or yod. For example, in Attic wefind the genitive plurals of the first declension regularly contracted in -��,but Attic shows the products of QM over a -w- hiatus still uncontracted.That is, nouns of the <��+� and > ��+� type never show contractionin the genitive plural, e.g.:

over an s-hiatus:

∗hedrahom > J�� (x h. Apollo) >∗hedreon > J�� (x Od.) > Att. J��

over a w-hiatus:

∗hippewon > <��:�� (x Il.) > Att.-Ion. <��

We can compare the situation in Attic of the s-stems, which show con-traction in the genitive, e.g. Q��#�, versus the -�#- stem gen. sg.> ��. Unlike the a-stems, the -�#- stems do not show the reductionof, for example, <�� to <�� by contraction (or monophthongiza-tion). The different treatments of the two types of hiatus suggest at leasttwo separate rounds of QM and a date for the first round of QM beforethe loss of intervocalic -w-, which was itself quite early in Attic and Ionic.The rule ordering is presented in table ..

Thus, Ionic singers should have had access to the innovative formswith QM from an early date, but these forms, again, do not appear at all



brandtly jones

frequently in epic language. I suggest that this situation can be plausiblyaccounted for within a continuous Ionic tradition and in fact demonstratesa broader pattern of poetic practice which renders the counting of archaismsversus innovations of little value for chronology.

Singers incorporate new linguistic features where they are useful forversification; this technique, as Hainsworth, Hoekstra and others havemade plain, often involves modification of older inherited diction. Whenthe modification makes use of innovative forms, the process is oftenobservable. For the genitives showing QM, we can compare such pairsas the clearly innovative genitive ;��.� 7��>��'�� (Il. .) with ;�9�ST�>��'�� (Il. ., etc. x.); the pair seems to constitute a case ofa declined formula, and the innovative nature is confirmed both by theending with QM and by the analogical Ionic genitive ;�� for A��. Atleast one example with QM has gained clear formulaic status: the line end-ing U��# �0� �4#��':�� | occurs x in the Iliad in non-repeatedlines; as such, the phrase qualifies as formulaic language by virtually anydefinition. The phrase is related to the line ending U�� 4#��':��(Il. .), and modified by the insertion of another element. Interestingly,this formula breaks the formulaic economy of Homer, who has also theequivalent nominative designation ��V� �� 3�� .

8 archaism of metrical shape versus phonology

I propose that the paucity of the innovative variants in -��(�) correlateswith the early date for QM in the following way. Ionian epic poetry hadaccess to Ionic forms in ∗-�� ∗-�� (for forms inherited in -�, -��),and such forms could well have proliferated further. At the point thatQM affected Ionic, these older sequences were not simply obsolete butphonotactically in opposition to the singer’s colloquial speech. Certainfigures were certainly part of the pan-Hellenic epic tradition, and as suchwe can assume that certain forms had analogues in both Ionic and Aeolic

This is, of course, Hoekstra’s focus in his Homeric Modifications of Formulaic Prototypes (). Thecorollary phenomenon is more difficult to discern. That is, actual archaic material provides thetemplate for other material which may not in fact be old at all; but because the new material ismodelled on old, it may be indistinguishable from genuinely old material.

Cf. Chantraine (: ): ‘Cet exemple enseigne que la langue epique a admis des formules de typenettement ionien.’

Cf. Hoekstra (: –), Mendez Dosuna (: ). Shipp and Hoekstra both agree that this phrase is ‘clearly older’, though no innovative variants ever

arise for any of these words in epic, or indeed in later Greek. The phrase may well be old, andprobably is, but there is nothing diagnostically archaic about it. Cf. Shipp (: ).




singing, e.g.��B��,W#��B��,�B��,��'��, etc. At an early pointin the Ionic tradition, foreign phonology was used to repair a metrical lossfor a specific easily analyzable category. The endings -� and -�� becamerecognizable tokens of epic diction itself and were thus easily spread.

The hymn to Demeter similarly illustrates the possibility of over-representation of the a-stem genitives in alpha in a way which does notfollow Janko’s scheme at all. H. Dem. shows examples of -�� and just

in -��/��, or .% – a higher frequency than the Odyssey, but few wouldconsider pushing the date of the hymn up to that of Homer on that basis.The poet of the hymn clearly understands the forms in alpha to be a partof the traditional singing practice. The most reasonable time for this tohave established itself is the period when QM was rendering the genitivesin eta obsolete.

9 conclusions

The evidence for an Aeolic branch of epic poetry is strong. The evidencethat this Aeolic branch preceded Ionian epic is not. Even if one were per-suaded that the Aeolic tradition predominated for some time while theIonian tradition remained less productive, the evidence for a punctualtransfer of this tradition is non-existent. Janko’s method of determiningrelative chronology among the songs of the epos depends crucially uponAeolic forms being de facto archaisms, but we have seen that this is notsound practice. Any transfer would have taken place over some time. Also,this transfer of poetic material would have been ongoing after contactbetween Aeolic and Ionic traditions. A period of diffusion ensued in AsiaMinor after contact, whereby a number of living Aeolic forms replacedobsolete forms in the Ionian tradition. While we have not touched onall of the forms, or often absence of forms, that Aeolic phase proponentsinvoke to bolster their case, we have seen the process by which certainmorphological categories come to have an Aeolic appearance while othercategories retain Ionic phonology. Szemerenyi’s formulation of the princi-ple proclaimed that ‘whenever the normal Ionic development would haveled to forms non-existent in Ionic, the speech-form of neighbouring Aeolic

The many epic features in Lesbian lyric poetry attests to the reciprocal nature of the exchange. Most prominent are the Aeolic personal pronouns and the lack of assibilated ∗�� for ��.

Regarding the latter, Wyatt points out that we have no direct evidence that the form ∗�� evergained any foothold in Ionic, though there is convincing evidence that the guaranteed monosyllabic�� is a recent innovation (Janko ). We find the disyllabic form C�� frequently in the epics;��may have enjoyed the same fate of frequent restoration within epic language (as Ionic generalized��).



brandtly jones

was adopted. This is to assume a certain amount of influence from Aeolic,but not a full-scale Aeolic stage in the development of epic poetry’ (:). This formulation accounts well for the facts surrounding the a-stemgenitives, as we have seen.

It is no surprise that Meister chose to organize his Die homerische Kun-stsprache () by morphological category. Factors beyond the regularchanges in the vernacular affect epic diction, and these factors tend to havea morphological and metrical/formulaic component which can supersedethe colloquial habits of the singer. This fact complicates any chronologicalclaims based on the proportions of archaic and innovative features. Thecase for a punctual transfer of an Aeolic phase of epic to Ionian bards, anecessary condition for Janko’s definition of archaism for the a-stem gen-itives, is not convincing. Diffusion of epic material, even if only in theperiod after ‘first contact’ between Aeolian and Ionian bards, provides thebest framework for understanding the development of epic language, evenif it does undermine some proposed chronological benchmarks.



chapter 3

The other viewFocus on linguistic innovations in the Homeric epics

Rudolf Wachter

1

This study is about the question how the epic language of an oral poetwho was active for some time and has left us more than one work shouldbe expected to have shifted. There are hardly any new detailed resultscontained in it; its main goal is rather, as the title suggests, a change ofperspective.

Two suppositions have to be made in order for this question to makesense at all: the identity of the poet of the Iliad and that of the Odyssey(I will call him Homer) and an Aeolic phase preceding the Ionic one as theexplanation of the linguistic mixture we are confronted with in the text wehave of the two epics.

First, it is not difficult to imagine that a genius who was able to composean Iliad was equally able to compose an Odyssey, albeit as different fromthe Iliad in its literary character as it is. Richard Janko’s recent figures (thisvol.: ) have shown how similar the Iliad and Odyssey are from a linguisticpoint of view, more similar to each other than each of them, even theOdyssey, to any other archaic epic text preserved. In particular, the fact thatthe sloping lines for the different features in his diagram are almost alwaysparallel between the Iliad and the Odyssey, not crossing each other, speaksfor one poet. For the constellation of frequencies may be expected to berather different between different individuals, whereas between differentpoints in the life of one individual the constellation (and the moves inbetween) can be expected to be fairly synchronous. But the two poemsstill differ. And if there is something like a consensus among scholars thatthe Odyssey on the whole contains more modern linguistic features than theIliad, we have to ask what exactly that means. The differences are surely not

The synchrony of the data and changes is also more likely to reflect a chronological linguisticdifference in the poet than simply a difference of needs of genres (kleos poetry as opposed to nostospoetry).



rudolf wachter

too great for the assumption that they are by one poet who, for example,composed and put down one long poem about the young hero and warriorAchilles in his forties and another one about the mature, intelligent andfar-travelled Odysseus in his sixties. For an oral poet would not have beenlazy in the meantime. We may imagine him to have – just like his newhero – travelled round the Greek world (even maybe to Ithaca), to havesung in very many places and for very different audiences. And we mayimagine him picking up many new ideas and stories, even some shorter orlonger bits of oral epic text, and further developing his art of composinglong epic poems by giving them a more complex structure.

As for the second supposition, it seems very difficult to understand howHomer, who must have been very well-acquainted with, most probably anative speaker of, Ionic dialect, could have used Aeolic forms, in formulaeand normal text and even where they were metrically unnecessary, unlesshe had been living and studying in a strongly Aeolic linguistic contextfor some years, be that a group of singers or a geographical dialect regionor – most likely – both. The unnecessary Aeolisms are well known, ifunderestimated in their significance for our problem. The main ones

are the conjunction � instead of ��, the particle ' XR� instead of ':�, theinfinitive C''�� instead of �)��, and pronouns in the following passages:Il. . Y�� # P''� |, . �5 � 9� P''�� | (instead of '0�; also Od.., .); Il. . | '0� . . . � Z''�� |, . � Z''�� | (insteadof S'�0�); . | �� @ Z''�� , Od. . .� Z''�� |

(instead of S'0�). In view of these cases and some formulaic Aeolisms itseems difficult to get around the assumption of an Aeolic phase precedingthe composition of the poems of Homer the Ionian. More generally, theparticular dialect mixture of the Homeric language only seems plausible ifHomer was, as the tradition has it, a North Ionian (he is said to be fromSmyrna and later to have lived on Chios) since such a mixture was best, ifnot only, acceptable to a population at a dialect boundary.

See for this question, and for the reason for the particular dialect mixture in the Homeric language,Wachter ().

Most Homeric examples in this article are already contained in my ‘Grammatik der homerischenSprache’, Wachter ().

It seems particularly interesting in view of Georg Danek’s new view on the Doloneia (this vol.: ff.)that two of these rare cases are preserved precisely in this song. Of course, the cases preserved are likelyto be only part of the total originally contained in the text and to have been missed by grammariansor copyists who normalized the text where an Aeolic form was not metrically important. So if thepoet of the Doloneia was, as is Danek’s impression now, more of a poeta doctus who imitated ‘Homer’than an oral poet of the traditional type, he may have fancied exaggerating the Aeolic pronouns ofhis model and the two Doloneia examples may reflect the original higher frequency of unnecessaryAeolisms.

See Wachter () for the full argument on these questions.



Linguistic innovations in the Homeric epics

2

But if we have a Homer who learned his art in Aeolis and has left us twobasically Ionic works, of which one is slightly more modern linguisticallythan the other, there is the question, what ‘modern’ means in such asituation; or, to turn the argument round, how would we expect the epiclanguage of a singer with such a biography to change in the course of his life?

And there is another important question. Since we do find modernismsin great number in Homer’s text, what was the attitude of the poet and hisaudience towards these modernisms, and is our attitude the same? I thinkit is not, and this is what this study is about.

A typical example which underlines this is the very exceptional mono-syllabic accusative �� ‘the ship’ instead of normal �� in Od. . | ��'�� '�� [� �� 7�� 3��. The form �� is of course exactlythe Ionic form we know from Herodotus and others. Nevertheless scholarshave tried to get rid of it by all sorts of conjectures, but without success.We will see later that its monosyllabic use was perfectly possible, if not nor-mal in the Ionic dialect at Homer’s time and I cannot believe that anybodyamong his audience would have been in the least shocked hearing it; in fact,they would not even notice it. We, however, as non-native speakers, judgethe Homeric language on strictly quantitative grounds and, as historians,find him fascinating because of the traditional, often archaic features in hispoems, both of form and content. But as far as language is concerned, weare normally happy saying, when looking at variants, that form A is olderthan form B, or, that form A is Aeolic, form B Ionic, assuming that theAeolic variant belongs to an earlier layer of the tradition. We take for nor-mal what is most frequent. And if a form is rare but particularly developedwe tend to suspect it of being a later intrusion or adaptation. After doingthat we call the Homeric language an artificial, mixed, literary dialect. Thissort of description creates, I think, a picture too diffuse for a good startingpoint for a better understanding of the Homeric language and its mixturein particular.

So we may ask the following question: What kind of Greek did Homerspeak when he stopped singing epic verse? If we can establish a realisticpicture of Homer’s spoken language we may gain a better understandingof the other forms he uses and the reasons for which he uses them. Forexample, if we can plausibly argue that monosyllabic �� was the normalform of the accusative of ��8� in the Ionic dialect already at Homer’s time,we will no longer be embarrassed by the single occurrence of this form in

Even Chantraine (: §).



rudolf wachter

the Odyssey and will, at the same time, be in a better state to judge thenormal epic form ��.

3

The reasons for which the epic poet used non-everyday forms are wellknown:� the linguistic restrictions imposed by the metre (below, )� the existence of non-everyday forms in traditional formulae or other

kinds of prosodic structures he had learned (below, )� the need or desire to gain flexibility by means of metrically different

synonyms (below, )But it is by no means easy always to find out what the everyday form was.Let us look at some examples.

4

As for the metre, its importance cannot be overrated. Chantraine (:) tells us clearly that in studying Homeric language we must never forgetthat the whole morphology is governed by the metre, and subsequentlygives a series of instructive examples for the use of synonyms with mutuallyexclusive forms.(a) There is, for instance, ��#3�'�3 ‘we learn’ in Homer because

∗�#�3��'�3 was metrically impossible, but, on the other hand,only �#�3��'��, since ∗��#3�'�� was equally impossible to use.This is straightforward. The problem however starts when we wantto attribute the two forms to different linguistic layers. The unani-mous opinion, also found in Liddell and Scott, is that ��+3�'� isthe older, �#�3��'� the younger, form, but this is not an appro-priate characterization. It is true that thematic ��+3�'� correspondsto skt. bodhami ‘I notice’, but the athematic present formation withn-infix, �#�3-, is just as well-attested in the Indo-European languages,for instance in Celtic, or in Lithuanian bundu ‘I wake up’, so we haveno right to say that it is the younger form. What is true is the factthat �#�3��'� is the normal form in post-Homeric Ionic and Attic,whereas ��+3�'� is almost exclusively epic and poetic, so we may betempted to claim that�#�3��'�, given its rarity, may be due to laterreworking of the text, whereas ��+3�'� represents the normal use at

‘Toute la morphologie est commandee par des preoccupations metriques et nous aurons a chaqueinstant a faire appel a cette consideration.’

See Rix et al. (: –). Only | �#�3��'� (Od. .), | �#�3��'�� (.).




Homer’s time. Yet, since neither form can be said to be younger thanthe other, ��+3�'� cannot just be a predecessor of �#�3��'� inIonic but must originate from a different language variety, which, in thecase of Homer, we call Aeolic. But why then is �#�3��'� so rare inHomer? And what was this Aeolic, from which ��+3�'� is supposedto be taken? Are we to claim that Aeolic had lost the present forma-tion in -��'� and preferred the simple thematic formation with fullgrade in the root, of the type of ��? This is implausible. For it isprecisely in Sappho ., L–P that we find the only occurrence inearly literature of the rare present ��'��, reminiscent of the n-infixpresent preserved in Latin linquo and Sanskrit rin. ajmi, whereas allother Greek dialects here have the simple formation, ��. So we canbe quite certain that the Aeolians used �#�3��'� just as all otherGreeks did. The reason why this present formation is so rare in epic

is therefore much more likely to lie in its prosodic structure, whichmakes many forms impossible to use: the nd and rd person plural ofthe indicative ∗(7)�#�3�� 3�, ∗�#�3��, ∗(7)�#�3��, thewhole subjunctive and optative, the participle, the infinitive, etc. Andthere are similar restrictions in the active, so the other verbs in -��(Risch : ) are extremely rare too. There is only one form thatfits the metre perfectly, the rd sg. imperfect active. But even that couldnot prevent the present formations in -�� from being shunned bythe poet, and indeed only one such rd sg. imperfect form, \��(from ]��), which had found its way into several formulae, is wellattested. On the other hand, ��+3�'� was of a desired prosodic struc-ture, like ��'�, 1��'� etc., and therefore much easier for thepoet to handle. So, the so-called ‘Aeolic’ present ��+3�'� has hardlyanything specifically Aeolic to it, but was just a traditional elementof the epic language. No doubt Homer the Ionian adopted it witheverything else when he learned the art of epic singing. No doubteither that it was easily understandable to the well-trained audience ofepic poems even in Ionia, who in fact will have considered it slightlywhimsical rather than archaic or poetic. On the other hand it seemsclear that Homer in his daily speech would use �#�3��'�, the formwhich was to be the normal one in all later Greek, and its sporadic usein his poems would go completely unnoticed.

See Rix et al. (: –). Once precisely this st person sg. occurs, and once the same form from ��+3�'�, a ‘flexibility’ case. Also used in formulae, e.g. line-initial ��+3�� 4,� ‘for (s)he learned’, attested once each in the

Iliad, the Odyssey and Hesiod’s Theogony.



rudolf wachter

(b) A similar and well-known case is the genitive ^'�� ‘of the day’ whoseparallel form from '��, viz. ∗ '��, did not fit the hexameter. Infact, '�� occurs only six times in Homer – twice in the Iliad, fourtimes in the Odyssey, three times in the nominative singular, three timesin the nominative plural – whereas M'� is used dozens of times inmost case forms. Now, it is true that from the point of view of itsformation, M'� is more archaic than '��, but there is no doubtthat '��, which is attested in the whole of post-Homeric Greek,particularly in Ionic–Attic, was the current form at Homer’s time.Hesiod already formed a compound from it, '�� ‘sleeping byday’, talking of thieves (Op. ), where he could have easily coined∗I'��, had '�� still been a rare word at his time. Surely thiswas not the case. Its only handicap was that it was against the metre;apart from the two nominatives only the dative singular was usable.

These examples teach us three things. First, if one of two variants israre in Homer but is going to be the usual Greek form in post-Homerictimes, its rarity in Homer need not be due to its having been of recentformation or even unusual at the time of the poet. It could as well havebeen quite normal. Secondly, the frequent variant need not be frequentbecause of a strong poetic tradition which favoured and preserved archaicforms but simply because the usual form was ill-suited to the metre. And,thirdly, whole lexemes or even morphological categories can be more orless banned from epic even if single forms of their paradigms would beusable and rarely are used.

5

(a) Our first example to illustrate the preservation of a prosodic structureas the reason for using non-everyday forms is the ‘digamma’ (i.e. thephoneme /w/). Clearly Homer did not pronounce a /w/ in the lineends �5 �� (G)5�� | (Il. .) or �� (G)P�� | (.),since this would have lengthened the short syllable. On the other hand,line ends like �5 ��# =� �� #<.� C�� H�� (G)5�� | (.) or�� (G)P�� | (.) and �� (G)P�� | (.,) must be read with unaltered hiatus, i.e. a ‘ghost digamma’, because

Of course not ∗I'�� or ∗I'�� (nor any gen. pl. form of '��). A form like ��+3�'� need not even be inherited, despite the Sanskrit parallel (and German bieten)

and the existence of the active verb in Doric (Crete, see LSJ with Suppl.), but could be a ‘recent’analogical formation some time back in the epic tradition.

This corresponds to the widespread principle of linguistic economy.




originally there had been a /w/ at the beginning of the words ��0� andP�[. There is some debate whether or not we should assume that inthese latter cases Homer actually spoke a /w/. I do not think this islikely. At least he could pronounce these lines without the digammajust as well as we can. On the other hand – and this is the crucialpoint – we can be certain that he knew this sound existed in otherdialects and that he even knew exactly where those fellow-Greeks pro-nounced it and where they did not. For he neglects it many times, butthere is, as far as we can see, not a single case in which he wronglyinserts such a digamma-based hiatus. On the other hand, when usingdigamma words and forms, he allows for the hiatus not only in for-mulae, but also in ad hoc lines. In fact, in the course of his career heeven coined entire formulaic lines without digamma, like �5�� @7�3�'�� '�� M'� (G)�� 3� (Od. ., ., .).

This etymological accuracy of his prosodic use of original /w/, whichwe may call ‘prosodic habit’, can be used as an argument that Homerhad had personal contact with Aeolic singers who pronounced thesound, and that he even knew their dialect. For through an intermediatestage of Ionic teachers this accuracy would very probably have sufferedmore damage; and on the other hand it is unlikely that in the Ionicdialect itself the generation before Homer still fully pronounced thesound and Homer’s generation lost it so rapidly that Homer couldneglect it as frequently as he did. Sound changes of this sort usually takeat least years. So we may claim that Homer is personally responsiblefor the modern, Ionic features in his poetic language, indeed for theentire mixture of his particular ‘Homeric language’.

(b) A very similar example of prosodic habit is the so-called diek-tasis (Chantraine : – §–). Verbs like I4�� |

(Il. .), �� | (.), I4�� 3� | (.), �� 3�(.) are of a historically implausible structure. Etymologically, weexpect ∗I4��, ∗�� , ∗I4�� 3�, ∗�� 3�, andfinally they were contracted to I4��, �� , I4�� 3�,�� 3�. But, in fact, the Homeric forms are the contracted forms,save only that they were against the metre and were therefore adapted totheir original prosodic structure. This is proof, of course, that Homerin his spoken language pronounced the contracted forms. That con-traction had largely taken place in Ionic by Homer’s time is confirmed

The formula �� %"��, mentioned by Chantraine (: ), does not prove anything; it is anold formula with /h-/.

See Chantraine (: ). See Wachter ().



rudolf wachter

by numerous cases of many different sorts. Here are a few examples(more will follow below): | ��>�0 �(�� 1�>�0�� (Il. .), and –at the end of a line – '�0� | (.), 7�3�0� | (.). Even crasis iswell attested (Chantraine : – §), e.g. in | '� �#�� @ P��P�� (.), | ��Z�� (.), probably also | ^ _� �� < �I'�0�� '+��'�� (.).

So, both the digamma and contraction show how prosodic habit was astrong factor conserving older linguistic features. But we must not thinkthat the more modern equivalents, digamma unobserved or contractionaccomplished (or even diektasis), would in any way have been problematicto use or unacceptable to Homer’s audience. Otherwise they would not beso frequent.

6

This brings us to the third reason for using non-everyday forms, viz. theneed or desire of the oral poet for linguistic flexibility. In fact, already theexamples we have been looking at partly show this tendency.(a) First the digamma. In most of the above cases without formulae it

would not have been necessary to account for the hiatus left by the lost/w/ sound, since the forms would have been perfectly usable withoutit. But, of course, a form like 5�� was much more useful if itcould be treated both as beginning with a vowel and beginning with aconsonant. This was as good as a pair of synonyms of equal length butdifferent beginning (consonant vs. vowel) according to Edzard Visser’sprinciple for his variables (e.g. �� vs. 7�:��) (: passim,e.g. –) or Milman Parry’s principle for (epithets in) formulae (e.g.Il. . >�V� �43.� -�� | vs. . ��B1�� -�� |).

(b) Next, contraction could be used to enhance flexibility, e.g. | � ��# .��'�� (Il. .), but | ` ��# .� 1��# ��.� 7��[�� ':�� |

(Od. .; here, for once, it is the Odyssey that shows the ‘older’form!), or | �� 3��1�� F�� (��0 �� $�� 1�� (Il. .),but | �#�'�� a� 3 @ F�� 3��1�� 9� $�� 1�� (.). Forthe purpose of the present question we have to stress that in mostvowel combinations known to have been contracted e.g. in Attic, thechange must have been accomplished in the Ionic dialect already byHomer’s time. This, too, is overlooked or at least underestimated inmost Homeric grammars. It is just that Homer, for several reasons,

Therefore we have no right to restore ∗ '�� e.g. in A�� @ '�0� �5 �'�� P�� | (.). Including my own, Wachter ().




chose to use the uncontracted variant more often, partly because thecontracted one would not fit the metre, partly for lack of ‘prosodichabit’ because contracted forms had not yet been provided for andpractised in the traditional oral technique. But if a contracted formfitted his purpose, why should he not have used it? And he did use itand will have been glad to have it.

(c) A very similar phenomenon is synizesis. Here, too, the phonologicalchange can be clearly checked, since it has led to the loss of a syllable.It is mostly the combination of a short [e] + a short [a] or [o] thatis reduced in this way. We have clear examples that can only be readmonosyllabically, e.g., with [e + a], L �� +�� | (Il. .), or��[�� 3�� | (.; a frequent formula, though normally inthe nom.: 3��:� |), both at line end, | '�� 9� P��#� (.), |

��': ��, b�� 7�� # �� (.; if not�� cR#� aspreferred by West); or, with [e + o], | �� @ �� @ �(��4�� 1��#(Il. .), �� )�� | (.), | S'0� '�� 3�� 0��(.). In the case of [o], the manuscripts often give <�#> insteadof <��>, e.g. | 7[ @d��>�#� (Il. .), | ��# '�� 3�� #� (.),�3�� 8�� | (.); we do not know exactly how oldthis spelling is, but the first examples are found in inscriptions of theearly sixth century bc, so it may be a very old reading aid precisely forthese monosyllabic cases.

Here, too, we find the disyllabic versions, of course, and for thesame reasons as before, namely out of prosodic habit, e.g., with [a],in C�� +� | (Il. .; a formula, also containinga case of contracted, i.e. monosyllabic, -D |, from ∗-�), | ��'�'��, �� @ 7�� 3��#� (.), or, with [o], |

��+'��, '�� . . . (Il. .), | �8� � @ 7'�� (.);or because the monosyllabic version would not fit the metre, as in |

P�� # '�� (.) or '��'�� 1�� | (.).Again we can see clearly how both versions served the poet perfectly.

So why should he not have used them both? And again he did, nobody

It is true that this form is (weakly) attested in . and . (not in . and ., ., .,, Od. ., .). But we may be sceptical of its being an original Homeric form rather thana later import into some of the Homeric traditions, even e.g. by a grammarian who looked for ameans to distinguish the cases with monosyllabic ending (synizesis) from the ones with disyllabicending. For it seems difficult to see where at Homer’s time this third form – besides ��+�, thenormal Ionic form (× Il., × Od.), and �� without synizesis, i.e. most probably the Aeolicform (trisyllabic �� Alcaeus . L–P = . LGS; this is yet another unnecessary Aeolismin Homer, by the way, occurring Il. ., ., , , , ., ., ., , , Od..) – should have come from.

First attested on a Corinthian vase with a scene from the Iliad (.ff.), in which Kleoboulos,clearly written [U]�. �.+.>�.#��, is being killed by Ajax; see Wachter (: – COR ).



rudolf wachter

minded, and we can conclude that synizesis was the normal way topronounce these forms in everyday Ionic at his time.

(d) A long vowel preceding a vowel in word-interior position was shortenedin Ionic. The principle is often called vocalis ante vocalem corripitur or‘prevocalic vowel shortening’. A frequent combination affected by thischange was a long [e] before an [o] or [a], e.g. in the genitive pluralof a-stems, which had been ∗-asom originally (see Latin -arum), thenbecame ∗-ahon in Early Greek, ∗-eon in Early Ionic, whence -�� inHomer and later Ionic and -�� in Attic. The Homeric form of theIonic genitive plural is mostly monosyllabic, that is, pronounced withsynizesis, e.g. e�� @ �( �:3�� @ 71��'�� | (Il. .), rarely evencontracted, e.g. | ��3�� 7 �� (.). The disyllabic versionis very rare, but not unattested, e.g. |f� ��g��#�� 7[� #�� (.).Something similar happened to the genitive singular of the masculinea-stems, which had -ao in Mycenaean and in Ionic–Attic eventuallybecame -�� by a similar shortening, and at the same time a lengtheningof the second component. Therefore the phenomenon is here knownas metathesis quantitatum or ‘quantitative metathesis’. See, e.g. theformula U��# �� 4#��':�� | (Il. .), to be read with amonosyllabic ending, that is with synizesis.

Now, the Ionic vowel shortenings in these endings were so frequent anddestroyed so many formulae that a poet had to find some way out. Againhe decided to retain the Aeolic prosodic variants as a means of flexibility,only that in this case he used them almost exclusively. This is why we alsohave gen. pl. - XR�� (mostly at line end, e.g. . ��'�� |) and sg. -D�(ditto, . ��B�� |), both even with non-Ionic vowel quality, a long[a], because this was an existing form (in Aeolic), whereas the intermediatestage that had led to Ionic -��-, viz. ∗-��-/∗-��-, was not.

The same radical sound change had happened to an important noun,�D��, and the names of two important gods, Hermes and Poseidon. As for�D��, forms in -��- (from ∗-D�-) are lacking almost completely, exceptionsbeing three proper names, �� *:�� M�� | (Il. .; Il.x, also in the other cases), �4�� (Od. ., ; to be pronouncedwith synizesis), and, with metrical lengthening, *�� (Il. .;which may be called an unnecessary Ionicism). They show that their ∗[e]from earlier ∗[a] (the latter preserved in Aeolic) was short in Ionic everydayspeech in Homer’s time. As for the two gods, their names are attested

This is a recent, Ionic formula. On the other hand, the compensatory lengthening of the following o-vowel cannot be proven in

any of the Homeric passages (so ��?); nor can it be excluded.




in Mycenaean as /Hermahas/ and /Poseidaon/ and had become !d�'��and �� in contemporary Ionic, impossible to use in the hexameterunless pronounced in the final, radically shortened form with synizesis. Allgood old epic formulae for these gods had become impossible to use. Sothe Ionic poet decided to use the two names in a form closer to the Aeolicone: �� XR�� and !d�'��D� (gen. -D� etc.). Nevertheless Hermes quiteoften occurs in his recent prosodic form too, e.g. | !d�'�� 7[:44�� (Il..), or 7��+�� !d�'�� | (.). Poseidon, somewhat surprisingly, doesnot occur in his recent form in Homer, only in Hesiod, 3+�� @ 7��3�� | (Theog. ; an undisputed passage). But there is no reason,even in the case of Poseidon, to doubt that the monosyllabic pronunciationwith a shortened first vowel was fully achieved in Ionic by Homer’s timeand that the poet used the longer forms mainly out of prosodic habit.

The principle of linguistic flexibility is of course a general one in Homerand can be observed also in other domains of the language. Here are a fewexamples: �/�� vs. P�, �� vs. ��, S�� vs. S��, ��/�� vs.��, �'1�(�), P��(�)/'��(�), �� vs. �� /�� ,��'�� vs.��'�#,h�� vs.��,;�� vs.A�, #i/#<� vs. #<��,4�#�� vs.4�+6��, ��>: �� vs. ��>�, �+ �'�� vs. �+ �'��, C3�� vs. 3��,C>� vs. C>� �, �� and �� vs. �� and ��,��:'>�� vs. �1�'��, the pairs �� ( )��, C ( )� 3�, �� ( )�,@&�# ( )�+�, ��(�)�+�, /�(�)��, C�(�)>�, �(�)��'��, +� vs. [+�,P�, ��, �� vs. ��, ��, ��, and so on.

7

To sum up, the starting point for a linguistic analysis of the Homericlanguage should be as clear-cut and comprehensive a picture as possibleof the contemporary Ionic at Homer’s time, which must – and can –be gained from Homer himself. We have to be aware that some of therespective modern forms are rather rare or otherwise difficult to detect inthe text Homer has left. We have also seen the reasons for this strangesituation: first, metrical restrictions, secondly, prosodic habit, and, thirdly,reluctance to give up the traditional forms or to renounce the modern onescompletely because the choice increased linguistic flexibility. But we shouldnot think that, as far as the poet or his public were concerned, anything

The long [a] of !d�'��D� may be due to that of �� XR��, the gen. in -D�, and a general desireto give the name, since it had to be altered, a traditional aura; the -��- is probably due to metricallengthening.



rudolf wachter

was wrong with using a modern form. If that had been the case, we wouldnot have so many of them.

In terms of relative chronology we may state the following: If a poet, inthe course of his career, altered his poetic language, which I am convincedcould easily and recognizably happen, in particular with light, unconsciouslinguistic features, this must normally lead him away from any peculiari-ties that did not conform to his or his audience’s vernacular dialect towardsthat dialect. In the case of Homer, this means away from Aeolisms anddisyllabic pronunciation towards Ionic and monosyllabic pronunciation.

And if we believe that modernisms were fully acceptable to the poetand his audience, there is a second thing we may learn, namely that weshould be rather cautious in restoring pre-forms that are weakly attested orunattested in the manuscripts and papyri we have. This does not mean, ofcourse, that restoration of pre-Homeric forms e.g. in formulae should beconsidered a futile thing to do. But we should take that for what it is, andnot rush to claim that Homer himself put those archaisms in his text. Ourmotto should be: in dubio pro textu.

8

And yet, there are cases where we may claim that Homer wrote differentlyfrom what we have in the text now. We have to ask therefore, how themodern Ionic we are trying to reconstruct was transcribed at his time. Wehave two relevant testimonies, both extremely important for many reasonswhich cannot all be mentioned here. One was found far away from whereHomer lived but is contemporary or even slightly earlier, the other closergeographically but slightly later (depending on how we date the poet). Thefirst is Nestor’s cup, the famous geometric skyphos from Ischia inscribedaround bc:

=� �� : C.[�� ]�. : �Z�� : �� XR��2h.� � @ j� �o�� R� � : �� R��[�] : (�� k��h XR�'�� h�� XR� �� : �� [1�]�o : �1�� XR�� R��.

Nestor had a cup from which it was good to drink. But whoever will drink fromthis cup here will be seized with the longing of Aphrodite of the beautiful wreath.

Tmesis, studied by Dag Haug (this vol.) is probably already too ‘conscious’, too ‘stylistic’ a featureand should be treated differently.

In the case of Hesiod, on the other hand, this may have been different, although we do not knowenough about his life and his place(s) of residence and work. Some of Richard Janko’s figures (thisvol.: ) seem to indicate that this poet may have tended towards a dialect with less contraction(more 7#- than �(- and more 40 than 4� in Op. than in Theog.) – a feature notably of Boeotian.




This inscription, which is in the Euboean alphabet, teaches us several thingson the notation of epic texts at Homer’s time; first, that punctuation wasused (mainly in the caesurae), secondly, that poetic text was written instichic order line after line, and thirdly, that one knew how to write longconsonants with geminate letters. All three features are, of course, valuablereading aids and testify to what we should call a remarkable didactic con-cern. I cannot see how this inscription can be denied to reflect written epictexts. And as far as our modernisms are concerned, it presents an excellentexample for the principle in dubio pro textu, in that it contains a modernform, as we have it in our Homeric text, not the earlier, reconstructedone, as is sometimes claimed for the original text, namely �� 1�6��# �1��, not �� 1�� @ �1�� (i.e. elided -��). So weshould not try to claim Il. . ��D @ �� either (Chantraine : §). Surely Homer used the Ionic genitive form in all these cases.

This leads us to the second testimony to be mentioned here, the highlyarchaic Nikandre kore from around bc, which was found on Delos buthas to be attributed to the Ionic dialect and alphabet of Naxos:

=�� ' @ ��3 R�� h(�)�>�� ,l XR�� A��|�� o =h ��, Ch �� (�):o�,A��'�� 4� XR��, | m��h o � @ P�� . (8�).

Nikandre dedicated me to (Artemis) the far-shooting (?) caster (?) of arrows,daughter of Deinodikes the Naxian, pre-eminent among all (women), sister ofDeinomenes, and now (?) wife of Phraxos.

This inscription uses the letter heta/eta as both a consonant and a vowelsign, and – what is quite particular – uses it as a vowel sign only where thelong [e] in question continues a former long [a]. What is important forour purpose is the fact that the gen. sg. and pl. of the a-stems and the gen.sg. of an s-stem are measured monosyllabically, i.e. with synizesis.

In fact, in the case of the gen. pl. fem. of P�� the inscription, althoughof heavily epic diction (see the formula in line and the noun �4�:��),is younger linguistically than Homer in that the poet, both in the Iliadand in the Odyssey, uses the trisyllabic Aeolic form �� XR�� (Il. .; Od.., .).

What is also remarkable is the fact that the inscription considers the[e] shortened by the principle vocalis ante vocalem corripitur, more worthnoting with a sign which rendered the correct quality but a wrong quantity

See e.g. Chantraine (: §). h = the same sign as � (< ∗[a]); h = a slightly different sign. Not at line end, which is not the position for words of three long syllables.



rudolf wachter

of the vowel, i.e. eta for the open, but normally long, [e], than with thesign for the short [e], the epsilon, which must have seemed too closed.Of course, if one knew that the combination of two vowel signs had tobe pronounced monosyllabically anyway, the quantity of the vowels wasless important than the quality. It seems quite likely, therefore, that in theHomeric text also these cases were originally written with eta, the signfor open [e] and therefore the standard sign for the former long [a], andthat this was only later changed into the classical Attic spelling <��>.In fact, there may be a small hint of that, for in our very first example,Od. ., with the unusual monosyllabic and line-initial accusative ��'�� '��, which is confirmed by some manuscripts and Aristarchus, mostmanuscripts show the usual form �� '�� '��. It seems quite likely,therefore, that this form �� had been in the Homeric text right from thebeginning and had been overlooked by the grammarians who otherwiseadapted the forms with synizesis to the later standard spelling with epsilon,just because �� looked so normal.

We know for certain that in some cases the original, ‘Homeric’ spellingmust have been modernized, notably in the case of the so-called spuriousdiphthongs <��> and <�#> resulting from contraction or from com-pensatory lengthening of short [e] and [o], which in the most archaicinscriptions, e.g. in the two just cited, are always written with epsilon oromicron only. So, we may ask, may there have been more such instancesof modernization? Very likely candidates are the cases cited above in whichflexibility was gained through contraction (of vowels of different qual-ity). For it seems more coherent to assume that the now contracted formswere still pronounced and written as diphthongs in Homer’s time, andcontraction was introduced later into his text in order to make readingeasier. So Il. . could have been written | � ��# .� ∗��':��, justlike | ` ��# .� 1��# ��.� 7��[�� ':�� | (Od. .; whichthen would have the same form, only pronounced not with synizesis butaccording to prosodic habit), or | �� ∗��3��1�� F�� (��0 ��$�� 1�� (Il. .), just like | �#�'�� a� 3 @ F�� 3��1�� 9�$�� 1�� (.), or C�� ∗�� +�� | (.), or | ��3��7 ∗�� (.), or 7��+�� ∗ !d�'�� | (.). The same may beassumed for the ‘crasis’ cases like | ^ _� �� < ∗I'�0� �� '+��'��(.; now �I'�0�). These ‘contraction’ and ‘crasis’ cases would then bejust two types of ‘synizesis’.

The reading �� '�� '�� is preferred since van der Valk (: ). Disyllabic counting was perhaps retained longer in a short word, but this does not mean that

monosyllabic counting was not possible.




Especially the two most extreme cases of contraction I have come acrossin Homer speak in favour of such a post-Homeric modernization. First,the hapax legomenon �D�� in Od. ., a fully Attic form (as Ar. Pax) and expected to be ∗�� in Homer (with spurious <��> due tocompensatory lengthening from ∗�G� ��, like Homeric 1�� from∗1G� ��, but Attic 1D��), may originally have been written �8� � @ �?�� [+� ��, �� (with [ae] pronounced monosyllabi-cally). Secondly, the gen. pl. fem. �� in | �� @ b3��>�� S4�.� C�� (Od. .), confirmed by Aristarchus, whichin its expected epic form ��(G)� �� or ��(G)� �� would notfit the metre and is now normally spelled and counted ��# ��, mayoriginally have been written in the normal way �� (or indeed�� :��) but pronounced with two synizeses.

We will perhaps never know, but the crucial point about these contrac-tions is again that already in Homer’s prosody they were pronounced inthe modern, monosyllabic Ionic way, as were many of the other cases inthe Iliad, and even more in the Odyssey, that we have discussed here.



chapter 4


Margalit Finkelberg

In the Preface to the second edition of his Studies in the Language of Homer,G. P. Shipp wrote:

That speeches tend to be later in language and to have more other abnormalitiesthan the narrative has been noted by readers of the first edition and has nowbeen stressed in the analysis of several books. The project would no doubt repay asystematic treatment. (: vii)

In the subsequent discussion Shipp from time to time highlights the specialstatus of the speeches vis-a-vis the narrative. Although far from the system-atic treatment of the language of the speeches that he recommends, this isstill the fullest treatment available. To adduce some examples, he notes thatwhile ‘[t]he freedom from exceptional features in the narrative is typical ofsingle combats’ that open Iliad , ‘[t]he linguistic character changes with thespeech of Helenus to Hector’ at .ff. (); that ‘Nestor’s speeches alwaysmake us expect neologisms’ (); that ‘[a]n important general observationin regard to the duel [of Hector and Ajax in Iliad ] is the difference linguis-tically between the narrative itself, which has few features to be noticed,and the speeches, which are often marked by many abnormalities’ ();that ‘[u]nusual features thus cluster especially in these extra-narrative por-tions’ (); that the great speech of Achilles in Iliad ‘is for the most partcharacterized by features that reflect contemporary Ionic’ (); that ‘thelinguistic contrast [of the speech of Menelaus in Iliad .ff.] with thenarrative itself is very marked’ (); that ‘[t]he tendency for late featuresto occur in speeches rather than in the narrative is especially impressive’ inIliad (), and so on.

Unfortunately, as distinct from Shipp’s thorough treatment of the lan-guage of the similes, his ad hoc remarks relating to the speeches have exertedlittle influence on the subsequent study of the language of Homer. This isnot to say that these remarks came as a surprise. Students of Homer havelong been aware of the fact that the language of Homeric speeches differs




from that of the narrative in many and various ways. O. Jorgensen in

and M. P. Nilsson in pointed out that Homer’s characters speak ofthe gods differently from the poet himself. P. Krarup in and HermannFrankel in called attention to the fact that abstract nouns and person-ifications are much more frequent in the speeches than in the narrative.In an important article published in , T. B. L. Webster showed thatclusters of late features, including linguistically late formulae, are especiallycharacteristic of the speeches. Numerous neologisms, anachronisms, andother peculiarities of language and vocabulary have been registered for thespeeches in many a commentary on the Homeric poems. Yet, to the best ofmy knowledge, the first to treat the distinction between Homeric speechesand Homeric narrative in a thorough and systematic way was Jasper Griffin.In a ground-breaking article published in , Griffin stated unequivo-cally that ‘in important senses the Homeric epics have two vocabularies’,one for the narrative and the other for the speeches (: ). As we shallsee, the conclusions that he drew from this observation were quite differentfrom those made by Shipp and others.

In fact, two main approaches to the speeches have crystallized over time.Older scholars tended to account for the speeches’ linguistic and otherpeculiarities by applying to them the interpretative methods of Analysis.The culmination of this approach was reached in Shipp’s Studies, firstpublished in . According to this approach, the fact that innovations inlanguage and vocabulary tend to concentrate in the speeches indicates thattheir composition is later than that of the main narrative; consequently,the passages in which these innovations are especially numerous should betreated as interpolations. However, although it cannot be denied that moreoften than not interpolations are indeed concentrated in the speeches, andespecially in their concluding parts (Shipp : ), this does not meanthat all the speeches lend themselves to this kind of treatment. The reasonis simple: in so far as the speeches constitute about per cent of Homer’stext, this would mean that, if consistently applied, the Analyst approachwould culminate in the conclusion that about half of Homer should beregarded as interpolated.

As distinct from this, the neo-Unitarian approach that became pop-ular in the second half of the twentieth century tended to regardthe speeches’ idiosyncrasies as due to self-conscious stylistic strategiesdeliberately employed by the poet. Although occasionally applied by other

The special status of Homeric speeches was already well recognized in antiquity, see Nunlist (). Jorgensen (: –), Nilsson (: –), Krarup (: –), Frankel (: ), Webster

(: , ).



margalit finkelberg

scholars as well, the first systematic treatment of the speeches from thispoint of view was incontestably Griffin’s ‘Homeric words and speakers’,mentioned above. Griffin’s starting point that the speeches in Homer have‘important distinctions of vocabulary, and of style, from the rest’ (:), is identical to that of Shipp and other Analysts; but the conclusionswhich he reaches are diametrically opposite. According to Griffin, ratherthan being indicative of interpolation, the distinction between speech andnarrative, including the distinction in language, is entirely a matter of style:

It therefore seems appropriate to expect that the later stages of the tradition willnot simply have been introducing more contemporary linguistic modes into thespeeches without reflection, but on the contrary allowing them into the speeches,and excluding them from such narrative as they composed themselves, in accor-dance with a feeling that they were more appropriate there. (: )

While I agree with Griffin that the argument of style may effectively accountfor such features of Homeric language as the avoidance of specifying thenames of gods involved in a given action or the use of terms of moralevaluation that are absent from the narrative, it is doubtful that the stylis-tic interpretation he proposed would account equally well for the purelylinguistic characteristics of the speeches. It is indeed difficult to envisagea traditional poet deliberately employing, say, quantitative metathesis asa means of stylization. This would become even more evident if we takeinto account another important aspect of Homeric language, its formulaicidiom.

The role of the formulae as indicators of chronologically different strataof Homeric language was aptly summarized by Bryan Hainsworth:

A linguistic development of the vernacular quickly penetrated the fluid and non-formular part of the Kunstsprache (where it differed least), or took effect atthe junctions between formulae: next the development would appear in mod-ified formulae, ‘formulae by analogy’, and other derivatives of primary for-mulae: last of all would the development be found attested among regularformulae. (b: –)

That deviations from the formulaic language can be used with profit forthe identification of later strata of epic diction has been emphasized byother scholars as well, notably Arie Hoekstra () and Richard Janko

See e.g. Dodds (: ), on the difference between the poet and the characters in the identificationof divine interventions: ‘For it is the poet’s characters who talk like this, not the poet: his ownconvention is quite other – he operates . . . with clear-cut anthropomorphic gods such as Athenaand Poseidon, not with anonymous daemons. If he has made his characters employ a differentconvention, he has presumably done so because that is how people did in fact talk: he is being“realistic”.’




(: ). On the whole, however, surprisingly little attention has beenpaid to the fact that the special status of the speeches vis-a-vis the narrativebecomes particularly manifest when Homer’s formulaic language is takeninto account. As far as I can see, one of the reasons why this phenomenonhas not drawn as much attention as it deserves is that, in spite of the evidencethat has accumulated since Milman Parry’s theory of oral composition firstbecame known, many adherents of oral formulaic theory still proceed fromthe assumption that per cent of Homer consists of traditional formulae.Accordingly, Homer’s language is still regarded by many as essentially amonolithic phenomenon. It is symptomatic in this respect that one ofGriffin’s purposes in ‘Homeric words and speakers’ was ‘to suggest that thelanguage of Homer is a less uniform thing than some oralists tended tosuggest’.

This is not to say that the belief in the -per cent formularity ofHomer is shared by all oralists. Thus, by counterposing charts of formulaicdensity in a routine battle scene on the one hand and in the Lamentof Helen over the body of Hector on the other, Hainsworth effectivelydemonstrated in that the formulaic density in the latter is ‘sharplyreduced in comparison with the battle scenes’ (: ). In , JosephRusso studied two Homeric passages, the exchange between Odysseus andEumaeus in the Argos episode in Odyssey (–) and Hector’s rebuke ofPolydamas in Iliad (–), and came to the conclusion that their totalformulaic content ‘falls far below the % we have been led to expect’.

That it is first and foremost in direct speech that Homer’s non-formulaicexpressions are concentrated was further argued by this author in an articlepublished in . To quote its conclusions, ‘Thus, the so-called “isolated”expressions differ from the formulaic expressions in several respects: theycannot be shown to have been modelled on formulaic patterns; and theytend to occur in direct speech (rather than the main narrative)’ (Finkelberg: ). On the whole, however, it cannot be denied that the study ofthe formulaic aspect of Homer’s speeches is still a largely neglected field.

As Janko demonstrated in Homer, Hesiod and the Hymns, the evidenceof language is only significant when it comes in clusters: ‘[w]hen we find acluster of results, this implies that the poet concerned is using the traditionaldiction “naturally”, at more or less the stage at which he found it’ (:, –). In other words, only when we find peculiar features pertainingto the language, vocabulary and, last but not least, the formulae of Homer

Griffin (: ). See now also Finkelberg (). Russo (: ). Cf. also, on the Odyssey passage: ‘By the criteria normally used by the orthodox

Parry school, this passage would be the creation of a literate poet’ (ibid.: ).



margalit finkelberg

assembled in one and the same passage, can we say with certainty thatthe passage in question belongs to a different chronological stratum fromthe rest of the text. In what follows, I apply this approach to three Iliadicpassages: Hector’s reference to the former wealth of Troy in his rebuke ofPolydamas in Il. .–; the opening lines of the speech of Menelausin Il. .–; and a story of the persecution of Dionysus by the kingLycurgus as told by Diomedes in Il. .–.

il. 18.288–96

�� '�� 4,� ��'�� '�� P�3�� '#3� �� +��# �� +��2�8� �� V 7[�� '�� ':�� ,��, �� V m�#4�� -�� 7��V��:'� ��'�� @ F��, 7�� '�4� n�+ �� ;�+�.�8� �@ /�� '�� C�� U��# �� 4#��':��8�� 3@ 7�� # �, 3�� @ C� � ��+�,�:��, '�� 8� ��:'� 10� @ 7�� :'�2�( 4�� W�� 7�� 2 �( 4,� 7� �.

The passage is famous for its brevis in longo in '�� P�3�� at .In the above-mentioned formulaic analysis Russo showed that it containsa surprisingly high number of expressions that occur only one other timein Homer. He wrote in this connection: ‘In other ways, too, the passagepresents the analyst with some very “marginal” phrases’ (: ). MarkEdwards remarked in respect of '�� P�3�� and ��+��# ��+��: ‘There are two violations – or innovative uses – of normalformular conventions in this couplet [.–]’ (: ). Shipp com-mented on the passage’s linguistic lateness and suggested that ‘the twospeeches [Polydamas’ and Hector’s] were originally in a simpler form, e.g.merely a suggestion by P. that the Trojans should retire to the city and H.’sreply, unless the deliberations as a whole are late’ (: ).

Il. .– contains no less than five uncommon or linguistically lateformulae:. '�� P�3�� at . Kurt Witte long ago pointed out (a:

) that this highly irregular expression was prompted by analogy with'�� 3��, which occurs nine times in Homer, and Milman

Since the characteristically Odyssean language and formulae within the Iliad are usually regarded aschronologically marked, this makes the Iliad a more convenient target of analysis in terms of relativechronology. What should be regarded as relatively late within the Odyssey itself is of course a separatequestion.




Parry later showed (in A. M. Parry b: –) that it is a result ofthe juxtaposition of two Homeric formulae,�� (or��) '��3�� and��'�� (or��). Both the irregularity of theexpression and its derivative character indicate that it could not belongto the stock of traditional formulae.

. ��+��# �� +�� at . The expression belongs to whatRusso (above, with n. ) defined as a marginal group of formulae thatoccur only one other time in Homer. Although both ��+��# �� and��+�� are well-attested Homeric epithets, each of them beingoccasionally employed for the description of a city’s wealth (��#��+6 �� -#:�� ×, Q�� #��# Od. .), they are as a ruleused separately. The only other occurrence of the combination foundhere (nominative case) is in the description of Dolon in the linguisticallylate Doloneia (Il. .).

. ��':�� at . The expression belongs to the same group ofmarginal formulae as��+��# �� +��. The only other occur-rence, also in direct speech and in a rather similar context (the bootythat Odysseus brought with him from Troy), is in the Odyssey (.).The expression seems to have been created by analogy with the for-mula ��':�� 0�� in order to express the meaning of another estab-lished formula, ��':�� , � 7 3�� (x), after the fourth-footcaesura.

. m�#4�� -�� 7��V� at . The only other occurrence(genitive case), also in direct speech, is at Il. .. Occasional emergenceof Phrygia and Meıonia in Homer is a well-established anachronism.

On the whole, Phrygia is mentioned five times and Meıonia twice (thePhrygians and the Meıonians three and four times, respectively), only inthe Iliad. In view of this, it is especially noteworthy that the epic poetsalready had at their disposal a formula for the description of the two(but cf. also *#�� -�� 7��V� in Hymn. Hom. Ap. ).

. U��# �� 4#��':�� at . Although this expression occurs noless than eight times in Homer, not only is it linguistically late (quanti-tative metathesis in �4#��':��) and, based as it is on the formulaeU��# �� and U�� 4#��':��, formulaically derivative: it isalso equivalent to the much more frequently occurring formula, ��V�� 3�� (x). Note also that, though U�� 4#��':��as such occurs three times in the Homeric corpus (Il. ., Hymn. Ven., ), it is an unmistakenly ‘Hesiodic’ formula (x in the Theogony).

See esp. Kullmann (a: ), including a discussion of the present passage.



margalit finkelberg

All these indicate that U��# �� 4#��':�� is a rare example of alinguistically late formula. To quote Shipp’s assessment of the contextsin which it occurs:

Whereas the latter [��V� �� 3�� ] is always in narrative ourformula is often not, and mostly in contexts that may not be old: B commenton army discipline, B (ath. Ar.) in the portent,A simile, M ath., in anexplanatory verse, leaving I (Diomedes’ speech), � (between a simile andthe dialogue between Zeus and Hera which Zen. omitted), Q (Hector toPolydamas). (: –)

Contrary to appearances, the fabulous wealth of Troy is far from beinga well-established Homeric theme. This can be seen first of all from thecity’s epithets: as distinct from such cities as Mycenae or Corinth, noneof the constant epithets of Troy emphasizes its exceptional wealth. Thetheme recurs only one other time, in Iliad , again in direct speech (–, Achilles to Priam): although Phrygia is mentioned there as well, thewording is different, and the rest of the passage does not resemble Iliad .This seems to indicate that the wealth of Troy is a relatively recent epictheme which has not yet attained a formulaic fixity. If correct, this wouldmean that the passage under discussion should be considered late not onlyin respect of its language and formulae but also in respect of its thematiccontent.

il. 13.620–5

��T�� 3�� 4� �� A�� #��W�� S��1�� o��,P�� '�� >�� 5 �� ( 7��#�0�`� 7'� ��>: 3� � +��, �(�� 3#'�;��.� 7��>��'�� V� 7�� '��[��#, /� �� @ Z''� ��13�� :�2

Although ancient scholars apparently did not throw doubt on itsauthenticity, in modern times Menelaus’ speech in Iliad has drawnmuch negative attention. August Fick, Walter Leaf and G. P. Shipp, amongothers, thought it interpolated, whereas Bernard Fenik went so far as to callit ‘Menelaos’ unhappy excursus’ and ‘a particularly unsuccessful example of

Cf. Chantraine (: ): ‘Cet exemple enseigne que la langue epique a admis de formules de typenettement ionien.’ On the formulaic lateness of the expression see also Hoekstra (: n. ).

Troy and/or Ilios is usually characterized, in a rather banal way, as �(��'�� (×) or �? ��'��(×) and, more specifically, as �Z�� (×), �(�#�4#� (×), and �(�� (×).

It seems significant in this connection that after Homer the theme of Troy’s wealth becomesincreasingly popular. See e.g. Aesch. Ag. –, Eur. Tro. –. Cf. below, n. .




the “expansion technique”’ (: ). Janko seems to be the only one tohold that ‘[m]ost critics misjudge this speech’ and that ‘[i]n fact it weavesseveral motifs in an integrated whole’ (: ).

Shipp wrote about the passage under discussion: ‘– have a strongmoralizing flavour, and the linguistic contrast with the narrative itself isvery marked.’ And indeed, the opening lines of Menelaus’ speech containno less than four late linguistic features: contraction in 7��#�0� at ,monosyllabic quantitative metathesis in 7��>��'�� at , plus twoadditional recent Ionisms – �� at and ;�� at . Since thesefeatures have often been commented upon, most notably by Shipp andJanko, in what follows I will concentrate mainly on the formulaic dictionof the passage.

Il. .– contains four uncommon or linguistically late formulae:. �� A�� #�� at . A�� #�� as such is

of course a standard Homeric formula, which occurs ten times in theIliad. However, its combination with ��, resulting in the emergenceof a new expression for ‘ships’ in the second half of the verse, is unique.Comparison with two equivalent expressions emerging in the Odyssey,�� #��#� at . and �� 1��#� at .

and ., both in direct speech, shows that we have here a seriesof attempts at making use of the recent Ionism �� in order to createa formula for ‘ships’ fitting into the second half of the verse. As weshall see immediately, the fact that none of these expressions became astandard Homeric formula is of considerable chronological importance.But first let us discuss another linguistically late formula occurring inthis passage.

. ;��.� 7��>��'�� at . I can hardly improve on Janko’s character-ization of this expression: ‘;��.� 7��>��'�� is trebly novel: -�� and;. are recent Ionisms, and the phrase is a unique variation on ;�9� ST�6>��'�� (× Hom.), duplicating ;��.� 7��4��+��# (.).’ Janko’scomment on the latter is also relevant: ‘As we can expect of a recentcreation, two equally new equivalent phrases appear elsewhere, ;. 7��6>��'�� (.) and ;. 7�� 3�� (Erga )’.

Shipp (: ). Cf. Webster (: ). �� Janko (: ); 7��#�0� (also Il. .) Shipp (: , ), Janko (: , ); ;��.�

7��>��'�� below, n. . The expression �� 1��#� belongs to a passage (Teiresias’ prophecy) which is repeated

verbatim in Odyssey and Odyssey . Cf. Alexanderson (: ). Cf. Janko (: , this vol.), Chantraine (: , –). On Homeric formulae for ships see

also M. Parry (in A. M. Parry b: –), Hoekstra (: –), Alexanderson (). Janko (: , ). On ;. 7��>��'�� see also M. Parry (in A. M. Parry b: ), Chantraine

(: ), Hoekstra (: ), Shipp (: ).



margalit finkelberg

In his discussion of the presence of quantitative metathesis in Homericformulae Hoekstra wrote:

It appears, then, that the evidence for the existence of formulae originally builtupon quantitative methathesis is extremely slight. This strongly suggests after themetathesis had begun to develop in East Ionic, oral composition came to an endso soon that hardly any substantial expression created out of the new materialprovided by the evolution of the spoken dialect had time to attain a formulaicfixity. (: )

This would of course be true also of the other late features under dis-cussion, ;�� and ��. As we have seen, the fact that linguistically lateequivalents of the expressions �� A�� #�� and ;��.� 7��6>��'�� are attested elsewhere in the epic corpus strongly suggests that, asdistinct from U��# �� 4#��':�� discussed above, the expressionsin question had no time to attain a formulaic fixity. Hoekstra’s assertionthat ‘after the metathesis had begun to develop in East Ionic, oral com-position came to an end’ will be discussed in the concluding part of thisessay. At this stage, it is sufficient to emphasize that by all standards both�� A�� #�� and ;��.� 7��>��'�� of Menelaus’ speechshould be associated with the latest layer of epic diction.. �� :� rather than ��0� at is obviously a metrically con-

venient substitute for ��+�. This is the only occurrence in the Iliadof this distinctly Odyssean formula (×). To quote Chantraine’s assess-ment, ‘[i]t is remarkable that this form is found in the parts that do notseem very ancient’.

. Finally, the expression W�� S��1�� at . As distinct from theformulae discussed above, linguistic commentaries on Menelaus’ speechusually pay no attention to this expression. The reason is obvious: on theface of it,W�� S��1�� is an inconspicuous expression, not charac-terized by late features or other linguistic peculiarities. Yet it is not onlyunique in this specific position (three modifications occur elsewhere,see below) but it is also metrically equivalent to W�� S��3#'��, thewell-established Iliadic formula for the Trojans (six times in the sameposition in the verse, of which three are in narrative and three in directspeech, plus one direct-speech formulaic modification). And while itis true that S��1��, ‘arrogant’, and S��3#'��, ‘great-spirited’,though close in meaning, cannot be considered exact synonyms, thisdoes not alter the fact that what figures here is the introduction of an adhoc variation on the standard formula W�� S��3#'��. Significantly,

Chantraine (: , my translation). Cf. Shipp (: ), Janko (: ).




this variation offers an unambiguously negative evaluation of the Tro-jans, which is absent from the formula W�� S��3#'��. This is thereason why Milman Parry, who studied S��1�� as an epithet ofthe Trojans in his L’Epithete traditionnelle, subsumed it under the cate-gory of the so-called ‘particularized’ epithets, emerging ‘when the poetwanted to include an adjective for its sense rather than for its [metrical]convenience’.

The case under discussion seems to fit Griffin’s stylistic approach quitewell. W�� S��1�� in the mouth of Menelaus clearly lends itself tobeing interpreted in terms of characterization – or, as many would say today,‘focalization’ – of the Trojans from the standpoint of Helen’s offendedhusband. It is significant in this connection that the modifications ofW�� S��1��, all of them in Iliad , belong to direct speech andare put in the mouths of such bitter enemies of Troy as Hera, Athenaand Achilles after the death of Patroclus, and that the same would betrue of the other unambiguously negative characterizations of the Trojansencountered in the Iliad. That the situation in the Odyssey is different canbe seen from comparison of the distribution of the adjective S��1��in both epics.

While in the Iliad S��1�� occurs only in direct speech, in theOdyssey it is distributed almost evenly between narrative and speeches. Thisobviously reflects the well-known difference in the ethos of the two poems.Namely, although the Iliad characters often criticize themselves and eachother, its poet consistently avoids judging their behaviour by the standardsof ‘good’ and ‘bad’. This withdrawal of moral judgement, which resultsin the famous impartiality of the Iliad (in the apt formulation of Simone

Cf. Hall (: ): ‘“Great-spirited” (huperthumos), on the other hand, is relatively frequent(seven occurrences) but is not confined to the Trojan side, and always seems to be approximatelysynonymous with megathumos . . . A critic who believes that the Trojans’ epithets portray them assignificantly more arrogant than the Achaeans therefore betrays his or her own pro-Achaean bias,not the poem’s.’ Among those to whom the epithet S��3#'�� is applied in the Iliad are Diomedes(twice), Heracles, Achilles (twice) and Nestor.

M. Parry (in A. M. Parry b: ); see ibid. (): ‘But S��1��, which has the same metricalvalue, clearly shows the particularized meaning given by the translation “arrogant”.’ Cf. Sale (:).

On focalization in Homer see de Jong (, ), Nunlist (). Il. . (Achilles), (Athena), (Hera). Cf. M. Parry (in A. M. Parry b: ). Sale (: , –). Note, however, that in his speech in Iliad Menelaus only says that it

is the sons of Priam rather than the Trojans as such who are S��1�� P�� (.), acharacterization which would obviously square much better with the general tenor of the poem.

Cf. Griffin (: ): ‘No feature of Homeric style is more important than this. The narrator depictsevents in a way which leaves the understanding of their moral significance to the audience – anaudience whose presence is never acknowledged.’



margalit finkelberg

Weil, ‘[i]t is difficult to detect that the poet is Greek and not Trojan’(in Holoka : )), is entirely alien to the Odyssey. The wrongdoersof the Odyssey – Aegisthus, Penelope’s suitors, Odysseus’ companions –are explicitly identified as such by both the poet and the characters: theOdyssean formula '�� S��1�� , appearing after the second-foot caesura no less than nine times, five of them in the narrative, readilycomes to mind in this connection. In view of this, it seems reasonableto suppose that, rather than purely a matter of style, Menelaus’ W��S��1�� reflects the erosion of the impartial attitude to the Trojanscharacteristic of the earlier tradition. The term S>�� , applied to theTrojans later in Menelaus’ speech (–), obviously expresses the sametendency.

Be that as it may, it cannot be denied that the context in which theexpression W�� S��1�� emerges is both linguistically and formu-laically late. With this in mind, let us turn to the title of Zeus as Xenios,attested here for the first time. It is significant in this connection thatboth Pierre Chantraine and Hugh Lloyd-Jones based their interpretationsof justice in the Iliad on this specific passage. Both interpretations werepolemically directed against Dodds’s assertion in The Greeks and the Irra-tional () to the effect that he finds ‘no indication in the narrative ofthe Iliad that Zeus is concerned with justice as such’. Either assessmentis based on solid evidence: on the one hand, Chantraine and Lloyd-Jonesare correct in that Menelaus does appeal to the authority of Zeus Xenios;on the other, Dodds is also correct in claiming that the narrative of theIliad never credits Zeus with this title. What seems especially important,however, is that, like W�� S��1�� discussed above, Zeus’s title asXenios emerges in a context that abounds in late linguistic and formulaicfeatures.

On the generic meaning of this formula see Sale (: ). According to Sale (: –), the negative epithets for the Trojans reflect the inherited ‘Achaean’

attitude: this attitude was changed by Homer who, being sympathetic to the Trojans, only allowedthe negative epithets in the mouths of the characters. Yet, the distribution of the formulae W��S��3#'�� and W�� S��1�� as described above suggests a different chronological picture.See also Hall (: –), on the increasing ‘barbarization’ of the Trojans in post-Homeric poetry.

Chantraine (: –), Lloyd-Jones (: ). Cf. Janko (: –). Dodds (: ); cf. Chantraine (: –), Lloyd-Jones (: ). Another Iliadic passage in

which Zeus’s responsibility for xenoi is implied (Janko : ) is also part of a speech, seeIl. .ff.

Cf. Rutherford (: n. ): ‘The idea that the gods constantly watch for and punish mortalwrongdoers is undoubtedly current – it figures in speeches in the Iliad, as well as in the Odyssey andin Hesiod – but the main narrative of the Iliad presents the gods as capricious and little concernedwith justice.’




il. 6.130–7

�(�� 4,� �(�� A�+�� #<.� ��.� *#��4��V� M�, /� _ 3��0 �� 7��#�� C��H��2/� �� '��'�� A��+ �� 3:�� 8� �@ I4�3�� =# :��2 p �@ q' �� 3+ 3� �'� ��#� S�@ ��1�� *#�+�4�#3��'�� >�#��4�2 A��# �� 1�>�3��+ �3@ ]�.� �, 8', e�� @ S��[��

��2 ��.� 4,� C�� '�� .� r'��.

This is one of the few occasions on which Dionysus makes his appearancein the Homeric poems. To quote G. S. Kirk’s comment on the passageunder discussion, ‘[R]eferences to Dionusos are rare in Homer – in Il. onlyotherwise at . (incidentally to his mother Semele whom Zeus hadloved), and in Od. in relation to Ariadne and Thetis at . and . –and only in contexts which are allusive and incidental. His membership ofthe Olympian pantheon is marginal at this stage’ (: ). In addition,the passage contains two hapax legomena, 3+ 3� at and >�#��4�at , as well as a rare use of S�� and contraction *#�+�4�# at .

Finally, the layer of formulaic expressions in it is remarkably thin.Il. .– contains five expressions that are based on formulaic associ-

ations:. 3��0 �� 7��#�� at . Although the expression is found twice

in Homer, both occurrences belong to the same passage (Il. . and), which makes it unlikely that this is a well-established formula. Atthe same time, the related expressions 7. 3��0�, also in Iliad (), and�� 7��#�� 3�� 7 �� (Od. .) allow us to suggest that 3��0 ��7��#�� is either an underrepresented formula or one in statunascendi.

. �'� ��#� at : a unique expression, probably created by asso-ciation with �+�� '� �� (x Il.).

. ��1�� *#�+�4�# at . This is the only time in the epicsthat the word order attested in the widespread formula % d��1�� (x Il.; cf. also h�� . Il. .) is reversed. Emen-dations have been proposed to avoid contraction in *#�+�4�#.

. �+ �3@ ]�.� �, 8' at . The expression can be compared withS�. 8' 3�� (�� @ C�# � at Il. . (of the Nereids). Cf.also S�. (or 7�) �� 7�+ �� (x Od.).

On the two latter see Shipp (: ); cf. ibid. (). West (–: I. ad loc.): ‘-1��# *#��4�# Heyne (-��4�# Brandreth)’; Shipp (: ):

‘��1��# *#��4�# with Nauck?’ and ibid. ( n. ).



margalit finkelberg

. e�� @ S��[�� at . The hemistich is repeated in fullin Il. ., Thetis receiving Hephaestus thrown down from Olympusby Hera. Since the Hephaestus story is obviously more popular withHomer (cf. Il. .–), it can be suggested that what is being dealtwith is an ad hoc adaptation of this epic theme.

It is true of course that C�� '�� at , which occurs two additionaltimes in the Iliad, can with right be considered a formula and that thenoun-epithet combination ��.� *#��4�� at is built on a well-established formulaic pattern (cf. . A��':�� x Il.). On the whole,however, it cannot be denied that, as far as Homer’s formulaic diction isconcerned, Il. .– is almost devoid of formulae. When taken againstthis background, the efforts to eliminate contraction in *#�+�4�# at

should be recognized as misguided: the linguistic, the formulaic and thethematic evidence concur to create a consistent picture of the lateness ofthis passage.

To be sure, in so far as one’s methodological position demands treatingany passage containing linguistic and other deviations as issuing frominterpolation, there is good reason to discard the passages discussed in thissection as late additions. It seems, however, that although the concentrationof late and deviating features in Il. .–, .– and .– isperhaps somewhat above the average, they highlight what we find in otherHomeric speeches as well. It follows, then, that to the degree that thepassages discussed are representative of Homeric speeches as a whole, theconsistent application of the approach in question would, as observedabove, lead to assessing some per cent of the text of Homer as resultingfrom interpolation. Needless to say, this can hardly be regarded as a realisticsolution. On the other hand, as I hope to have shown, the linguistic andformulaic lateness of the speeches cannot be ignored or explained away asa purely stylistic phenomenon. Accordingly, my next task is to suggest analternative way in which the status of the speeches may be approached.

It is generally recognized today that before they were fixed in writing,Greek epics about the Trojan War had circulated for centuries in oraltradition. It is also generally recognized that this tradition was permanentlyin a state of flux. Not only is the language of Homer a Kunstspracheconsisting of different historical layers of Greek language, but it has alsobeen shown, above all in the studies of Hoekstra and Hainsworth, that

As well as other, similar attempts; cf. Rudolf Wachter’s caveat in his contribution to this volume(pp. f.).




Homer’s formulaic idiom too is characterized by a high degree of flexibilityand adaptation, and therefore should also be approached diachronically.The same mixture of different historical periods can also be found inHomer’s depictions of material culture, social institutions, moral valuesand religious beliefs.

Like *#��4�� and *#�+�4�# within one and the same passage, theold and the new exist side by side in Homer, for the simple reason thateach successive generation of poets retold anew what had been bequeathedto them by their tradition. Since the traditional subjects dealing with theHeroic Age were not only universally known but also accepted as historicaltruth, the poets were not allowed to mould them in a free and independentmanner: the Trojan War will end with Trojan rather than Achaean defeat,Hector will be killed by Achilles and not vice versa, and so on. This is whydissonances between the plot of the poems and what is expressed in thespeeches are so important: while the plot is fixed in tradition, the contentof the speeches is not; accordingly, the speeches are fit to express not onlythe opinions of the characters but also the poets’ attitude to what theyreceived from their tradition.

As I have argued elsewhere (Finkelberg b), the double perspectivethus adopted would often result in one and the same episode beingsimultaneously delivered from two points of view, the traditional and thepoet’s own. It comes as no surprise that the latter would as a rule expressthe attitudes of the poet’s own time. Owing to Homer’s extensive use ofdirect speech, it was possible to incorporate these attitudes into the textof the poems without changing their plots. And although this would ofcourse be true of other elements of the poems too, by their very naturethe speeches would present the most appropriate vehicle for furnishingthe traditional diction with elements of the world to which the poet andhis audience belonged.

It is no accident, therefore, that Homer’s non-formulaic and metricallyfaulty expressions, linguistic innovations and the like are concentratedin direct speech. This does not mean that the contexts in which non-standard expressions and ideas occur were much less prominent in epicpoetry before Homer. Though everything suggests that expressions of thiskind belong to the latest layer of the Homeric poems, this does not entailthat epic diction before Homer contained no speeches or consisted mostly

The poet’s famous comment on the inequality of exchange in the Glaucus–Diomedes episode inIliad provides a good example. Cf. Seaford (: ): ‘the implicit criticism of the increasinglydangerous institution of gift-exchange from the new perspective of commodity-exchange, in whichinequality is more surprising’.



margalit finkelberg

of standard scenes cast in traditional formulae. Hoekstra’s observation tothe effect that formulaic introductions to Homer’s speeches indicate that‘even at very early stages dialogue existed alongside narrative’ saves memany words here. Yet it is reasonable to suppose that while the formulaewere preserved in the stock of traditional expressions, the non-formulaicand irregular expressions were ephemeral creations that varied from onepoet to another. In other words, even if epic poets before Homer alsocomposed long speeches abounding in non-traditional expressions whichreflected the attitude of their own times, these were not likely to survive.Janko’s formulation, although it does not address the speeches as such, isappropriate here:

In a tradition that consists of oral improvisation rather than accurate memorisation,it is inevitable that, in those fields where the tradition hands down no ready-madediction, the improviser will draw on the only other diction he knows, that of hisvernacular.

But how late can these late features of the Homeric epics be? In Hoek-stra’s words quoted above, the expressions that did not attain a formulaicfixity – and it should be kept in mind that this includes the overwhelmingmajority of linguistically late expressions in Homer – should be seen assimultaneous with the end of oral composition. Hoekstra appears to implythat at the moment when the Homeric poems were composed the epiclanguage became frozen and the oral tradition came to an end. But surely,as neo-Ionisms and similar phenomena strongly suggest, oral tradition assuch did not die out with the fixation of the Homeric poems in writing.

This is why I see the now classical formulation by Adam Parry to the effectthat ‘the name “Homer” . . . must be reserved for the poet who composedthe Iliad at the time when it was put into writing’ (: ) as providinga more satisfactory explanation for the phenomena discussed.

That is to say, if the epic language indeed ceased to develop, this hap-pened only in respect of those individual products of oral tradition thatwere fixed in writing. At that time, epic diction already presented an amal-gam, or rather something on a par with an archaeological site whose upperstratum represents the last stage of the site’s continuous occupation. Andsince, as I hope to have shown, Homeric speeches by definition belong to

Hoekstra (: n. ); cf. Finkelberg (: n. ). On Homeric speech introductions seeEdwards ().

Janko (: ); cf. ibid. (). Cf. Foley (: ): ‘There is no reason to believe that, even if the Iliad and Odyssey that have

reached us were fixed in the sixth century B.C., oral composition of the Homeric ilk immediatelyceased.’ On neo-Ionisms see West (a: , , –).




this upper stratum, they were also most prone to preserve the realia, thelanguage, the institutions and the mores as they existed at the momentwhen the poems were fixed in writing. As far as I can see, this wouldaccount satisfactorily enough for the peculiarities of the speeches as dis-cussed in this essay. Indeed, it seems not to be a mere coincidence thatthe linguistic and formulaic irregularities observed above occur in thoseparts of the Homeric text which are innovatory in content: Dionysus andZeus Xenios, as well as the fabulous wealth of Troy and the unambiguouslynegative evaluation of the Trojans themselves readily come to mind in thisconnection. If this is correct, then linguistic and formulaic irregularitiesemerging in the speeches should be regarded as by-products of the lastpoet’s intervention in the traditional idiom.

To recapitulate, although the speeches undeniably form an inseparablepart of Homeric diction, they belong to its latest stratum, the one thatcoincided with the fixation of the Homeric poems in writing. It follows,then, that it would be dangerous to base a comprehensive theory regardingHomer on the material of the speeches alone. To return to a passagediscussed above, Diomedes’ use of the Dionysus myth or Menelaus’ appealto Zeus Xenios cannot be seen as adequately representing Homer’s religiousbeliefs in their entirety, but, rather, as representing such beliefs that, likemonosyllabic quantitative metathesis, only became available at the lateststage in the poems’ history. In view of this, it would be welcome if Homericdictionaries began to specify the types of context in which a given word orexpression normally appears in Homer, for Homeric words and expressionsfound in the narrative are not necessarily identical to those found in thespeeches, and vice versa.

To my knowledge, the only database that allows this kind of search is The Chicago Homer (Kahaneand Mueller).



chapter 5


Dag T. T. Haug

1 the phenomenon and its place in the epic kunstsprache

It is a characteristic feature of Greek epic diction that compound verbs,which would be integral words in classical Greek, can be split by so-calledtmesis. Consider the following example:() �.� � -�� .� '83�� C�� (Il. .)It is obvious that in this line, �� does not form a prepositional phrasewith '83��. Rather, it belongs with C�� and forms what we from thestandpoint of classical Greek usage would call a compound verb. Andstill, this compound verb is discontinuous, because the poet has chosento separate the verb and the particle. We can tell that the poet chosethis version and did not simply follow the constraints of the metre, sinceit would be easy to compose the same line without separating ��.� andC��:() ∗�.� � -�� '83�� It could even be said that version () is prosodically preferable, since Il.. as given by our manuscripts violates Wernicke’s law in having afourth biceps which is long by position. But of course, it is likely that Ionic�� has here replaced Aeolic ��, and that the verse had a resolvedfourth biceps in the original version.

Besides illustrating tmesis, this example shows the importance of prefer-ence hierarchies in the study of the Homeric dialects. As we know, Homerprefers Ionic forms to Aeolic ones, except when the latter are metricallynecessary. Thus, the text has the Ionic form �� which gives a correct,

Diachronically, however, it is not impossible that �� belonged to '83�� as an object predicativecomplement, the whole sentence meaning ‘he said the word forth’. But such an analysis is notpossible in historical Greek. For the analysis of ‘place words’ in Homer, see Haug ().

A terminological note: in this article, the two words will be called the ‘verb’ (C�� in example ) andthe ‘particle’ (��). Note that I will use ‘particle’ as a cover term for preverb, postverb, preposition,postposition and adverb – that is, for words such as ��, irrespective of what function they have.

See Janko ().




though slightly irregular, hexameter. The mere coexistence of Ionic ��and Aeolic �� (as well as numerous other dialectal doublets) in Homeris less important than the fact that there is a preference hierarchy whichwe can give a historical interpretation in terms of phases in the evolutionof the epic diction: �� is preferred whenever the metre allows it, so it isthe form of the later generation of poets.

Is it possible to establish the position of tmesis in such a hierarchy?Consider Od. . (the formula is repeated in Od. . and .):() �� @ P� ��4�� '��, 7� �@ �)�� C��#�� (Od. .)As it stands, this line has tmesis, but it does not respect the digamma of�)��, which is otherwise a high priority: it is respected about times andignored about times according to Chantraine (: ). It would havebeen easy, though, to compose this line as a metrically correct hexameterwithout tmesis and respecting digamma:() ∗�� @ P� ��4�� '��, 7��#� �� (G)�)��In other words, it seems as if the poet has voluntarily produced tmesis,at the cost of violating digamma. This shows that even at a late stage ofthe epic diction, after poets had started neglecting digamma, tmesis wassought after and deliberately used by epic poets. If we believe that preferencehierarchies were mechanically applied by poets in the epic tradition, linessuch as () must mean that tmesis was generally preferred to digamma.However, we will see that the question is perhaps not so simple, and itmight not be coincidental that () comes from the Odyssey, as this poem’spoet seems particularly fond of tmesis.

() and () are just two examples, but there are many more where thepoet has used tmesis even where he did not have to. We can concludethat while tmesis might be handy for composing hexameters, it is notconditioned solely by metrical factors. Forms with tmesis are similar to theIonic forms, which are allowed even when the metre does not enforce theiruse, and different from the Aeolic forms, which are only permitted whentheir metrical structure differs from that of their Ionic counterparts.

Most scholars would agree that epic poets tended to replace archaismsin their language with vernacular forms whenever possible. Underlyingthe stylometric approach of Janko (), there is even a claim that thereplacement proceeds at a more or less constant rate which allows us to drawchronological conclusions. But Janko’s research is based on morphologicalfeatures rather than syntactical ones, and as we shall see, there are a numberof differences between the two areas. Replacing unfamiliar desinences byfamiliar ones in a correct way is an altogether easier affair than changingthe syntactic rules of your language. And so, even if it is clear that the



dag t. t. haug

last poets, far from trying to avoid tmesis, actually sought it, we shouldnot jump too lightly to the conclusion that tmesis was a feature of theirvernacular.

2 the historical linguistic background of tmesis

There can be no doubt that the freedom of word order which allows theseparation of preverb and verb is inherited. The same phenomenon is foundin the earliest texts of several other IE language groups, and so we mustassume that at one point in the history of Greek, the later preverbs hadfreer positions within the sentence. The term tmesis (‘separation’) wouldbe meaningless as a description of this phenomenon: instead it seems as ifthe particle and the verb had not yet coalesced in Homer’s vernacular.

On the other hand, there can be no doubt that in Hellenistic poetry, forexample, what we witness is really tmesis: a dividing in two of a synchronicunity. In Moschus (Europa ), we read:() �# �'��:� �� '��s �, 1�� '�sAll scholars would agree that Moschus is here not following the syntacticrules of his own vernacular, but rather imitating a feature of the epic style.However, if we take Homer’s language as the norm, Moschus has made amistake: while Homer does in some cases put the particle after the verb, healways lets it follow immediately upon the verb in such cases. This rule isbroken by Moschus, who lets a dative adjective intervene between the verband the particle. At some stage, then, it seems as if the phenomenon oftmesis was interpreted as a license to cut up any compound verb and placethe results where you want in the metre. On its way to becoming a stylisticfeature, tmesis became mannered. We see the end point of the evolution insuch monstrosities as saxo cere comminuit brum, transmitted under Ennius’name. In general, archaic Latin style seems to have favoured such ‘wild’tmesis, witness sub uos placo (for uos supplico, in an archaic prayer cited byFestus). Since supplicare is a denominative from supplex, it is very unlikelythat this reflects a stage of Latin where tmesis was still allowed. ArchaicLatin poetry belongs to a very different context from that of early Greekhexameter poetry, of course; but the point is just that at some time, tmesisceased to obey any linguistic constraints.

Now before we interpret the frequency and distribution of tmesis inearly Greek epic, it is important to know what kind of phenomenon weare dealing with at this stage. Does tmesis in early epic poetry reflect thevernacular of the poets? Do our texts reflect syntactic change in progress?Or are we rather dealing with a stylistic feature without any basis in thespoken language, even at this early stage of the tradition?




Scholars found little room for doubt before the decipherment of LinearB. Wolf concluded with a ‘nondum coaluisse’. Pierson, in his major studyon tmesis (), said that tmesis in Homer is not figure, but idiom. Wacker-nagel (: ) was convinced that early Greek shows a real independenceof the preverb.

But since , many scholars have come to think that the fusion ofpreverb and verb had taken place already in Mycenaean, because there areno cases of tmesis in the Linear B texts. This view was strongly argued byHorrocks () and supported by Morpurgo Davies (). Since then ithas figured as a stock argument for the pre-Mycenaean roots of the Greekepic tradition. But the Mycenaean evidence is far from compelling: thereare few finite verbs in Mycenaean, and only nine of them are compoundverbs. Moreover, the most frequent type of tmesis in Homer is that wherethe preverb precedes the direct object (the type �, ��# ��# ).This is only possible with transitive verbs – and only four of those ninefinite compound verbs in Myceanaen are transitive. That none of theseshow tmesis could be due to chance, or even have a linguistic reason inthe way information is presented in the Mycenaean texts, as I have arguedelsewhere (Haug : –). At any rate, the evidence is too weak to allowthe conclusion that tmesis was disallowed by the syntactic rules of Greekin the Bronze Age.

We should not be surprised that we cannot extract convincing negativeevidence from such a small corpus as the Mycenaean one. Indeed, it is notaltogether easy to establish the status of tmesis even in much later times: wefind things like ��. 4,� b��8'� (Nub. ) in Aristophanes; Herodotusalso has a number of cases of tmesis, but always with only an enclitic elementintervening between particle and verb; and similar examples are found inother Ionic authors such as Hipponax and Hippocrates. Wackernagel wasno doubt right in claiming that this limited kind of tmesis by a cliticremained possible for a long time, especially in Ionic, but ‘long-distancetmesis’, where a full prosodic word intervenes between the preverb andthe verb, such as can be found in the tragedians, is at that time a poeticartifice.

So all we know is that tmesis disappeared from Greek some time betweenProto-Indo-European and the classical period. But an interesting fact aboutparticles in Homer allows us to fix the moment more precisely. Whereasthere are only cases of ��/7� in tmesis, we find cases of 7�(�) intmesis. We know that 7� and ��/7� are related historically: Common Greekinherited from Proto-Indo-European a local particle ∗en ‘in’ whose Latin

In fact, only 7� occurs in tmesis in Allen’s editio maior, which is the one used by TLG.



dag t. t. haug

cognate in has preserved the original syntax: as a preposition with accusativeit has directional meaning, and with the dative it has locational meaning.

In what we could call the ‘peripheral’ Greek dialects – Arcado-Cypriote,Thessalian, Boeotian and Northwest Greek – 7� has preserved this syntax.Ionic-Attic, Lesbian and Doric, on the other hand, innovated a form ∗en-s(with various later dialectal developments yielding 7� and ��) which aloneis used with the accusative in the directional meaning, leaving the old form∗en with the locational meaning and the dative case only. Thus, the twomeanings were disambiguated by the form of the preposition, and not bycase alone.

That tmesis is so much more common with 7�(�) than with ��/7� wouldsuggest that, as a linguistic phenomenon, tmesis belongs to the periodbefore �� was created, or at least before ��/7� became usable in the epicdialect. This is corroborated by the fact that 7� in tmesis can have thedirectional meaning that would normally demand �� or 7� in a prepositionalphrase in Homer. Consider the following example from Od. .:() 7� �� , '�� >�� 7>: '��The force of the particle is clearly directional: Odysseus and his companionsmade the sheep go onto the ships. Still, the poet uses 7�. It should also benoted that even in Attic, compound verbs show a clear preference for 7�instead of �� (Chantraine ). So it is likely that the univerbation ofparticle and verb happened before the creation of ��/7� and therefore thatthe poets of the last generation before Homer, who no doubt had ��/7� intheir vernacular, did not have long-distance tmesis.

A closer study of the cases where �� is used in tmesis further supportsthis conclusion. In out of the cases, �� is separated from its verb onlyby the particle ��. As we have seen, this is exactly the kind of tmesis thatis possible even in later Ionic. With two exceptions (Il. . and .),then, this mild kind of tmesis is the only one allowed with ��, whereas 7�allows for long-distance tmesis and has retained its old semantics in suchconstructions. We can therefore retain the conclusion that long-distancetmesis had disappeared from the spoken language in Homer’s time.

3 defining tmesis

This means that in examining tmesis in the epic tradition, we are forthe most part dealing with a defunct phenomenon, an archaism like thegenitive in -� or the short vowel subjunctives. Still, as we saw in thebeginning, Homer uses tmesis even where it is not metrically necessary.From this there is one conclusion to draw: not only is tmesis no longer apossible construction in the vernacular – it has also become a figure that




the poets use not only for their metrical needs, but for stylistic reasons. Assuch, it probably reflects the individual choices of poets rather than theevolution of a whole tradition, which is why a study of the frequency oftmesis is likely to be an interesting complement to statistic study of otherfeatures whose use and replacement seems to have been more mechanical,such as the genitives in -� and -��.

But studying the frequency of tmesis is also more difficult than studyingdifferent genitive endings, since it is harder to define what we are lookingfor. For the study presented here, the data were gathered in the followingway: First, all occurrences of words that are liable to function as preverbs, inall their forms, were retrieved from the TLG texts of the Homeric poems,the Theogony, the Works and Days, the Shield of Heracles, and the hymnsto Demeter, Apollo, Hermes and Aphrodite. Examples from the Hymn toApollo were divided into the two well-known groups. It was my intentionto divide examples from the Theogony into two groups as well, before andafter line . But as it turned out, there are no cases of tmesis in thedisputed latter part of the poem.

This search yielded a large number of attestations where particlesappeared separated from the verb of the sentence and so potentially intmesis. On the other hand, there is also a possibility that the particle func-tions as a pre- or postposition – and this is by far the most frequent case.Finally, the particle can also be an independent adverbial. The followingexamples show the possibilities:() Preposition

�(�� @ 7� m3�� 7��>�� >��.� 7��: ��@ (Il. .–)

() Postposition��8 4,� *#�� O��3� C�� :�� (Il. .)

() Preverb in tmesis�� @ ��@ t�� J��'�� (Il. .)

() ‘Postverb’ in tmesisu&1�@ �p ��9� 7��H�� @ C�� '�'�� (Il. .)

() Adverbial4�� 3g� (Il. .)

Examples ()–() are quite easy to categorize, but this is not always thecase. In particular the boundary between adverb and verbal particle can beunclear. Consider Il. .–:() 7�� @ 53�� )�� 0>�· (Il. .–)

Here is the complete list: �'1�, �'1@, ��, ��@, ��, ��, ��@, �1@, ��, ��@, ��, 7�, 7, 7[, 7�, 7��, 7��,7�@, 71@, ��, �@, �, 3@, '��, '��@, '�3@, ��, ��, ��@, ��, ��, ��, ��, ��, +�,[+�, S��, S��, S�@, S1@.



dag t. t. haug

Looking at this line in isolation, it is tempting to conclude that 7�� is herea free adverbial. But in Od. . a compound verb 7��>�� appears, andIl. .– could have the same verb with tmesis. Alternatively, one couldargue that even free adverbials could appear to the left of their verbs andthat the compound form 7��>�� is due to later univerbation. But suchspeculation is outside the scope of this study.

Instead, the approach adopted here was to count tmesis and free adver-bials together, even if they are traditionally kept apart in linguistic analysesof Homer’s language. The rationale behind this choice was that even ifthe two phenomena are not the same linguistically speaking, they are bothforeign to the classical norm and we cannot even be sure that a poet in clas-sical times, who had neither construction in his vernacular, would be ableto keep the two apart. Both were manifestations of a stylistic trait properto epic and other higher poetic genres. ()–() may illustrate phenomenawhich are different from a linguistic point of view, but similar from thepoet’s perspective.

The only problem was thus to decide for every case whether the particlewas an adposition (pre- or postposition) or not. In some cases, the choicehas already been made by the editor, cf. Il. .:() �� C�@ 7 �+�� (West’s apparatus: C�@ Aa: 7�@ Hdn Ac G: 7�� -

Z v∗)If the editor had chosen the text 7�� +��, the example would noteven have turned up in the query. But fortunately, it is without importancefor our research, since on either analysis there is no tmesis here.

The difficult cases, then, are the ones where the particle is separatedfrom the verb, the particle and the verb constitute an attested or possiblecompound verb, but the particle also appears next to a noun that it couldgovern:() �'1� �@ P�@ w'�� >�� [�1�� 4#��2 (Il. .)There exists a compound verb �'1�>��, so Il. . could be countedas tmesis, but on the other hand, it is also possible that �'1� governsw'�� . In such cases, my approach was to count every construction asa prepositional phrase if the particle stands next to or is only separated byclitics from a noun which it could govern according to the syntactic andsemantic rules of epic Greek. Here is another illustrative example:(a) 3�, �@ 7� ��'� �� (Od. .)(b) ;��.� �@ 7� ��' � �� (Hes. Th. )(a) is counted as tmesis, since 7� cannot govern the accusative in Homer,nor in any other Greek dialect when no motion is involved. On the otherhand, the very similar example (b) is not counted as tmesis, since here the




Table . Frequency of tmesis in early hexameter poetry

Work Cases of tmesis Freq. per lines

Iliad ,

Odyssey ,Theogony – ,

Theogony –

Theogony – ,

Works and Days ,Shield of Heracles ,To Demeter ,

To Ap. – ,

To Ap. – ,To Hermes ,

To Aphrodite ,

analysis of 7� ��' � as a prepositional phrase is possible. According tothe principle set out above, this analysis is preferred and the constructiondoes not count as tmesis.

4 results

On the definition above, tmesis distributes in the corpus of early Greekhexameter poetry as portrayed in Table .. The frequencies that we observein the hymns vary considerably, and are in fact not statistically significant.In the Shield of Heracles, the frequency is extremely high, . cases of tmesisper lines. This is so because the poems contain many cases of 7� inlines like these:() 7� �� #�� 4�� +�� C � I�� (Asp. )We can also test the significance of the numbers for the hymns on a casewhere there is universal agreement, namely the Hymn to Apollo. Everyoneagrees that this hymn should be split in two parts and the numbers show arather marked difference in the frequency of tmesis. But on closer inspec-tion, the numbers reveal themselves too small to allow for any conclusion.Knowing that the frequency of tmesis in the whole Hymn to Apollo is .

per lines, we can, using the Poisson distribution, find the probabilitythat any given lines will contain exactly cases of tmesis – and infact there is a per cent chance for this, so the result could easily bedue to chance, and the same is true of the other hymns. This means thatwe should concentrate on the four larger works, the Iliad, the Odyssey, theTheogony and the Works and Days. It would perhaps be possible to achieve



dag t. t. haug

significant results even for the shorter texts, if one also counted the numberof compound verbs without tmesis, since this would allow us to study abinomial distribution instead of variation in frequency.

In the four larger epics, Janko () got an impressive correlationbetween the frequencies of the eleven phenomena, though the results forthe a-stem genitives need to be modified (Janko and Jones in this volume).Generally, there is a recurrent pattern where the frequency of the moremodern variant increases slightly from the Iliad to the Odyssey, then thereis a large jump to the Theogony and a further small increase in the Works andDays. But since we have already seen that tmesis is a stylistic feature even atthe earliest known stage of the Greek epic tradition, there is every reason toexpect that the numbers will be different from those of the morphologicalcriteria, and that is what we find indeed.

Given that we seem to be dealing with the choices of individual poets,it is reassuring to see that the Theogony and the Works and Days patternso closely together, since there is almost universal agreement that the sameauthor wrote these two poems. This result is particularly evident if we lookat the result for the whole Theogony, but even if we disregard the last

lines, the figures are rather similar and this can be confirmed by statisticaltests.

And indeed, the numbers on tmesis do support the conclusion that arather large part at the end of the Theogony does not belong to Hesiod.The last example of tmesis is found in line ; there are no examples fromthe last lines of the text. Statistics by themselves cannot show whereto draw the line, but we can make some sample calculations using West’sproposal to put the end of Hesiod’s genuine work at line . The frequencyof tmesis in the whole text of the Theogony is . per hundred lines; usingthe Poisson distribution we can once again calculate the chance that

lines contain no examples of tmesis: in fact, this is a mere . per cent chanceif the distribution is random. In other words, it is highly unlikely that theabsence of tmesis in the last part of the poem is due to chance – thereshould be some underlying cause. On the other hand, statistics cannot tellus what that cause is. But to my mind the data supports the conclusionthat the end of the Theogony does not belong to Hesiod, although moreprecise tools are needed in the search for the exact boundary.

The fact that the results for Hesiod’s usage are uniform stands in contrastto the most interesting finding of the examination, namely that there is alarge difference between the Iliad and the Odyssey. And in fact, it is theOdyssey which is the more archaic poem, with . cases of tmesis per

lines. It would be perverse to claim, on the basis of this evidence, that theOdyssey is the older poem, but there are other consequences.




5 conclusions

It is of course no longer a very controversial claim that the Iliad and theOdyssey were composed by different authors – this is rather a majority viewby now. Furthermore, there is no doubt that the diction of the Odyssey is onthe whole slightly more innovative than that of the Iliad. However, Jankoshowed that the Odyssey is not only more advanced, but also consistentlyso throughout his ten criteria. Be it digamma, quantitative metathesis orephelcustic nu, the Odyssey shows a slight increase in frequency – and asJanko observed, this is obviously compatible with the hypothesis that theIliad and Odyssey are the work of one man whose diction advanced withhis age.

On the other hand, Janko also showed that the change of frequency inthese criteria was rather uniform in the whole tradition; so we are dealingwith a change in the language of a genre, not with the choices of anindividual poet. In this respect, as we have seen, tmesis is different andtherefore, I think, more relevant to the question of authorship.

The numbers are clearly statistically significant. The frequency of tmesisin both Homeric poems taken together is . per lines, which meansthat on average we would expect cases in a corpus of , lines. Againusing the Poisson distribution, we see that there is only a . per centchance of getting examples or more if the distribution is random, sowe should look for an explanation. Once more, statistics cannot tell uswhat the explanation should be; perhaps the poet of the Iliad developeda taste for tmesis before writing the Odyssey. But such a hypothesis seemsless plausible than postulating two authors.

The case of tmesis shows us that stylometrics is not always a certain guideto chronology: what is important for chronology is not the frequencyof single traits, but correlations of several features. The frequency of aphenomenon like tmesis tells us rather less than expected about the relativechronology of texts, and rather more about the phenomenon of tmesisitself. But what frequency tells us about tmesis is interesting in itself,since it shows that stylistic predilections and false archaisms played a rolealso in the earliest Greek hexameter poetry, and that even at this stage,the evolution of the diction cannot be wholly explained as mechanicalmodernization by the singers.



chapter 6


Georg Danek

In the discussion of relative chronology in archaic Greek epic the Doloneiaplays a prominent role. When we try to define the relationship of singleworks without extra-textual evidence, we are limited to observing implicitreferences: intentional or unintentional adoptions of phrasings, citations,allusions. But by this method we can never be sure if what we label a‘model text’ of a quotation is identical with the state of the text whichhas been handed down to us, as the model may as well have been a muchearlier condition of the same text. Thus conclusions concerning relativechronology in archaic Greek epic must remain controversial.

The Doloneia is the only single extended passage within the Iliad whichhas been labelled a ‘late addition’ or ‘not authentic’ by most Homericscholars, starting from the famous note in the scholia. In my doctoralthesis I tried to show that the Doloneia, though referring to the Iliad,and aiming at being an integral part of it, differs significantly from theIliad ’s style and is no part of its larger poetical conception, but a secondaryaddition. Only with regard to the question of ‘authorship’ did I refrainfrom making a strong judgement, for the following reason.

What does talking about ‘authorship’ in an (originally, or predominantly)oral epic tradition mean? The conception of ‘author’ and ‘authenticity’becomes problematic when we talk about an oral tradition in its strongsense, as it was outlined by Albert Lord () following Milman Parry.

For similar discussions concerning the ‘continuation’ of the Odyssey see S. West (), Oswald (),Danek (b: ff.). The case of the unauthenticity has been made recently again by Cantilena(: xxiii–xxix and passim). We will see that, with the end of the Odyssey, things are different fromthe Doloneia.

Schol. T in Il. .: 1 � �V� _T�� S1@ !&':��# �� 3� � 'V �)�� '�� @x��, 7�� # ��3� �� V� �� (They say the rhapsody has been placedseparately by Homer and was no part of the Iliad, but has been placed into the poem at the times ofPisistratus). Cf. Montanari .

Danek (: –). Some scholars took my book as final proof for the unauthenticity of theDoloneia, cf. West (a: –); Cirio () votes for the authenticity once again without discussingmy arguments.




Within the ‘oral period’ of an epic tradition, strictly spoken, epic singerstake over their songs from other singers, change them in the course of theircareer and pass them on in a fluent state. As long as no written fixationtakes place, the only ‘author’ we can identify is the singer of the versionwhich is performed on a single occasion. Under such conditions, the firstsinger who added the lay of the Doloneia to the (still fluid) epic concept ofthe Iliad must be simply called the ‘author’ of this new version as a whole.Any effort to discuss the authenticity of the Doloneia might be questioned,then, with a reference to the oral tradition.

For that reason I want to pose two questions anew: first, how far canwe posit that the Doloneia differs from the style of the Iliad as a whole?And, second, what does this mean for its ‘authorship’ within a cultureon the fringes of orality and literacy? My essay will first follow the threemain chapters of my dissertation, and only afterwards will I try to shednew light on the second question through a comparison with the Bosnianepic tradition: what happens when a singer within a predominantly oraltradition adds a part of his own to a ‘model text’?

∗ ∗ ∗

In my first chapter I discussed quantitative aspects of the language of theDoloneia. Starting from Richard Janko’s documentation (: –)which shows that the Doloneia linguistically does not differ from the Iliad,I tested several more aspects of formularity which didn’t see significantdepartures from the Iliad as a result, either (Danek : –). Buthere we may restart our scrutiny on a more basic level. Stylometricalstudies betray an individual author’s style by exposing unobtrusive features,such as high-frequency word-types like connectives or particles. This canbe scrutinized much more easily nowadays with the simple help of theMicrosoft Word Search function.

But my new sample controls show disappointing results: in the Doloneia,there are no significant divergencies in the frequency of the most frequentconnectives ��, �, '��, ��,4��, or the particlesP�, �:, �?�. Divergenciesfrom the norm of the Iliad can be found only with expressions that do notoccur often enough to produce statistical significance. For example, I found occurrences of the word C��, while in the Iliad we expect only .occurrences in a book of the same length, or occurrences of the particle^ (or), compared with . calculated occurrences in the rest of the Iliad.

Thus recently Grethlein (: n. ): ‘Es ist fraglich, inwiefern es angesichts der oralen Traditionuberhaupt sinnvoll ist, von einem Autoren der Ilias zu sprechen.’

Close to significant seems the case of ��@ ( times against . times in the Iliad).



georg danek

Now these numbers tell us more about the content of the Doloneia,and less about its author’s style, and offer no proof against its authenticity.I tried to explain already in my book (Danek : –, –) why weshould not expect significant deviations from the Iliad by counting wordfrequencies: for one thing, the poet of the Doloneia very much sticks tothe formula language of the Iliad, using identical idioms and syntacticalclusters because he wants his poem to sound like the Iliad. On the otherhand, the Doloneia is simply too short to admit statistical results of anysignificance. So we might be content with stating that the poet of theDoloneia is well acquainted with the Iliadic formula language and uses itselements to a sufficiently high degree to make any differences disappearwhen they are tested by quantitative analysis only.

∗ ∗ ∗

My second chapter concerned the formula style, properly spoken: in arunning commentary of pages I concentrated on formulations whichI thought to be specific of the Doloneia, trying to define criteria forexploring the ways in which the Doloneia differs from the Iliad ’s ‘normstyle’. I judged as typical for the Doloneia cases where the poet, on theone hand, takes over a formula or formulaic expression from the Iliad andplaces it in an unusual context, so that its traditional connotation getslost; on the other hand I considered cases where he uses an Iliadic formulajust to evoke its traditional context, so that we get the impression of anabbreviating, allusive style. I expose here one passage which I noted onlynow when reading the Doloneia again and which has not been commentedon until now.

We are right in the middle of the Doloneia. Diomedes and Odysseus areon their way across the battlefield, when they face the Trojan spy Dolon.They hide away from the path to cut off his flight back to the Trojan camp.Dolon passes by and proceeds on his way to the ships. The two Greeks waita moment and then start after him. Dolon stops first as he supposes themto be Trojans who want to call him back. But when the Greeks approach herealizes that they are enemies and runs away. The last stage of this processis expressed as follows:

��@ /�� : _@ P�� #�� y � C� ��,4�� _@ P�� B�#�, ��T��, �� 4�+��@ 7��'1�#4�'��2 �� @ )T �� z�':3� �.

(Il. .–)

This part of my book found small resonance. In his commentary, Hainsworth () seldom refersto my critique of the formula style of the Doloneia.




For a precise understanding of these lines we need a verbatim translation:

but well, when they were at a distance of spearshot or even less,well, he recognized the men as his foes, and [at the same time]was [already] pumping his swift knees,in flight; and they immediately hurried up to follow him.

When reading this passage, I was puzzled by the verbal aspect of 7��',because in this context the imperfect tense can only describe the back-ground of what was already going on when Dolon suddenly made hisdiscovery. Literally spoken, the sentence means: ‘He recognized the men –and even before he recognized them, he was already running’, a meaningthat contradicts the logic of the events. What we need here is a phrasemeaning ‘he began to move’ – which can be expressed only through theaorist, not the imperfect.

We can see better what has happened here when we look at the secondinstance of the expression ��T��, �� 4�+��@ 7��', Il. .–:

% d�� @, z� 7�� , L�� '��2 �(�@ P{ C�@ C��?3� '��, b�� +�� , >� �� 1�>�3��.��B�� @ 7��# � �� 0 � ��3��2I|�� $�� 1��, 7�1�� ,_�� 5'� � '��, ��:�� ,

` �� 3@ ��3 1�>�0��, N �@ 744+3�� b[9 ��g��1�@ 7�B ��, J�� J 3#'.� ��4��2f� P{ / 4@ 7''�'g� �39� ��, �� @ % d��0�� W��, ��T��, �� 4�+��@ 7��'.

Here things are described as we expect them to be: Achilles approaches Hec-tor; Hector realizes the situation and starts running away; Achilles takes upthe pursuit. Up to this point, all references are given in the aorist tense. Thenwe get an epic simile, which illustrates start and process of flight and pur-suit, and then this image is projected back to Achilles who, by that time, isfollowing Hector in the imperfect tense, and to Hector who starts his flightin the direction of the walls ( �� , aor.). It is only at this point whenHector’s flight has been well established in the narrative that the expression��T��, �� 4�+��@ 7��' follows. Here the idiom is a descriptive ele-ment which conforms to the verbal aspect of the imperfect tense. The sameis true for the more elaborated expression ��T��, �� 4�+��@7��':

The comment by Cirio (: ) is linguistically imprecise: ‘l’imperfetto mette in risalto l’azionedurativa.’



georg danek

f� ��9� ��T��, �� 4�+��@ 7��'(.)

f� %d�� T��, �� 4�+��@ 7��'(.)

In both cases the expression immediately follows an extended simile whichhas already helped to establish the idea of the start and the resultingstable condition of physical movement. Thus in both cases the expression4�+��@ 7��' must be understood as a descriptive element. Here again,the verbal aspect of the imperfect tense is exactly what we need.

Now, what has happened in the Doloneia? The poet took over the half-line, as it were, as a successful phrasing, but used it just as a metaphor forthe meaning ‘hero X (started to) run away’ and put it in a context where wedo not need a descriptive element in the imperfect tense, but a phrase in theaorist tense, like >� �� 1�>�3�� (.). With the use of the imperfect,we get the impression that the poet left out a short phase of the actionin his narrative (or at least its substitution by a simile) and immediatelypassed over to the description of Dolon’s running style, without telling usthat Dolon first started to run.

It may be just a matter of taste if we call this kind of verbal style incapacity,mannerism or even a postmodernist citation of the traditional Homericformula style. I collected pages of more or less similar examples, whichturn out to be characteristic of the Doloneia. What irritates most readersof the Doloneia is the impression that the poet does not simply use thetraditional formulaic language in a natural way but scatters his preciousfindings of Homeric phrases all over his rhapsody in the way of a cento justto signal how very Homeric his language is.

∗ ∗ ∗

In my third chapter I discussed two aspects of narrative technique which, asI argued, operate on a more conscious level and thus let us better judge thepoet’s intentions. First, I showed that the Doloneia poet not only ignoresthe common principles of constructing speeches and dialogues as they areused in the Iliad, but systematically employs techniques that remind us ofthe composition of speeches in the Odyssey.

I got the same result in my analysis of dressing and arming scenes (Danek: –): while the poet of the Iliad restricts his use of arming scenesin a rigid way in order to construct a poetically meaningful pattern, the

In both cases the beginning of the movement is expressed only in the simile, and not in the mainaction: as in many other cases, the similes here switch over, and in that way substitute, a small partof time which is left out in the narrative of the main action.

Danek (: –), based on Lohmann (: ).




Doloneia poet ignores these patterns and builds up his own pattern ofarming scenes structuring the plot of the Doloneia.

Here I want to discuss a third category, the treatment of time andsimultaneity. Starting from Zielinski (), scholars noticed that Homernever represents actions which happen on different scenes as taking placeat the same time, but one after the other. So the narrative knows onlyforeground actions, while events which happen in the background arereduced to stable situations without change or development. Even ifthe so-called Zielinski’s Law has become the object of debate anew, mostscholars agree about the following facts: (a) the prime narrator never jumpsback in time to recount actions taking place at a strand of the plot whichhas already been introduced in the narrative; (b) when two or morestrands of the plot get announced at a ramification point, these actionsare presented in the narrative as taking place one after the other (Krischer: –). When we take a look at the South Slavic epic tradition, werealize that individual singers within a single tradition may handle similarrestrictions of time management quite differently. So we may concludethat Homer’s consequent adherence to ‘Zielinski’s Law’ was his deliberatechoice to economize his narrative requirements.

The Doloneia systematically differs from the Iliad in this respect, aswe see when we follow the plot. In the beginning, Agamemnon cannotsleep and decides to consult Nestor about what to do. We get a shortdressing scene including a catalogue of clothes. Only afterwards the narratorswitches to Menelaus:

f� �@ Z�� -�� C�� '�� – �(�� 4,� (�� 7�� >��1�� 71�H�� – (Il. .–)

and in the same way Menelaus was frightened, because even he could not sleep.

We are told how Menelaus rises, puts on his clothes (here again with acatalogue) and goes off to wake up Agamemnon. When he joins his brother,we learn that Agamemnon is just putting on his clothes:

This part of the analysis conforms with how Erich Auerbach () commented on the permanentpresence of the Homeric narrator.

The debate has been reopened by Rengakos (). See now de Jong (). Cf. Steinruck (). The Odyssey maybe differs from the Iliad in this respect, cf. Rengakos ();

but see my remarks in Danek (a: –). Danek (a). Cf. Danek (). I made a first suggestion in Danek (a: n. ). Cf. Jens (: ).



georg danek

�.� �@ �}{ �'1@ w'�� 3:'�� C�� (.)

Following this presentation, Menelaus has risen not simultaneously withAgamemnon, as is suggested in the scholia, but even earlier, althoughthe poet gives it the other way round. This conclusion may sound tooscrupulous, but when Agamemnon meets Nestor, he tells him:

�8� �@ 7'�� '��@ 7��4�� '�� 7�� 2(.)

But now, [Menelaus] woke up much earlier than I and joined me.

So, the narrator does not only step back in time, contrary to ‘Zielinski’sLaw’, but almost comments on his violation of the rule, signalling that heenjoys the transgression.

But let us go back to the two brothers’ encounter again. Agamem-non orders Menelaus to call, i.e. to wake up, Ajax and Idomeneus, andannounces that he himself will wake up Nestor and force him to go togetherto the guards outside the camp:

‘ . . . ��@ 53� �8�, t5�� @x��'�� _�'1 3�� , ��2 74g �@ 7�� =� �� 0��)'�, � b��#�� :'��, 5 @ 73�� 7�3�0� 7� 1#�� <��.� �� I�@ 7��0�� . . . ’

(.–)

Here we are given the announcement of a plot ramification with twoseparate strands of actions. Following Zielinski, we are forced to expectthese two actions to be described one after the other. But again, it comesthe other way round:

�.� �@ I'��>��@ C�� >�V� �43.� -��2‘�� 4�� '�� '+3� 7�� I�� +��~?3� '�� '��, ��0 � ��4'��, �� / �� C�3��,M� 3�� '��, @ ?��, 7�V� �? ��0� 7��~’

(Il. .–)

Menelaus presents alternative plot possibilities. He could either ‘stay atplace’ together with the two companions waiting for Agamemnon (who,simultaneously, would arouse Nestor), or come back to Agamemnon afterhaving fulfilled his mission. The second choice might be operated in accor-dance with the usual technique of the Iliad: Menelaus might arouse Ajax

Schol. A ad .a: f� �@ Z�� -�� C�� '��2 �, �.� (�.� ��.� �� 4'�'��. The wording of the phrase is rather imprecise, cf. Danek (: ).




and Idomeneus and come back to Agememnon, who would only then starton his way to Nestor, possibly together with Menelaus. But Agamemnonvotes for the first option:

‘?3� '��, ': �� >��[�'�� :��7��'��2 �� 4,� ��, �� #3�� . . . ’

(.–)

Only now we realize what Agamemnon intended to say in his first speech:Menelaus should accompany the two companions to the guards and waitfor him there. Agamemnon in his reasoning reveals to us that even in thesecond option he would not observe ‘Zielinski’s Law’, but set off for Nestorsimultaneously with Menelaus, so that they would miss each other runningdifferent ways across the camp.

So again, the narrator comments on how the two actions might berepresented in ordinary Homeric narrative, and emphasizes that he willviolate these rules by letting Menelaus’ activities be performed in thebackground, without representing them in the narrative.

The same pattern occurs once again. Agamemnon arouses Nestor, andthey both go to wake up first Odysseus and then Diomedes. WhenDiomedes awakes, Nestor tells him to arouse Ajax the Locrian and Meges.

Diomedes starts on his way, wakes them up and takes them with him:

‘ . . . ��@ 53� �8�, t5�� 9� � m#�� #<.�P� �� 2 9 4�� 7 � ��, �5 '@ 7��.’f� 1�3@, N �@ �'1@ w'�� J� �� ' ��53�� '�4�� , �F�� @ C4��.>� �@ ��, ��9� �@ C�3�� : � P4�� \��.

(.–)

In an implicit way, here again we get two simultaneous actions announced:Agamemnon, Nestor and Odysseus will go the direct way to the guards,

We may compare this with the double assembly of the gods in Odyssey and . Athena announcestwo goals: Hermes shall go to Calypso, and she herself will go to Telemachus; she immediatelystarts for Telemachus and accompanies him on his journey; only when she comes back to Olympusis Hermes finally sent on his way. Similarly in Il. , Zeus announces two goals: Iris shall go to stopPoseidon fighting, and Apollo shall arouse Hector; but only when Poseidon has left the battlefielddoes Zeus finally send Apollo to Hector.

Homer’s different technique may be seen in Iliad : Hector leaves for the city of Troy; while he ison his way, Glaucus and Diomedes meet on the battlefield; when the action on the battlefield hasfallen into a stable pattern, Hector arrives at Troy. Here the audience can rest assured that Hectoron his way to Troy will not do anything which is worth telling.

In this dialogue, too, we are not informed where Diomedes is expected to go afterwards, as Nestordoes not even mention the guards. Both the dialogue and the ensuing narrative are abbreviated andallusive, as they presuppose what the audience already knows.



georg danek

while Diomedes will fetch Ajax and Meges on his way to the same des-tination. Here we cannot distinguish which one the narrator understandsas foreground action and which one as background action, because in thenext sentence all of them already arrive at the guards:

�p �@ /�� V 1#�� 7� �4��'�� C'��3�� . . .(.)

The poet here uses a verse of Iliad , where Menelaus and Odysseus meetwith the Trojan assembly. Hainsworth (: ad .) comments on thewording of ., ‘as if the pickets were at assembly, as the Trojans wereat .’, and criticizes the verse as ‘not very appropriate here, althoughthe commanders are all found together’. In fact, the Doloneia poet wantsto say something different: those who assemble are not the guards but thekings who just come out of the camp, and they do not so much mix withthe guards, but with one another, because the three groups who have actedseparately up to now here come together for the first time. Thus the poetuses an Iliadic formulation in an imprecise manner. But more importantly,here again he points at the fact that the three strands (one foreground andtwo background actions) reunite to form a single strand of narrative.

In the assembly which follows, Diomedes and Odysseus volunteer fora spying mission and arm themselves. The poet describes this in a doublearming scene and lets us understand that the armings of Diomedes andOdysseus take place simultaneously (.–):

f� ��3@ /�� C�� 0 �� 7�+��.W#��B�� '�� '��'�� e� #':��

1� 4�� P'1�� (�. �@ J.� ��, �� )� ��2 �'1� �� < #�� 1��1�� C3��#��, P1�� P��1��, \ �� 0�#[��, _+�� 3�� H��.-�� @ @&�# �� # >�.� I�� 1��

� [�1��, �'1� �� < #�� 1��1�� C3�� . . .

Thrasymedes gives to Diomedes a sword, a shield and a helmet, which isexpressed in the aorist tense; then the narrator switches to Meriones, whogives Odysseus his bow, his quiver and his sword. This action is meantto take place simultaneously with Diomedes’ arming, in the background,as may be judged from the imperfect ��# (). But then the poet

�p �@ /�� V W�� 7� �4��'�� C'��3�� (Il. .). This technique of implicity conceals a symmetrical structure: the three groups consist of three

heroes each (Menelaus, Ajax maior, Idomeneus; Agamemnon, Nestor, Odysseus; Diomedes, Ajaxminor, Meges). But the four heroes who have been called on in the background remain passive asthey are not mentioned again in the narrative.




switches to the aorist tense ( C3��), and tells us what else Merionesgave to Odysseus. So we have two simultaneous actions which are explicitlymarked as taking place at the same time, but which are described one afterthe other.

Homer avoids this pattern, as we can see in the only other instance ofan extended double arming scene, in Iliad : first the narrator tells us inthe aorist tense how Paris puts on his arms one after the other:

(�,� / 4@ �'1@ w'�� 7�+ �� +�� ,�0�� @t��[��, !d�� Io�'��.��'0�� '�� :'� �� C3�� . . .

(.–)

and afterwards we just get the summary remark

f� �@ Z�� -�� :�� C��7 C�#��.(.)

and in the same way, Menelaos was putting on his arms

which means that nothing remarkable needs to be said of his arming becauseit is identical with Paris’ arming (f� �@ Z��) and takes place in thebackground (C�#��, imperfect). Homer avoids describing simultaneousactions even when he signals that they do take place, but prefers to presentthem as foreground and background. The Doloneia poet, on the otherhand, enjoys breaking this rule and puts his finger on his own violationsof the rule.

The parallel actions continue. When Diomedes and Odysseus havestarted on their way across the battlefield towards the Trojan camp, thenarrator switches to the Trojan side where Hector starts an assembly withthe same goals as the Greeks:

�(�� '�� (�� W�� 4:�� 5 �� %d��, ��@ P'#�� : �� #�,/ �� C � W�� 4:�� I�� '��,��9� / 4� #4�� #��V� I��+�� >�#�:�2

(.–)

By now we have learnt that the Doloneia poet likes to step back in timewith his narrative. So the phrasing �(�� '�� (�� makes us assume thatthe Trojan assembly takes place at the same time as the Greek one. Ourassumption is confirmed a little later: when Dolon finally flees from the

So Rengakos (: –), with reference to schol. A b ad .: �(� z� �� 7�� C��[��, �� , ��4'�. �( 4,� ��#3�� ^�� @&�# � ��0 ��9��>�+��#� r %d��, ��, 3@ N� ��.� � r �4'�'��.



georg danek

Greeks, he gets almost as far as the ships. This makes us conclude thathe has started from his post at the same time, at least, as Diomedes andOdysseus; in other words, the poet has stepped back in time, a lasttime.

After all this said, my conclusion is short: the Doloneia poet not onlysystematically violates ‘Zielinski’s Law’, but demonstrates that he does sodeliberately. He differs from the Iliad poet’s style not for lack of poeticmastery, but because he wants to do so. I see no reason why the poetof the Iliad would have taken so much trouble to avoid this in the othertwenty-three books just to do it in one and only one book, again and again.The Doloneia poet is not identical with the poet who is responsible for theuniform style of the Iliad.

∗ ∗ ∗

How may we use this result for evaluating the relationship between the Iliadand the Doloneia? If we were to deal with the work of two authors wholived in a stable literalized society, the case would be simple: the poet of theDoloneia added a text segment of his own to the text of the Iliad which wasthe product of a different author. But how can we be certain that the text ofthe Iliad had already won a stable unchanging shape at the time when theDoloneia was added? Can we be certain that at that time the Iliad was nolonger subject to the process of permanent modification and adaptation,which oralists claim is a typical feature of purely oral traditions? How canwe exclude the assumption that the author of the Iliad we have was justthe singer of its final version which was fixed down in writing – maybe thePisistratean singer who was responsible for the addition of the Doloneia,too? The question, in other words, must be: was the Doloneia added tothe Iliad at a point when the Iliad had already been written down, or ata point when it was still passed on orally and thus subject to continuingchange by the singers?

In order to exclude the ‘strong oralist’ model I will show what suchan ‘addition’ looks like in the predominantly oral South Slavic tradition.What happens if an ‘illiterate singer of tales’ takes over a specific song froma specific identifiable source and makes it part of his personal repertoire,while adding a substantial part of his own? In which ways does the model getadjusted, and in which way does the additional part get integrated into the

The scholiast (.) sees this as proof of the fact that the two assemblies take place simultaneously:�� 4,� � J#��0� #'�� 8�� < �� '��.

See now Colakovic (b), who postulates a substantial difference between ‘traditional’ and‘post-traditional’ singers, but assumes that both types of singers adopt, i.e. change, their modelssignificantly.




narrative, concerning both stylistic integration and narrative adjustmentto its context?

I take my examples from three songs of Avdo Mededovic. In each case,Mededovic uses a source which we can read and control today, but addsa substantial part to the plot. The three sources are of varying characters,but Mededovic’s method of supplementing always remains the same. AsMilman Parry recorded all three songs as a kind of scientific experiment,I will comment on the ‘prehistory’ of the songs and the circumstances oftheir recordings.

Our first example is taken from a song called ‘Beciragic Meho’. Whileworking with Mededovic, Parry made sure that he did not know this specificsong, i.e. that it was not part of his repertoire. Then he asked another goodsinger, Mumin Vlahovljak, to perform this song in Mededovic’s presence.Immediately afterwards he asked Mededovic to sing the same song afterhaving heard it just once. Mededovic agreed and, without preparation,sang the song sticking closely to the plot of his model, but expanding thenarrative from Mumin’s , lines to , lines. He achieved these biggerdimensions chiefly by means of ornamentation and a more intense use ofdialogue, but made no substantial changes to the plot. Only at the end ofthe story, in a long series of scenes, he expanded on events which in hismodel text are not narrated but may be understood as implicit backgroundactions. And he invented an additional character, which forced him toadd a whole strand of actions, consisting in three short scenes which are,however, well interwoven with the main narrative.

As far as sketched out above, all this can be judged from Albert Lord’sThe Singer of Tales, who in an appendix presents a detailed comparison ofthe two versions. Only recently Mededovic’s version was published as anonline edition of the transcript of the original audio tapes, available in thebeautiful handwriting of Parry’s assistant Nikola Vujnovic. Thus I couldread the text of Mededovic’s version and assess, apart from the narrativestructuring, the style of his retelling. Now it springs to one’s eye that thissong as a whole is flat in its wording, and not concentrated in its narrative

For the relevance of Mededovic for Homer studies see now Colakovic (a), with the Introductionof Robert Fowler (vol. .–). Colakovic and Rojc-Colakovic () and Colakovic (a)include the editio prima of six songs of Mededovic, a partial transcript of Parry’s working diaryin Bijelo Polje, and several important discussions on the relationship between Mededovic and histradition.

vv. ,–,. vv. ,–,, ,–,, ,–,. See Lord (: –, –, –). For more background information and discussion of the

poetological problems involved, see now Colakovic (, b). See pds.harvard.edu:/pdx/servlet/pds?id=&n=&buttons=y.



georg danek

structuring. But what concerns the language and narrative technique ofthe added strands, I did not notice any change of style. When reading thepoem, we can sketch out the joints of the added scenes, but the style asa whole remains exactly the same for the whole song – it is, after all, justMededovic’s typical expansive style which distinguishes him from all hiscolleagues.

This may sound like a rather subjective, and irrelevant, judgement ofstyle. But importantly for this analogy, most Homerists are convinced, evenon a single reading of the Doloneia, that they can spot the difference instyle from the Iliad; by contrast, no similar difference of style strikes onewhile reading Mededovic’s text.

Thus, Mededovic integrated Mumin’s version of the song into his ownrepertoire by adapting it to his own characteristic style. Differences of stylebetween the two versions, easily evidenced by the difference of length, arethe result of each singer’s individual style, even within the constraints ofuniformity of the traditional formula language. Regrettably, we cannot givefurther support to this point, as Mumin’s version has not been publishedyet. On the other hand, within Mededovic’s version there is no recognizablestylistic gap. If we did not know that Mededovic added these three scenesto his model, we would not be able to spot them.

Our second example is taken from Mededovic’s most famous song, ‘TheWedding of Smailagic Meho’, which Lord and Bynum published in

together with an English translation. We know that Mededovic used ashis model a printed version that had been reprinted just a few years beforeand which, as he was illiterate, had been read to him by a friend severaltimes. Here, too, Mededovic sticks to the plot of the model, but expandsit from , to , lines by using an extremely expansive style, becauseParry asked him to perform ‘the longest song of his repertoire’. Here, too,he added a completely new scene sequence to his model just at the end ofthe plot, introducing the sultan himself and thus lending the story moresignificance. But here, as well, he integrated the song into his personalstyle so that it is not possible to spot out the additional scenes by using

See Colakovic (), who expands on Mumin Vlahovljak’s severe criticism of Mededovic’s version.To give one example, in the main hero’s long flashback (–,) Mededovic switches from thefirst-person narrative to the third person and returns to the first person only more than one hundredlines later (vv. ,–,). Lord does not mention this and similar flaws.

Mededovic (a). For the obscure archeology of the model text see Semsovic (); for a German translation of

its original version see Grober (). For an even shorter version of lines see Danek (a:–, German translation).

vv. ,–,. Parry recognized Mededovic’s ‘addition’ to his source text, but thought that thesinger had combined, i.e. contaminated, two separate songs.




criteria of style. So we see that for Mededovic it made no difference if helearnt a new song from a colleague’s sung performance or from a printedtext which was read to him.

Our third example is taken from ‘The Wedding of Vlahinjic Alija’, asong which Mededovic performed twice, in a sung and a dictated version.Parry’s experiment consisted in making Mededovic perform the two ver-sions alternately, first the first half of the dictated one, then the completesung one, and then the rest of the dictated one. As a result we can watchthe singer working on his song. David Bynum edited both versions in ,and you may use my German translation of the dictated version.

As is to be expected, here the two versions are even more similar to oneanother than in our first two examples: they not only follow the same plot,but have about the same length and correspond, up to a certain degree,even in the exact wording. Only at the end of the second (dictated)version, Mededovic adds a small detail, which causes him to change thewhole course of this part of the song. Here again the addition is wellenough integrated in the narrative structure, and here again we can see thejoins, but cannot spot any difference in style between the ‘older’ and the‘younger’ portions of the poem.

It is evident that Mededovic in all three cases understood his changesas necessary supplements which helped him to round off the story, to tieup threads of the plot which had been left untied in his models, and tolend more significance to the story. For him, it made no difference if herevised a song he took from a colleague’s performance, or from a printedtext, or from his own repertoire. In each case he first made the song part ofhis own repertoire by making it sound like his own, as far as the languageand narrative technique are concerned. Any change he made should beunderstood as the singer working on his own repertoire, i.e. improving hisown song.

Mededovic (), Danek (a, with comments and additional materials. , and , lines. The numbers for the complete songs are misleading, as for each half of the

song, Mededovic is significantly longer in his second version. Most important is the general difference between singing and dictating style: When dictating a

song, Mededovic was usually more concentrated and precise in both his wording and narrativestructuring, than in a singing performance.

The changes start with v. : the main hero not only abducts his two fiancees from the enemies’capital, but (against all traditional typology) takes together with them the (typical) female innkeeperwho had helped him, in order to save her from the enemies’ vengeance.

For the narrative structuring of this passage, see Danek (a: –). In both versions, Mededovicstruggled with the synchronization of the narrative strands: see Danek (a: –).

We may compare what Colakovic (: –) reports of the excellent singer Murat Kurtagic,whom he met in : ‘When Kurtagic, together with his wife Amira, was our guest in Zagreb



georg danek

Here we may ask if our observations conform to Zlatan Colakovic’stheory of Homer and Avdo Mededovic as ‘post-traditional singers’. Iprefer to work with the distinction between ‘traditional’ and ‘creative’singers on the one hand (with ‘creative’ singers still working within, andusing the traditional techniques, but excelling, outdoing and absorbingthe whole tradition within one or few songs), and the distinction between‘creative’ and ‘re-creative’, i.e. rhapsodic, singers on the other hand (withrhapsodic singers learning their songs more or less by heart, in most casesusing written sources, and keeping closely to their wordings).

I agree with Colakovic that ‘post-traditional [i.e. creative] singers’ haveno high impact on a living tradition, because their pupils in most casesreduce their songs to a ‘normal’ format again. Sung versions by ‘creativesingers’ which exceed the average format survive only if they get fixed, usu-ally in writing. This was the case with the great singer Mehmed Kolakovic,whose highly original and creative songs were collected by Luka Marjanovicin , but left almost no traces in the living tradition, as far as we canjudge from what Milman Parry found in the same region only fifty yearslater.

Thus what Colakovic labels ‘post-traditional’ is just ‘creative’ in my ter-minology, for Homer as well as for Mededovic and Kolakovic. These greatsingers composed epics which surpassed the whole tradition by includ-ing much new material and new ideas in the old songs. Concerning theDoloneia poet, on the other hand, our comparison with Mededovic provesthe point that he was no ‘post-traditional’ singer, but a ‘post-post-traditionalsinger’, at least, as he no longer tried to recompose the plot of the Iliad as awhole in a creative way. He was a rhapsode who learned the Iliad by heart,using a written text, and decided to add to it a new part of his own (evenif he used an old story).

Mededovic’s creative additions to his traditional plots can teach us onefurther lesson. As I already said, the joins between the added parts are easy tospot. Mededovic sometimes abruptly springs from one strand of the actionto another. In the added parts his style sometimes becomes abbreviated,

and we could watch him the whole day, we realized that Kurtagic incessantly worked on his songs.Quietly, totally unintelligibly, he performed parts of the song which he intended to perform. Hemurmured in a low voice, without any cognizance of what was going on around him. Wheneverwe asked him what he was singing about, he would say that he was thinking over parts of the songwhich we intended to record, and that he was “renewing” some parts of it’ (my translation).

Colakovic (, b, more fully expanded throughout a). Marjanovic (–). See Danek (a: –) for a short introduction and one song in German

translation. This holds true for both when Mededovic’s corrects himself and when he ‘corrects’ his colleagues.

But, after all, the experimental context of all three performances left Mededovic no time forpremeditating his ‘new’ versions, as Kurtagic (see n. ) was able to do.




condensed or summarizing. But these style features can occasionally befound in other parts of Mededovic’s, and other singers’, songs too. ThusI am not so much reminded of the style of the Doloneia, but of the‘continuation’ of the Odyssey. There too the pace of presenting the narrativegets more and more accelerated, the causation of the story line is less and lessmotivated, and still we cannot say that these style features are characteristicfor this part of the Odyssey alone. The analogy of Mededovic thus suggests,in my eyes, that the ‘continuation’ was the last addition which the Odysseypoet himself added to his own epic.

With the Doloneia things are different, as we can detect here distinctivemarks of stylistic difference. That means that the Doloneia poet had notmade the Iliad part of his repertoire in the same way as Mededovic haddone with his model texts – otherwise the Iliad would look more likethe Doloneia. He learned the Iliad by heart, in all probability by using awritten text, and did not integrate it into his personal style, but left it asit was while just adding a single passage of his own. He was well awareof the differences in verbal style and narrative intentions, but adopted anironic stance to his master text and did not shrink from showing off thedifference between his text and the Iliad.

It will have become clear by now why the Doloneia is such an importanttest case when we discuss matters of relative chronology in archaic epic:the Doloneia was added to the Iliad when it had already been set downin writing. This must have happened in the early sixth century bc, at thelatest, because otherwise we should have hints at a more precise knowledgeof how it was added. At that time, the Iliad was no longer subject to ‘oral’changes of narrative structure, narrative technique and verbal style (if itever was). For that reason, the Doloneia proves the case that we are entitledto discuss chronological relations between different works in archaic epic,i.e. between texts with a fixed wording, and not just story traditions.

See Danek (b: –, ad Od. .–), Cantilena (: xxiii–xxix).



chapter 7


Stephanie West

Very many of the topics once subsumed under the general heading of theHomeric Question have been completely reformulated, or even abandonedas meaningless, as a result of universal recognition that behind the Homericepics lies a very long tradition of heroic narrative poetry composed to beheard, not read. Nowadays any attempt to distinguish earlier and morerecent strata within the two poems (apart from the two almost certainlater additions, Iliad and the end of the Odyssey (.–.)) willprobably be judged old-fashioned. But if we accept that each of the twoepics is likely to represent the consummation of its composer’s work,the fruits of his experience in telling and revising its various elements,we have to allow that a very long time must have been spent on itscreation – years rather than months – and it is not unreasonable to suggestthat some parts of the epic may belong to an earlier, some to a later, stagein the poet’s development of his theme. Nowadays the preservation of awriter’s drafts and working papers sometimes allows us to trace the genesisof lengthy compositions over many years, and we see how the author’sconception of his work evolved in the course of time. Such study generallyenhances critical appreciation. If we suspect that the Homeric poems maypreserve clues to their own development, it cannot be a waste of time toinvestigate. It must, however, be conceded that the problems encounteredin the course of such inquiry do not encourage optimism about establishingrelative chronology within the wider environment of early Greek hexameterpoetry.

It would be simplistic to suppose that the later books of the Odysseymust necessarily come from the later years of the poet’s work. We mightindeed suspect the opposite. At the heart of the Odyssey lies the tale of theHusband’s Return, a type of story found all over the world and repeatedly

As is well emphasized by Goold (); see also Reece (: –), Heitsch (: ). Cf. Focke (: –) on constructive analysis. Thompson (–: N ); see also Hansen (: –).




gaining fresh life from actual cases; the aftermath of the Second World Warbrought many tragic instances. In the Odyssey this fertile theme has beenenhanced by combination with the motif of a competition between suitorswith a bride as the prize. Odysseus’ home lies further from Troy than thatof any other Greek hero, and to reach it he had to round Cape Malea,the most southerly point of the Peloponnese and a notorious problem forseamen. So it would not be surprising if it took him longer than anyother hero to get home from Troy; but his failure to return a year afterTroy’s fall would raise the suspicion that he had perished on the way, thatPenelope was a widow, and that her remarriage would bring her husbandthe lordship of Ithaca. But the story told in the epic’s latter half does notessentially require the preceding nine years of adventures which for us nowseem to typify the Odyssey, the speciosa miracula, Antiphaten Scyllamque etcum Cyclope Charybdin.

Emulation of the Iliad, I would argue, has led the poet to extend to nearlyten years the period of Odysseus’ nostos, thus bringing to manhood the sonwhom the hero left as a baby and allowing the elaboration of a colourfulnarrative of Odysseus’ adventures on the way home. In Telemachus’ pleaon Phemius’ behalf we should see the fruit of the poet’s own experience:�V� 4,� ��V� '�� 7��# @ P�3�� | \ �� #�� '1�� (Od. .–). While we recognize that epic stylebelongs to a traditional art developed over many centuries, style and sub-ject matter must be distinguished, and there is no compelling reason whythe story of Odysseus’ adventures should be older than the compositionof our Odyssey. Romantic notions of the conservatism of oral poetry andtraditional storytelling should not pass unchallenged. By highlighting inthe proem Odysseus’ experiences on his homeward journey the poet drawsattention to a part of the poem where he displayed his capacity for inde-pendent creation. By contrast, Odysseus’ ultimate reunion with Penelopeand the savage vengeance by which, preferring honour to material com-pensation as the heroic code requires, he reasserts his position as lord ofIthaca were, at least in outline, familiar.

However, we repeatedly see that the poet knew more than one way totell his story. Inconsistencies, misdirections and oddities which to theanalysts appeared to indicate multiple authorship, the reworking of anolder poem (or poems) by a poet (or poets) whose intentions and outlook

Thompson (–: H ). Used in by the Corcyraeans, according to Herodotus (..), to account for their absence from

Salamis. I have learnt much from Georg Danek’s exploration of this topic (b).



stephanie west

differed significantly from the original composer’s, are better interpreted byreference to a reluctance to sacrifice what had proved successful. Concernfor the effectiveness of individual episodes overrides overall consistency, apreference shared with, among others, Aristophanes and Plautus, and themore readily understandable if we remember that episodic performancemust have been the norm, performance of the whole epic, to a stableaudience on a single occasion, extremely rare. It is much too often assumedthat the poet envisaged an audience who would absorb his narrative likea serial story, in a regular sequence of instalments, rather as nowadayswe might listen to a novel read on the radio. But the episode must havebeen the norm (as it is when Phemius and Demodocus perform), and weshould do better to see as a model the kind of familiarity with the Bibleonce enjoyed by regular churchgoers – enough to recognize the majorfigures and to contextualize the lesson, but not encouraging detection ofinconsistencies with other parts of the Bible. The poet’s audience may besupposed sufficiently familiar with the general outline of the story and itsmain characters to provide a context for whatever episode was performedwithout being troubled by any difficulty in reconciling what they heard withsome other part of the narrative. The bard could exploit the assumptionsof an informed audience.

A lack of concern for consistency is perhaps most obvious in the presen-tation of Penelope, where a version of the story in which she is Odysseus’principal ally has not been wholly discarded despite her displacement inthis role by the now adult Telemachus; the result is an unpredictable, enig-matic character, well suited to a major role in a modern novel and offeringvast scope for feminist literary criticism. We need to recognize the influenceof earlier versions of Odysseus’ story if we are to appreciate the poet’s skillin adapting and unifying his materials.

To the modern reader much of books – looks like the stuff of fairytale,and it is customary to regret a deviation from the austere rationality of

I have taken Sir Denys Page’s The Homeric Odyssey (a) as the outstanding anglophone propo-nent of the case against single authorship; his book made easily accessible much that otherwisemight be patronizingly dismissed as obscurantist pedantry. But he was better at demolition than atconstruction, and his own view of the genesis of the Odyssey has found little support.

Burkert (: –) succinctly sets out the weakness of the widespread assumption that theopportunity to hear both epics in full was offered at the Panathenaia: ‘To recite the whole of theIliad alone, not to mention the Odyssey and all the other works still attributed to Homer, would takethirty to forty hours, more than the time available for all the tragedies and comedies at the GreatDionysia, which clearly was the more important literary event in Athens. Rhapsodes could onlyproduce selections from the huge thesaurus that remained in the background.’ The Panathenaic lawof which we hear from Lycurgus (Leoc. ) and Diogenes Laertius (.) was simply meant to ensurea proper order of recitation in the performance of selected episodes.




the Iliad. But the subject matter of these books is better classified astravellers’ tales. Distance in an age when there were numerous blanks onthe map made credible ‘the Anthropophagi and men whose heads do growbeneath their shoulders’ (Shakespeare, Othello I iii), along with Sir JohnMandeville’s vegetable lamb and many other marvels. In Herodotus’ timean adventurous traveller should be prepared to encounter flying and two-headed snakes, griffins, monocular pastoralists, epidemic lycanthropy, andpeople who slept for half a year. Exaggerated reports of the dangers to befaced on the next stage of a journey have often led explorers to turn back;in particular, rumours of cannibalism as a socially approved practice, outof all proportion to its actual occurrence, have been a powerful deterrent.

Odysseus’ encounters in books – test to the utmost his resourcefulnessand endurance, the qualities particularly needed for successful venturesin strange environments, and thus peculiarly appropriate in the age ofcolonization. The storm which drives Odysseus’ contingent westwards andoff the map (.ff.) thus brings about a much more challenging homewardjourney than the realistic itinerary, via Crete, Egypt, and Thesprotia innorth-west Greece, described in the cover-stories which he tells on hisreturn to Ithaca (.ff., .ff., .ff., ff.; cf. Eumaeus’ referenceto a report of Odysseus repairing his ships in Crete (.–)) and bettermatching the proem: �� @ ��3�� 5�� P �� C4��(.). I believe that these cover-stories are the relics of an option whichthe poet discarded in favour of the stranger adventures which constitutebooks –. Menelaus could be given an interesting homeward journeyby way of Crete, Egypt and Phoenicia, while Odysseus is taken to terraincognita.

Within this section book fits oddly. If books , and representsecond thoughts, we have third thoughts in book . Here I cannot avoidrepeating old arguments advanced by the analysts. My purpose is to argue

Cf. Privitera (: –): ‘Cio che oggi, per gli studiosi del folk-lore, e una fiaba, era per Odisseoe per i Feaci l’ignoto, il mondo non frequentato dagli uomini, ma con gli uomini e con gli deisaldamente connesso. Omero non ha mai pensato che Odisseo si fosse perduto nel mondo dellafiaba: si era perduto in un mondo sconosciuto e remoto, ma reale, posto ai margini del mondofrequentato dagli uomini.’

See S. West (a, b: lxxxiii–xc). This approach was pioneered by Woodhouse (: –,–), though he weakened his case by presenting the hypothetical earlier version as historical,‘the real experiences of the real Odysseus on the way home from Troy’. The theory has recentlybeen revived: see Reece (). A brief but stimulating exploration of the topic by Colin Hardie(), published in a Festschrift, failed to attract much attention.

Well set out by Page (a: –). The recent monograph by Tsagarakis (), though it offersseveral useful observations, fails to deal adequately with the most serious problems; see furtherHaubold (), Danek ().



stephanie west

that this episode is not the work of a later composer, but that the featureswhich once fuelled the case for multiple authorship throw some light onthe poet’s procedure in the construction of his long epic.

Obvious problems arise from the loose connection of this episode withthe surrounding narrative. Most serious is the lack of adequate motivation.

Circe responds to Odysseus’ request for leave to go home with the newsthat he must first go to Hades and consult Teiresias, information to whichhe reacts as if it were a death sentence. Only after lengthy instructionsabout the route and ritual procedure (.–) does she give a reason.Teiresias will give essential guidance on the route to be followed home toIthaca: r�.� � '�� +3�# | �� 3@, z� 7�� 7��+ ��3#�� (.–). Theban legend had made Teiresias familiar to theOdyssey’s audience, topographical knowledge is seen as one of the skills tobe expected in a prophet, and it is a sign of Odysseus’ distinction that,though without the advantage of any Theban connection, he will be ableto consult such an authority.

Odysseus now rouses his men, rather brusquely (.–). The narra-tive of their embarkation does not run quite smoothly, being interruptedby Odysseus’ account of the death of Elpenor, who was not very brightor particularly brave, and having gone to sleep on the roof while drunkgot up in a fuddled state and fell off, with fatal consequences (.–).

Odysseus’ surprise when he encounters Elpenor’s ghost awaiting admissionto the underworld (.–) indicates that he had not realized that Elpenorwas missing when he left Circe’s island; what is here related about thecircumstances of the young man’s death is to be understood as learnedlater, though the poet does not make this immediately clear. While we mayacquit Odysseus of neglect of funerary rites, we might judge him negligentin that, with only one ship’s contingent under his command, he failed tonotice that he was short of a man. But Elpenor has an important narra-tive function in anchoring book to its surroundings; his fate eases thetransition from the world of the living to that of the dead and his ghost’s

As with Telemachus’ journey. Cf. .–; Menelaus’ encounter with Proteus (.–) is in many respects comparable. He

is likewise advised by a supernatural female to consult a clairvoyant person, following a difficultprocedure according to her instructions, and receives not only (relatively brief ) travel advice butalso information about his life’s end.

Thus Calchas guided the Greeks to Troy `� ��, '�� +��, �:� �< �� m�0>�� (Il. . –).

On this passage, which looks very much like an afterthought, see further Burkert (: –). A realistic detail; in August the Times reported that fifteen people had died in south-eastern

Turkey after falling off rooftops on which they had been sleeping.




plea for a proper funeral (.–) ensures that Odysseus must return toCirce’s island.

Book itself is a curious synthesis, consisting for the most part of necro-mantic conjuration at the edge of the world, a procedure which bringsbefore Odysseus a long parade of the great dead, but concluding withmaterial appropriate to a conventional katabasis, a descent to the under-world, with a view of the exemplary punishments suffered by the greatsinners. But the demarcation between the two is not altogether clear-cut.Such oscillation between rather different conceptions is not simple care-lessness nor indicative of multiple authorship. The poet skilfully exploits alicence not available to a writer guided by the presumed expectations of areflective reader.

Odysseus’ necromantic ritual attracts the ghosts in swarms. It is perhapsinappropriate to describe anything as ‘normal’ in this eerie procedure, butusually the necromancer addresses his conjuration to an individual anddoes not gain access to other spirits. This feature better suits katabasis thannecromancy. After Elpenor’s ghost Odysseus identifies that of his mother,Anticlea (.–), who, as he now learns, has died since he left Ithaca.But his interview with Teiresias must come before all other considerations.

Teiresias does not need to be told Odysseus’ purpose. He immediatelyhighlights the expectation of danger arising from Poseidon’s anger, the seri-ousness of which Odysseus might be inclined to underestimate after his yearof rest-and-recreation on Circe’s island. But as regards Odysseus’ home-ward journey he concentrates on warning of the dreadful consequences tobe expected if Helios’ cattle on Thrinacia come to harm. We note thatTeiresias does not forecast what will happen; he sets out alternative out-comes. The comrades who assisted in his necromantic sacrifices, Perimedesand Eurylochus (), are now forgotten; the homeward journey would verylikely have been rather different if they too had overheard Teiresias’ solemnwarning, but once Elpenor’s ghost disappears Odysseus is treated as if hewere on his own. However, Teiresias is quite certain that back in IthacaOdysseus will find his wife beset with suitors, who must be killed. After

Some scholars have seen an aetiological significance in Elpenor’s role. ‘The old conjecture ofWilamowitz and Von der Muhll is very tempting that the motif originally had an aetiologicalmeaning and stems from the Argonaut legend; for Elpenor demands a monument in memory ofhimself (l ff.) as if this could be the origin of a future city. The monument motif goes well with asource that provided the motif of the journey to the land of the dead’ (Kullmann c: –).

Considerable caution must therefore be exercised if this text is treated as a source for archaic Greekideas about the afterlife.

Odysseus’ evocation of Teiresias is well discussed by Ogden (: –). The name occurs only here: what is it supposed to mean?



stephanie west

that, Odysseus must undertake a journey overland to propitiate Poseidon.That satisfactorily accomplished, Odysseus may expect a gentle death inold age, among a prosperous people.

We must wish that Odysseus had displayed a little more curiosity aboutthe manner of his death. What is the relationship between what is said hereand the account of his death at the hands of Telegonus given in the Telegonyas summarized by Proclus (W��4�� 7�� H:�� 8 ��.� ��,��>,� �� V� @x3�� '�� V� �� · 7>��3: � �@ @&�# �9�S�. ��8 ��.� ��0�� @ P4��)? This is not easily recon-ciled with 3�� . . . �>��.� '�� 0�� (.–). The poet whetsour curiosity with this enigmatic phrase. Eugammon, the composer of theTelegony, was not, it seems, so overawed by Homer’s authority that he felthis narrative must satisfy Teiresias’ description.

Odysseus receives Teiresias’ instructions with remarkable composure(.), but we may find them surprising, since the start of the poemhad conveyed the impression that once he reached home Odysseus neednot worry any more about Poseidon’s hostility. Zeus has assured us, ashe and Athena plan Odysseus’ homecoming, that Poseidon’s wrath willpass: �� '�3: �� | N� ��· �( '�� 4�� #�: �� | �3�� 3�� 7��'�� )�� (.–, cf. .–);similarly Athena’s reassurance to Telemachus (.–) implies that oncehome Odysseus will not have to face any more travel. With this prospectof further journeyings for his hero the poet stimulates demand for a sequeland this may explain why he did not set before Odysseus’ return to Ithacawhatever measures might be necessary fully to propitiate Poseidon. TheOdyssey is often treated as a story with a happy ending; hence its cliche-isticdescription as the first European novel. But Penelope, after one extendednight of renewed happiness, must wait a long time before her husband isback for good and they can both live happy ever after.

We think of Poseidon as primarily god of the sea, but, as is indicated byhis titles 7�� 3�� and 7�� 4��, he controls earthquakes, a hazardto which the Ionian Isles are particularly exposed. Odysseus could not besupposed immune to the god’s fury provided he avoided the sea. The awfulconsequences of provoking Poseidon’s wrath were peculiarly apparent toOdysseus’ fellow islanders.

Odysseus has now achieved the purpose for which Circe sent him, andmuch more. He has got some guidance regarding his homeward journey

Contrast his reaction to Circe’s revelation that he must journey to Hades (.–). For specific details see Bittlestone et al. (: esp. –, –). Bittlestone argues that Homer’s

Ithaca is Paliki, now western Cephallenia, and that the poet was familiar with the terrain.




(though Circe herself will be able to give him much more detailed directions(.–)). More important, he knows what he must do to make his peacewith Poseidon, a problem which called for a prophet’s peculiar insight. Buthe spends less time in conversation with Teiresias than with his mother’sghost, who had appeared even before Teiresias’. Here I believe that wehave a relic of the more realistic itinerary which brought Odysseus homeby way of Thesprotia, where there was a famous oracle of the dead, towhich, according to Herodotus (.�), Periander, tyrant of Corinth,sent a delegation to consult the ghost of his murdered wife Melissa. ForOdysseus his mother’s ghost, reliable and devoted to his interests, would bean ideal source of information regarding the situation in Ithaca at the timeof her death. Of course, what he learns from her here, some eight yearsbefore he gets home, is of little practical use to him with the epic’s presenttimescale, and Athene must bring him up to date (.ff.). Odysseus’encounter with his mother now serves a different purpose, and in theirmutual affection we see an aspect of the hero previously little in evidence.The restraint which he shows in postponing this reunion until he has talkedwith Teiresias is very much in character. He displays such self-control, ina heightened degree, when he delays making himself known to Penelopeuntil he has disposed of the suitors, and because we have already observedhis behaviour here we shall admire his strength of will and not thinkhim unfeeling. Such recycling of components designed to serve a differentfunction in the narrative is highly characteristic of the Odyssey. We should,however, note that Odysseus ignores what he has just learnt from Teiresiaswhen he asks his mother whether Penelope has remarried (.–); if heis to find her beset by unscrupulous suitors on his return (.–), thequestion is silly.

Consultation of Teiresias provides a pretext for Odysseus to follow theexample of Orpheus and, more importantly, Heracles; the poet more orless acknowledges the latter as his model in Heracles’ speech to Odysseus(.–). But it is a pretext. The procedure which allows Odysseus toconverse with the great Theban prophet is necromantic; Odysseus doesnot need to pass into Hades’ realm to achieve his objective. The lack ofa sufficient motive for visiting the world of the dead is most apparentin what follows Odysseus’ conversation with his mother, the parade of

Cf. Pausanias (..): 4�� e� �� C �� '�� # � P�� 3�� P[�, <�� A�.�7� A�� <��, ��8 3��8 1�4��· ��.� �� U��+�� '�� 7 �� # � ��#'�� '.� ��, _�0 �� U�#�.� �� . %&'�� '�� 0 �8�J��g� C� �� V� P�� '� � �� 7� ��# � �V � �, b��'� ��0��'�0� ��. �� 7� e� �� 3� 3�. See further Hammond (: –).



stephanie west

fourteen noble (or notorious) female ghosts who present themselves beforehim (.–). The transition is awkward, particularly since Anticlea hasurged Odysseus to leave without delay. No explanation is offered for hisfailure to follow her advice. Curiosity apparently gets the better of him.

The women’s stories are told, on the whole, so briefly and allusively thatthe poet must have assumed this material to be familiar to his listeners.

Some relationship between this section and the Hesiodic Catalogue ofWomen is indicated by the similarity between Poseidon’s farewell to Tyro(.–) and Hesiod F; but it is difficult to draw a more definiteconclusion about this apparently random selection. These women haveno particular relevance to Odysseus, though Anticlea’s status is enhancedby association. We are surely meant to admire his ingenuity as a storytellerin thus elaborating material with a particular appeal for Queen Arete, whoresponds appropriately by urging the company to make generous gifts toOdysseus. We may reflect on the skill with which the poet borrowed fromoral hexameter poetry of a rather different type to enrich his narrative withmaterial in keeping with the greater emphasis on women characteristic ofthe Odyssey, the song celebrating Penelope (as Agamemnon’s ghost saw it(.–)).

This intermezzo instructively reflects the reality of the poet’s profession:some skill lay in choosing a suitable point for a break in an extendednarrative, with the suggestion that the performance has gone on longenough, so that a further incentive is appropriate if it is to continue. Thepoet of the Odyssey brings his craft to our notice in a manner quite alien tothe Iliad; here we should observe the form taken by Alcinous’ complimentto Odysseus: �� @ C�� '�� '��1V 7��, C�� 1�� 7 3��, | '83�� @z� /�@ ��.� 7�� '�� [�, | �� 4�� @ (��8:�� #4�� (.–). The blurring of the distinction between truthand verisimilitude should be noted.

The parade of famous women also serves to delay what we are eagerlyexpecting, Odysseus’ reunion with his dead comrades from the TrojanWar. It is rather characteristic of the poet of the Odyssey to postpone whatis expected, a relatively naive way of heightening tension and increasingsuspense, observable at the outset in the delay in bringing Odysseus himself

As with his exploration of Cyclopean territory (cf. .–, –) and his aborted reunion withAjax (.–).

As is implied by Antinous’ allusion to such a tradition (.–); see further Danek (b: ,–).

Osborne (: –) well highlights ‘the very different tone’ of the Odyssey’s catalogue, in particular‘a more general suggestion of female responsibility’; see also Rutherford (: –).

Eumaeus similarly stresses Odysseus’ narrative skill in recommending him to Penelope (.–).




into the narrative and an important element in the latter part of the poemas he postpones revealing himself to Penelope.

In his encounter with the war veterans we are back on more familiarOdyssean ground. The post-war fate of the heroes who fought at Troy is arecurrent theme. The Atreid paradigm, as seen from Olympus, started theaction (.–); now we hear about the affair from Agamemnon’s pointof view (.–). To Odysseus it comes as a shock. Agamemnon dwellson the breach of hospitality at the banquet where the butchery took place,and we now hear of the slaughter of Cassandra at Clytaemestra’s hands(–). There is a peculiar pathos in Agamemnon’s longing for news ofhis son (–), but Odysseus cannot help.

With Achilles’ arrival (.ff.) we note that the poet now envisages akatabasis rather than necromancy; Achilles’ ghost does not need to drink theblood of Odysseus’ sacrifices to recognize and converse with him, and hisgreeting confirms that Odysseus has now penetrated the world of the dead: ��, ��@ C�� '�0H�� 7�� 1�� ': �� C�4��; | �� C�� h�� 3�'��, C�3 �� | �1�� # �, >�� 5�� '��6��; (–). Odysseus’ conversation with Achilles is unforgettable, aboveall for Achilles’ unexpected reaction to Odysseus’ admiring estimate (–) of his status among the ghosts: 'V �: '�� 3�� 4� ��+�, 1��'@@&�# �8. | >�#��'�� @ 7��#�� 7g� 3��#�'�� P��, | �� {��:��, 'V >�� 9� �5�, | y �� +� � �13�'�� (–). Remembering that in the Iliad (.–) Achillesknew he could choose between a short life bringing everlasting renownand a long life in obscurity, his violently angry response sounds like anadmission that he made the wrong choice: Odysseus, the survivor, has hadthe better fate.

Achilles’ language is strangely untraditional, and we might wonderwhether the poet is rather emphatically distancing himself from a rosierpicture of the hero’s fate after death such as was offered in the Aithiopisof Arctinus of Miletus, in which, according to Proclus’ summary, Thetissnatched him from the funeral pyre and conveyed him to Leuce, the WhiteIsle (7 �� #�� e�� .��0� �� V� *�+�� '�H��). It is frustrating that we do not know what Arctinus actuallysaid. The Black Sea’s only proper island, c. km east of the northernmouth of the Danube, and now in Ukrainian territory, was in antiquityknown as Leuce, well qualified for the name by reason of its white cliffs.

Cf. .–, .–, –, .–, .–, .–. Ostrov Zmeinyy, the Island of Serpents. The recently published monograph by Okhotnikov and

Ostroverkhov () includes photos and a plan of the island. A problematic Athenian black-figure



stephanie west

Here Achilles was to enjoy a long-lasting hero cult, under the protectionof the Milesian colony of Olbia. But did the Milesian poet have this realisland in mind, or did he envisage a mythical paradise in the northernsea, comparable to Elysium and the Isles of the Blessed, and subsequentlyidentified with Leuce by the colonists of Berezan and Olbia? Was Achilles’translation to this remote uninhabited island Arctinus’ own invention andthe Aithiopis the inspiration of the cult?

Alcaeus’ reference to Achilles as ‘lord of Scythia’ (F) indicates thatthe cult was already well established among the Greek colonists of hisday. But legends do not develop everywhere uniformly; at Miletus itselfthe cult of Achilles attracted no following. I do not see how we can tellwhether the Odyssey reflects unease that, for all his distinction as a warrior,there awaited the half-divine Achilles no better fate post mortem than theshadowy existence appropriate to humankind in general, or offers a protestagainst a development out of keeping with the stern philosophy of heroicepic. In the second Nekuia, surely the work of a later poet, Achilles isundoubtedly in Hades (.–). We cannot use this passage to clarifythe relationship between the Odyssey and the Aithiopis.

Achilles’ concern for his father (cf. Il. .–) and son contributes tothe affirmation of family values which is an important element in book (by contrast with and ). Odysseus can give no news of Peleus, butoffers a heart-warming (and tactful) report on Neoptolemus, and we seeAchilles’ ghost stalking off through the fields of asphodel 4�3� +�� / �<#<.� C1�� )�� (.). In the dead warriors’ interest in thesons who have grown up unknown to them we see variations on the themedeveloped most fully in Telemachus’ role.

We see why the poet quietly abandoned the necessity for the ghoststo drink from Odysseus’ trench before they could recognize and speak tohim as the focus shifts to Ajax, standing somewhat apart, still harbouring

amphora, dated c. , now in the British Museum (LIMC i.2 s.v. Achilleus, plate ), depicting awinged warrior flying across the sea (there is a ship below), watched by a raven (Apollo’s bird), hasbeen thought by some to depict Achilles’ translation northwards; but it might be better connectedwith his role as helper to those in peril on the sea, on which see Arrian, Peripl. M. Eux. .

See further Hommel (), Hooker (), Hedreen (), Bravo (a: –), Burgess (:–, : –), Buiskikh (), Hupe ().

Special treatment for Menelaus might be interpreted as basically a privilege granted to Helen asZeus’ only mortal daughter.

This strangely belated encounter between the ghosts of Agamemnon and Achilles was most likelyborrowed from or modelled on an episode in the Cyclic Nostoi in which, as we learn from Pausanias(..), a scene in Hades figured. On the end of the Odyssey see further S. West ().

On the relationship between the Aethiopis and the Iliad see M. L. West (c). It would have given Achilles no joy to hear of Neoptolemus’ slaughter of Priam at the altar of Zeus

Herkeios.




resentment against Odysseus over the Judgement of Arms. Odysseusimplores the ghost to come and talk to him, but the other makes noanswer, and leaves without a word. Still, it might have been otherwise:C�3 �@ r'�� 1� ��'��, ^ �� 74g ��· | �� '�� ^3��3#'.� 7�� :3� � 1�� | �� P�� T#�,� �� 3��.(–) Page comments tartly: ‘The silence of Ajax, then, was accidental,imposed by the requirements of a timetable. Given another moment hewould have spoken. And Odysseus’ plea, that Ajax might forgive and speakto him, was nothing but formal politeness: Ajax was about to reply, butOdysseus is in a hurry, he cannot wait for the answer; another day, perhaps,but just now time is pressing’ (a: –).

The weakness of the transition, motivated simply by Odysseus’ curiosity,is obvious. We are not actually told that Odysseus has left his sacrificialtrench to stroll round Hades and view the traditional Underworld figures,but the necromantic framework has been abandoned. In Heracles’ greetingwe see the culmination of the poet’s efforts to enhance his hero’s status.

Heracles recognizes that they have much in common: K ��@, M �� 9 .� '�� 4��H��, | /� �� 74g� b�� S�@ (4,�I�� (.–). On his mission to get Cerberus Heracles had thehelp of Hermes and Athena: !d�'�� '@ C��'T�� 4�#�� 3:��(; cf. Il. .–). Odysseus has managed his visit to the world of thedead without apparent divine support. Heracles’ acknowledgement thatOdysseus’ career has much in common with his own is highly significant.Endurance, determination and resourcefulness win for the trickster fromthe Western Isles the respect of Zeus’s son, the archetypal hero.

Odysseus was hoping to meet more of the ancient heroes when his visitwas brought to an appropriately uncanny conclusion: ��, �� 7�� C3��@�4�� '#�� | I� 3� �� · 7'� �� .� �� \��, | ': '��4�� 1�V� ��0� ��# | 7[ h�� 'T�� 4#V �� 61�� (–). The weird atmosphere evoked at the start of the episodeis thus revived. The scholium on (very likely displaced from two lines

Goold (: –) interestingly compares the conclusion of the duel between Hector and Ajax(Il. .–): ‘What the poet is doing in these two passages . . . is to effect a juncture betweentwo blocks of different material. He does not want to drop Ajax and abruptly introduce Minos,nor does he wish to kill off inexpendable heroes, but has elected to secure a transition in the onepassage and a suspension in the other by means of a contrary-to-fact apodosis: from a sympatheticviewpoint no one could reasonably imagine that Ajax was willing to forgive Odysseus or that theduellers-to-the-death were really shamming.’

The problems created by .– are notorious; already suspected in antiquity, they are bestregarded as an interpolation.

Their affinities are well brought out by Finkelberg (). Presented in a rather different light at .–.



stephanie west

earlier) tells us that the passage describing Odysseus’ underworld sight-seeing (–) was judged spurious (almost certainly, by Aristarchus):��3�+�� '�� 8 f� ��g� r '�� ?3�� C�# ��'�� h�� 5 � ().

Persephone’s eerie intervention would have made a fine ending to Odysseus’encounter with Ajax’ shade.

The problems of book are familiar enough, and I have not attemptedto gloss over them. The situation in some ways resembles that of Iliad. There both the way in which the book is worked into the poem andthe way in which Phoenix is worked into the book have long seemed tochallenge most severely a unitarian approach. Yet in this book we haveso much that goes to the heart of the Iliad and illuminates Achilles thatwe cannot happily label it a late interpolation as we may Iliad . BryanHainsworth well sums up: ‘Reinhardt’s hypothesis . . . that the compositionof our Iliad was a long drawn-out process with repeated revison over manyyears rests on a persuasive premise: the poem cannot have been created in amoment of inspiration. It is in such places as book that we have a glimpseinto the workshop of Homer.’ The difficulties which once fuelled thearguments of analysts are better seen as clues which may reveal somethingabout the poet’s methods and view of his subject. The production of awritten, relatively fixed, text of a lengthy composition was still a novelty.

We can, I believe, roughly trace in book four strata in the Odyssey’scomposition. (My geological analogy should be supposed to allow somefolding and shifting; the strata do not invariably lie neatly one abovethe other.) The earliest involves Odysseus’ conjuration of his mother’sghost, at the Thesprotian nekuomanteion, an episode belonging to themore realistic nostos through known regions outlined in his cover-stories.The poet has left a tantalizing indication that the audience for that narrativewas originally Penelope. Anticlea’s ghost concludes her utterance with thewords �8� �� | 5 3@, F� � '�� 3� �� 5�� 3 4#��(.–), natural enough if Odysseus is now addressing his wife, otherwisecuriously pointless. That itinerary was discarded in favour of a journeythrough quite uncharted territory, in the course of which his dealingswith a savage ogre provoked Poseidon’s wrath and thus became pivotalto the plot. This is the second stratum; at this level Circe can provideOdysseus with all the advice on his route home that he needs, and it isassumed that once home he has nothing more to fear from Poseidon. But

Cf schol. Pind. Ol. .; see further Garbrah (). Hainsworth (: ); see also S. West (). It is not irrelevant that it is in book that we have

(at –) one of the clearest indications of the date of composition to be found anywhere in theIliad: see further Burkert ().




with the passage of time the poet saw a need for Odysseus to make hispeace with Poseidon, for which specialist guidance was required: hence theconsultation of Teiresias, the prophet par excellence. Ghost-raising does notin itself call for a journey to the world of the dead, but a visit to Hades hadbeen Heracles’ supreme exploit, and such an adventure would enhanceOdysseus’ status. The insertion of the greater part of what we know asbook entailed slight changes to the end of book (–) and to thestart of book (–). (But the poet allowed Odysseus to appear to haveforgotten what he had learned from Teiresias about Penelope’s suitors whenhe got back to Ithaca (. –).) Finally, as the last stratum, we have whatfollows Odysseus’ encounter with Ajax, culminating in his meeting withHeracles (.–); here, unquestionably, the framework is a katabasis,and that is how Circe viewed their adventure on their return: ��, �pH�� S�:�3�� '@ �B��, | �� 3��, /�� @ P�� q�[ 3�� # @P�3�� (.–).

In book we get an unusually strong sense of work-in-progress. In thisessay I have assumed a much greater element of innovation and creativityon the poet’s part than is nowadays commonly supposed, encouragedby his own emphasis on the importance of novelty (.–), which issupported by Phemius’ pride in his own originality: (�� @ ��'�,3�.� �� '�� 7� 1�� 5'� | �� 7��1# �� (.–). Modernstudy of traditional oral epic has emphasized its conservatism, and theenvironments in which such traditions survive – or are revived – do notencourage innovation. But it cannot always have been so. Whoever had theidea of giving the dignity of hexameter verse to Odysseus’ encounter withPolyphemus (whether the poet of our Odyssey or another) could not haveproceeded by simply adapting elements from roughly similar narratives;too much here is peculiar, without close congeners in heroic narrative. Inbook I believe we can trace more clearly than elsewhere the sequence inwhich the poet’s ideas developed.

If this analysis is on the right lines, it may be worth looking for indi-cations of similar developments elsewhere in the Odyssey. This is not theplace to set out the grounds for believing that it was the possibility of mak-ing a written record that stimulated the composition of long epics such asour Iliad and Odyssey. Those who had been trained in the oral poet’s artwould not change their style or narrative technique because script couldnow offer a degree of permanency; they did not envisage reflective solitaryreaders who might be troubled by inconsistency or misdirection whichwould not worry a listening audience. When oral theory was a novelty,many oddities which had troubled analysts and inspired some amazingly



stephanie west

ingenious defences from unitarians appeared sufficiently accounted for asnatural elements in a style which did not nicely weigh every word. Thisapproach could too easily lead to a lack of attention to detail. We shall dobetter to consider the ways in which is revealed a poet’s adaptation of histechnique to meet the demands of a lasting written record.

We too readily assume that in antiquity the production of a written textwas as definitive an act as printing is nowadays. But the unmarked endingsof very many ancient works reflect a certain indecisiveness about formallyconcluding. The end of an episode is marked as such, but the authorleaves open the possibility of going further. The last chapter of Herodotus’Histories (.) well exemplifies this feature. Of course, extension at theend of a work is a more obvious idea than expansion midway, but (literal)cutting and pasting presented no great difficulty (though systematic minorcorrection at a series of points in the narrative was more demanding), andso long as the roll was the usual format, alteration and expansion werephysically easy. Writing could give durability to what was previouslyephemeral, but did not preclude improvements. Modern conditions ofpublication make it highly unlikely that a literary work will be disseminatedto the reading public before its author judges that it has reached a definitiveform; if death overtakes the writer before finishing, that circumstance isunlikely to be overlooked. But book production was very different beforethe invention of printing.

If the composition of the Iliad and Odyssey extended over many years –decades, even – questions of mutual influence, interference and intertextu-ality in early Greek hexameter poetry gain added interest but become muchharder to answer. Singers enjoy a place in Eumaeus’ list of peripatetic pro-fessionals (.–); in archaic Greece this is more likely to reflect the wayof life normal for a singer who made his living by his art than the situationof Phemius and Demodocus, attached to a court or wealthy household.Though we may doubt whether our poet and Hesiod really met on Delos(Hesiod F ), it is pleasing to speculate about opportunities for singersto become acquainted with one another’s work before texts circulated.But once we take this scenario seriously it becomes much more difficultto assess what might look, at first sight, like evidence of the influenceof one work on another; questions of intertextuality become much more

See further van Groningen (: –), S. West (). The situation was interestingly different where the normal writing material was the clay tablet; once

baked its text was unalterable. Or perhaps exercised his skill, in the evenings or on holidays, as a useful sideline to a peripatetic

day-job.




complicated. We can trace connections between our Odyssey and a widerange of other hexameter poetry – not only with poetry dealing with theMatter of Troy (Memnonis/Aithiopis, Little Iliad, Iliou Persis, Nostoi) butalso with the Argonautica, the Catalogue of Women and poetry celebratingHeracles. With some of these titles we may be reasonably confident thattexts already existed, but for others the poet more likely drew on whatexisted only in memory and performance. The poet was unmistakablyin debt to Argonautic poetry, as he acknowledges (.–), perhapsindeed to a version of Jason’s adventures already stabilized in writing; butthe relationship with the Cyclic epics (as opposed to the complex of Trojanthemes which formed their subject matter) is less clear.

Scenes set in Hades were clearly popular in early Greek epic, notrestricted to heroes who succeeded in returning from the world of the dead.From Pausanias (..) we learn that they featured in both the CyclicNostoi and the Minyas. As in descriptions of scenes among the gods, thepoet who described the underworld told of matters beyond human knowl-edge, with an authority derived from his Muse (cf. .–) but originallyrooted in shamanistic song. Where Heracles had gone, Odysseus, his sta-tus greatly enhanced by our poet’s skill, might also venture. Inconcinnitiesin the narrative should not distract us from the grandeur of the Nekuia’stheme.

See further M. L. West (). See further Burkert (: –).



chapter 8

Older heroes and earlier poemsThe case of Heracles in the Odyssey

Øivind Andersen

In the Iliad, Heracles’ P�3�� for Eurystheus are taken for granted (Il..–, .–, .–). That ‘even Herakles, dearest of all to Zeus,succumbed to fate and the wearisome anger of Hera’ (.) is a comfortto Achilles. Heracles is said once to have wounded Hera, though (.–),and Hades as well (.–), and also to have attacked Pylos (.–). Heracles’ dealings with Laomedon and his sack of Troy are proudlyreported by his son by Astyocheia, Tlepolemus (.–), who is leader ofthe contingent from Rhodes (.–). Along with the barely mentionedPhidippus and Antiphus, sons of the Heraclid Thessalus (.–), Tle-polemus may have been introduced into the action at Troy by the poet whohas him die at the hands of Sarpedon (Il. .–). Heracles as a father figureand a formidable hero looms large in the horizon. Whether Homer for hisHeracles references in the Iliad relied on a specific Heracles epic or drew on acomposite tradition of epic song and mythological narrative, of which theremust also have been a great deal, we cannot know. Also, we cannot knowhow much in the way of invention and adaptation that went into the fash-ioning of the hero at relevant points in the Iliad. While Heracles – Hera’shero of the labours – will have been a figure about whom everybody knewthe essentials, he will at the same time have lent himself to being associatedwith ever more adventures and to being exploited for a variety of purposes.

That will hold true also in the case of the Odyssey. There, the older herocomes into the picture on three occasions: first, in Scheria, as Odysseus

M. L. West, who in his contribution to this volume does not deal with hexameter poetry aboutHeracles, has maintained (a: –) that poems about his deeds were current before bc andthat the allusions in Homer show that there must have been ‘a poem or poems’ covering the labours.That the poet of the Iliad relied on a specific epic poem about Heracles was argued by Kullmann(: –) within the framework of a developmental hypothesis for the Gotterapparat.

In addition, Heracles is mentioned in the catalogue of women in the Nekuia, where Odysseus sees‘Amphitryon’s wife, Alkmene, who after lying in love in the embraces of great Zeus, brought forthHerakles, lion-hearted and bold of purpose’ (.–). (I quote or adapt Richard Lattimore’stranslation throughout.)



Heracles in the Odyssey

identifies himself as the next best archer in the world, while adding that hewould not contend with earlier men like Heracles and Eurytus (.–);second, at the end of the Nekuia, when Odysseus reports his encounter withthe terrifying �5�� of Heracles (.–); finally, in Ithaca, as Penelopegoes to fetch Odysseus’ bow and we get to learn about its provenance in astory which features Eurytus and Iphitus as well as Heracles (.–). Myinterest in these passages, relevant to the question of relative chronology,is threefold. First, concerning matter: what do the passages in Books , and refer to and rely on in terms of contents? how should theybe contextualized and filled out? Second, concerning method: how canwe know what is old and what new? For example, does opaqueness ofcircumstances or details imply the existence of, and even the audience’sfamiliarity with, a clearer and more complete version? Third, concerningfunction: to what extent and in what way do the three episodes contributeto the Homeric (and the Homerist) audience’s appreciation of what isgoing on in the Odyssey? We shall mainly consider the incidents told inbooks and , which involve Eurytus as well as Heracles.

∗ ∗ ∗

After his sudden discus throw and some hearty encouragement by Athena,the as yet unrecognized Odysseus assures the Phaeacians that he is willingto compete in both boxing and wrestling and the foot race, and boastsof his athletic all-round prowess (.–). He then identifies himselfemphatically as a bowman: ‘I know well how to handle the polished bow’(.). He was the first to hit his man with an arrow in a throng of enemies,however numerous were his fellow archers around him; Philoctetes was theonly one to surpass him with the bow in the land of the Trojans wheneverthe Achaeans shot with bows; all others Odysseus declares himself to outdoby far. As Odysseus throughout the Iliad appears as a spearman and in fullarmour, it is remarkable that the poet here has him introduce himself asa bowman before Troy. By simply asserting the fact, the poet, I submit,

Danek (b: ) asserts not only ‘daß die Odyssee nicht nur auf eine vage “Herakles-Sage”rekurriert und die Einzelheiten jeweils neu erfindet, sondern auf breit ausgestaltete Versionen derSage verweist, die man sich in epischer Form vorzustellen hat’, but also that ‘der Horer der Odysseemit Herakles-Epik gut vertraut sein mußte’.

‘These allusions help to expand the range of human possibilities the poem embraces and the rangeof ethical judgements it invites, . . . [the Odyssey’s mythological allusions] contribute to the sense thatthere are other ways to evaluate the individual achievements and social institutions that the poemrepresents’ (Schein : ).

Odysseus does not even take part in the archery contest in the games for Patroclus; Meriones andTeucer are the contestants (Il. .). Meriones lends Odyssey his bow in the Doloneia (), butOdysseus has no use for it. On archery in Homer and Odysseus the archer, see Dirlmeier (:–), Holscher (: –), Danek (c: –).



øivind andersen

establishes the hero in that capacity. Whatever ‘Odysseus the bowman’may owe originally to tales about Odysseus the unheroic trickster, hisbowmanship at Troy as it is presented here in book enhances Odysseus’heroic stature. Odysseus’ boast is the poet’s post festum amplification of hisOdyssean archer in the light of what will happen at his homecoming.

Odysseus is only the next best of his generation, though. There is nobeating Philoctetes. It is admittedly not so easy to envisage the two of themfighting together as archers during the siege of Troy. The Iliad seems toreckon with the tradition according to which Philoctetes was brought toTroy in the final phase of the war (cf. Il. .–). The defining featurefor Philoctetes’ status as the greatest archer before Troy is his role at thesack of the city. Here in book , however, Philoctetes is not associatedwith any decisive feat, but appears as a generally good archer, ‘whenever weAchaeans shot with the bow in the Trojan country’ (.). For the Odyssey,of course, Odysseus is eminently the ��3��. As Philoctetes in theprevailing version will have been absent for most of the duration of thewar, until he was finally fetched from Lemnos, we again can observe howthe poet has Odysseus convey a picture that suits the immediate context,although it hardly tallies with the received story. What the Odyssey offershere is not a glimpse of earlier traditions or lost poems, but a reflection ofthe action of the Odyssey itself.

In addition to comparing himself to Philoctetes, Odysseus contrastshimself with two others. For

I will not set myself against men of the generationsbefore, not with Herakles nor Eurytos of Oichalia,who set themselves against the immortals with the bow, and thereforegreat Eurytos died suddenly nor came to an old agein his own mansions, since Apollo in anger against himkilled him, because he had challenged Apollo in archery.

(Od. .–, tr. Lattimore)

The context implies that Odysseus could stand comparison to Heracles andEurytus in archery. The fact that the two of them were good archers goeswithout saying; indeed, if they had not been good archers, they would nothave competed with the gods. By stressing their vying with gods Odysseusforegrounds the contrast between his own wise ways and the objectionable

Garvie (: ), on .–: ‘But with the climax of the poem already in mind H. may haveattributed to him a skill that is quite untraditional, at least in the context of Troy.’

In the Odyssey, we briefly learn about his successful return (.). ‘The fact that Heracles and Eurytus competed in archery even with the gods is mentioned in the

first instance to prove how skilled they were with the bow’ (Garvie : , on Od. .–).




conduct of the older heroes, which for Eurytus had fatal consequences.The mentioning of the older heroes thus is transformed into a warningexample. Odysseus has made a point of himself standing ‘far out ahead ofall others such as are living mortals now and feed on the earth’ (.–); hewill not compete even with men of earlier generations, let alone with theimmortals. It may also be significant that Odysseus had just issued a generalchallenge to the Phaeacians, to all of them ‘except Laodamas himself, forhe is my host; who would fight with his friend?’ (.–). There is apreoccupation here, on behalf of poet and hero, with proper conduct. Itis hardly accidental that Odysseus introduces himself as the man with thebow within a morally charged context involving guest-friendship. I wouldagree with the scholiast (to .) that the present passage foreshadows(��'��) the massacre of the suitors. Archery at Troy is not veryrelevant in the Phaeacian context, where no archery has taken or is goingto take place. Odysseus otherwise refers to athletics that is actually goingon: He is willing to compete in boxing and wrestling and the foot race(.), and he refers to contests generally (.). Archery does not figureon the programme in Scheria; the prominence that it achieves in Odysseus’speech refers us forward to his wielding of the bow amongst the suitors inIthaca.

Both Heracles and Eurytus will have been well established as bowmenfor poet and audience alike. But to which story or stories does Od. .–

refer the listener? For Heracles competing with gods we have no furtherevidence. His vying with the gods simply is asserted here and it will, Isubmit, have been accepted accordingly. Heracles is the sort of hero aboutwhom a poet could easily assert, and an audience would readily believe,such things. After all, he did fight with Apollo over the tripod in Delphi,

‘Odysseus implies that he himself is more sensible’ (Schein : ). Danek (b: ) sees the contrast as being essentially between the Trojakampfer and older

heroes. Pace Hainsworth (a: ) on Od. .– that ‘the episode does not need the support of

so distant and incidental a comment as this’. Many scholars have sided with the scholiast, e.g.Thornton (: ) (‘The decisive part played by the bow at the end of the poem is foreshadowedthroughout’), Dirlmeier (: ), Crissy (: ).

Odysseus mentions also throwing the spear (.), which equally is not on the agenda, but it isinterestingly made relevant for him as an archer: he can throw the spear further than others canshoot an arrow. The fact that the spear figures prominently in the killing of the suitors (., )may have contributed to the mentioning of spear-throwing here.

Schischwani (: ) thinks that if Eurytus were not already known as a master of the bow, hecould not have served as a comparison to the archer Odysseus, as ‘it would be absurd to bring in anunknown person in the comparison’. But it is only to the internal audience of the Phaeacians thatEurytus needs to have been known as an archer – and even they would have taken Odysseus’ wordfor it!



øivind andersen

and he did shoot at Hera, and even at Hades. The fact that his vyingwith the gods is not elaborated upon and that it apparently did not haveserious consequences for him, as it had for Eurytus, also makes it easy toassume an ad hoc innovation in this context: a past fact is created to servethe narrative context.

As for Eurytus, we do not otherwise know of his over-optimism inrelation to the gods. His challenging Apollo is, I presume, established as apast fact here by being told. Challenging a god within the area of the god’sspecific competence and being punished for it in some relevant way is a notuncommon motif. The fact that the Eurytus incident follows a typicalpattern will have made it even more easy for the poet to generate it and foran audience to accommodate it. The motif of his death by the infuriatedarcher god may even have been prompted by another tradition about himmeeting a violent death by the bow for acting unfairly (see below). Eurytusis killed by Apollo if the poet says so, just as Odysseus was a great archerbefore Troy if the poet says so. Uppermost in the poet’s mind is a thematiccluster of archery and of the proper use of the bow.

Heracles and Eurytus, who appear together in our passage, both have aspecial link to the two allegedly best archers before Troy, whom Odysseushas just compared to each other: Philoctetes and Odysseus himself. ForPhiloctetes inherited the bow of Heracles, while Eurytus’ bow eventuallybecame the bow of Odysseus. It is uncertain whether or to what extentthis parallel genealogy of the bows is part of the picture here and is meantto contribute to the audience’s appreciation of the passage. It may havebeen general knowledge that Philoctetes was needed to take Troy becausehe had inherited the bow of Heracles; it was the bow that Philoctetes had,not Philoctetes himself, that really was needed. But that Philoctetes hadthe bow of Heracles may not even have been part of the tradition at thisstage. The bow as such is at any rate not highlighted in our passage in

Some commentators and the LfrE (s.v. 7��, 7��H�) refer the reader from Od. . to Il..ff., and it cannot be ruled out that the verb (‘contend with’) in . should be translated ‘fightwith’ and not, as in ., ‘compete with’.

Andersen () develops the concept of ‘instant past’ and the principle that ‘the present hasprecedence over the past’.

One may think of Thamyris, Marsyas, Niobe. It may be relevant that Thamyris suffered hispunishment ‘as he came from Oechalia and Oichalian Eurytos’, according to Il. ..

‘Now, while Homer nowhere explicitly states that Philoctetes inherited Heracles’ bow, it seems clearthat both he and his audience knew that tradition well’ (Clay : ); also Danek (b: ).

Thus Gercke (: ). Along with Heracles and Eurytus, Philoctetes is probably an old archer:‘Wenigstens ist dieser oder sein Vater Poias schwerlich von vornherein als Erbe des Herakles in dasEpos eingefuhrt worden: Ilias und Odyssee berichten davon nichts, und die Ubergabe von Bogenund Pfeilen zum Danke fur die Hilfsleistung bei der Selbstverbrennung auf dem Oita ist mit dieserjung.’




Odyssey , where Philoctetes simply appears as a generally superb archerduring the war. As for Odysseus, if one believes that the audience wasthoroughly familiar with the story of the transfer of the bow from Eurytusvia Iphitus to Odysseus, one may consider as of little account the fact thatthe transfer itself is told only in book (see below). But then Odysseus,who is second only to Philoctetes as an archer, did not bring with him toTroy the bow of Eurytus. Did the audience of book know that as well?And what did they make of that?

Heracles’ and Eurytus’ joint appearance in this passage has been seen bysome in the light of the fact that the two heroes are associated in a famousstory, in which archery indeed played an important part: Heracles came toEurytus in Oichalia to compete with him (and/or his sons) in order to winthe hand of his daugther Iole; although Heracles won the contest, Eurytus(and/or his sons) would not give him Iole in marriage; thus Heracles sackedOichalia and killed Eurytus (and/or his sons). The story is unattestedin the Odyssey, and that Homer is unaware of this tradition, i.e. that itpostdates the Odyssey, seems to be the more commonly held view. ThusClay (: , n. ) finds that Homer ‘seems to betray no awareness ofthe tradition surrounding the Sack of Oichalia’. Danek (b: ), on theother hand, thinks that Homer by the joint mentioning of Heracles andEurytus puts himself ‘geradezu in Gegensatz zu einer gelaufigen Version’and that he seems ‘die Geschichte vom Konflikt zwischen Eurytos undHerakles um Iole regelrecht auszublenden’ – with the addition that onemay not of course conclude from that that the poet did not know themyth. Schein (: –) thinks that, ‘by referring to Herakles andEurytos in the same line [Od. .] and to Eurytus’ death at the hands ofApollo, Odysseus both alludes to and contradicts a myth that presumablywas told in the Sack of Oichalia’, a version of the myth which ‘would havebeen well known to Homer’s audience from the oral poetic tradition’. I find

To Clay (: ) the parallel genealogy and parallel rivalry – Heracles/Eurytus–Philoctetes/Odysseus – suggests that Odysseus ought to be viewed as both a parallel and a rival toHeracles, but also as a contrasting figure. ‘An elaborate comparison between Heracles and Odysseusis thus established, though its precise terms remain obscure.’ Crissy (: ) suggests that ‘a like-ness is implied between these two characters [viz. Odysseus and Heracles], not a contrast’. Surelythe relevance of Heracles to Odysseus is obvious even if one reckons with less elaborate parallelismsor inverse parallelisms.

Apollod. Bibl. ..– and .. References to other ancient sources in Frazer’s notes ad locos inthe Loeb edition. For thorough discussion of the sources and older scholarship, see Davies (:xxii–xxxvii) and Burkert ().

Also Davies (: xxvi, with reference to Od. .ff. and .ff.) cautions against equating un-and post-Homeric, and thus to underestimate ‘Homer’s own amply documented addiction toidiosyncratic innovations in myth.’



øivind andersen

it problematic to speak of an allusion here to ‘the bride contest with bowsbetween Eurytus and Heracles’, as we are explicitly presented with a versionof Eurytus’ death (by Apollo’s hand) which does not square with that story.And I find it even more problematic to take the mentioning of Eurytus andHeracles together here as in any interesting sense a negation of that other,untold story. It is just ignored and does not come into the picture. It isreasonable to assume that the introduction into the poem of an ‘Eurytus ofOichalia’ (., also Il. . and ) as a great archer will owe somethingto the story that above all made Eurytus into a great archer, viz. the bridecontest at Oichalia. I would content myself, however, with the assumptionthat in some form and in some stage of its development, that story informsour passage in Book in so far as it is the basis for pairing Heracles andEurytus. Perhaps that story – which included challenges and unseemlybehaviour and contests of the bow – inspired the poet’s presentation ofEurytus as a man who behaved in a thoroughly unacceptable way, evenvying with gods.

So I would refrain from (re)constructing as the telling background ofOdysseus’ words here a specific version of a specific story that everybodyknew and so would be referred to. The Phaeacians listening to Odysseus inthe Odyssey would know – or at least they would nod knowingly, impressedby the hero’s name-dropping. Whatever the Phaeacians made of it, I findit hard to believe that Homer meant his audience to think of the bows ofHeracles and Eurytus respectively in this context. Philoctetes’ bow mainlyis ‘Heracles’ bow’ when it comes to the fall of Troy – while here, Philoctetesis a generally good archer during the siege, and the link to Heracles’ bowmay not even have been established for the Homeric audience. Eurytus’bow is not Odysseus’ bow at Troy. Odysseus boasts of his skill in archery andraises the topic of the just use of the bow with none too subtle implications.The story of the conflict between Heracles and Eurytus and the sack ofOichalia, even if it were well known, is not ‘alluded to’ here by the poet ofthe Odyssey in order for it to resound.

∗ ∗ ∗

Heracles and Eurytus and his kin appear again at the beginning of book

as Athena puts it into Penelope’s mind to go and get the bow and grey iron

The allusion-cum-negation point of view is developed by Bruno Currie in his contribution to thisvolume, but with respect to more specific texts (hymns to Demeter) and a more cirumscribedcontext (Eleusis).

It is quite another thing to speculate whether the story of the bride contest in Oichalia informsthe main plot of our Odyssey, as does Kirscher (), following in the steps of Gercke (). Thatwould not rule out such ad hoc glimpses of Eurytus as we get in our Odyssey.




to set up in the halls of Odysseus. The presentation of the bow (.–),focalized by Penelope, highlights the bow’s significance by revealing itsprominent provenance and its status as a token of guest-friendship. By itslength and detail and well-wrought form, the passage in itself lends weightto what is going on and signals that we are at an important juncture in thedevelopment of the plot.

The bow that Penelope fetches is the one that once had belonged toEurytus and which he, dying (��3�� ) in his high halls, left to hisson, godlike Iphitus. Iphitus gave it to Odysseus as a token of guest-friendship when the two met in Ortilochus’ house in Messene; youngOdysseus was abroad to seek compensation for sheep and shepherds thathad been stolen from Ithaca by the Messenians, and Iphitus to collect somelost mares and their mule offspring. Odysseus gave a sword and a spear inexchange for the bow and arrows, but the two men never came to entertaineach other as [�0�� (.–, –). These parts of the excursus associateOdysseus from his early youth with the bow that he needs in order to fightthe suitors. At Troy, he could have used any bow (and as he left Eurytus’bow at home, he must have used another one when or if he was shooting).In Ithaca, however, a regular bow will not do as the revelatory functionof the archery resides less in the aiming and shooting through the axes asin the actual stringing and wielding of the very special bow. If we go bywhat we learnt in Od. .–, this will be the bow with which Eurytuschallenged Apollo, something that led to the archer being killed by thearcher god. If we go by the Oichalia story, this will be the bow of the manwho lost the archery contest and refused to keep his promise to Heracles,something that eventually led to his being killed by the greater hero. Howsignificant is this, in the action and for the audience? On the basis of the‘allusion’, albeit negated, to the story of the untold bride contest, wherethe bow was used by the loser, Schein (: ) finds that ‘the allusioncasts an odd shadow on Odysseus’ victory and subsequent slaughter of theSuitors by associating him both with Eurytos’ transgressive behaviour andwith the loss of a bride to a better archer. Indeed, in figurative terms, italmost makes Odysseus one of the Suitors.’ The main idea here may be therather simpler one, that Eurytus was a great archer, the closest a human

It has often been said, e.g. by Davies (: xxv) and Gantz (: –), that what we hear aboutEurytus’ death here in Book squares well enough with what is said in . about him beingkilled by the enraged Apollo so that ‘old age did not come to him in his house’. But the atmospheresurrounding the death of Eurytus is rather different in the two passages, and the overall impressionin . is not of a man being shot dead by Apollo. The poet creates the situation, or that versionof a situation, that suits him.



øivind andersen

can approach Apollo in archery, on a par with Heracles, and thus a suitableformer owner of the bow which is now going to be used by Odysseus.

The story of how Eurytus’ bow became the cherished bow of Odysseusexplains the convenient fact that the bow is available in Ithaca, so thatPenelope may prepare the contest. For Odysseus left it behind in his hallsin memory of his guest-friend (.–). The story’s patent explanatoryfunction has been taken to indicate invention on behalf of the poet of theOdyssey, as has the fact that the poet seems almost to go out of his way toprovide a background for the encounter. Both the underlying criterionof ‘Ausfuhrlichkeit der Angaben’, and its application here, raise questions.Even exhaustiveness is relative. There is always more to be known, inthis case, e.g.: why should the very young Odysseus be entrusted with adiplomatic mission? What about its outcome? On the other hand, howevercontrived a story seems and however circumstantial it is, it cannot bedeemed to be untraditional on that account. Both the old and the newmay be presented in great detail, just as both the old and the new maybe briefly sketched and hinted at. I would question the assumption that abrief reference to, or the incidental remark about something – some wouldsay: the allusive character of a passage – speaks for the audience’s familiaritywith what is evoked. In many cases this will be the case. But new ‘facts’ areestablished all the time by being told, both in the main narrative and in thebackground. This obviously has implications for the way we consider therelative chronology of narrative elements and mythological incidents.

Crissy (: ) makes the point that Odysseus ‘rather than challenge Apollo with the same bow,will claim the god as his ally when he begins his attack’ in Od. ..

Iphitus’ all too obvious function as the conveyor of the bow has led to the suggestion that he has beenintroduced or even invented by the poet of the Odyssey or his predecessors (‘vom Odysseedichteroder seinen Vorgangern’) for the sake of this function, see Prinz (: ). But if the encounterwith Iphitus had been part of the tradition for, say, a couple of generations, then there would be noneed for the poet of the Odyssey to spell it out. For the audience of our Odyssey, what would be thestatus of an ‘innovation’ made a couple of generations earlier?

On exhaustiveness (‘Ausfuhrlichkeit der Angaben’) as a criterion for invention, see with respectto our passage Danek (b: ‘reichlich konstruiert wirkt und nicht nach alter Traditionaussieht’), Holscher (: ‘Schon die Doppelung der “Suche” . . . macht nicht den Eindruckdes Ursprunglichen’), Schischwani (); for general discussion of this and the complementarycriterion of brevity as indication of traditionality, see Kullmann (: –, esp. ); also relevantStoessl (: –, esp. ).

See Friedrich (: –) for formal analysis and his comments on the ‘komplizierte, ja fast artifizielleStruktur’ and the enhanced ‘Wirkung der Retardation’ achieved thanks to the bulk of the excursus;also de Jong (: ). Kirk (: ) postulated for our passage ‘a convoluted style suggestinga more extensive poetical model’.

See Andersen () for a fuller statement and Scodel (: passim, esp. –) for discussion,with due recognition that ‘[a]bbreviated narratives claim to be and may often actually be compressedor short versions of tales that circulated in fuller form. But, this may not always be so. The tools ofanalogy and recombination that created the main stories of the tradition can still operate’ ().




∗ ∗ ∗

To account for audience response in relation to the composite and incom-plete picture that a poem offers, the distinction between a narrative audi-ence and an authorial audience, as developed by Ruth Scodel, is helpful. Anarrative audience is one ‘for whom the text would be true, the one whichknows exactly what the text takes for granted and does not already knowwhat it explains’. An authorial audience on the other hand is ‘the audienceto which the text actually addresses itself’. Many readers seem to assume‘that authorial and narrative audience are effectively identical in Homer’. Inthis view, ‘exposition is not really necessary for comprehension, but is partof the traditional style; hence the narrator often repeats what the audienceknows already, but never pretends that they already know something theydo not’. There is an assumption of transparency. But traditional oral nar-rators do not, Scodel argues, always rely on audience familiarity with theirstories. And the authorial audience can live with gaps: they will accept thepoet’s version, fill in the picture or hold elements in suspense, according towhat is called for. Deficiency of information invites the audience to takethings for granted, and assumes that it can cope with what there is. Forthe poet’s exploitation of this information gap, Scodel has developed theconcept of ‘pseudo-intimacy’.

Also the concept of ‘Unbestimmtheitsstellen’ (‘points of indetermi-nancy’) is helpful when we deal with this kind of patchy and incompleteinformation. Brief mention particularly of things to come or of things pastinvites the reconstruction or even the construction of what is not told.

When new items or unexpected details are introduced into a story, theaudience is challenged to establish a context and to fill out the picture. Theaudience is ready to do so on the basis of the few clues given, and will recon-struct or construct contexts as need be. Just as the poets base innovationson traditional materials and story patterns, to the degree that the poetsthemselves need not be consciously aware of their own innovations, anaudience steeped in oral poetry would find both construction and recon-struction easy. ‘In a rich tradition crammed with variants, and a mixedaudience who are familiar with different versions in varying degrees, anindividual would in any case find it hard to be sure that an unfamiliar ele-ment was actually new. The narrator has tremendous power over any areasof the stories that are not completely standardized – which probably means

Scodel (: ). See also Scodel (: ). Although ‘pseudo-intimacy’ seems not to figure inScodel (), the ‘assumption of transparency’ is discussed throughout.

I draw on Schmitz (: esp. –). For this point and the following quotation, see Scodel (: ).



øivind andersen

all but the basic outlines.’ As long as it does not vitiate the Faktenkanon, aco-operative audience is able to make sense not only of the odd detail butof whole new episodes, in the present, surface action of the poem and inits past background.

∗ ∗ ∗

We should bear this in mind when we now turn to that part of the excursuswhich concerns Iphitus and Heracles (Od. .–, f ). The mares andmules that Iphitus was out looking for, were baneful to him, for as hecame to Heracles’ place, the son of Zeus invited him home and recklesslykilled his guest in defiance of the gods and of the norms of hospitality,while he kept the horses for himself. The lack of an explicit context for theencounter and the sketchy presentation of the incident have led many toconclude that the story only makes sense within a fuller picture. There areindeed questions: Where does Iphitus meet Heracles? What is Heracles’relation to the mares that Iphitus is collecting? Why does Heracles killIphitus at the table? However, the ‘allusive’ character of the story does notnecessarily mean that Homer relies on familiar material, a traditional storyof Heracles and Iphitus, a story which the poet wanted to accommodatebut which he did not need to fill in because both content and contextwere sufficiently familiar to the audience. We do not know what storythat would have been. Some philologists, ancient and modern, opt forthe story about Autolycus having stolen Eurytus’ mares and sold them toHeracles – or perhaps Heracles stole them himself? Some think that thekilling of Iphitus is somehow related to the bride contest in Oichalia,

although Iphitus, in later versions of the myth, comes forth as the mostfair one in the household of Eurytus. Danek concludes his discussion ofthe sources with a non liquet, but maintains that the audience would havehad to know. But perhaps what the poet said was enough – sufficient tomake sense for his audience and even for us.

Unlike the Iphitus/Odysseus episode, the Iphitus/Heracles episode hasno explanatory function in the Odyssey. But it has relevance, and it isprominently placed at the heart of the excursus as Penelope sets eye on thevery weapon that signals ‘the beginning of slaughter’ (.). As Odysseus’

Clay (: ), Schischwani (: ), Davies (: xxcii ff.). For Fernandez-Galiano (: –) the story is ‘an afterthought’ and ‘[t]he whole interpolation,

and particularly –, would have been inserted to relate the passage to the epic Sack of Oichalia;the story of Eurytus has already been mentioned in viii –’. Crissy (: –) also suggests thatHeracles’ murder of Iphitus may have had a motive ‘similar to that of Odysseus, namely, revengefor a bride’.

Danek (b: ): ‘Somit lassen sich keine prazisen Angaben uber jenes Verhaltnis zwischenIphitos und Herakles ermitteln, dessen Kenntnis beim Horer aufgrund des Anspielungscharaktersdes Textes vorausgesetzt sein muß.’ Also Crissy (: ) is reluctant to reconstruct and acceptsthat ‘[t]he poet feels no need to explain the motive for the murder’.




bow makes its appearance, the audience knows that the bloodbath in thehalls of Odysseus is drawing near. The poet is working towards a situationin which a host kills ‘guests’ who live unjustly off his property. By thematicmirroring and recombination the poem offers us a glimpse of a situation inwhich an unjust host slays a guest and keeps his property. The ‘source’ forthis episode is not to be found in tradition but in the cluster of ideas thatengage the poet’s mind as he works out the Odyssey. The story providesa sinister background to foil the Odyssean foreground. That is not tosay that the poet sketches an incident with nothing to go by. Like theassociation of Heracles with Eurytus in book , the association of Heracleswith Eurytus’ son in book is likely to depend somehow on tradition.The poet may take his clues from tradition, without ‘alluding’ to it.

The Heracles/Iphitus episode makes Odysseus stand out in contrast tothe savage hero of old, the Heracles whom we already know as a chal-lenger of gods. Some pin down more specific parallels and elicit a subtler,more ambivalent message. For example, Crissy (: , , –) finds ‘acurious reversal’ coming to light in our passage as Heracles’ deed bears astriking resemblance to Odysseus’ act of revenge against the suitors; thusthe story ‘evokes the mutual savagery of both parties in the revenge story’of the Odyssey; moreover, because Eurytus in book is the one specifi-cally portrayed as the transgressor in challenging Apollo, while Heraclesgets away with it, ‘a certain successful hybris which is like that of Herak-les and in rivalry with Eurytos’ failed attempt seems to be implied here’.Thalmann (: –), with a keen eye for parallels, finds that if thesuitors parallel Iphitus, ‘Odysseus is rather uncomfortably in the posi-tion of Herakles’; ‘[t]he roles of the suitors and Odysseus in the plot ofthe Odyssey conform to those of Ipihitos and Herakles respectively, andeven though the ethical positions are reversed (or so they are portrayed) itremains true that Odysseus does what the narrative of Iphitos’ fate roundlycondemns’. Schein (: ) finds that the ‘explicit mention’ of Heracles’murder of Iphitus at the table, which is ‘somehow analogous to the feastOdysseus prepares for the “guests” at his tables’ (cf. .–, .–),makes Odysseus’ revenge on the Suitors ‘ethically more complicated thanit might otherwise seem to be, and thus challenge[s] interpretation’. Onthe other hand, Danek (b: ) has drawn attention to the fact thatHeracles – otherwise the bowman – apparently does not use his bow to killIphitus. Precisely in a situation that represents a perversion of Odysseus’just revenge, Heracles is less of a parallel/contrast to Odysseus than he easily

For Heracles as a foil and contrast to Odysseus, see Galinsky (: ), Clay (: –), Crissy(), Danek (b: ), Thalmann (: ), de Jong (: ).



øivind andersen

could have been with the bow – for a poet intent on schematic relations.Thus the bow as a weapon is dissociated from unjust murder; the very bowthat Odysseus is going to use is a '��' [�� 1�� (.), which isnot going to be tainted by unfair use even by association. The bow isalso the one that has already been employed in the contest for a woman atOichalia. We may see significance in the fact. But even if Heracles’ wooingof Iole was already part of tradition, and even if it were well known to theauthorial audience, we may still wonder whether those who knew wouldmake the connection and find it significant. Whether or to what extent wemay expose and exploit implicit parallels in a picture of explicit contrastsis a methodological problem akin to Clay’s question ‘how can we knowwhether an allusion is rejecting a tradition or incorporating it’ (: ).And how, may we add, could the audience know? There are cases wherethe poet simply ignores a tradition, rather than emphatically rejecting it,and cases where there is no narrative tradition at all behind the ‘allusion’.

∗ ∗ ∗

In the Odyssey, Heracles’ reputation suffers from his being a foil to Odysseus,as the trespasser in archery (book ) and guest-friendship (book ) thatOdysseus is not. Also the encounter between the two heroes in the Nekuiais full of contrasting parallels. By his very descent to Hades Odysseusemulates Heracles’ greatest P�3��. The poet – or the interpolator whofurther developed the contrast – has made the two heroes ‘a pair in fate’.

He sets great store by Heracles being available for an encounter as he dulyexplains that it is in fact Heracles’ �5�� that is in Hades, while Her-acles himself ((��) enjoys life among the immortals with Hebe (Od..–). A Heracles happily married in heaven is not irrelevant to theOdyssey. The (�5�� of ) Heracles appears, like dark night, ‘holdinghis bow bare with an arrow on the bowstring and forever looking, as one

Tracy (: ) even sees Odysseus’ killing of the suitors as ‘indirect revenge for Iphitos’ treacherousmurder’; he also thinks that, since the poet mentions the death of Eurytus (.–), ‘we are meant toremember that Eurytos was killed by Apollo for challenging him to an archery contest’ in .–.

The fact that the encounter with Heracles is the final episode of the Nekuia has been seen as a kindof source quotation on the part of the poet of the Odyssey; see Heubeck (: ), on .–.For aspects of the encounter between Odysseus and Heracles, see esp. Clay (: –), Crissy(: –), Thalmann (: –).

Thornton (: ): ‘The poet here presents Odysseus and Heracles as a pair in fate: the two whowent to Hades alive, the two great archers.’

One may disagree about the wider implications. For Schein (: ), ‘Odysseus’ reference to thedeification of Herakles is one example of how the poem asks its audiences and readers to hold inmind apparently contrasting realities.’ His unspecified ‘audiences and readers’ are asked, on thebasis of these Odyssean verses, and Il. .–, to say both: ‘[Y]es, Odysseus differs from Achilles,as a nostos hero differs from a hero of kleos poetry, a survivor from one who is committed to hisown and everyone else’s death. No, Odysseus does not differ from Achilles in any freedom from thelimitations imposed (and the opportunities offered) by being human.’




who shot, with terrible glances’ (.–). Like the sinner Orion in theprevious episode, Heracles does what he typically did in life, as if undera curse. The bow-wielding Heracles is especially relevant to the Odyssey,of course. His appearance is amplified by the description of the wondrousworks on the golden belt that crossed his chest: bears, lions, boars, bat-tles and quarrels, murders and manslaughters – 1�� @ �� (.–). Odysseus in an aside expresses the hope that its maker maynever again fashion anything like it – it is as if the poet or interpolatorhere (and one may compare the aside on Heracles in Il. .–) distanceshimself from a whole tradition, which is not that of the heroic sacker ofOichalia even, but of the superhuman action hero who fights giants anddwarfs, beasts and monsters. Nonetheless Heracles himself in his addressto the younger hero draws a parallel between the two of them (note � +,.) in so far as he states that both have suffered some evil doom. Bothhave P�3�� to perform. Thus Odysseus, in his last encounter in Hades,halfway through the poem, is elevated to the same level of suffering as thegreat Heracles, who had to suffer ‘although I was the son of Zeus’ (.).

Heracles went to Hades to fetch Cerberus. He tells Odysseus that hisworst work was as he was sent to Hades ‘to fetch the dog back’ and that hesucceeded, with the help of Hermes and Pallas Athena (.–). Thisspecific divine support may or may not have been part of the story. It hasbeen suggested both that the two divinities appear here in order to makeHeracles more of a parallel to Odysseus, who is well served by Hermes andAthena in the Odyssey, and that they have been introduced to enhance thecontrast between the older hero assisted by gods and an Odysseus actingalone. Odysseus’ encounter with Heracles in the Underworld cannot bepart of any Heracles tradition. Heracles’ extended sojourn there by proxy,so to speak, is due to the role that he is allotted in the action of theOdyssey. Whatever the mythological background and the epos tradition,the Heracles we meet in the Odyssey is very much a product of the Odyssey.The audience would accept that, as they would accept and relish so muchelse that was sung to them in the course of an epic performance. The poetwould give the audience enough impulses to build a background from, tothe extent that they needed one, even if he did not and could not referthem to specific, familiar epics and stories.

On P�3�� see Thalmann (: ); Schein (: ): ‘Herakles’ diction at the end of Book , atleast as reported by Odysseus, seems to establish him as similar to, even a paradigm for, Odysseus.’

See Clay (: –) for the limits of parallelism and for the main contrasts.



chapter 9

The Catalogue of Womenwithin the Greek epic tradition

Allusion, intertextuality and traditional referentiality

Ian C. Rutherford

1 introduction

This chapter sets out to investigate the relationship between the HesiodicGunaikon Katalogos (GK) and the early epic corpus. It looks not only atcases where there is a correspondence of theme but also at cases wherethere is a striking difference, because ‘difference’ is itself an important typeof relationship. It also asks how we decide whether or not such pointsof correspondence or dissonance are significant, and explores the issue ofrelative priority between them.

The poem

The GK is unusual because it is by its very nature a compendium, a com-pilation of narratives about figures from Greek mythology, particularlystories about women. The poem is notoriously difficult to reconstruct:what survives is mostly isolated fragments, either papyri or testimonia,and often assignation of these to the poem is tentative and uncertain.Thanks to the labours of Martin West (), we know that it had a loose,genealogical structure, in outline much like that of the Bibliotheca ascribedto Apollodorus, beginning with the great family of Aeolus, which occu-pied book and probably most of book ; then moving on to the otherheroic families. The Inachids in books and , with Argive mythology,as well as the Arcadians and Atlantids. In book the Asopids, the Athe-nians and the Pelopidai and in book the wooing of Helen and declineof heroes. In this way, the GK surveyed the whole of Greek mythology,going through the stemmata, following branches to the end, and some-times jumping between then. In some cases, the ‘entry’ for someone was

Thanks to all in the conference, and particularly to Bruno Currie. This arrangement is not quite certain: Meliado () drew attention to a new piece of evidence that

suggested that the story of Atalanta, who ought to be part of the family of Aeolus, came in book .



The Catalogue of Women within the Greek epic tradition

expanded into a short myth summarizing the life of a mythological figure.Myths relating to women sometimes start y �F� (or such as she), whichgives the work its alternative name. It had its own style, its own reper-toire of favoured formulae and idioms, such as long patronymics witha genitive in -� (ones in -�� are by contrast avoided) and frequenta-tive verbs in - �. The form is essentially loose. There has been a lot ofdiscussion recently about ‘multiforms’ and ‘canonical’ epic, but the GKwould be a very good candidate for the ‘multiform’ category. It is the sortof text that would become fixed, if at all, rather late; in fact, we know ofanother text, the Megalai Ehoiai, which may be precisely a multiform of theGK.

The basic form was open to adaptation in two main ways: first, agenealogy could be changed, and second, which entries were expandedinto mini-myths could be controlled. The GK is likely to have a close, evensymbiotic, relationship to other early epic because material from otherpoems could easily be incorporated into it. Alternatively, sections of theGK could easily be adapted by other poets and made into longer poems.Either of these developments must have occurred in the case of the poemknown as the Shield of Heracles, because the first fifty lines or so actuallyappear in the canonical GK as the ehoie of Alcmena.

The view that the GK is by Hesiod himself is an unusual one today. Theorthodox view is that the canonical form arose much later than Hesiod,being composed around bc, the date of the First Sacred War, since themember-states of the Delphic Amphictyony are supposed to correlate wellwith the Aeolid family which makes up the first and largest part of thepoem. This is Robert Fowler’s view (), and also that of Jonathan Hall(). He argues that the context was the emergent idea of Greek identityin this period; I worry, however, that the Greek notion of hellenicity, whichalready seems to be there in Homer, cannot have developed as late as

bc. West (: ) puts the canonical form of the GK even later, say inthe mid sixth century bc, stressing the presence of a few apparently Atticelements in the GK (e.g. the participle �? �). A second point on whichthere is broad agreement today is that the GK went through a long processof development. The latest version may have been Athenian, but there

See Finkelberg (). See D’Alessio (). See Martin (). Janko (this vol.) and Malkin () have suggested that parts of the GK reflect early Greek colonization

in the West, e.g. in the Hesiodic fragment preserved in the scholion to Ap. Rhod. Argon. . inwhich Circe came to the island over against Tyrrhenia, ‘and she called it Hesperian’ (fr. M–W(though the clause translated is not included) = fr. Loeb (not in Hirschberger (), as far as Ican see)). See Malkin (: ).



ian c. rutherford

could have been a sequence of earlier versions: the proto-catalogue, thecanonical version or the post-canonical version, or various proto-versions,e.g. the proto-version with just the Aeolids. I have some misgivings aboutthat hypothesis, which I fear may be a scientific way of saying that wedo not know when it was written, and of keeping the option open for itsrelation to other poems, so that, whenever we date Homer, we can alwayshave some version of the GK before it and also some version after it.

Critical concepts

Two things about early Greek poetry are generally agreed:(a) Early Greek poets or oral singers composed against the background of

a tradition of poems by other poets or singers. Like all ancient writers,they may have engaged in ‘creative imitation’ of the model of otherpoets’ or singers’ compositions, or they may have reacted against apoem or tradition they knew, generating an original and distinctivecomposition by an act of distancing.

(b) Poets intended audiences to identify relationships (similarities or dif-ferences) between the poems to which they were listening and otherwell-known poems or poetic traditions with which they were familiar.They may have deliberately included signposts to this end. It is alsopossible that audiences will have identified relationships other thanthose intended by the poet: the reception of a poem has a life of itsown.

In theory, we could have either (a) or (b) without the other. Poets mayhave used models without intending the audiences to appreciate thisfact; and audiences may have appreciated poems in the context of otherpoems they knew when the poets themselves had in fact composed denovo (inspired by the Muses, so to speak). But in practice (a) and (b) gotogether.

What is less widely agreed is how to describe such relations. Mod-ern scholarship has a panoply of terms: among the more traditional onesare ‘allusion’ and ‘(creative) imitation’, the latter at least to some extentgrounded in classical texts; and among the more recent ones are ‘intertex-tuality’ and ‘traditional referentiality’. Thirty years ago, the general viewwas that reference by one poet to the works of another was, first, a phe-nomenon found principally in ‘writerly’ traditions where poems could beconsulted in written form, and, second, a special literary effect, not part of‘ordinary’ poetic language. One poet might ‘allude’ to another to displayhis or her place in a literary tradition and his or her learning (Thomas




). Adaptation of another writer’s lines might be seen not as lazy theftbut as an improvement on the tradition (Russell : ).

This model has been challenged on two fronts in more recent scholarship.First, there is the model of ‘intertextuality’ which takes into account allsimilarities between texts, whether they were intended by the author ornot, and whatever the chronological relationship between the texts; andnot just between texts, but between all instances of linguistic utterance.Intertextuality claims universality both in so far as its scope is any text orutterance, but also in so far as it claims to be a distinctive feature of theway that semiotic systems work, whereas ‘allusion’ has traditionally beenseen as an ‘add-on’ (D. P. Fowler : ). Taken to the limit, the modelof intertextuality implies a different way of reading – of all texts togetherrather than one at a time. It follows that the criteria for intertextuality arecompletely different from those for ‘allusion’: you do not have to provethat the author intended us to appreciate it, and it is permissible to ignorefactors such as the apparent chronological order of composition. Some ofthe claims made for intertextuality defy common sense, and are meant to.Anyone who tries to work with the model will inevitably have to limit itsscope.

The second challenge comes from the model of ‘traditional referen-tiality’, a term coined by Foley (). It means ‘the power of traditionalelements to evoke a set of connotations determined by the entire tradition’

and is meant to apply to oral traditions, where we cannot talk of specific‘texts’ but of evolving traditions realized in individual performances. Thisseems, on the face of it, quite different from text-based allusion, creativeimitation and intertextuality (although, as we have seen, intertextuality isnever confined to texts). It shares with intertextuality the assumption thatlooking for clear chronological relations between different works is futile,because the works are simultaneously evolving. There remains, however,a clear watershed between writerly poetics, where one poet can allude toa specific phrase in a specific work by another, and oral poetics, in whichsuch specific references are supposed to be impossible. I wonder, how-ever, whether the claims made for the essential difference between oraland written poetics are sometimes exaggerated. Even in oral traditions theform of one particular poem, even if not wholly and permanently fixed,may still be stable enough over a given period of time to become recog-nizable for an audience, and thus to be the object of allusion or creativeimitation.

Heath (: ), see Kelly (: –).



ian c. rutherford

2 points of correspondence between gk and

the epic tradition

General considerations

The GK is likely to stand in a closer or more remote ‘intertextual’ rela-tionship to all forms of early Greek epic. For obvious reasons it is eas-iest to talk about its relationship to the Trojan War Cycle, particularlythe canonical Homeric epics. It also stands in a special sort of intertex-tual relationship to the works of Hesiod. For the poorly attested ThebanCycle, and the hypothetical archaic Heracleid and Argonautica, it is almostimpossible to draw conclusions. To take the case of the Thebaid, the GKcovered some of the main themes – the funeral of Oedipus ( M–W =

H), for example – and there is a striking resemblance between the lineused to describe Amphiaraus at . M–W = . H: /� _@ �43.� '��C�� 4�� 43.� �� '�� 3� (who was good at speech and good atfighting) and a description of him in one of the few lines attested fromthe Thebaid: �'1�� '�� @ �43.� � ��#�� '�� 3� (botha seer and good at fighting with the spear, B). More significant per-haps is the suggestion of Beck (), on the basis of an interpretation of M–W = H, that the GK uses the name ��4�� for Argia, thewife of Polynices, and that this usage, also found later in Antimachus ofColophon, is likely to have come from the Thebaid. In that case the name��4�� in the GK might be ‘resonant’ for contemporary audiences,although it is worth bearing in mind that the GK also uses the same epithetfor Helen of Argos (a M–W, cited below).

Occasional intertexts are also found between GK and the HomericHymns, the composition of which many scholars would argue was closein time to that of the canonical GK. For example, the account of thedaughters of Leucon in the GK contains a description across several lines ofthe course of the River Cephisus through Boeotia. The description begins:/� �� *�� B�� (which sends forth fair-flowingwater from Lilaia (. M–W = . H)) and then mentions the river’seastward course past Panopeus, through Orchomenus, and presumablyinto Lake Copais. Line occurs verbatim (except with *��3��) in theaccount of the young Apollo’s journey through Boeotia in the Pythian

The task of surveying points of similarity or dissonance between GK and the epic tradition has beenmade a lot easier thanks to Martina Hirschberger’s recent commentary (: esp. ff.). Meier() is also a useful source in this area.

See Hirschberger (). Not accepted in Hirschberger (). For the date of the Homeric Hymns, see Janko ().




part of the Homeric Hymn to Apollo (), where, stretching geographicalcredulity, he crosses the River Cephisus (north-west of Lake Copais) ashe comes from (the later sites of ) Thebes and Onchestus and before hereaches Ocalea and Haliartus (all to the south of Lake Copais). The twocontexts in which the line occurs have little in common except that theyare both itineraries – one the itinerary of a river, and the other that of agod who crosses the same river. If a derivational relationship exists betweenthem (which is by no means certain), it is surely less likely that the authorof the GK is exploiting a locale already associated with Apollo in the Hymnthan that the author of the Hymn is excerpting from an established epicekphrasis about the Boeotian river.

The Hesiodic corpus

The GK was believed to be by Hesiod in antiquity and it begins as acontinuation of the final lines of the Theogony (lines which some haveargued are not integral to the original Theogony). This means that anyallusions to Hesiod’s Theogony in the GK would have a special status: beingcases not of intertextuality but of intratextuality. Even if the GK is not byHesiod, it may well be that early audiences considered it to be ‘Hesiodic’, sothat allusions to Hesiodic works within it would be interpreted as markersthat it belonged to that tradition.

Having said that, comparatively few allusions can be detected to theTheogony and Works and Days. West (: –) finds another Hesiodicinter-/intratext in the fragmentary description of how the daughters ofPorthaon were dancing around a fountain when Apollo carried off one ofthem to be the bride of Melaneus:

F _ ��. @ �.[5]�� 4�[��'�� ]�� '1� �� .�. .. c . . . [��# ��4]#��I�� 〈�〉0.>�. [� (.– M–W = .– H)

. . . who then delighting in their beauty and ignorance went among the mist (?)around the source of silver-eddying Achelous

He argues that this is an imitation of the lines describing the Muses in theopening of the Theogony (–, –, ), and a rather clumsy one at that:

The Aeolid section of the GK begins with the impressive line: t�� @ 74�� 3�'� ��> �� . . . (The sons of Aeolus, kings concerned with just rulings, were . . . (a. M–W = .

H)). The epithet 3�'� �� occurs otherwise only in the Homeric Hymn to Demeter (,), and in rather different applications. See Richardson (: – ad loc.) on the issue of therelationship between this epithet and the near-synonym �� .

Hirschberger (: ff.).



ian c. rutherford

for example, the adjective I�� is an echo of Theog. (��#''�� I�� – veiled in �:�, i.e. invisible). West’s position on these lines hasrecently been challenged by Burgess (: –), who sees this as a casewhere it is appropriate to talk more of a general relationship to the epic tradi-tion than of one to a single text in particular. On the other hand, West’s posi-tion receives support from the fact that the alleged intertext here comes fromthe opening lines of the Theogony, the part of a poem generally believed tobe most likely to be the object of allusion (Garner : and –).

There is little sign in the extant fragments that the GK referenced eventhemes in the Hesiodic works that might be thought to be of specialrelevance to the GK: Pandora, whom the Works and Days (–) andTheogony (–) present as a proto-woman of ambiguous significance,seems to have been demoted in the GK to the role of a daughter ofDeucalion ( M–W = H), unless the GK gave that name also to themysterious mother of Deucalion by Prometheus (, M–W). Even morestrikingly, there is no trace of the myth of the Four Races, except for anapparent allusion to the fate of the race of heroes/demi-gods (Op. ) inthe ‘twilight of the heroes’ section (. M–W = . H); Koenen() takes this as a reworking of the Hesiodic Races in the light of NearEastern myths about cyclic destruction, in which the end of the previousworld age is marked by the departure of the gods from the Earth andthe company of men; the author of the GK has replaced them with thedeparture of the heroes (Rutherford : ). The fragmentary proem ofthe GK ( M–W = H) seems to have set out the special conditions ofthe heroic age, where gods and men enjoyed ‘common feasts and seats’, inother words a closer relationship than they do today, but although this is ageneral feature of the Golden Age in Greek mentality, it has no specificparallel in the Theogony or Works and Days.

The Trojan Cycle

The Trojan War is for the most part a terminus that the GK does not cross;the climax of the GK was probably the ‘wooing of Helen’ in book , whichculminated in a description of the end of the heroic age. Occasionally the

See Hirschberger (: –). See Hirschberger (: –), with secondary literature cited in n. . Tsagalis (: ) sees the mention of the fact that Achilles was too young to woo (.–

M–W) as standing in an intertextual relationship to the passage in the Cypria where Achilles seesHelen (for which Gantz : ). He also sees evidence for the bridal competition for Helen beingreprised in Il. .




poem looks forward to the War, e.g. in M–W = H, where Sarpedongoes to war, despite negative omens, and in M–W = H, where theGreeks are repulsed by Telephus in Mysia. Another reference forward tothe Trojan War comes in a M–W = H, which described the sacrifice ofAgamemnon’s daughter Iphigenia (here the variant name Iphimede is used):

@x1�'�� '�� 1�[� 7o�:[']�� >�'�[� C� @ ��'�� # ��]�. �[�#] ��^'�[� �� /�� # �� ]��. u x�� .[5 ��V[� �� '�� ]1+��# ��4��. [��]�.�

�5��[��2 (�V� � @ 7�1�>�]��. � ��_�0 '�� @ 7[� �[� �� '>�� ]�� [7�]�. �.[��V� ��[� �, ��[3�� F� �< �]�.g�. [C]'.��[�]�. [�] �.[5��3�� @ �3��[�] � �4:�]�� ^'[� ��.�V� �V �8� ��[# �� 7�� ]3.�� 18� @ ��. [3��

h��'�� [�� #]��8 �[�]�[�]��[��.

The well-greaved Achaeans slaughtered Iphimede on the altar of golden-spindledresounding Artemis on the day when they sailed to Troy taking vengeance for thebeautiful-ankled Argeione, an image. Her the goddess saved easily, and drippedlovely ambrosia over her head so that her flesh was unchanging, and made herimmortal and ageless for all days. Her the tribes of mortal men now call ArtemisEinodie, the attendant of the glorious arrow-shooter.

In the Cypria Iphigenia is transported by Artemis to the Tauri at themoment of sacrifice. The myth also occurs in Stesichorus ( PMG), whohas Iphigenia becoming Hecate, a mytheme also attributed to the Ehoiaiand Hesiod (b M–W, from Pausanias and Philodemus); as far as theGK is concerned, Hecate must represent an interpretation of the epithet‘Einodie’ (Gantz : –). In addition, there is a striking intertext withthe passage towards the end of Odyssey , where Odysseus sees Heracles,or rather ‘an image, but he himself (�5��2 (�.� ��) was on Olympus,married to Hebe’ (.–). This looks like the same formula as in fr.a., although it should be noticed that the supplement ��[��# in thegenitive, in apposition to ��4��[��]�, is also possible.

In some cases the GK even points beyond the war: it alludes to therevenge of Orestes, and may have related the union of Telemachus andPolycaste ( M–W = ∗ H). But mostly the Trojan War is the terminus,

In this passage Telephus has the rare epithet �� , ‘of the family of Arkas’, an epithet nowrestored in a description of Telephus in the new narrative-elegy attributed to Archilochus (P.Oxy.): does this coincidence suggest that the formula W:��1�� occurred earlier in theCypria?

See Hirschberger (: ad loc.).



ian c. rutherford

which could be seen as reflecting a poetic strategy not to compete with theTrojan Cycle, just as it purports to be dealing with the lives of women,rather than with �� which is the theme of epic. Thus, the themesof the GK would be chronologically and thematically distinct from heroicepic, as I argued in an earlier essay.

The Iliad

To turn to the Homeric poems themselves, the best place to hunt for inter-textuality between the GK and the Iliad is in the final section, which nar-rated the wooing of Helen and the ‘end of the heroic age’. In M–W =

H, line we have an account of Zeus’s plan to cast (��) theheads of many heroes to Hades. The text is a little unclear, but it seemslikely that we have an echo of the verb �� at the start of theIliad. The list of suitors in the ‘wooing of Helen’ itself recalls the IliadicCatalogue of Ships, a topic that has been explored recently in an articleby Cingano (). Numerous heroes come to take part, including Ajaxof Salamis: t5� � @ 7 Q�'0�� '�'.�� '. �. . �V� | '�� (Ajaxfrom Salamis, the blameless warror, came wooing (.– M–W)).Ajax promises to give a dowry comprising a number of towns on themainland, mostly in the area of the Argolid, were he to win. The firsttwo words remind us of the entry for Ajax in the Homeric Catalogue ofShips: t5� � @ 7 Q�'0�� P4�� +� � �� | �� @ P4�� F� @�3�� 5 �� 1��44�� (Ajax from Salamis brought twelve ships,and set them where the battalions of the Athenians were standing (Il. .–)). This is likely to have been a celebrated couplet at certain points in thedevelopment of the early Greek epic tradition. The relationship betweenthe two texts seems even more complex because the cities of the Argolidlisted in GK seem to correspond to some extent to the contingent of Argosin the Catalogue of Ships, led by Diomedes and Sthenelus, which comesimmediately afterwards. Some have thought that the entry for Ajax in theGK signifies the growing power of Athens at the time of its compositionbecause Ajax was a hero of special significance to Athens. Others haveargued that Ajax’ entry preserves an earlier stage in the epic tradition; itwould be the Homeric version with a small kingdom for Ajax which is later.Another view is that Ajax is just boasting: he does not own these cities atall, but boasts that he will conquer them if he can. In that case, the poetof the GK is not after all at variance with the Homeric Catalogue, and the

Rutherford (). Also at Il. .. See Cingano ().




formula t5� 7 Q�'0�� serves to alert the audience to the Homericparallel. If that is right (and it is a big ‘if ’), this would be a case where theGK creatively alludes to the lliad.

The Odyssey

The richest of all intertexts for the GK is the Odyssey, particularly book ,where Odysseus relates his own 4#�� 4��, comprising famous(and infamous) women of the past. There are strong verbal echoes betweenthe accounts of some of the women Odysseus sees and the correspondingaccounts in the GK, particularly apropos of the women: Tyro and Chloris.It is as if the poet of the Odyssey knows such a work, and is reproducingsections of it.

According to the GK Aeolus has six sons and six daughters. One ofthe sons is Salmoneus, described as P�� (a. M–W = . H) and�� 3�� (. M–W = . H), who behaves like a god, hanginganimal skins from his chariot, and claiming that he makes thunder. Even-tually, Zeus strikes him down, together with his house and city, leavingonly his daughter Tyro, because she has taken issue with her father.

��8 � @ P�] �0� 7�� 1�� '�� 3��0 �W#�g 7o�]��'�� [�]# �� 1��[�]��[�� ]�� . . �. [� �] Q�'�� #�� (]� @ �5 � 3��0� [>��.� � ]�1��H�� . . .

(.– M–W = .– H)

His daughter was left, dear to the immortal gods, Tyro of the fair tresses, similar togolden Aphrodite, because she quarrelled and disputed with Salmoneus continuallyand did not consent to set a mortal on a par with the gods . . .

So Zeus saves her, and takes her to the house of her uncle Cretheus. Whenshe grows up (reaches the end of ��#:�� \>�, a formula much usedin GK), Poseidon falls in love with her (7�� ), and he seduces her whenshe is walking on the banks of the Enipeus River. Then the text breaksoff, and there is a prophecy by Poseidon: ‘you will bear splendid children,for the embrace of the gods is not in vain . . . care for them . . . and concealthe matter’ ( M–W = H). She goes home, and gives birth to Neleusand Pelias. Zeus sends Neleus to Pylos, where he marries Chloris, andthey have twelve sons including Nestor and a daughter Pero (a M–W =

H). But things do not end there. Heracles decides to go to war on the

See S. West (this vol.).



ian c. rutherford

Neleid family; one of the Neleids, the shape-shifter Periclymenus, managesto delay fate, but finally Heracles wins, and all the sons are killed, except forNestor, who escapes because he is ‘among the Gerenoi’ (, . M–W =

a, c H). After this comes the story of the wooing of Pero by the seerMelampus, after he steals the cattle of Iphiclus. And finally the narrativefolds back to Tyro’s other son Pelias in Iolcus ( M–W = H).

There are strong parallels with the ghosts of the famous women of thepast that Odysseus sees in the Underworld. Tyro, the first woman that hesees after his mother, tells her own story in words that strongly resemblethe GK ’s account (.ff.): again Poseidon falls in love with her but onlyafter she herself has fallen for the River Enipeus; Poseidon comes to her indisguise, she bears him Neleus and Pelias, and then she has three sons withCretheus. Significantly, she, or Odysseus, omits the story of Salmoneus,who is connoted with the surprising epithet �'+'��, just like Aegisthusearlier on in the Odyssey. A few female ghosts later Odysseus sees Chloris,who is said to have married Neleus and given birth to a mere three sons,not twelve, and the daughter Pero, which leads to a brief account of thewooing of Pero by Melampus (.ff.). The Odyssey omits the story of thewar with Heracles, although it is narrated in the Iliad (.–) as oneof Nestor’s reminiscences.

The key question is: what is the relationship between these two texts?On the face of it either one draws on the other, or they draw on a commonsource, perhaps a hypothetical ‘Neleid’, from which many of Nestor’sreminiscences in the Iliad might also derive. One point that countsagainst the last possibility is that the narrative about the Neleids in Od.

comes in the context of a catalogue of famous women seen by Odysseus.Unless there was a similar catalogue in the ‘Neleid’, it follows that the linkbetween Od. and the GK was a particularly close one.

If the GK dates to the sixth century bc, and Homer is considerablyolder, one possibility would be that the author of the GK took sectionsof Od. (and part of Il. ), adapted them slightly, and expanded them,filling them out with additional material to create a sort of ‘back-story’.One factor that points in this direction is the line about Nestor havingescaped because he was among the Gerenoi (. M–W), which looks likeit must be an explanation of Nestor’s Homeric epithet ��:��; on theother hand, if we allow that ‘Gerenian Nestor’ was a formula with a generalcurrency in the epic tradition, that would not point to Homer particularly(see table .).

For the formula, see Parry (). For the ‘Neleid’, see Bolte (). Brillante () thinks this explanation is historically accurate, pointing back to the political

geography of Messenia in the Late Bronze Age.




Table . The story of Salmoneus and Tyro in GK and Odyssey 11 compared

GK Od.

a. Arrogance of Salmoneus .– M–W –b. Punishment of Salmoneus .– M–W –c. Tyro saved .– M–W –d. Poseidon and Tyro .– M–W –

cf. M–We. Genealogical prophecy of Poseidon

about Neleus and Pelias M–W –

Birth of Neleus and Pelias –

Children of Tyro by Cretheus –

f. Neleus and Chloris and their twelve sons .– M–W – ( sons)g. Periclymenus the shape-shifter .– M–W –h. Periclymenus and Heracles a.– M–W

b M–W–

i. Sacking of Pylos and death of eleven sonsof Neleus

.– M–W cf. Il. ., .

j. Children of Nestor .– M–W M–W (a double?)

–

k. Wooing of Pero .– M–W –

l. Family of Pelias .– M–W –

The alternative hypothesis that the Odyssey draws on the GK is suggestedby three factors. First, the use of the epithet �'+'�� for Salmoneus: whileone could argue that this is insignificant since (or if ) this is oral poetry,where the metrical properties of epithets are more important than theirmeaning, it is surely more likely that the Odyssey knows all about thereputation of Salmoneus, and uses the epithet deliberately: Tyro is gamelydefending her father, like any loyal daughter would. The second factoris the list of the sons of Neleus and Chloris: the GK has the full list oftwelve names over four lines, while the Odyssey has just one line of threenames, identical to one of the lines in the GK (=� �� '�� +'�� @ �4��). It is surely a more reasonable hypothesis thatthe Odyssey excerpts from a longer list familiar from a model than thatthe GK augments the shorter list in the Odyssey. And thirdly, there is thereference to the story of the wooing of Pero by Melampus in Od. , whichis obscure and allusive.

If we want the GK to be a model for the Odyssey, then some ver-sion of the GK must be earlier than Homer. This could be either a

�,� � @ �)�� S�� '�� '+'�� | 7[��2 ��V �� 3��8 �, '�0� �� | �� '�� @��4�� >�#�� 4��. (These the flawless seer alone undertook to drive off; but aharsh fate of the gods ensnared him, hard bonds and the country herdsmen (Od. .–).)



ian c. rutherford

proto-GK, distinct from the final version that was written in the sixthcentury (and which might in turn have been influenced by the Homerictext: cf. ‘Gerenian Nestor’), or it could be the canonical GK, which Jankoargues is by Hesiod himself. Either way, the pre-Homeric GK could beeither an oral or a written poem. It might be argued that only in the lattercase is it permissible to talk about ‘creative imitation’ of a written model,and that in the former case we should use the term ‘traditional referen-tiality’, but should we really imagine that an early, orally transmitted GKwould have been so unstable and transient that the author of the Iliad andOdyssey could not have used it as a model?

Contested genealogemes

There are a number of points where the genealogy of a hero seems to differbetween the GK and Homer, some of them involving some of the mostimportant families in the poems. Take, for instance, the family of the Trojanfounding-hero Dardanus, which may be linked to Samothrace in the GK( M–W = H]), but not in Homer (Il. .). Then there is the familyof the Atreidae. If you read Homer, you get the distinct impression thatAgamemnon and Menelaus are the sons of Atreus, but we find attributedto the GK the opinion that they were the sons of a certain Pleisthenes( M–W), who, according to one of our sources, was a hermaphroditeor lame, or wore women’s clothing. Possibly the GK made Pleisthenes theson of Atreus, and for the genealogist this had the advantage of making thestemma of the Atreid branch of the Pelopids one generation longer, whichhelped to even things out with other stemmata. But what about Homer?Did he know this tradition but not deliberately exclude it? Or did he notknow it?

Then there is the family of the Aeacidae. It looks as if the GK knowsthe tradition familiar to us from Pindar, which included the followingelements: (a) the family came from Aegina; (b) Ajax and Achilles were firstcousins since their fathers were brothers; and (c) Patroclus was related tothem also (in fact the GK seems to have said that Patroclus was anothercousin). The Iliad ’s position on the family of Aeacus is quite different.It says nothing explicitly in contradiction to any of this, although .– seems to suggest that the three heroes are unrelated. But it does not

Hirschberger () discusses the fragment on p. . For the background, Peersman () andGantz (: , ).

Aeacus and Aegina: M–W = H; Menoetius a brother of Peleus: a M–W. The story occursfirst in the Alcmaeonis (fr. Bernabe).




Homer GK

Aeolus

Sisyphus

Bellerophon

Hippolochus Laodamia = Zeus

Sarpedon

Atlas

Zeus = Europa

Minos Rhadamanthus Sarpedon

Glaucus

Glaucus

Fig. . Genealogies of Sarpedon.

confirm it either, and one must remember that there are many points inthe Iliad where it would have been relevant to report that Ajax, Achillesand Patroclus were closely related. Again, the question arises whether theclassical version of the Aeacidae-saga had not taken shape at the time theIliad was composed, or whether Homer knows the other tradition, butignores it (perhaps for dramaturgical reasons).

Another interesting case is the family of the prophet Melampus, whichincludes Theoclymenus, and seems to have been different in the GK, tojudge from West’s brilliant reconstruction of the rather difficult fr.

(: –). West illuminates this passage with reference to Il. .ff.,where another seer, Polyidus of Corinth, warns his son Euchenor not togo to Troy. West suggests that the son Euchenor appears in the stemmaof Melampus in the GK, and the GK also has the warning of Polyidus,except that here Polyidus is not the father. Here is a promising text for thepurposes of exploring intertextuality between GK and Homer, then, butunfortunately it is too scrappy to support much argument, and noticealso that Hirschberger () doubts whether it belonged to the GKat all.

The last case I want to consider is the genealogy of Sarpedon of Lycia(see figure .). In Homer, he is a grandson of Bellerophon, in a complexfamily tree that includes his friend Glaucus. This is told in Il. , in thecourse of the famous dialogue between Glaucus and Diomedes. But in GKthis is all different: Sarpedon is a son of Zeus and Europa and brother of

Niemeier () is a reminder of how ancient the traditions of Aegina could have been. She classes the fragment as ‘of uncertain attribution’ (∗).



ian c. rutherford

Minos. This is narrated in M–W = H, an important fragment whichdescribes Europa and the necklace that Hephaestus made for her, later tobecome the necklace of Eriphyle, and then goes on to describe her sonsby Zeus: Minos, Rhadamanthus and Sarpedon. Zeus grants him to livethree generations (ll. –), and he goes to fight in Troy, despite his fatherZeus sending a bad omen (�� , :'� 1�� (l. ), similar to aformula that occurs a few times in the Iliad ). He rules Lycia, and muchhonour follows him (��V �� < C �� ': (l. )), a rare formula thatis used in the Theogony () to describe someone who sacrifices to Hecate,but in the context of Sarpedon inevitably brings to mind Il. ., whereSarpedon asks Glaucus rhetorically ‘why are we honoured?’ (��8� ��V �� ':'� 3 '�� ;).

Notice also that the genealogy Sisyphus–Glaucus–Bellerophon appearselsewhere in the GK in the Mestra ehoie ( M–W = H), with theadditional detail that Bellerophon’s father was really Poseidon who hadabducted Glaucus’ wife.

It is interesting that the Iliad seems to go out of its way to tell us about thegenealogy of Sarpedon. The statement of it comes in the famous episodewhere Glaucus tells Diomedes about his genealogy, the genealogical passageof the Iliad par excellence. Elsewhere, the Iliad is quite explicit that Zeusand Europa had two sons only, Minos and Rhadamanthus, as we hear fromthe mouth of Zeus himself in the Dios apate scene (.). So there is nomistake: Sarpedon is without doubt a son of Zeus, only not by Europa,but rather by a female descendant of Sisyphus. Of these two versions ofthe genealogy, can we say which is likely to be earlier? One factor mightbe that the Homeric stemma of Sarpedon suits the plot of the Iliad betterthan the one in the GK, in so far as, if Sarpedon was the brother of Minos,he would be very old by the time of the Trojan War. (That is why in someversions of the myth he was allowed to live three generations.) But that sortof superhuman lifespan would not suit the taste of the poet of the Iliad,so, I suggest, he opted for an alternative genealogy, grafting him on to thestemma of Bellerophon. It may be precisely because Homer has changedthe genealogy of Sarpedon that he explains it in detail. If that approach isright, it has the consequence that the version of the myth preserved in theGK is the older one. That is not to say that the GK itself is older, of course,and the situation is a complicated one, because the canonical GK seems tohave combined the earlier Cretan genealogy of Sarpedon with themes ofthe Iliad.

See the supplements in Evelyn-White (: ); for the myth, see Apollodorus, Bibl. ...




3 conclusions

Since the GK is a compendium of mythology, it is no surprise at all that itoverlaps with the ground covered by other early epic poems, and predictablywe find plenty of plausible cases of intertextuality. The chief problems arethe fragmentary nature of the material – both the GK and the other poems –and our ignorance about the sort of poetry we are dealing with: developingoral traditions, fixed written texts, or something in between. In some caseswe seem to be dealing with contestation not between different poems butbetween narrative or even mythological traditions, e.g. with respect to someof the ‘contested genealogemes’ I discussed; in these cases it is probablybetter to speak about ‘traditional referentiality’. Equally, however, there arecases where it makes more sense to speak in terms of relations between(specific) poems. That would certainly go for cases (if there are any) wherethe author of the GK reacts to Homer; the Ajax-entry in the Wooingof Helen fragment might be one such case. ‘Relations between (specific)poems’ also seems a better way of describing the hypothetical use by thepoet of the Odyssey of a ‘proto-GK ’, which may have been written but wasjust as likely orally transmitted.



chapter 10


Jonathan S. Burgess

Intertextuality is problematic for early Greek epic because it was performedin the largely oral culture of the Archaic Age. An orally composed poemcannot easily engage with another oral poem in a detailed manner. Evenif one assumes that some epics were recorded as texts at an early date,their normal publication would be by performance, where specific allusionbetween poems is not easily discernible. Anachronistic assumptions aboutepic-to-epic allusion in the early Archaic Age fail to consider whetherthere was any motivation, in terms of both composition and reception, fordetailed intertextuality. More plausible is interaction between epic poemsand mythological traditions as they were generally known. But reflectionof shadowy tales might be thought a rather nebulous and disappointingtype of intertextuality. General correspondence of motifs is one thing; whatabout words and phrases? Is there such a thing as quotation in early Greekepic?

In this study I explore the possibility of a textless intertextuality inearly Greek epic that would involve specific epic phraseology. By this Ido not mean an oral poem reusing words that have been composed fora previous oral poem, but rather an oral epic reusing phraseology thathas become associated with specific mythological situations as they weretraditionally articulated in the oral epic tradition. Two Homeric phrases

I thank the conference organizers for inviting me, and I am grateful to the Social Sciences andHumanities Research Council of Canada for financial support. Special thanks to Jennifer Phenix,Tim Perry, Lee Sawchuk and Michal Dziong for their work as research assistants. Communicationwith Ettore Cingano, Bruno Currie, Ian Rutherford and Oliver Thomas sharpened my thinking.

In a recent publication (Burgess , revised as ch. of Burgess ), I limit Homeric allusionto motifs at the mythological level, and question the neoanalyst interest in ‘quotation’ (Burgess: ). Here I seek to explore the issue of epic quotation more optimistically, though withinoral parameters. Cf. Nagy (: –), with a statement that has gained much attention: ‘[W]henwe are dealing with the traditional poetry of the Homeric (and Hesiodic) compositions, it is notjustifiable to claim that a passage in any text can refer to another passage in another text’ (; withfurther development at Nagy : –, : –). For (general) agreement, see Jones (: );for critique, see Clay (: –), Cairns (a: –), R. B. Rutherford (: ).




will be considered as candidates for such intertextuality without text. One isthe phrase '�4� '�4�� , or ‘great in his greatness’, found in referenceto Achilles in book of the Iliad. Neoanalysts have argued that the phraseis derived from a pre-Homeric poem about Achilles, perhaps a prototypeof the Aethiopis of the Epic Cycle. A second test case involves a line of afragment of the Little Iliad of the Epic Cycle that describes Neoptolemustaking Astyanax from the breast of his nurse. The language is reminiscentof the famous scene in Iliad in which Astyanax recoils into the breastof his nurse at the sight of Hector’s helmet, and scholars have usuallyassumed that the Cyclic poem has taken its phraseology from the Iliad.I will argue that in neither case is there a direct connection betweenone epic poem and another. Instead, I suggest, the Iliad reuses context-specific phraseology to remind a mythologically informed audience of thenarrative context in which it usually occurs. By such intertextuality withouttext the book passage alludes to the traditional story of Achilles’ death,and the book passage alludes to the traditional story of the death ofAstyanax.

1 methodology

The following comments are meant to clarify the assumptions that I willemploy in my subsequent discussion. Though I am gratified to see thatsome contributors to this volume have responded thoughtfully to mypresentation at the conference, my points are not intended as ‘rules’. Theyare designed, rather, to provide a heuristic strategy for my explorationof the possibilities of intertextuality in an oral performance culture. I donot insist on an everlasting and pure performance culture, and I readilyacknowledge that a wide range of circumstances involving a mix of textsand oral performance can easily be imagined. But since the relationshipbetween early epic poems is textually conceptualized by default, it shouldbe useful if I stake out a position that is more attuned to performanceculture.. ‘Intertextuality’ commonly means little more than allusion of one text

to another. This does not seem to be appropriate for the oral nature ofearly Greek epic, and a more theoretical conception of intertextuality isdesirable.

For the Epic Cycle and its tradition, see Burgess (, ). In my usage ‘cycle’ and ‘cyclic’, whencapitalized, refer to the specific poems eventually collected into the Epic Cycle, whereas withoutcapitalization they refer to mytho-poetic traditions whose content is also found in the Cycle.

For more detailed discussion, see Burgess (), with Tsagalis ().



jonathan s. burgess

. For traditional narrative, a distinction can be made between mytho-logical traditions (represented by various genres and media) and epictraditions (the manifestation of mythological narratives in epic verse).

. Oral epic traditions were not constituted simply of formulae and typescenes, but also of fluid yet identifiable mythological narratives. Oftencertain phraseology would be employed typically in the context of spe-cific mythological episodes.

. The early epic tradition was not limited to surviving or attested epics,with the addition of a few oral prototypes. There were countless epicperformances that narrated mythological traditions by means of oralcompositional techniques. The poems of the Epic Cycle are representa-tive of, but not equivalent to, the non-Homeric epic tradition.

. Homeric and cyclic epic can be distinguished, though this does notisolate the former so much as establish a context for it. The Homericpoems are ‘meta-cyclic’ in the sense that they transform cyclic motifsand phraseology into new contexts.

. Correspondence of material or even phraseology in the early epic tradi-tion need not be conceptualized as poem-to-poem intertextuality. Mostoral poems would not have been performed so often as to influence otherpoems, or be the object of allusion.

. When performed to a knowledgeable audience, early epic is poten-tially allusive to shared aspects of mytho-poetic traditions, includingmythological narratives and the epic phraseology commonly employedto express them.

But what of relative chronology, the issue with which this volume is con-cerned? Of course, it is of great interest if an epic contains datable aspects,whether linguistic, historical or cultural. But even when we are confident

Cf. Finkelberg (a: –) on Homeric poetry as ‘metaepic’. Intertextual correspondence between the Iliad and Odyssey may be plausible since the two Homeric

poems’ shared characteristics suggest some commonality of composition and reception. The bibli-ography for Homeric intertextuality (usually the influence of the Iliad on the Odyssey) is enormous;I might single out Pucci (), Usener (), R. B. Rutherford (), Schein (, ), Currie(: –).

Of course some poems became influential, at least eventually; see, for example, Bruno Currie andIan Rutherford in this volume for careful and nuanced explorations of this possibility. The majorpoint of my Cycle book (Burgess ), however, was that the influence of the Homeric poemsoccurred later than is commonly assumed. ‘Obvious’ early examples of textual influence are worthre-examining; cf. Martin (: –, : –), Burgess (: –).

See Andersen (), Scodel () for more sceptical approaches that raise important questions. E.g. Achilles’ afterlife at the White Island in the Aethiopis has often been linked to Milesian colo-

nization in the northern Black Sea, which suggests it is a post-Homeric narrative. See Burgess (:–), where it is concluded that the mythological content of the narrative may be pre-Homericeven if the Aethiopis presented a Milesian perspective.




that we can date two texts, we cannot necessarily extrapolate on that basis anintertextual relationship between them. One poem considered older neednot have influenced a second poem considered younger, even if they dis-play correspondences, for these could have been inherited independently.The issue becomes more complex when we consider the possibility that apoem existed fluidly in performance for some time before it was recordedas a text. In such circumstances a text with relatively ‘late’ details may failto indicate the longevity of the performance tradition from which it isderived. The relative date of two texts may not well replicate the relativedate of their respective performance traditions. Even if it is established to asatisfactory degree that one text has influenced the other text, treating thetwo poems as one-off textual compositions fails to consider the potentialcomplexity of the relationship between the performances that lay behindthem.

2 '�4� '�4��

My first test case is the phrase '�4� '�4�� , which may have regularlybeen used in epic traditions in reference to the death of Achilles. At Odyssey the shade of Agamemnon uses the phrase as part of his description tothe shade of Achilles of a battle over the corpse of Achilles. The shade ofAgamemnon reports to the shade of Achilles, ‘you lay in a whirl of dust greatin your greatness, forgetful of your horsemanship’ ( 9 � @ 7� ��1��44�� | �0 � '�4� '�4�� '�� <�� #�� (Od. .–)). Scholars have long found it striking that similar phraseology is foundat the beginning of book of the Iliad after Achilles has learned of thedeath of Patroclus. He is described as lying stretched out in the dust‘great in his greatness’ in grief: (�.� � @ 7� �� '�4� '�4�� # 3�� | �0�� (Il. .). The phrase '�4� '�4�� recurs, thoughin a different metrical position, and again we have reference to lying in thedust. Neoanalysts have most notably argued that the phraseology in Iliad was present in a pre-Homeric description of the corpse of Achilles.

At Burgess (: –) I summarize (in rather rhetorical fashion) critiques of the dating of earlyepic in Janko (). Hesitation about the conclusions of Janko () does not imply doubt aboutthe impressive linguistic and statistical skills on display; the critics rather seem to feel that thedata fail the methodology. It should be added that Janko would stress relative chronology overthe controversial issue of absolute dating. As it happens, the argument below does not depend onthe chronology, absolute or relative, of specific texts. My concern is not with the borrowing ofphraseology from one poem by another poem, but rather the reuse of traditional phraseology in asecondary fashion, with an allusive and textless type of intertextuality resulting.

See Pestalozzi (: –), Kakridis (: –), Kullmann (: –, , : n. , ,–), Schadewaldt (: ), Schoeck (: ). Those that have found the argument attractive



jonathan s. burgess

The argument is all but irresistible, especially since the whole scene at thebeginning of book seems to reflect the death of Achilles. The Nereidsmourn in a way that is reminiscent of Achilles’ funeral, Thetis holdsAchilles’ head in the manner of funeral ritual, and soon enough the heroand his mother will openly talk of Achilles’ coming death. The phrase'�4� '�4�� occurs in the midst of all this foreshadowing, and it ishard to avoid the impression that it contributes to the effect. Followingclassic neoanalyst methodology, one would view the Odyssey passage asrepeating phraseology in its usual pre-Homeric context, whereas the Iliad passage is transferred from its natural, or primary, context and placed inan unusual, secondary context.

It is interesting that in the edition by West the relevant lines of book are bracketed as an interpolation, for reasons having to do with therather vague conception of how Antilochus and Thetis have physical con-tact with Achilles as he lays on the ground (West a: –). ThoughI do not agree that there is an interpolation, his argument underscoresthat the phraseology is out of place. From a neoanalyst perspective thisis not the intrusion of external verse, but rather its incorporation. TheIliad has not taken words from the Odyssey, or vice versa, and presumablythe Odyssey passage faithfully preserves, except for the change of voice,phraseology that existed in pre-Homeric epic. The source of the phrase-ology has sometimes been conceived as the Aithiopis, but it has been morecommonly described as a prototypical Achilleis or Memnonis. Recently oral-ists attracted to neoanalysis have more generally described the source aspre-Homeric.

But before speculating on the nature of this apparent intertextuality,we have to recognize that there has been one major block to accepting theneoanalyst argument. The longer version of the phraseology used in Odyssey is also used to describe the corpse of Cebriones at Iliad . The passagedescribes the battle over Cebriones, and at lines – the narrator

include de Romilly (: –), Edwards (: –, cautiously), Dowden (: , : ),R. B. Rutherford (: –), Willcock (: ), Danek (b: –, b: ), Currie(: ), Burgess (: –). Dihle (: –) by contrast argues for the priority of the Il. example on syntactical grounds, favouring the use of a participle ��# 3�� with the adverbin the book passage; for a response, see Kullmann (: ). Other examples of parachesis(��3�� )��, .; ��3�� .) suggest that '�4� '�4�� can function as a relativelyself-contained phrase.

Schein (: ) provides an excellent analysis of the scene as a whole. Garner (: –) proposes that a newly found fragment of Stesichorus (, available in the

appendix of PMGF) relates the battle over the corpse of Achilles; if so, ��1�[��44� (line ; cf. Od.., Il. .) suggests another example of comparable phraseology in the context of Achilles’death.




reports, N � @ 7� ��1��44� �� | �0�� '�4� '�4�� '��<�� #�� (he lay in a whirl of dust, great in his greatness, forgetful ofhis horsemanship). This third Homeric occurrence of '�4� '�4�� phraseology raises the issue of typology. From an oralist perspective, itmight look as if the phraseology belongs to the typological scene of a battleover a corpse, and not to any one character. In other words, the phraseologymay be context-specific, but it is not certainly character-specific. One mighttherefore conclude that the phraseology in book did nothing more thanmeet metrical needs, perhaps triggered by the dust that Achilles poursover his head, which is not uncommon in Iliadic battle scenes. In thatcase the phraseology in book would be a meaningless by-product of oralcomposition, or at best an evocation of a generic scene. One might concludethat the book passage reuses context-specific phraseology in a ratherinappropriate manner, but not that it reflects any specific mythologicalscene featuring Achilles.

To withstand the oralist perspective, the neoanalyst argument needs toestablish that the phraseology is not just context-specific, but that it ischaracter-specific; that is, that it belongs to Achilles. This is not so easywhen we reflect on the specification of horsemanship in the longer versionof the '�4� '�4�� phraseology in Odyssey and in Iliad . Ithas not gone unnoticed that horsemanship seems especially appropriatefor the charioteer Cebriones, and some have in fact concluded that thephraseology actually belongs to Cebriones.

This is something of an embarrassment for the neoanalyst position,but a defence can certainly be made. Kullmann (: –) has correctlypointed out that Achilles is closely linked to his divine horses. Scholarshave also pointed out that <�� +�� can be attributed in the Iliad to anywarrior for whom the chariot is employed. The distinction between a driverof horses and a chariot-using warrior is therefore elided. Long ago Scottaffirmed that horsemanship was appropriate for Achilles, declaring that‘[t]o deny to Achilles this particular skill is to miss the whole tone of theIliad. The heroes themselves, not their drivers, were the skilled horsemen,and Achilles above all others.’

If the extended phraseology that refers to horsemanship is not inappro-priate for Achilles, then the Odyssey passage with its reference to Achilles

Cf. Erbse (: –), Heubeck (: ), Usener (: –). Willcock (–: I.)thinks that the reference to horsemanship phraseology is not appropriate for Achilles, but since hefeels the phraseology is too impressive for Cebriones he locates its origin with ‘horsemen of theprevious generation’, like Peleus.

Scott (: ); see Janko (: ), Burgess (: ).



jonathan s. burgess

forgetting his horsemanship is unobjectionable. We can also understandwhy the longer version of the phraseology is not used to describe Achillesin book , for the grieving Achilles cannot be said then to have forgottenhis horsemanship skills. If horsemanship is not inappropriate for Achilles,however, it is certainly not inappropriate for Cebriones. The questionremains, then: how does the phraseology ‘belong’ to the story of Achilles,and in what way is it secondary for Cebriones? It should be first observedthat poetically speaking the phraseology’s application to Cebriones in Iliad is nothing less than magnificent. Its success has been well described byJanko (: , ), who demonstrates that the phraseology is partic-ularly effective as a dramatic ending to the extended description of thebattle over Cebriones. Besides the full description of the opposing soldiers,there is an impressive succession of similes. First Hector and Patroclus arecompared to two lions confronting one another, and then a more gener-alized nature scene portrays two opposing winds causing the branches ina forest to knock together. The energy of the animal and nature similes isthen interrupted by the focus on Cebriones. In Adam Parry’s words, it is a‘sudden vision of the single man in the eye of the storm who has left it allbehind’.

In book , the essential focus is not on horsemanship but on the loss ofhorsemanship as a consequence of his death. Part of the effect is contrast:amidst the energetic battle Cebriones is completely still. The situation ismuch different with Achilles in book , where in contrast to CebrionesAchilles in his mourning very actively disfigures himself. But much of theeffect of the Cebriones passage depends on the negation of horsemanshipskill by death. Some of this effect is inherent in the parallel passageabout the corpse of Achilles in Odyssey , but Agamemnon’s account ismuch more concise and not as dramatic. Though it is possible that inpre-Homeric epic, descriptions of the corpse of Achilles were as poeticallyskillful, we must admit that in Homeric poetry the '�4� '�4�� phraseology is most successfully employed in book .

Usener (: ) thinks that the reference to horsemanship does not specifically match Achilles’last battle, even if horsemanship is a general characteristic of warriors. Danek (b: ) arguesthat Achilles must have employed the chariot in his rout of the Trojans just before his death. Iconsider the Odyssey phraseology generally appropriate to Achilles.

A. M. Parry (a: liii), with an excellent analysis of the book passage, made in the context ofa discussion (pp. lii–liv) of his father’s remarks on the '�4� '�4�� phraseology (M. Parry, inA. M. Parry b: ). The father admired the line in book but cited comparanda to demonstrateits traditionality, a view which the son finds limited.

The pathos has been well described at Griffin (: ), where it is linked with other passages inwhich the wastefulness of death is emphasized. On the tragic feeling of the Cebriones passage, seealso de Romilly (: –).




If horsemanship is not inappropriate for Cebriones, and the poetics ofthe passage describing his corpse are excellent, what is secondary about thephraseology in the book passage? From the neoanalyst perspective, it isthe central phrase '�4� '�4�� . The issue is the status of Cebriones.There is no reason to think of him as '�4� in a physical sense, and heis not a major hero. He is a bastard son of Priam who upon his death ismocked abusively by Patroclus. Many have therefore concluded that thephraseology, though it suits the context, is inappropriate for the character.‘Great in his greatness’ might be thought very fitting for Achilles, especiallyif one thinks of iconography of Ajax groaning under the weight of his hugecorpse.

So the phraseology probably belongs to Achilles, normally employedin the context of a battle over his corpse. In book the context is cor-rect, a battle over a corpse, but the character is different. This shift to adifferent character could be seen as accidental and therefore insignificant;arguably the context of a battle over a corpse triggered the employmentof phraseology usually reserved for the battle over Achilles. Oralists mightprefer this interpretation, and something like it was argued by Schoeck, inhis theory of multiple motif transference by association. From the per-spective of classic neoanalysis, the use of Achilles-specific phraseology forthe minor character Cebriones is a kind of narrative problem, comparableto the application to Patroclus of motifs usually associated with Achilles.Neoanalysts cite such ‘problems’ as evidence of the poet’s inventive trans-ferral of motifs; more recently, scholars attracted to neoanalyst argumenthave tended to regard such problems as allusive signals to other narratives.

It is probably too strong, however, to say that the Cebriones passageevokes the traditional death of Achilles, for Cebriones in no other wayresembles Achilles, certainly not in the way that Patroclus is an alter ego ofthe great hero. But the application of the '�4� '�4�� phraseology toCebriones might be considered an anticipatory doublet, since the Achilles-specific phraseology in book prepares for Achilles’ enactment of hisdeath at the beginning of book . As well, the Achilles-esque corpse ofCebriones prepares for the imminent rehearsal of Achilles’ death on thepart of Patroclus. Cebriones is the last major victim of Patroclus, and

Schoeck (: passim); at Schoeck (: ) the Iliad passage is described as gravitating to atraditional Achilles formulation.

On anticipatory doublets, see Edwards (: –). Usener (: ) and Danek (b: )note that the death of Cebriones directly precedes that of the Achilles-doublet Patroclus; Usenerargues that the Odyssey passage is thereby influenced by the Iliad passage, whereas Danek moreplausibly sees the Iliad passage as a secondary reflection of a context that belongs to Achilles.



jonathan s. burgess

Patroclus himself will soon be a corpse in the middle of opposing warriorsafter resembling Achilles in his aristeia and in his death.

Kullmann () has used the phrase ‘motif transference’ to describe themovement of a motif from its primary situation to a secondary situation,and that phrase generally describes the phenomenon of Achilles playingthe role of a corpse in book : a mytho-poetic episode, including some tra-ditional context- and character-specific phraseology, has been transferredinto a new context within the Homeric poem. But here and below I will tryto employ some more specific terms in order to describe the intertextualitywith more precision. In the book passage there is what may be calledmotif alienation which involves displacement (change of scene from battle-field to the camp of Achilles) and anachronism (enactment of death longbefore it occurs). Besides this variance, there is correspondence between thesecondary manifestation of the motif with its regular, or primary, manifes-tation, notably with the phraseology that we examined. Modified iterationis the basic recipe for intertextuality, certainly for intertextuality that ispotentially significant and allusive. In narratological terms the allusion inbook is external prolepsis (reference to a future event outside the bound-aries of the poem). In book the usual context of the '�4� '�4�� phraseology would quickly come to mind for a mythologically informedaudience, especially one which has recently heard it employed in book

for a battle over a corpse – which itself serves as an intratextual anticipatorydoublet.

3 astyanax

The preceding analysis of a much-discussed phrase has established somecriteria for examination of the intertextuality of epic phraseology. Mysecond text case is the correspondence between a line of the Little Iliad anda line in book of the Iliad. The fragment reports on the aftermath of thesack of the city:

�0� � @ J�g� 7 ��# 7o��'�� 3:��_0T� ��.� ��4g� ��. �+�4�#� �.� �� C��>� ��1+�� 3�� '�0� ��:

And taking the child from the breast of the fair-haired nurse he gripped him bythe foot and hurled him from the tower, and dark death and strong fate seizedhim.

Fr. .– Davies = .– Bernabe = .– West (Tzetzes ad Lycoph. Alex. ).




Much scholarly attention has been directed to the attribution of the secondhalf of the eleven-line fragment to Simias by a scholion to Euripides (adAndromache ); I will leave that controversy aside and focus instead onlines – of the fragment. This brief description of the death of Astyanax isvery similar to the famous passage in book of the Iliad in which Astyanaxrecoils into the breast of his nurse in horror of his father’s plumed helmet:

jT � @ r �� .� �� 7oH�� 3:��7��3� �� .� 1��# $T�� #�3��>: � �� 1�� <��.� �� @ �� #3�� +�� : �

(.–)

But the child leaned back towards the bosom of his well-belted nurse, crying out,upset at the sight of his father, in fear of his bronze and the horse-hair crest, as hesaw it nodding terribly from the top part of the helmet.

The correspondences are obvious. Most generally, both passages concernAstyanax. More specifically, line of the fragment corresponds extensivelyto line of book ; each line refers to a child and a breast fold of anurse (with a different case for �0�, and a different epithet for the nurse).And that is not all; other parts of the fragment correspond to differentpassages of the Iliad. Most significantly, a different form of the mainverb at the beginning of line in the fragment, along with the subsequentprepositional phrase, recurs at line of Iliad , all in the same metricalplace. This passage also concerns Astyanax, for Andromache is ruminatingabout the possible fate of her son (as well as her own) now that Hectoris dead. It has not gone unnoticed that a Cyclic passage about Astyanaxemploys phraseology also found in Homeric passages about Astyanax, andthe consensus view is that the Little Iliad has reworked phraseology fromthe Iliad.

Davies prints the text with a space between lines and ; Bernabe brackets lines –; West separatesinto two fragments. At Davies (b: ) the issue is deemed ‘one of the great insoluble mysteriesassociated with the Epic Cycle’. It has also troubled scholars that the eleven lines seem to moveabruptly between three distinct actions: the conveying of Andromache to the ships (imperfect tense),the previous murder of Astyanax and the acquisition of Andromache and Aeneas by Neoptolemus.Recent suggestions include Huxley (: , rearrangement of the lines), West (a: n., Tzetzes quotes two unconsecutive passages of the epic), Bravo (b: –, the fragment isa unity, with the succinct reference to the death of Astyanax a narratological flashback), Debiasi(: –, Simias reused a passage of the Little Iliad). For further discussion and bibliography onthe origins of the verse see Bernabe’s apparatus (: ad loc.), Powell (: –), Griffin (: n. ), Bravo (b: ), Debiasi (: –).

The apparatus at Little Iliad fr. Bernabe helpfully gathers some of the comparanda. Cf. schol. Il. .; Griffin (: –), Curti (: ), Richardson (: ), Debiasi (:

–). Kullmann (: –) argues that the Cyclic fragment is post-Homeric, believing that



jonathan s. burgess

I have previously resisted this idea, suggesting instead that the apparentcorrespondence is best explained as resulting from independent use ofshared poetic traditions (Burgess : ). Though the fragment can befaulted for some awkwardness of metrics, word placement and repetition,its language is not obviously post-Homeric, and Notopoulos (: –)demonstrated that all eleven lines of the fragment are formulaic in nature,at least by analogy. Most of the difficulties occur in the (disputed) secondhalf of the fragment, though the section concerning Astyanax has someaspects that might raise concern. In line the genitive form of ��necessitates metrical correption and hiatus, as the accusative at Il. .

does not. But this does not mean that the fragment’s formulation must beuntraditional, or derivative from the Homeric line. Though there seemsto be no qualitative reason to prefer the Homeric 7|H�� to the Cyclic7o��'�� as an epithet for the nurse, the form of the word in thismetrical position is anomalous. If the Homeric database reflects normalepic practice, the fragment may have been innovative with a non-integralpart of traditional phraseology – perhaps in order to avoid repetition withthe occurrence of 7|H�� in line , assuming the fragment is a unity. Thisslight lapse of economy is nothing extraordinary for oral compositionaltechnique in early Greek epic.

It is not enough to claim that the Little Iliad passage is independentof the Iliad, however; if the Iliad is not the source for the Cyclic passage,nonetheless it must be more than a coincidence that parts of the fragment’sdescription of the death of Astyanax are also present in two Homericpassages about Astyanax. These passages are linked at a thematic, as wellas lexical, level. If there is no direct connection between the poems, theonly explanation could be independent manifestation of phraseology thatwas traditionally used in reference to the death of Astyanax. In this casethe phraseology would be context- and character-specific (the death ofAstyanax), with the Cyclic passage more directly describing the primarycontext, and the Iliad passages more (in book ) or less (in book ) evokingit. Using the terminology established above for the '�4� '�4��

Hector, and by extension Andromache and Astyanax, were Homeric inventions; for refutation ofthis argument see Burgess (: –).

For philological and metrical data of the Little Iliad fragments, see Bernabe (: xi–xii); for analysisof fr. Bernabe, see Dihle (: ), Davies (a: , b: –), Curti (: –, , ),Anderson (: –), Bravo (b: –), Debiasi (: –). Debiasi defends the fragment’sunity; Anderson well explains the significance of the genealogical emphasis in the phraseology.

Janko (: ) suggests that direct imitation may be indicated if a poet violates his own formulareconomy; unfortunately not enough of the Cycle poems has survived for them to be subject to thelinguistic and formulaic guidelines concerning exemplum and imitatio at Janko (: –).




phraseology, we could view the book passage as an example of motifalienation. There is the same context, the walls of Troy, with Hector’shelmet representing an attacker, at least as it initially seems to the child.But there are obvious changes. The motif of the death of Astyanax isanachronistic here. Hector is certainly not the child’s murderer, and thechild is actually in no danger. Astyanax shrinks back into the safety of hisnurse’s bosom, instead of being snatched from it, an inversion underscoredby the modification of the phraseology. The correspondence, coupled withvariance, produces intertextuality, and the traditional death of the princeis foreshadowed through external prolepsis.

Andromache’s premonition in book , on the other hand, does notconstitute motif alienation, since the death of Astyanax is predicted muchas it traditionally occurs. There is narratological modulation, however,for the traditional event is foreseen and described by a character, insteadof the narrator. This external prolepsis, though explicit in its denota-tion, nonetheless has a hazy air about it; Andromache is not an omni-scient reporter and has no certain knowledge of the future, even if sheis well aware of the dire choices awaiting the women and children of asacked city. Skeptics about the pre-Homeric nature of Astyanax’ deathpoint out that Andromache offers alternative scenarios for the fate of herson, and also that she seems mistaken in her musings on the motiva-tion of the future murderer of her son (.–). But it is exactly herlack of prophetic knowledge when she happens to specify the traditionalfate of her child that produces a chilling effect for the mythologicallyinformed audience, especially since she uses context- and character-specificphraseology.

To what degree, however, can we say that the phraseology actually‘belongs’ to Astyanax? Some of the diction is apparently character-specific.A nurse, ��3:��, would be employed for care of a child of high social status,as is the case with Astyanax. Not all children thrown off a wall would have anurse, and in the Iliad ��3:�� is almost exclusively connected to Astyanax.

Another seemingly significant word is ��, which also occurs in con-nection to Astyanax in Iliad (., in modified correspondence to thefragment’s phraseology; also at . and .). This is not a rare Homericword, however, and there are different yet oddly echoing usages of ��elsewhere in the Iliad. At . Thetis receives the infant Dionysus intoher protecting �� (after the young god’s nurses have been chased off ),

Il. ., , , and .. The word does not occur in the Odyssey, but in the Homeric Hymnto Demeter Demeter acts as Demophoon’s ��3:�� (, cf. ). All the Iliadic occurrences refer tothe nurse of Astyanax except the first example, which refers to the nurses of the infant Dionysus.



jonathan s. burgess

and at . Thetis receives another victim, Hephaestus, into her ��(after his fall from the heights of Olympus), with the same formula that isused at . when Thetis rescues Dionysus, and with similar phraseologyto that used at . when Andromache receives the frightened Astyanaxinto her ��.

It appears that phraseology featuring �� in reference to Astyanaxcan also occur in the recurring motif of Thetis as rescuer of distressed divini-ties (cf. .–, her rescue of Zeus). And the motif of Thetis as rescuerat least on one occasion intersects with a motif that involves the hurling ofa divinity (notably Hephaestus) from Olympus (.–, .–, .–, .–, .). This hurling-of-a-divinity motif itself often employsthe phraseology used to describe the attack on Astyanax in the Little Iliadfragment. The phraseology in question is not altogether character-specific,we must conclude; much of it belongs to a general context of hurling some-one from a great height. The lexical evidence might therefore be labelled‘hurling-from-a-height’ phraseology. Allomorphs of the phraseology areemployed, mutatis mutandis, for the hurling of Astyanax and the hurling ofdivinities like Hephaestus. Though these are very different circumstances,it is possible that hurling in the divine realm provided an ironic foil to thetragedy of the Trojan prince’s death.

If the ‘hurling-from-a-height’ phraseology does not belong exclusivelyto Astyanax, it must have been employed often enough in the context of hisdeath so as to be recognizably associated with it, especially when some keywords like ��3:�� or (to a lesser extent) �� are present. When usedin connection with the Trojan wall, ‘hurling-from-a-height’ phraseologyshould certainly be allusive to the death of Astyanax. The Astyanax linesin the Little Iliad fragment certainly are not an accidental conglomerationof free-floating formulae. Their composition by chance is ruled out by thepartial reoccurrence of the fragment’s phraseology in two Iliadic passagesabout Astyanax; in particular, the occurrence of the rare word ��3:�� inthe fragment and the book scene at the city walls confirms that there isa connection.

Is the connection one of Homeric influence, as many suspect? Thecommon assumption that the Cycle reshuffled pilfered Homeric phrasesshould be resisted, since it is predicated upon an untenable conceptionof the Epic Cycle and of the epic tradition of the Archaic Age in general(Burgess ). There is nothing that proves Homeric priority, and itis probable that the phraseology – though employed in other narrative

I examine this issue more fully in Burgess .




circumstances – commonly occurred in traditional descriptions of the deathof Astyanax. Some of the diction is especially relevant to Astyanax, andthe phraseology in lines – of the fragment seems to have become attachedto the traditional epic description of the scene. If this is so, a reconceptionof the intertextuality between the fragment and the relevant Iliad passagesis required. The fragment would need to be seen as containing phraseologyas it was regularly employed in narration of the mythological scene ofthe death of Astyanax. When Andromache in book imagines this samescene, there is a narratological modulation from an actual description ofthe child’s death to fearful speculation about it. In book of the Iliad theallusion is more indirect and complex. There has been alienation of thephraseology, transferred as it is to a new context and time, and there is alsoinversion: the child recoils into his nurse’s bosom instead of being snatchedfrom it, and instead of a murderous Greek it is his own father who reachesfor him (though a father whose visage is hidden by a warrior’s helmet). Thisis iteration with variation, resulting in significantly allusive intertextuality.The passage thus serves as an excellent example of the meta-cyclic natureof Homeric poetry. A motif with its characteristic phraseology has beentransferred to a new, Homeric context; as a result the scene is deeplyresonant of the traditional, cyclic context.

From a mythological perspective the fragment is a manifestation of theprimary motif, and the Astyanax passages in the Iliad passages are sec-ondary. This is not to suggest that the Homeric passages allude to theLittle Iliad. Nor do I think that they allude to a prototype of the Lit-tle Iliad, or to a prototype of the Iliou Persis for that matter; by using‘Astyanax’ phraseology the Iliad is availing itself of phraseology commonlyemployed in the epic tradition when the scene of the death of Astyanaxwas narrated. The date of the Iliad and the Little Iliad is not relevant to myargument, for I conclude that the Iliad and the Little Iliad independentlyinteracted with traditional myth about the death of Astyanax. This tradi-tion certainly would have been flexible – Odysseus was often named themurderer, for instance, as in the Cyclic Iliou Persis, according to the sum-mary by Proclus. But the basic elements of the myth would have remained

Cf. Anderson (: –), where it is allowed that the phraseology may be pre-Homeric, or thatthere was mutual influence between Homeric and Cyclic traditions. He counts the conciseness ofthe Little Iliad fragment as an indication of imitation, but it is more likely that traditional epicresembled Cyclic narrative pace more than it did the leisurely expansion of Homeric narrative. Cf.the description of the battle over Achilles in Odyssey . On the expansion and compression of epicnarrative, see Nagy (a: –), and on the smaller scale of formulaic composition, Martin (:–).



jonathan s. burgess

stable, and in its epic manifestations the tradition typically included certainphraseology.

4 conclusion

I began by wondering at the possibility of an intertextuality of phraseologyin circumstances where there were no texts, or even any identifiably promi-nent epics to which to allude. My first test case, the '�4� '�4�� phraseology, has attracted wide attention, not least because of persuasivearguments by neoanalysts about its use in Iliad . The second test case wasthe Little Iliad phraseology, which is usually seen as imitative of the Iliad.

In both test cases I concluded that there were mythological traditionsabout the death of Achilles and Astyanax to which the Iliad reacts, and thatthese traditions were often represented in epic manifestations which regu-larly employed a certain context-specific and/or character-specific phraseol-ogy. A passage in Odyssey seems to represent faithfully '�4� '�4�� phraseology in its usual form, though with narratological modulation. Apassage in Iliad similarly represents phraseology of the Astyanax deathscene, and we are lucky to have a Little Iliad fragment that provides aneven more direct manifestation. Recognizing the primary status of thesepassages enables us to appreciate the poetics of the relevant yet secondarypassages in the Iliad that contain motif alienation, with attendant displace-ment and/or anachronism. Such motif transference is especially fascinatingwhen it includes the textless intertextuality of language traditionally asso-ciated with specific scenes: the '�4� '�4�� phraseology in Iliad

as Achilles collapses to the ground in intense grief at the news of Patroclus’death, and the ‘Astyanax’ phraseology in Iliad when Hector meets hisfamily on the walls of Troy. The effect is ominously foreshadowing, as thedeaths of these characters are evoked by external prolepsis. Neither Achillesnor Astyanax are in any immediate danger, and indeed they will not perishwithin the course of the poem, but to an audience that was sensitive tomytho-poetic traditions, their deaths are signified.

It has been noted that the traditional phraseology examined in thisstudy is not exclusively character-specific. In book the '�4� '�4��6 �� phraseology was used in its normal context, a battle over a corpse, butapplied to a character other than Achilles. This I considered an anticipatorydoublet; wonderfully effective in its immediate context, the phraseologyalso prepares for the death of Patroclus and especially for the later '�4�'�4�� passage in book . As for the ‘hurling-from-a-height’ phrase-ology, it became apparent that it could also be employed in scenes of divine




violence. Employed in certain circumstances, however, the phraseologymay effectively allude to the death of Astyanax; it is also possible that thedivine scenes provide a significant contrast to the death scene of Astyanax.

The source for the phraseology that I have explored is the general epictradition, not specific poems. This has allowed me to propose the exis-tence of a textless intertextuality that involves not one text influencinganother, but the traditional articulation of an episode being reflected bya secondary articulation of it. The passages considered ‘primary’ for thepurposes of my investigation simply contain phraseology commonly usedfor the situations that they describe; they need not represent the only orbest or earliest compositions of these scenes. Nonetheless these ‘primary’passages provide testimony for the traditional employment of phraseologyin specific circumstances. With this evidence in hand we have been ableto uncover surprisingly detailed examples of the Iliad ’s textless yet allusiveintertextual engagement with the epic tradition from which it stems.



chapter 11

Perspectives on neoanalysis from the archaichymns to Demeter

Bruno Currie

1 introduction

A fundamental, if vexed, contribution to a relative chronology ofearly Greek hexameter poetry is made by the consideration of literaryrelationships. Put at its most general, and most simplistic: if a literarywork x can be shown to allude to a work y then y is older than x. Such anapproach has played a key role in arguments for the priority of the Iliad tothe Odyssey, of the Theogony to the Works and Days, of the Cyclic to theHomeric epics and, conversely, of the Homeric to the Cyclic epics. Here Ioffer some reflections, theoretical and practical, on such approaches.

The theoretical issues are familiar. The very notion of a literary history,of establishing a dependency between works, has an especially controversialapplication to early Greek hexameter poetry. There is even debate whetherthere are discrete works (‘fixed texts’) in this oral(-derived) tradition. Thecommon stock of phrases, motifs and themes, moreover, makes it prob-lematic to detect allusions in one work to another. Scholars nowadays are

This piece became a fixed text through performances at Oxford, Leeds and Oslo; I thank theaudiences of those respective occasions for their comments and suggestions. I have also profited muchfrom discussions with and communications from Jonathan Burgess, Margalit Finkelberg, AnastasiaMaravela, Robert Parker, Nicholas Richardson and Oliver Thomas, none of whom, however, shouldbe assumed to share my ideas; certainly none shares the responsibility for any errors of argument orfact.

See e.g. Janko (: ‘General Index’ s.v. ‘imitation’). I use the term ‘literary’ without wishing toimply that early Greek hexameter poetry was read (widely) as ‘literature’, rather than experienced inperformance.

This simple formulation ignores the complications raised by the possibilities that a literary workmay go through different editions, or have an oral dissemination prior or parallel to its written‘publication’; it also ignores the complications of some reader–response models of intertextuality,which allow for earlier texts ‘alluding’ to more recent ones: see e.g. D. P. Fowler (: ),Martindale (: –).

Cf. Burgess (: –). Cf. Lord (: , : esp. –, ); cf. Garvie (: and references in n. ), Nagy (a:

), Pelliccia (). E.g. Hainsworth (: ), Hoekstra (: ); cf. Janko (: –).



Neoanalysis and the archaic hymns to Demeter

on the whole happier to assume that poets might exploit their audience’sknowledge quite generally of the mythological tradition (‘mythologicalintertextuality’) or might manipulate the traditional uses of type-scenes

or formulae (‘traditional referentiality’). Accordingly, rather than seeingearly Greek hexameter poems as interacting with other specific poems, thereis a preference for seeing them as ‘resonating’ with an abstract overarchingtradition. In Saussurian terms, the preferred model of interaction is ofparole with langue rather than parole with parole. Related to this preferenceis the ideological opposition between oralists’ ‘typical motifs’ and neoana-lysts’ ‘transferred motifs’. Where neoanalysts are apt to identify one use ofa motif in early epic poetry as ‘primary’ and another as ‘secondary’, inspiredby the other, oralists would reject any direct dependence between the two,seeing each as independent manifestations of a traditional ‘typology’. Stillmore problematic, even for many who would accept transferred motifs,is the notion that individual phrases, or whole verses, may be transferredallusively from one specific poetic context to another. It is with theseproblems in mind that I offer the following perspectives on Homeric neo-analysis, a literary method committed to establishing relative chronologicalrelationships between early Greek epics (as its alternative title, ‘the studyof the history of motifs’, makes explicit). My investigation cannot avoidtouching on related fundamental controversies: whether there can be suf-ficiently fixed ‘texts’ in this tradition to ground allusion and whether therecan be ‘intertextuality’ (appropriately defined) within this tradition.

If the theoretical issues in this essay are mainstream, their practicalexploration is not: my test case is the literary relationship between twoearly Greek hexameter compositions, the ‘Homeric’ Hymn to Demeter

Burgess (: ). Cf. the approaches of M. W. Edwards (: –), Emlyn-Jones (: –). J. M. Foley ( and elsewhere), Graziosi and Haubold (). The term used by J. M. Foley (e.g. : , , : xiv, , and passim) and subsequently by

Graziosi and Haubold (). So Burgess (this vol.: ): ‘The source . . . is the general epic tradition, not specific poems.’ Cf. Danek

(b: ): ‘Ubertext’. Something along these lines is already adumbrated by Lord’s distinction of‘the specific song’ from ‘the generic song’ (Lord : ).

Cf. Ballabriga (: ), Clark (: ). Cf. M. W. Edwards (: –). This is not to imply that neoanalytical and oral positions are

intrinsically irreconcilable: cf. Kullmann (), Burgess (: ). Fenik (: , –), Hainsworth (: –: ‘If themes are common property, there can be

no borrowing, no imitation.’). Cf. Burgess (: ; this vol.: esp. , point ). See Kullmann (: ) for neoanalysis as ‘die motivgeschichtliche Forschung zu Homer’. On ‘intertextuality’ in early Greek epic: Korenjak (), Danek (b), R. L. Fowler (b:

–), Burgess (: esp. ; this vol., arguing for a ‘textless intertextuality’). Rather differentconcepts of intertextuality in Pucci (), Tsagalis ().



bruno currie

(HDem) and a fragmentary Hymn to Demeter ascribed to Orpheus. Ourknowledge of the latter poem is due to a papyrus text written in the firstcentury bc, now in Berlin. I refer to this as ‘the Berlin papyrus’ (P.Berol.).The standard edition of the papyrus was for a long time Otto Kern’sOrphicorum Fragmenta (), but it has been recently re-edited by AlbertoBernabe in the second part of his Poetarum Epicorum Graecorum Testimoniaet Fragmenta (). The papyrus contains verbatim quotations of thepoem (twenty-six verses are extant, in whole or in part) and a paraphrase inprose of much of the poem’s plot. The paraphrase relates a version of theRape of Persephone which is clearly different from that of HDem; however,the verses cited are all but identical with verses of HDem, and thus raisedifficult questions. What is their relation to the different hexameter verseswhich are cited as coming from an (early?) ‘Orphic’ Hymn to Demeterby other sources (the Derveni papyrus, Ps.-Justin Martyr and Clement ofAlexandria)? And, most critically for our purposes, what is their relationto HDem?

2 the status of the hexameter verses in the berlin papyrus

On this last question, two basic positions have been taken. First, thatthe verses in P.Berol. derive from a poem distinct from HDem. This

Or possibly Musaeus: cf. Janko (: n. ). P.Berol. . The text in question is on the verso of the papyrus; the texts on the recto are dated

to the second century bc (Buecheler : –). Fr. Kern, frr. – Bernabe. The editio princeps is Buecheler (); also still worth consulting

is Colli (: –, –); cf. too Scarpi (: –, –, –, –). Bernabe’s editiongreatly facilitates study of the papyrus, but does not remove the need to consult earlier editions.First, because Bernabe’s ordering of the fragments of the poem hinders a sequential reading of thepapyrus itself, which Bernabe has broken up as follows: P.Berol. – = fr. Bernabe, – = fr., – = fr. .(II), – = fr. .(I), – = fr. .(I), – = fr. , – = fr. .(I),– = .(II), – = .(II), – = fr. , – = .(II). Second, because Bernabeis much more confident than his predecessors on several textual points (see e.g. below n. , on therelationship of the versions of P.Berol., Ps.-Justin and Clement; and p. n. d on P.Berol. ).

How much of the plot? Scholars differ on whether the end of P.Berol. is the end of the paraphrase:contrast Wilcken in Kern (: ad P.Berol. ) with Richardson (: ).

Bernabe presents the verse cited by the Derveni commentator (fr. Bernabe) as coming froma different poem than the verses cited in P.Berol., but presents the verses cited by Ps.-Justin andClement (frr. , Bernabe) as coming from the same poem as the verses cited in P.Berol.(p. ‘ex eadem recensione venire possunt’). Here he follows the principle formulated (thoughmore cautiously) by Richardson (: ); cf. Bernabe (: ). See also on this question Colli(: –).

The distinctness of the two versions is assumed by Buecheler (: –), Kern (: –),Wilamowitz-Moellendorff (: –), Colli (: ), Janko (a), Bernabe (: , –,where P.Berol. is seen as providing ‘fragmenta’ (‘F’) – not ‘testimonia’ (‘T’) – of an Orphic poemon the rape of Persephone); cf. Graf (: –).




position invites further questions: whether this distinct poem is derivedfrom HDem, is itself a source for HDem or simply stands in a commontradition with HDem: these are questions to which we will return below.The second position is that the verses in the Berlin papyrus are simplyquotations of our Homeric Hymn to Demeter, in a wild or rhapsodic text.

This view is recommended by the fact that the verses quoted in the papyrusare virtually identical with verses of HDem, whereas the many fewer versescited by the other sources as coming from an ‘Orphic’ Hymn to Demeter arewholly unlike verses from HDem. Moreover, such divergences as are foundbetween the verses in the papyrus and their counterparts in HDem are(with one significant exception) consistent with what we would expectfrom ancient manuscripts of a single early hexameter text. This viewwould entail that the Berlin commentator saw HDem as the work ofOrpheus, not Homer; but for that, too, there may be parallels. Thisview, although it commands notable scholarly support, involves significantdifficulties.

First, the commentator purports to be working from one and the samepoem, whether paraphrasing it in prose or quoting hexameter verses fromit. If he is doing what he purports to be, he must be quoting from adifferent poem than HDem, for the paraphrase is incompatible with HDem.We must then assume that the poem that lay behind the Berlin papyruscontained several verses which diverged significantly from the verses ofHDem, even though the commentator does not (with a possible partialexception to be discussed presently) actually quote any such verses. Onthe alternative view, that the commentator was quoting the verses fromHDem, he must, contrary to his indications, have been following different

For the first of these positions, cf. Kern (: –), Buecheler (: ). Bohme (: ), cf. Bohme (: n. ). Allen et al. (: ), Kruger (: ), M. L. West (: , n. , b: ). Richardson (:

–) treats P.Berol. as a manuscript witness to HDem (‘pap. ’ in the ‘Sigla’ p. ), but he recognizesthat in some cases we have deliberate adaptations, not just textual variants, of HDem. In the viewof Cassola (: ), ‘I passi in esametri discendono da un’edizione dell’inno diversa dalla nostra[sc. HDem], ma non necessariamente orfica.’

HDem – and P.Berol. –: see below, p. . See in general Haslam (: –); cf. M. L. West (a: ). Compare, for instance, Thuc. .

and HAp –, –. Allen et al. (: ‘[P. Berol.] shows . . . that poems of the Homeric corpus were appropriated

at this period by the Orphics’), Richardson (: : ‘[T]he Homeric Hymn was taken over andattributed to Orpheus’). Comparably, P.Derv. xxvi.– suggests an ascription of Il. .– to‘Orpheus’, rather than ‘Homer’, assuming that at P.Derv. xxvi., ��0 is used personally, withOrpheus as subject (cf. P.Derv. xiii., xvi., xxi.). See Funghi (: ), Obbink (: n. ),Betegh (: , n. ). Differently, Kouremenos et al. (: n. , ).

Kruger (: ) is understandably suspicious of this coincidence. However, the problems do notseem fewer on alternative views of P.Berol.



bruno currie

sources for the paraphrase and for the verse citations. Something like thisis the assumption of scholars who see the prose paraphrase as testimonyto a lost ‘Orphic’ Hymn to Demeter, but do not see the verses quotedin the papyrus as coming from a lost ‘Orphic’ poem. This makes thecommentator’s procedure extremely hard to understand: it becomes eitherinept (if the commentator has failed to realize that the plot of HDem isdifferent from his paraphrase), or deceptive (if the commentator intendsthe reader not to realize; but how could such an intent ever succeed?); theredo not seem to be any grounds to impute a playful or humorous intentto the author. It is true that our interpretation of the Berlin papyrus is ingeneral hampered by our ignorance of the agenda of the commentator (asof his identity), but on this view it seems that we must resign our hopeseven of making sense of what he is about.

Second, it is not possible to argue that the Berlin commentator quotedfrom HDem because there was no other poem for him to quote from.Around the middle of the third century bc (terminus ante quem), theParian Marble attests the existence of ‘(hexameter) poetry on the rape ofPersephone and the search of Demeter’ attributed to Orpheus. From thescanty indications of the Parian Chronicle, the plot of this ‘Orphic’ poetryseems to have both diverged from that of HDem and converged in essentialswith that of the paraphrase of the Berlin papyrus. Given the availabil-ity, perhaps widespread, of such a poem on the Rape of Persephone –which was distinct from HDem, was ascribed to ‘Orpheus’ and had a plotapparently conformable to what the Berlin commentator required – it ishard to account for the failure of the commentator to use this poem. Or(to make the same point the other way round) the independent attestation,from at least the third or fourth centuries bc, of just such an ‘Orphic’ poemmakes it likely that the commentator was drawing on a poem distinct fromHDem, ascribed to Orpheus, and, if not identical with the poem referredto in the Parian Marble, then substantially similar to it.

See above, n. . This is the unattractive assumption of Kruger (: ). Cf. Buecheler (: n. ). FGrH A. The terminus ante quem can probably be pushed further back, to the fourth century

bc: Asclepiades and Palaephatus (FGrH F and FGrH F, both in Harp. s.v. A# +��),probably after an ‘Orphic’ poem; cf. Jacoby (–: I.a. on FGrH F), Richardson (: ),Graf (: –). In general, ‘Orphic Hymns’ are attested in the fifth–fourth centuries bc (P.Derv.xxii.; Pl. Leg. d–e), and still being read in the second century ad (Paus. ..). The archaic-classical ‘Orphic’ Hymn(s) to Demeter is (are) probably reflected in Eur. Hel. – (cf. Kannicht: ) and OA –, – (cf. frr. – Bernabe; Vian : ; Nelis : ). Different fromthese early ‘Orphic Hymns’ is the extant collection (from the th c. ad?) of ‘Orphic’ Hymns, of whichno. is to Demeter (‘Meter Antaia’); see Athanassakis (: vii), Betegh (: ).

The crucial divergence from HDem and convergence with P.Berol. consists in Demeter’s gift ofagriculture to the Eleusinians: Marm. Par. (FGrH A) �� S��['�� .� ��.

Cf. Buecheler (: ).




Third, it is impossible in all instances to regard the verses cited by theBerlin papyrus as coming just from a wild or rhapsodic text of HDem: wecome here to the significant exception alluded to above. Three verses inthe papyrus (P.Berol. lines – = fr. .– Bernabe) indicatea version in which Demeter came to Eleusis ignorant of the identity ofPersephone’s abductor, and thus embody a conception of plot incompatiblewith that of HDem – even though their wording is still extremely closeto lines – of HDem (more on these verses below, pp. –). At thevery least we would have to assume the commentator, if drawing on HDemfor these verses, has wilfully tampered with them. Again, this wouldinvolve the imputation of a deceptive or disingenuous procedure to thecommentator and, perhaps harder still, the assumption of a public capableof being duped. Such a scenario is hard to endorse without a convincingaccount of the commentator’s agenda and procedure.

These three points are weighty enough, I believe, to warrant the assump-tion that the verses cited in the papyrus derive from a poem distinct fromHDem. The assumption, I contend, is warranted; I do not maintain thatit is inescapable or without problems of its own, only that it is viableenough for its implications to be worth exploring, the more so as theseimplications are interesting ones. I hereafter make the assumption that thecommentator is working as he purports to be, in both the paraphrase andthe verse citations, with a poem distinct from HDem and call that poem ‘theHymn to Demeter in the Berlin papyrus’ (HDemPBerol ). This assumptioncan still accommodate the view that HDemPBerol consists to a large extentof verses adapted from HDem; it excludes only the view that the commen-tator is drawing directly and exclusively on HDem. For the rest of this pieceI forego further discussion of these first-order questions about the statusof the verses in the papyrus and address myself to second-order questionswhich are directly pertinent to the question of relative chronology: whichof these two poems depends on which and whether demonstration of aliterary relationship here yields a relative chronology.

3 hdem and hdempberol

Scholarship has provided absolute dates for both poems: a current con-sensus places HDem in the earlier sixth century, while HDemPBerol has

These three lines of the papyrus refute the view of Allen et al. (: ): ‘The passages quoted asfrom Orpheus or Musaeus show several variants from the text of the hymn given in M, but they donot amount to a new version’ (my italics).

Cf. Richardson (: ). With those scholars cited above, n. . Richardson (: –), Janko (: ), H. P. Foley (: –), West (b: ).



bruno currie

been dated variously to the late sixth and the later fifth centuries. Thedating of HDemPBerol in particular rests on slippery criteria. Probably weshould keep an open mind about both the relative dating and the relation-ship of the two poems. We are, though, in a position to say somethingabout the background of the Hymn to Demeter in the archaic period. Therewere evidently several hexameter compositions on the Rape of Persephonecurrent in Attica in the archaic (and/or early classical) period: apart fromHDem, ascribed to ‘Homer’, we know of early hexameter Hymns to Demeterascribed to Orpheus, Musaeus, Eumolpus, and Pamphos. Ancient tradi-tion (for what that is worth) regularly placed these poets before Homer.

In any event we are not entitled to assume that HDem is the oldest of theattested Hymns to Demeter. A couple of verses of Hesiod’s Theogony (seebelow) imply that the Rape of Persephone was a subject of hexameter verseas early as the eighth century, perhaps a century and a half before HDem.All the attested archaic Hymns to Demeter had associations with Attica.

There is a notable degree of conformity between them. In the first instancethis consists in certain stable elements of the plot: the abduction of Perse-phone by Hades, Demeter’s search for Persephone, Demeter’s hospitablereception by the Eleusinians, Demeter’s rewarding of the Eleusinians withthe gift of agriculture and/or her mysteries. Second, certain specific motifsrecur in some of them: Persephone’s gathering of flowers before the rape,

Demeter’s sitting down by a well in Attica as an old woman. Third (apoint of particular interest for us), there seems to have been at least some

Late sixth century: Malten (a: , pointing to the circle of Onomacritus), Kolb (: –);cf. other references in Graf (: n. ). Later fifth century: Graf (: –, dating thepoem to – bc, and arguing that the author is Lampon or one of his immediate circle). Cf.in general, H. P. Foley (: n. ), Parker (: n. ).

Graf () makes much of the transition from uncultivated to cultivated existence, relating it tolater fifth-century sophistic thought, and to Sophocles’ Triptolemus.

So Colli (: ). Orpheus: P.Derv. xxii. (= F Bernabe), Marm. Par. = FGrH A, P.Berol. –. Musaeus:

Paus. .., ... Eumolpus: Suda s.v. dZ'��. Pamphos: Paus. .., .., .., ... Eumolpus: Suda s.v. dZ'��. Pamphos: Paus. .., .., ... In general, cf. Hdt. .,

.. Cf. Linforth (: ). Cf. Malten (b: ). We must similarly assume the existence of earlier hexameter hymns before

HAp and HHerm. On the Attic context of HDem, cf. Richardson (: , , ). The Hymns to Demeter by Pamphos,

Orpheus, and Musaeus were used in the mystery cult of the Lykomid genos in the Attic deme ofPhyle (Paus. .., .., .., ..). Eumolpus is eponym of the Eleusinian hierophantic genosEumolpidai.

Pamphos: Paus. ... Orpheus: P.Berol. –, cf. –; cf. OA , Clem. Al. Protr. .. (=F. Bernabe). HDem –. For divine abductions of girls picking flowers as a traditional motif ofearly hexameter poetry, cf. Richardson (: –).

Pamphos: Paus. ... Orpheus: P.Berol. –. HDem –.




recurrent phraseology in the archaic Hymns to Demeter. This importantfact emerges from a confrontation of Hesiod’s Theogony, lines –:

(�,� r A:'�� #1��>�� 7� �� M�3��` �� 1�� #�� `� ��9�\�� , '�� C�� '�� ;�+�

with HDem –:

A:'��@ I|�'�� '�V� 3�,� P��'@ ��(�V� I�� 3+4�� + 1#�� `� ��9�\��[�� >�+�#�� (�+�� ;�+�.

The allusive character alone of the Hesiodic verses implies that the Rapeof Persephone was a traditional subject of hexameter poetry before theTheogony. But the detailed correspondences between them and the open-ing verses of HDem imply that this traditional poetry had, in addition torecurrent episodes and motifs, at least some recurrent verses or phraseology.This finding is particularly interesting when it comes to evaluating the sur-prisingly large number of shared verses between HDem and HDemPBerol.

These three levels – plot (myth), motif and phraseology – are all impor-tant to our inquiry. On the one hand, it is often granted by scholars thatHDem takes in general an innovative and allusive approach to the tradi-tional Attic myth; and it is granted that HDem redeploys at least sometraditional phraseology, in the light of the comparison of HDem – withHes. Theog. –. On the other hand, it is commonly assumed thatthe paraphrase of the Berlin papyrus provides testimony to the traditionalAttic myth with which HDem is engaging; but it is almost universallydenied that the verses cited in the papyrus constitute traditional verses towhich HDem may be reacting. Partly this is a consequence of the viewof the verses in the papyrus as mere quotations, with or without deliberateadaptation, from HDem. Partly too it is due to theoretical assumptions,for the asymmetrical treatment of myth and wording here mirrors thatprevailing in Homeric scholarship: while a number of scholars allow that(say) the Cypria, Aethiopis or Little Iliad may preserve pre-Homeric myth to

Hesiodic authorship has been denied by e.g. West (: –). Differently, G. P. Edwards (:–), Janko (: –, ), Drager (: –).

So Richardson (: –), Clay (: , –), H. P. Foley (: –); cf. Parker (:–).

Cf. Richardson (: , ), Clay (: and n. ); cf. Bohme (: , : –). In thispart of the Theogony, the catalogue of Zeus’ unions with goddesses (–), Hesiod apparentlyalludes to much traditional mythical material: Richardson (: ).

Cf. Bernabe (: , note on line ); and the references in n. . A solitary – if persistent – voice has been Bohme (: –, : – and elsewhere).



bruno currie

which the Iliad alludes, few allow that they preserve pre-Homeric phrasingto which the Iliad alludes. It is against this background that I mean toargue for two propositions. First, that HDem presupposes and alludes toearlier Attic hexameter poetry on the Rape of Persephone, and does so onlevels of plot, motif and phraseology that exceed the currently prevailingmodels of mythological intertextuality or traditional referentiality. Second,that the Berlin papyrus gives us access, direct or indirect, to the earlier Attichexameter poetry on the Rape of Persephone to which HDem alludes. Itis the language and concepts of neoanalysis that I find most serviceable todescribe the relationship that I discern between HDem and HDemPBerol;and my analysis can also be seen as giving some support to the method ofneoanalysis. I structure my argument around three related ‘techniques ofallusion’ in HDem, each paralleled in well-known neoanalytical argumentspertaining to the Iliad or Odyssey.

4 hdem alludes by transferred motif and

transferred wording?

My first technique is allusion by transferred motif and (more controver-sially) transferred wording. It can be exemplified by a well-known neo-analytical argument from Iliad .–: Achilles, Thetis and the Nereidslamenting Patroclus’ death is held to recall Achilles’ own funeral as this wasnarrated in earlier epic poetry (a ∗Memnonis, say, the hypothetical forerun-ner of the Cyclical Aethiopis). The detailed motifs of such a scene seemto be reflected at Od. .– and in Proclus’ summary of the Aethiopis(Bernabe : .–). But beyond detailed correspondence of motifs,the phrase '�4� '�4�� . . . | �0�� (Il. .–), applied to the pros-trate Achilles on learning of the death of Patroclus, closely corresponds toOd. . �0 � '�4� '�4�� , applied to the prostrate Achilles at hisown death. Several neoanalysts have thus argued that the Iliad has adopted(or ‘quoted’) wording used in earlier poetry of Achilles at his own funeral.This argument would find in the Iliadic passage both a transferred motifand transferred wording. Moreover, it assumes that the transference is allu-sive: the audience is to sense both the difference of the Iliadic scene fromthe scene of the Aethiopis (∗Memnonis) and its similarity to it; perception

Cf. Burgess (: –). Pestalozzi (: ), Kakridis (: –), Schadewaldt (: ), Kullmann (: –), Dowden

(: , n. ), Willcock (: ), Danek (b: –, b: ). Cf. Janko (: ).West (b) has recently challenged the neoanalytical view of the relative chronology of the Iliadstory and the Memnon story; but see Currie (: –).




of the ‘source text’ behind the ‘target text’ is crucial to the understand-ing of the scene. A final point worth emphasizing is that transferenceof the motif to a new context implies that the poet will not go on toemploy the motif in the traditional way: having conveyed – symbolically,allusively, proleptically – the death of Achilles, the Iliad poet is not goingsubsequently to narrate the death of Achilles (along traditional lines, or atall).

Consider now HDem –: Demeter converses with Hecate and Heliosand learns from the latter that Hades has abducted Persephone. The sceneis likely to have reminded an Attic audience of a functionally equivalentscene in earlier poetic treatments of the Rape of Persephone: a scene ofrevelation in which Demeter learned from the Eleusinians that Persephonehas been abducted by Hades. Such a scene is described in the paraphraseof the Berlin papyrus (P.Berol. – = frr. – Bernabe). It must,moreover, have been standard in the Attic version, where it was very likelyan important, climactic scene. The role of Helios as Demeter’s informantin HDem appears to be an innovation of the Homeric hymn; it should beconsidered as ‘secondary’, whereas the role of the Eleusinians as informantsis (in Attic tradition) ‘primary’. But it is important also to recognize thatthe version of HDem necessarily excludes the standard Attic version: a plotin which Olympians (Helios) act as Demeter’s informants rules out one inwhich Eleusinians (Triptolemus) play that role. That is already, in effect,an allusion: the Hecate–Helios scene in HDem derives part of its meaningfrom what it, by implication, excludes. Nothing less is at stake here thanthe 5�� of the traditional Attic myth: if Demeter is not helped by theEleusinians to recover Persephone, she cannot institute agriculture and/orthe mysteries at Eleusis out of gratitude for the Eleusinians’ help. HDemthus tacitly advertises a radically different treatment of plot – and so signalsa competitive relationship with the traditional Attic version.

The allusive relationship between these two scenes seems to be high-lighted by transferred motifs: comparison of HDem – and P.Berol.– (from the paraphrase) reveals a suggestively similar constellationof narrative details. Both narratives present an inconclusive preliminary

Cf. Burgess (: , this vol.: ). Pace West (b: –). A hexameter poem of ‘Orpheus’ in which the Eleusinians made the disclosure to Demeter and were

rewarded by her is explicitly recorded by Paus. ..; it is implied also by Isoc. Paneg. (where itis called a '#3�� . . . ��4��). Richardson (: ); cf. Kruger (: –).

On the identity of Demeter’s traditional informants, cf. Richardson (: , ). For problemsin general with the terms ‘primary’ and ‘secondary’, see Burgess (: –).

For the Eleusinians’ disclosure as the 5�� for the institution of agriculture/the mysteries atEleusis, cf. Isoc. Paneg. , Paus. ... Cf. Parker (: ), H. P. Foley (: ).



bruno currie

conversation between two females (Demeter and Baubo in P.Berol., Hecateand Demeter in HDem) followed shortly by the decisive revelation from amale (Triptolemus in P.Berol., Helios in HDem). Within this basic com-mon pattern there are inversions of detail. In HDem, Demeter is questionedby Hecate, whereas in HDemPBerol Demeter questions Baubo; in HDem,Demeter and Hecate go to meet Helios, whereas in P.Berol. (the para-phrase) Baubo and Demeter are themselves met by Triptolemus returningfrom the field. Inversions of source texts by target texts are a familiar featureof allusion in later classical literature, but they are also a significant featureof archaic poetry’s, and Homeric epic’s, transformation of its ‘models’;

our scene in HDem appears to be such a transferred motif with inversion.Here, the neoanalytical approach seems to impose itself strongly: the motifseems too specific to the Demeter myth to be regarded as a type-scene; andpriority for the motif seems to be assured: the context of the Berlin papyrusis primary, that of HDem secondary.

Apart from transference of motif, there is also arguably transference ofwording. Compare the three lines in which Hecate addresses Demeter atHDem –:

�� A�':�� z��1�� 4�� 3�� (�� I� 3�� 3��\�� 1�� .� 1�� ^�� 3#'��~

with the three lines in which Demeter addresses Baubo in HDemPBerol(P.Berol. – = fr. .– Bernabe):

��'� �� A�[']:�� z��1��[�� 4��]��.�� 3�.� �(�� I� 3�[�]��[� ��3��]��\�� m�� 1[�]�� [J.� 1�� ^�]1� 3#'��~

From a neoanalytical perspective, the motif of Triptolemus as witness to the rape and as Demeter’sinformant is transferred to Helios. The Sun is of course in his own right an appropriate enoughwitness of and informant on the rape (see Walcot : ; Richardson : –, ; Penglase: with n. ); for the Sun as witness, cf. Il. .; as informant, cf. Od. .. But Helios isalso a peculiarly appropriate stand-in for Triptolemus, for the two of them share the highly unusualattribute of being conveyed over the world in an air-borne chariot (Triptolemus: Soph. (TGF IVF); LIMC s.v. ‘Triptolemos’, passim; Helios: HDem –). For the possibility of iconographicconfusion between the two, cf. Mastronarde (: ), to which add Sen. Med. –.

Currie (: , –, –). A good example of opposition in imitation from fifth-century chorallyric is Bacchyl. .–, where the simile of Il. .– is transferred from the Greeks to theTrojans and 7�BH�� 3#'�� is transferred from the tenor to the vehicle.

Cf. Bohme (: ): ‘Die Frage als Ganzes ist primar die der suchenden Mutter: sie an Demeterrichten zu lassen – als ob diese es wusste – ist wenig sinnvoll und fraglos sekundar. Kein Zufall dassHekate von Demeter gar keine Antwort erhalt.’ But the adaptation of the motif in HDem shouldnot be seen as ‘wenig sinnvoll’, but as consciously allusive; even the inconsequentiality of the scenein HDem may be part of the point: cf. below, p. .




The vast majority of scholars who have discussed these two passageshave seen the papyrus as adapting the verses of HDem. This view hasits problems, not least the hiatus z��1�� 4�� (). In earlyGreek hexameter poetry hiatus frequently results from the adaptation of atraditional formula to an untraditional context. In this case the traditionalformula is evidently the nominative, A�':�� z��1�� 4��,the form we find elsewhere in HDem (). A second anomaly in thesame verse () is the use of the nominative A�':�� for the vocativeA:'��. On the two other occasions in HDem where a vocative of thegoddess is required, we do find A:'�� (HDem , ). These linguisticanomalies together seem to expose HDem as making untraditional useof traditional language. On the other hand, the papyrus’s verse ��'� ��A�[']:�� z��1��[�� 4��]�� conforms to what we would expectof a traditional use of the formula.

Various parallel arguments point in the same direction: in terms of plot,the papyrus’s version (where Eleusinians act as Demeter’s informants) con-forms to the traditional Attic version, whereas the version of HDem departsfrom it; on the level of wording, HDemPBerol exhibits a traditional useof a formula, whereas HDem deploys that formula innovatively. The con-clusions that suggest themselves are, first, that the verse ��'� �� A�':��z��1�� 4�� was a traditional verse spoken by Demeter inthe context of her self-revelation to the Eleusinians in (some) early Attichexameter poetry on the Rape of Persephone; second, that HDemPBerolreflects that traditional poetry; and third, that HDem has transferred (allu-sively) those traditional verses from the traditional context of the disclosureat Eleusis to its own innovative scene featuring the Olympians Hecate andHelios. If so, the Berlin papyrus in conjunction with HDem offers us aconcrete textual example of the kind of intertextuality that remains purelyconjectural in the case of the Iliad ’s '�4� '�4�� . . . �0�� (Il. .–). The resulting picture is of HDem engaging allusively with its tradition,on the level of wording as well as of motif; and P.Berol. seems to reflectthat tradition, on both levels.

Except Bohme (cf. above, n. ). It has been argued that P.Berol.’s ��'� �� A�[']:�� z��1��[�� 4��]�� is a combination

of HDem ��'� �� A�':�� with HDem z��1�� 4��: cf. Richardson (: , ,). Yet it would be odd for HDemPBerol to have produced a more traditional-seeming verse bysuch means.

E.g. Milman Parry in A. M. Parry (ed.) (b: –), G. P. Edwards (: –). Cf. HDem . The nominative is felt to be so anomalous by Cassola (after Ilgen) as to need emendation (Cassola

: ). So Bohme (: ). Cf. Richardson (: ).



bruno currie

The implications of this for relative chronology are not straightforward.A strong position would be that HDemPBerol is the very poem alludedto by HDem and is therefore older than HDem. But a weaker positioncan also be adopted: that HDemPBerol preserves at least some traditionalverses that are earlier than and are alluded to by HDem. This positionwould be tenable even if HDemPBerol were a more recent poem thanHDem, and it could be compatible even with the view that HDemPBerolis itself dependent in places on HDem, provided that this was not the onlydependence. This is akin to the neoanalytical argument that the Cyclicalepics, while drawing on pre-Iliadic material on which the Iliad draws, alsodraw on the Iliad. We will return to such questions below (p. ).

5 does hdem allude by inconsistency between ‘surface’ and

‘deep’ layers of narrative?

I pass to my second technique of allusion: allusion to a traditional versionhighlighted by inconsistency between ‘surface’ and ‘deep’ layers of narrative.Again, the technique may be illustrated by two well-known examples fromHomeric neoanalysis. First, neoanalysts have often held that the narrativeof Sarpedon’s death in Iliad is modelled on a traditional narrative of thedeath and immortalization of Memnon, as it is preserved in the Aethiopis.

The two narratives are compared in table ..

In the narrative of the Iliad there is a notable inconsistency betweenthe text’s ‘surface’ insistence that Sarpedon is not immortalized (. =

) and the simultaneous intimations of the text at a ‘deep’ level of hisimmortalization, intimations that are effected both by individual motifsand by the sequence as a whole.

A similar approach has been taken to another narrative sequence, fromOdyssey . Georg Danek and others have argued that the Odyssean narra-tive is modelled on a lost earlier poetic treatment of Odysseus’ homecom-ing in which Penelope recognized Odysseus and conspired with him in themurder of the suitors (see table ., in which the elements in the left-hand

This is the view of Bohme (cf. n. ). Cf. Kullmann (: ). See e.g. Janko (: –). A note on the reconstruction of the Aethiopis: stages and are from Proclus’ summary (Bernabe

: –.–). For stage , cf. Schadewaldt (: , ), Janko (: , ). For stage, cf. apparent reflections of the Aethiopis in ‘Aesch.’ (TGF III.) and in vase-painting (LIMCIII.i.–, VI.i.–, –).

Individual motifs: the translation of the body by Sleep and Death; the annointing of the body withambrosia; the clothing in immortal clothes (cf. Achilles at Od. .; Janko : : ‘Immortalclothes are reserved for those who are to be heroized . . . or for gods’). Sequence as a whole: seeCurrie (: –).




Table . Comparison of the deaths of Memnon and Sarpedon

Aethiopis Iliad .–, –

Achilles kills Memnon. Patroclus is about to kill Sarpedon. Dawn asks Zeus to make Memnon

immortal, and Zeus grants himimmortality.

Zeus asks Hera whether he should rescueSarpedon from imminent death; Heravehemently rejects the proposal.

(Not directly attested for Aethiopis:)Dawn bathes Memnon’s corpse,annoints it with ambrosia and clothes itin immortal clothes.

Apollo bathes Sarpedon’s corpse,annoints it with ambrosia and clothesit in immortal clothes.

Dawn (assisted by Sleep and Death?)carries Memnon’s corpse to Ethiopia.

Sleep and Death carry Sarpedon’s corpseto Lycia.

Table . Hypothetical reconstruction of the Homecoming of Odysseus

∗Hypothetical pre-Homeric poem on theHomecoming of Odysseus Odyssey .–

∗Penelope interrogates Odysseus. Penelope interrogates Odysseus. ∗Penelope washes Odysseus’ feet. Eurycleia washes Odysseus’ feet. ∗Penelope recognizes Odysseus. Eurycleia recognizes Odysseus; Athena

prevents Penelope from recognizingOdysseus.

∗Odysseus and Penelope plot the killing ofthe suitors.

Penelope has dreamt of the return ofOdysseus and the killing of thesuitors, but she does not believe it.

∗Odysseus tells Penelope to propose thecontest of the bow so that he may kill thesuitors.

Penelope proposes the contest of thebow believing that she must take anew husband.

column are hypothetically reconstructed, not actually attested, and henceprefaced with ∗).

On a ‘surface’ level, the Odyssey’s narrative insists that Penelope neitherrecognizes nor conspires with Odysseus. However, at a ‘deep’ level, indi-vidual elements of the narrative and its overall shape intimate a recognitionand conspiracy. Following Danek and others I find it attractive to see suchinconsistencies between ‘surface’ and ‘deep’ layers of narrative as an allusivetechnique, whose effect is to orient the audience simultaneously towardsthe traditional version and towards its reworking in the present poem: thetraditional narrative is both superseded and left vestigially present in the

Danek (b: ); cf. Rutherford (: , : ); Currie : –.



bruno currie

text. This view sees such inconsistencies not as the result of an improvisingoral poet caught involuntarily between the conflicting claims of traditionand innovation (Lord : ), but as evidencing, even advertising, thepoet’s intention of getting his audience to confront a new version with theold version with which it is in competition.

In HDem we may see a similar tension between ‘surface’ and ‘deep’ layersof narrative. Here, however, the evidence of the papyrus is deficient in acouple of places (see , and in the table) and has to be supplementedfrom other sources for the Attic myth of the Rape of Persephone (seetable .).

The Berlin papyrus seems to attest a version in which Hades descendedwith Persephone into the Underworld at Eleusis, whereupon Demeterfollowed Persephone’s cries to Eleusis, learned from the Eleusinians thatHades had taken Persephone into the Underworld, descended into theUnderworld to retrieve her daughter, and by absenting herself from theupper world caused a famine. Assuming such a traditional version, HDemwill have preserved most of the key elements in their expected order:Hades takes Persephone into the Underworld, Demeter arrives at Eleusis, afamine breaks out. But although most of the key events occur in the samesequence, the events are, explicitly or implicitly, very differently motivatedand consequently the significance of the narrative is transformed. Mostprovocatively in HDem, on the ‘surface level’, Demeter’s visit to Eleusishas become irrelevant to her quest for Persephone. Still, at a ‘deep level’,individual narrative elements and the general direction of the narrativeintimate a version in which Demeter’s sojourn in Eleusis is integral to herrecovery of her daughter. Again, I would see in this an allusive intent:HDem plays off the traditional version. And once again, the traditionalversion seems to be that of the Berlin papyrus. This technique of allusionimplies a detailed knowledge on the part of the audience of whole narrativesequences of earlier poetry.

6 does hdem allude by inconsistency between

narrator-text and character-text?

My third technique of allusion may be thought of as a special case ofthat just considered: inconsistency specifically between narrator-text and

Cf. Rutherford (: ), Danek (b: passim), Burgess (: ). H.P. Foley (: , n. ) well recognizes the ‘gaps in motivation’ as a means for the poet to

‘shape audience reaction and expectation even in the case of a traditional oral narrative’. For a different view of the irrelevance, cf. Richardson (: –), supposing that HDem has

imperfectly combined originally separate narrative strands. A similar view of narrative inconsistencies in HDem is taken by Clay (: –, –).



Table . Comparison of the Rape of Persephone and the Hymn to Demeter

Traditional Attic version of Rape ofPersephone (P.Berol., supplementedfrom other sources) Homeric Hymn to Demeter

Hades descends into the Underworldwith Persephone near Eleusis.a

Hades descends into the Underworld withPersephone, but there is no indicationwhere the descent occurs (–).

Demeter comes to Eleusis in search ofPersephone.b

Demeter comes to Eleusis, but there is noindication why she does so (–).

Demeter reveals herself to theEleusinians and is informed by themthat Hades has taken Persephone intothe Underworld.c

Demeter reveals herself to the Eleusinians,but receives no information from them(–).

(Attestation controversial) After therevelation at Eleusis, Demeterdescends into the Underworld.d

After the revelation at Eleusis, Demeterwithdraws into the temple which hasjust been built for her at Eleusis(–).

(Conjectural) Demeter’s absence fromthe upper world causes a famine,necessitating Zeus’s intervention.e

Demeter, secluded in her temple atEleusis, causes a famine by an act of herwill, necessitating Zeus’s intervention(–).

(Not in P.Berol.) After the recovery ofPersephone (?), Demeter instructs theEleusinians in the mysteries and/oragriculture in gratitude for theirinformation.f

After the recovery of Persephone, Demeterinstructs the Eleusinians in hermysteries, but not in the arts ofagriculture, and not out of gratitude forany information received (–).

a P.Berol. . Cf. Kruger (: ), Richardson (: , under (c)).b Cf. P.Berol. –, Clem. Al. Protr. .. (= F Bernabe).c Cf. P.Berol. –.d P.Berol. /3�� 3�� 4[�]�[�]. I take this to refer to a descent into the Under-

world of Demeter in search of Persephone, which is sufficiently justified by later allusions(Hymn. Orph. .–; Hyg. Fab. ; cf. schol. Pind. Ol. .c; see Richardson :, ). Differently, most commentators have assumed rather the descent of Perse-phone into the Underworld with Hades; Bernabe (fr. .–) actually supplements thepapyrus text to read /3�� U�3�� 4[�]�[� �� U�� ] (following Buecheler,who however left the supplement in the apparatus); cf. Richardson (: ).

e This motif is not attested, but is conjectured on the strength of various Near Easternparallels. First, the Akkadian Descent of Ishtar, lines –, –, in which Ishtar’sabsence in the Underworld brings an end to reproduction on earth: Foster (: –);see Graf (: –), Burkert (: , n. ), Foley (: –), Penglase (:–, cf. –). Second, the Ugaritic Baal and Mot, where Baal’s absence in theUnderworld causes famine and drought (Wyatt : –): see Richardson (:), Hooke (: –). Third, the Sumerian complex In the Desert by the Early Grass,where the dead Dumuzi’s mother searches for him and determines to descend into theUnderworld after him: Jacobsen (: – at ); see Penglase (: –, –, ).Note that on this reconstruction famine would have a place in the Orphic version, paceRichardson (: ), Foley (: ).

f This was apparently the standard Attic myth: cf. Isoc. Paneg. , Marm. Par. (FGrH

A), Paus. ...



bruno currie

character-text. I begin, as usual, with an illustration from Homeric epic:the speech of the soul of the murdered suitor Amphimedon to the soul ofAgamemnon in the last book of the Odyssey (.–). This character-text is consistent with the narrator-text of the Odyssey except in a couple ofkey points: in particular, Amphimedon’s soul claims Penelope’s complicityin the killing of the suitors. A number of considerations point to thisbeing a traditional version of Odysseus’ homecoming. Danek has arguedthat ‘[f]or the listener the discrepancies between the two versions [sc. thenarrator-text and character-text] present themselves as a citatory referenceby the [poet] to the tradition, from which his own version is made tostand out.’ The Deuteronekyia of book , like the Nekyia of book ,seems particularly interested in exploring the Odyssey’s relationship withother poetry. In so ‘metapoetic’ a context, the use of a character-textto highlight the relationship between the Odyssey and earlier poetry onOdysseus’ homecoming makes good sense.

In this light we may consider the speech of Persephone at HDem –:Persephone narrates to Demeter what befell her in the Underworld and inthe Nysian plain at the time of the rape (Hades’ ruses, respectively, of thepomegranate and the narcissus). Like the speech of Amphimedon’s soul inthe Odyssey this speech of Persephone throws up significant points of agree-ment and disagreement with the version of the primary narrator (HDem– and –). Some of these differences (as with Amphimedon’s soul’sspeech) may be explained with reference to who is speaking and to whom.Thus it is natural for Persephone, speaking to Demeter, to claim that shewas not merely tricked but forced into eating the pomegranate, althoughthe narrator-text mentions only guile (Richardson : ad ). How-ever, at least one difference may be explained as a pointed inclusion ofa traditional version. In the verse which rounds off the list of nymphsin attendance at the time of her rape, Persephone includes Athena andArtemis (HDem ). There was no hint of either of these goddesses in thenarrator-text of the rape in HDem, where only the daughters of Oceanuswere mentioned (HDem ). However, Athena and Artemis are mentionedas present at the rape in the paraphrase of the Berlin papyrus (P.Berol.–). They occur, moreover, in so many other sources that their presenceat the rape must be regarded as a traditional element in the Attic Demeter

For the terms ‘narrator-text’ and ‘character-text’, see de Jong (: xxvii). See e.g. Rutherford (: –). Danek (b: , my translation); cf. Rutherford (: ). Cf. Currie (: –, n. ). Cf. Richardson (: ad ), H. P. Foley (: ). Cf. Danek (b: –), de Jong (: –).




and Persephone myth. Persephone’s character-text in HDem thus divergesfrom the narrator-text of HDem and converges with what is to all appear-ances a traditional version. Once again, the narrator-text of HDem appearsconsciously to depart from tradition in its narrative of the rape (Richardson: ). And once again the traditional version appears to be reflectedin the Berlin papyrus.

This neoanalytically influenced approach is, of course, only one of anumber of possible approaches to the anomalous presence of the verseHDem in Persephone’s narrative. An alternative is an ‘analytical’approach, which regards the verse HDem as interpolated into HDemthrough the influence of the traditional Attic version. This would parallelanalytical approaches to the Iliad which regard certain Iliadic passages asintrusions on the ground that they included ‘Cyclical’ material (‘Cyclical’being taken as synonymous with post-Homeric). Another approach iswhat we might call an ‘oralist’ approach, according to which HDem isa blemish resulting from the poet composing in performance and failing toexclude a traditional detail when otherwise recasting traditional material inan untraditional form: the notorious ‘nod’ of the oral poet. This approachwould parallel that which has sometimes been taken to the duals of Iliad or Zenodotus’ ‘Cretan verses’ in Odyssey . This ‘oralist’ approach sub-stantially converges with an ‘older’ neoanalytical approach that explainsinconsistencies as a failure of the poet to assimilate fully a new version withan old one. The ‘newer’ neoanalytical approach propounded by Danek,

which I have been following, finds evidence rather of an allusive strategy:a different kind of ‘nod’ to the traditional version. If this is correct, theBerlin papyrus would provide once again an external textual control for

For the version where Artemis and Athena are present at the rape, cf. esp. Eur. Hel. –, Diod...– (after Timaeus), Claudian, De Raptu Proserpinae .–. Also, Paus. .. and otherreferences given by Richardson (: ). See Graf (: –), Bernabe (: note on fr..). Traditional by at least th cent. bc: Dale (: on Eur. Hel. –), Kannicht (: on Eur. Hel. a-), Richardson (: –).

So Kannicht (: n. , after Malten); cf. Graf (: –). Kullmann (: –) discusses this approach with reference to Bethe and Wilamowitz; a recent

example of the approach is West (a: and n. ). In general, Lord (: ). Il. .–: cf. e.g. Griffin (: ). Od. . and : cf. e.g. Burkert (). Kakridis (: ): ‘What if here and there unassimilated points escape [the poet’s] attention? We

should be grateful to him for them, as otherwise it would be impossible for us to prove the extent,and above all, the nature of his dependence on his predecessors.’ Cf. Richardson (: ): ‘Allthese features may have been omitted in the epic narrative of the rape, leaving however a trace atthis point.’

See in general Danek (b: Index s.v. ‘konkurriende Alternativversionen’). J. M. Foley (: –) argues similarly against the view of a ‘nodding’ Homer, but from the

different theoretical standpoint of traditional referentiality.



bruno currie

HDem of a kind that we wholly lack for the Odyssey, and would providea rare textual basis for the attractive, but otherwise unsubstantiated, the-ory that early Greek hexameter poetry could use character-text to ‘cite’ atraditional version.

‘Cite’ is the term used (after Kullmann) by Danek (b: passim),who is chiefly interested in ‘citation’ on the level of plot. However,we may wonder whether HDem cites on the level of wording: whetherHDem could be regarded as a verbatim quotation from earlier poetry.The list of nymphs at HDem – occurs in virtually identical form inHDemPBerol (P.Berol. – = fr. .– Bernabe); and although theBerlin commentator’s quotation of HDemPBerol stops before that verse,it is possible that HDem also occurred in HDemPBerol. We mightthen assume that HDem has taken over – ‘cited’ – verses – fromearlier poetry, and assume that HDemPBerol reflects that earlier poetry.That assumption would receive support from the fact that in HDemP-Berol the list of nymphs in attendance on Persephone at the time ofthe rape came in its ‘natural’ place in the narrative (in the narrator-text and at the start of the narrative of the rape), whereas its positionin HDem is anything but straightforward (character-text, and postponeduntil very late in the poem: actorial analepsis (de Jong : xi)); onceagain, comparison suggests that the context of the lines in HDemPBerol isprimary, and secondary in HDem. One could thus argue that HDemPBerol(P.Berol. –) preserves the traditional form of the narrative of the rape,to which HDem alludes (–) by means of extensive verbatim citationwithin character-text. Such an argument would parallel the argument above(pp. –) that HDemPBerol (P.Berol. –) preserves the traditional form

The title of his book is Epos und Zitat; Kullmann (: ) uses ‘zitiert’ of a verbatim quote, Il. .‘citing’ Cypria fr. . Bernabe; Danek (b) also explores the possibility of citation on the levelof wording (‘intertextuality’). Allan (: and n. ) takes exception to the use of the terms‘quotation’ and ‘intertextuality’ in Currie () and Danek (b, b): in my view they remainfully viable. Apropos Allan’s disagreement (ibid.) with Danek’s dating of the ‘Verschriftlichung’ ofHomeric epic, see now also Danek in this volume.

HDem P�3� ��'�� ’ finds a close echo in OA P�3� �� # �. TheOrphic Argonautica quite likely follows an early ‘Orphic’ hexameter source at this point: see Nelis(); cf. Kruger (: ). Could that early ‘Orphic’ hexameter source be HDemPBerol ? IfHDem + – and were in HDemPBerol, the likelihood increases that HDem was inHDemPBerol.

Richardson observes that the language of Persephone’s character-text appears more traditionalthan the language of the corresponding narrator-text of HDem: see Richardson (: , ),comparing HDem (narrator-text) and HDem – (character-text). However, if that showsthat Persephone’s character-text represents an earlier version than the narrator-text in HDem, it failsto show that HDemPBerol represents an earlier version than HDem, for the untraditional-seemingverse HDem also occurs in HDemPBerol (P.Berol. ).

The phrase is Buecheler’s (: ).




of Demeter’s self-revelation at Eleusis, to which HDem alludes (–) bymeans of transferred motif and transferred verbatim wording.

Finally, one should note how self-conscious HDem’s allusive engagementwith tradition is on this interpretation. HDem inscribes the traditionalversion within character-text, and narrator-text and character-text are madeto play off one another. The striking emphasis placed by Persephone onthe veridicality of her speech, at both its beginning and its end (HDem

and ), might also be felt to suit a poem that plays self-consciously withdifferent versions of the same events.

7 self-consciousness in hdem

This issue of self-consciousness – of a poem’s awareness of its own indi-vidual standing within a tradition – deserves attention, as it has beenlinked directly with the question of intertextuality in early Greek hexameterpoetry. The Odyssey shows the clearest signs of such self-consciousness.

HDem shows further arguable signs of self-consciousness in addition tothose already mentioned. First, there is the striking tendency of HDemto duplicate motifs. Persephone’s return occurs in two stages: from theUnderworld to Eleusis (HDem –) and from Eleusis to Olympus(HDem ). Two deities, other than Demeter, are involved in the return:Hermes (who brings Persephone from the Underworld to Eleusis); andRhea (who brings both Demeter and Persephone to Olympus). And thereare two mother–daughter pairs, the mother in each case bringing her daugh-ter back to Olympus: not only Demeter and Persephone, but also Rhea andDemeter. It is tempting to think that such motifs may have been employedsingly in a traditional version of the Rape and there is in fact some evidencefor a traditional version in which Demeter, having descended herself intothe Underworld, conveyed Persephone by chariot herself directly from theUnderworld into the presence of Zeus on Olympus. Such reduplication

The special status of character-text, as a vehicle for traditional material the Homeric narrator-textotherwise avoids, was recognized already by Aristotle (fr. Rose, apud schol. A Il. .).

Cf. Richardson () on HDem , , . On ‘telling the truth’ and the (oral) poetic tradition,cf. M. W. Edwards (: –); cf. Danek (b: ‘Wahrheitsanspruch’).

Cf. Burgess (: ): ‘we cannot be sure that performance traditions would have had theself-awareness about either themselves or other performance traditions to engage in allusive inter-textuality.’ On self-consciousness, cf. also R. L. Fowler (b: –).

See esp. Od. .–, with Rutherford (: ) and Danek (b: ). Poetic self-consciousness is also to be seen in the Odyssey’s numerous depictions of bardic performances:Demodocus in book , Odysseus in books –, and Phemius especially at Od. .–, .–.Cf. too Od. ., with Currie (: n. ).

Well observed in general by Clay (: –). Schol. Pind. Ol. .c, where ��4�� suggests a traditional version.



bruno currie

of motifs may suggest a poet self-consciously playing with motifs that anaudience expected to be deployed and whose traditional form was wellknown to them. (We might compare Euripides’ handling of traditionaltragic motifs in the Helen, where the most self-conscious of tragedianstreats us to two prologues, two recognition scenes and two supplicationscenes.)

Self-consciousness might also be seen in the fact that, as J. S. Clayhas put it, ‘the poet seems to take perverse pleasure in thwarting ourexpectations and rendering the narrative progress problematic’ (: ).We have already seen (above, p. ) how in HDem traditional episodesoccur without their traditional consequences ensuing. Instead of narrativeprogress, HDem gives us regress, ring-composition frequently returningus at the end of an episode to its beginning. Thus, Helios’ disclosure toDemeter, far from advancing Demeter in her quest, only gives her greater‘grief’ (P��) than before the meetings with Hecate and Helios (HDem, echoing ). The episode in Eleusis does not advance Demeter inher quest either, and again the end of the episode echoes its beginning(HDem –, in ring-composition with –). While the reunion ofDemeter and Persephone at Eleusis brings an apparent resolution to therecurrent theme of Demeter’s ‘grief ’ (HDem �� @ ��+��3#'��, harking back to both P�� . . . F�� 3#'�� and �� P��C��>��), that resolution is only apparent. If a traditional ending indeedhad Demeter bringing Persephone back by chariot to Zeus on Olympus,then the motif that HDem offers us (Hermes bringing Persephone backby chariot to Demeter at Eleusis, HDem –) evokes that traditionalending by comparison and contrast: it is, in fact, a false ending. To convertthe false ending into a real one requires the introduction of Rhea, whoserole (like that of Helios) may be an innovation of HDem. Her involvementcertainly comes as a narrative surprise; before Zeus sent Hermes to Hades,we were told he had sent ‘all the gods’ (–) to treat with Demeter:evidently Rhea was one goddess he had kept up his sleeve. Rhea’s last-ditchintervention in HDem is comparable to a tragic deus ex machina (comparethe appearance of Heracles in Sophocles’ Philoctetes to ensure the properoutcome to the play after the playwright has teased us with several falseendings). And Rhea’s untraditional role in HDem enables an untraditionalsignificance to be given to the ‘Rarian plain’: according to Attic tradition,this was the first place to be sown with cultivated seed by Triptolemus,after Demeter had instructed him in the arts of agriculture. In HDem,

Marm. Par. (FGrH A), Paus. ...




its distinction is to be the place where Rhea first alights on her mission toconciliate Demeter (–, ), with an explicit negation of the plain’straditional significance (–: the plain was already a 1�� >�� ?3��+��). Again, the traditional version is strongly evoked while beingnegated.

A last sign of self-consciousness could be seen in Zeus’ role in HDem. InHomeric epic gods or ‘fate’ may sometimes be taken as figures for either thepoet or the dictates of the mythical-poetic tradition. In HDem the actionof the poem is from the outset signalled as coming about by the design ofZeus. The second half of the poem, however, shows Zeus’ plan gettinginto difficulties and Zeus struggling to get events back on course. Theend of the poem finally shows Demeter reconciled with Zeus on Olympus,and the poem thus broadly realigned with its traditional ending. Thistension between Zeus’ plan and the actual course taken by events in thenarrative of HDem may be seen as a figure for the hymn’s negotiation ofits own relationship to the tradition.

8 conclusions

It is time to draw some conclusions. First, regarding the literary relationshipof HDem and HDemPBerol. My argument builds on the recognition ofother scholars that HDem engages allusively with its tradition on the level ofplot, and on the recognition that some of its phrasing is derived from earlierpoetry on the rape of Persephone (compare HDem – with Hes. Theog.–). I have gone further than many scholars in suggesting that the Berlinpapyrus allows us to track the allusivity of HDem to an unexpected degree.I have argued that we may see certain common narrative structures, motifsand not least phrases as having their primary contexts of use in HDemPBeroland their secondary contexts of use in HDem. If this identification ofprimary and secondary contexts is correct, then HDemPBerol emerges asa ‘source’, either direct or indirect, for HDem. The indeterminacy of this

Clay (: ). In general, cf. Currie (: –, on Iliad), Danek (b: passim, on Odyssey). Janko (: , ). Compare the ‘table of contents’ speeches spoken by Zeus in both the Iliad

and Odyssey: de Jong (: ); cf. Macleod (: n. ). In the latter part of the Odyssey, Athenafrequently directs the action as a kind of figure for the poet-narrator: cf. Olson (: –, ),de Jong (: , ). On similar uses of the gods in tragedy, cf. Easterling (); and in Virgil,cf. Kennedy (: –).

HDem �� . . . ;�+�~ A�.� >�#� �; A�.� 7�� ~ – 5�� . . . ;�+�, | /� '��C��’~ U�� #��V� ��, '��.

HDem �� 'V ;�9� 7�� ; – (Zeus obliged to send Hermes to Hades); – (Zeus obligedto send Rhea to Demeter).

Differently, Clay (: –) interprets Zeus’ plan in cosmic terms.



bruno currie

formulation is meant to mirror the indeterminacy found in neoanalyticalscholarship over whether the Aethiopis is a direct or indirect source for theIliad: the Aethiopis will be a direct source if it is earlier than the Iliad andis the very poem to which the Iliad alludes; it will be an indirect sourceif, though later than the Iliad, it reflects a specific pre-Iliadic poem (or,depending on one’s theoretical standpoint, pre-Iliadic mythical traditionor non-specific pre-Iliadic poetry) to which the Iliad alludes.

If HDemPBerol is a direct source of HDem then we should expect it toemerge as primary at every point where the two may be compared; and wewould expect HDem to be more developed linguistically than HDemPBerol.This has indeed been argued by Robert Bohme, though the case is notstraightforward. Alternatively, if HDemPBerol were an indirect sourcefor HDem, it would be possible for HDemPBerol to emerge as primary insome points of comparison and secondary in others: that would parallela model that sees the Cypria as an indirect source for the Iliad (as givingus access to pre-Iliadic mythology and, possibly, to pre-Iliadic phrasing),but which sees the Iliad in turn as being a direct source for the Cypria.Some such model may be more adequate to the complexities of the datapresented by P.Berol.

On its own, then, the literary method may not be sufficient to establisha relative chronology of these precise poems. While we can with a certaindegree of plausibility say that the use of a motif or phraseology is primary inone context and secondary in another, it remains possible that motifs andphrases may occur in their primary contexts in poems composed later thanpoems that employ those motifs and phrases in secondary contexts. Inother words, a relative chronology of motifs (a prime concern of neoanalysisas ‘motivgeschichtliche Forschung’) does not equate to a relative chronologyof the poems in which they occur. This is partly because conservatism andinnovation are both so firmly entrenched in the early Greek hexametertradition, and because they may coexist in combinations that are hard tofathom within a single poem.

However, if my test case illustrates how difficult it may be in practice todemonstrate a direct literary dependence between a specific extant poem(HDem or the Iliad ) and a specific fragmentarily preserved poem (HDemP-Berol or the Cypria), the discussion nevertheless suggests a conclusion of

Cf. Kakridis (: ‘if Homer’s source is not the Aethiopis itself, at least it is an epic which . . . isto be to a great extent identified with the Aethiopis’), Kullmann (: –), Dowden (:), Burgess (: ).

Esp. Bohme (: –, : ). Cf. the opening paragraph of Danek in this volume (p. ), and Burgess in this volume (p. ).




more general value: it may not be mistaken in principle to think in terms ofliterary relationships between specific poems in the early Greek hexametertradition. To think exclusively in terms of mythological intertextualityor traditional referentiality may be to limit unduly the possibilities of thispoetic tradition. If each of my techniques of allusion is accepted for HDem,then we must posit an audience of HDem familiar with a version of thestory of the Rape of Persephone on several different levels: they will haveknown (and have been able, at least passively, to recall) an earlier version ofthe Rape in which (a) certain specific events occurred in a specific sequence,(b) certain specific scenes and motifs were drawn in considerable detail, and(c) certain specific hexameter verses and phrases were employed. It willbe obvious that such a scenario goes beyond mythological referentiality:elaborate narrative structures and verbatim phrases are evoked. Nor arewe dealing with verses or phrases that were traditionally used to describea particular mythological event without being anchored in any particularpoem. Faced with evocations of whole narrative structures, of scenes andof verses, I do not see what is gained by refusing to speak of allusion to aparticular poem. For the sake of this conclusion, it makes little differencewhether HDem is alluding to HDemPBerol or to another poem, many of

Cf. Dowden (: ). Differently, Burgess (: ; : : ‘The Homeric epics . . . probablydo not allude to specific poems’). Allan (: ): ‘[T]he pursuit of specific dependence or influence(from Homer to the cyclic poems, or vice versa) is, in the pre-textual stage of early Greek epic, amisleading methodology’; but ‘the pre-textual stage of early Greek epic’ (again at Allan :

n. ) hardly describes any historical stage of Greek epic; cf. Pelliccia (: –), J. M. Foley(: ).

For (a), see above, p. ; for (b), see pp. –; for (c), see pp. – and pp. –. Contrast Burgess (: ). Contrast, again, Burgess (: ); cf. Burgess (: , this vol.: p. ). Compare and contrast Burgess in this volume (p. , point ). Burgess posits ‘oral epic traditions’,

whose narrative outlines and phraseology were sufficiently stable to form the basis of ‘textlessintertextuality’ in early epic. It is an important question for me when we should begin to recognize‘oral epic traditions’ of such stability as a poem (on the understanding that a poem in the earlyepic tradition may have a very largely, but not totally, fixed textual form: see below, n. ).There is thus both important convergence and important divergence between Burgess’s positionand my own. Whereas Burgess thinks of oral epic traditions with a very high degree of fixity, Iprefer to think of hexameter poems of less than total textual fixity. The difference is more thanjust terminological. To talk of poems will permit one to think in terms of individually authoredcompositions, and to entertain the possibility that certain poems may ‘stand out’ from theirtradition – as ‘classics’ in their own time and thus prime targets for later allusion (my wordinghere is chosen to chime with both Burkert (: ) and Taplin (: –)). By contrast,to speak of allusion to ‘oral epic traditions’ (even ones that were remarkably fixed on the lexicallevel) implies an impersonal and anonymous model of allusion, on which individual poems orpoets could scarcely acquire prominence within their tradition, and on which interaction remainsresolutely on the parole-to-langue level (see above, p. and n. ). A related consideration is thedegree of poetic self-consciousness vis-a-vis its tradition that an early Greek hexameter poem couldpossess: another point of divergence between Burgess and me (see above, pp. – and n. ).



bruno currie

whose features (plot, motifs and verses) are retained in HDemPBerol. Theargument, of course, presupposes a considerable degree of textual fixity inthe early Greek hexameter tradition, but not absolute fixity: early hexam-eter poems might undergo small-scale changes in reperformance or textualtransmission without that affecting their ability to function as the objectof allusion or intertextuality for other poems.

Throughout this essay I have argued for a reading of HDem that parallelscertain well-known neoanalytical readings of the Iliad and Odyssey. Thoseneoanalytical readings have provided interpretative paradigms by whichto understand the literary relationships between the two archaic Hymns toDemeter. In turn the two archaic Hymns to Demeter give these neoanalyticalreadings a textual footing which they otherwise lack. So circular anapproach does not of course prove the legitimacy of such readings ofthe Iliad and Odyssey. But it does suggest the viability of extending theseneoanalytical readings to other early hexameter texts. If a theory’s valueresides in part in how well (how comprehensively, how attractively) itenables one to make sense of the available data then these perspectiveson neoanalysis from the archaic Hymns to Demeter have a serious claimon our attention. If it is true that the real difficulty with allusion orintertextuality in early Greek hexameter poetry is not one of theory butof evidence (Dowden : ) then we cannot afford to ignore the dataoffered by HDemPBerol and HDem, its very considerable complexitiesnotwithstanding. Nor should it be thought very strange that the criticalissues surrounding Homeric neoanalysis are seen to replicate themselvesin miniature with HDem. Walter Burkert once observed (: ) thatthe critical issues in Homeric analysis are played out, on a small scale,in the ‘Homeric’ Hymn to Apollo. In a different but complementary wayBernabe’s decision to edit the Orphic fragments in volume II of PoetarumEpicorum Graecorum Testimonia et Fragmenta alongside the fragments of

Cf. on this question Danek in this volume (p. , concluding paragraph). HHerm. – and Hymn. Hom. .– are clearly two versions of the ‘same’ poem, the divergences

between which may be seen as the result of the vagaries either of performance tradition (Hainswortha: –) or of textual transmission (cf. Cassola : ). Either way, the degree of fixityexhibited by the poem would seem quite sufficient to enable it to function as an intertext: cf.Currie (: ).

See pp. and –. This extension itself has interesting implications: motif transference and intertextuality will not

just be Homeric. Differently, Burgess (: , ). About verses of HDemPBerol are preserved, and the complete HDemPBerol can hardly have

totalled more than a couple of hundred verses (HDem totals verses). Contrast the Cypria, ofwhich some verses are preserved out of an original books (totalling some , verses, on aconservative estimate).




the Epic Cycle in volume I invites us to expand the repertoire of early Greekhexameter material to which neoanalytical approaches may be applicable.

I am conscious of having based some large conclusions on controversialtexts and controversial methodologies: it is perhaps appropriate to endby reviewing the problems. First, the problems surrounding the Berlinpapyrus cannot be minimized; in particular, the ‘first-order questions’ thatdominated the first part of my analysis lack a decisive resolution. I haveassumed that not all the verses in P.Berol. have been generated by HDem,but in at least some cases (P.Berol. – ∼ HDem –, perhaps P.Berol.– ∼ HDem –) are older than HDem and are alluded to by HDem.Second, it is legitimate to doubt how transferrable conclusions drawnfrom the archaic Attic Hymns to Demeter may be to the Homeric epics.In the first place, the Hymns to Demeter may have had a quite restrictedcirculation: circumscribed temporally, geographically and contextually (ifall are products of sixth-century Attica and performed in the context ofmystery cults). So selective a public may have created optimal conditionsfor allusion between poems, especially such short poems. There is also aquestion of (relative) orality: one might suppose that the poets of the(sixth-century) hymns worked with written texts in a way that the (eighth-or seventh-century) poet(s) of the Iliad and Odyssey did not. However,an extreme contrast between the Homeric epics and the hymns alongthese lines is not attractive: probably we should think of the Iliad and theOdyssey as existing in written form at the moment of their conception;

moreover, the archaic hymns were oral in the sense that is most crucial forour purposes, that they were performed to audiences who (for the mostpart) can hardly have known them in any other form. Proper though itis to ponder the cultural differences that a century or so may have brought,it is proper also to remind ourselves that we are dealing with a far closercomparandum than more culturally remote oral traditions can provide.

On this question, see Janko (: –). Cf. A. M. Parry (), J. M. Foley (: ). Cf. Richardson (: ).



chapter 12

The relative chronology of the Homeric Catalogueof Ships and of the lists of heroes and cities

within the Catalogue

Wolfgang Kullmann

I should like to begin with some preliminary remarks. The topic is notnew and is much debated. I myself have set forth my opinions on it inseveral publications, but there are some new aspects which seem worthdiscussing. I presuppose that we are able to speak of relative chronologyonly if it is possible to detect recognizable differences in the age of certainparts of the text or of certain constituent elements of it; and I think this isthe case with the so-called Homeric Catalogue of Ships (Il. .–).

My thesis is that the poet of the Iliad took over the epic Catalogue ofShips in a fixed form as a whole from an extraneous source or from one ofhis own earlier performances of another epic and adapted it to this poem.

In a way the Catalogue belongs to the scenes in the first books of theIliad which in one way or another mirror the events which took place atthe beginning of the war, as I have pointed out at length in Quellen derIlias. But there are differences. Take for instance the famous Teichoskopiain book . Though we would expect that it took place in the first year ofthe war, the poet makes clear that a long time has passed since then. Helen,ashamed at the sight of her husband, whom she has deserted, and filledwith the deepest remorse, declares: ‘Would I had chosen grim death whenI followed your son here, leaving my bedroom and kin and my late-borndaughter and my delightful friends of my own age’ (.–). And pointingout Agamemnon to Priam she says: ‘and he was my brother-in-law, slutthat I am, if indeed he ever was’ ().

The motif probably does not come from a description of a similarsituation at the beginning of the war, but was taken over from the famousTeichoskopia in the epic of the Seven against Thebes. Nevertheless theconnection between this Teichoskopia and the situation at the beginning ofthe war cannot be doubted, no matter how we interpret it. However, in the

I should like to thank Hans-Wilhelm Nordheider and Antonios Rengakos for helpful suggestions. See especially Kullmann (: ff., : –). Cf. Kullmann (: –). Cf. Kullmann (: ).



The Homeric Catalogue of Ships

Catalogue of Ships the strict concentration on the situation at the time ofdeparture is extremely surprising. We mainly hear of the number of shipsat the disposal of the leaders when they were embarked. Yet there are someallusions to the battlefield. These are restricted (a) to the contingents forwhich new leaders are introduced, Podarces instead of Protesilaus (.–), Medon instead of Philoctetes (.–): in both cases we find theword � '� �(�) (he arranged the warriors); (b) to political additions,which are possibly, but not necessarily, interpolations, in .– (thePhocians have a position near the Boeotians) and .– (the SalaminianAjax arranges his warriors near the Athenians); (c) to the most prominentleaders Agamemnon (.–) and Menelaus (.–); (d) to Achilles,because he refused to fight (.–). In all these cases the allusions areadditions to the stereotypical bulk of the Catalogue which, if consideredas an episode of the epic, is composed in a very rigid style. Apparentlysome of the additions have caused the loss of the leader’s patronymic (i.e.of Protesilaus, Philoctetes, Ajax, Achilles), which confirms the fixed formof the original catalogue. This implies that it was taken over as a wholefrom another epic context in which the departure from Aulis was describeddirectly. Its poet may have been the same as the poet of the Iliad.

There is another irregularity in the Catalogue, which I shall now discussin detail: the contingent of the Boeotians, as the first described in theCatalogue, has five leaders, whereas the maximum in the other entries isjust four (in the case of the Epeians). Even more astonishing is the factthat these five leaders are not especially prominent in the myth. WhereasPeneleos and Leitus in were at least suitors of Helen in Apollodorus’catalogue (Apollod. Bibl. ..), Arcesilaus, Prothoenor and Clonius aremere fillers destined only to be killed by the Trojans later in the Iliad.On the other hand, there is a very prominent Boeotian not mentionedin the Catalogue or in the rest of the Iliad but important in the legend.This is Thersander, son of Polynices and legitimate successor who becamethe legal king of Thebes after its conquest by the Epigoni according toPausanias .., ... Although the Cadmeia may have been destroyedin the war, the Catalogue mentions Hypothebai, which is called a well-founded city (7o��'�� 3��). Perhaps Hypothebai was not theoriginal version. In the Cypria it is said that Thersander fell at the handsof Telephus, the son of Heracles (Proclus § according to my numeration

Possibly he was mentioned in the Ps.-Hesiodic Catalogue of Women fr. . West (= F . H) asassumes Cingano (: ).

Wathelet (: ) suggested that the original list contained a verse like �F �@ P� e:>� �)��7o��'�� 3��. Cf. Cingano (: and n. , ).



wolfgang kullmann

(Kullmann b: –)), during the expedition of the Achaeans toMysia or Teuthrania, which they destroyed by mistake assuming it wasTroy.

In Quellen der Ilias in I propounded the hypothesis that the poetof the Iliad knew of the Teuthranian expedition and the second departurefrom Aulis after the sacrifice of Iphigenia eight years later, but concealedhis knowledge (Kullmann : , , ff.). I argued (a) that Helenspeaks of twenty years’ absence from her homeland (Il. .–), whichincludes the expedition to Teuthrania; (b) that Neoptolemus was begottenat Scyros (.), which seems to presuppose Achilles’ stay on the islandwhen he travelled home from Teuthrania; (c) that the sacrifice of Iphigeniais referred to when Agamemnon addresses Calchas as a seer of evil whonever prophesies anything favourable (.ff.); and (d ) that Arcesilaus,Prothoenor and Clonius are substitutes for Thersander in the Catalogue ofShips. My hypothesis was neglected by my colleagues apart from DenisPage and Uvo Holscher who doubted it in their reviews of and

respectively (Page : ff., esp. ; Holscher : –). Fortunately,it is now supported by the new Archilochus papyrus, published by Obbinkin volume of The Oxyrhynchus Papyri (Obbink ). In the fragmentthe battle between the Achaeans and Telephus in Mysia is narrated. Tele-phus seems to be called �� , which alludes to his mother tZ4�,so being the great-granddaughter of ��, the eponym of the Arcadians.We are told how Telephus put the Achaeans to flight and drove themback to the coast, and that the River Caicus was filled with corpses. Thename Teuthras is mentioned as well as a cry and gratitude to a father. Themythological background can be reconstructed from the Cypria (Proclus§ according to my numeration) and, in part, from Apollodorus withrespect to the relationship between Telephus and Teuthras, but above allfrom Hesiod fr. West (= F H). In the Catalogue of Women it is toldhow the gods brought the little Auge to Teuthras, who adopted her andbrought her up like his own daughters. Later she encountered Heracles,who was on the way to Troy, and she bore him Telephus. From ApollodorusBibl. .. we can supply the fact that Auge married Teuthras, who adopted

Kullmann (: ff.). With respect to argument (d ) cf. especially , , , ; and Cingano(: ff., esp. ff.).

In Kullmann (: ff., esp. , n. ) I have argued that this version fits to the role of Heraclesin the Trojan myths better than that of Hecataeus FGrH I F a (= Fowler : ). Accordingto him Auge was engendered by Heracles already in Tegea and banished by her father Aleus andexposed in an ark which landed in Mysia. But this looks like a doublet of the story that Danaetogether with her son Perseus was exposed in an ark by Acrisius (cf. further versions in Apollod. Bibl.... and ..).




Telephus and made him his successor. Perhaps the cry happened at anepiphany of Telephus’ father Heracles.

The context of the new fragment of the elegies of Archilochus is opento speculation. But an attractive suggestion has been made by Parsons(Obbink : ). He believes that Telephus functions as an exampleof Archilochus’ own fate when he lost his shield in a bush (fr. IEG).According to the Cypria he was wounded by Achilles (Proclus § ) becauseDionysus made him stumble over a vine shoot (Cypria fr. Bernabe,confirmed by Apollod. Epit. .) and perhaps lost his shield (see Philostr.Her. .–, ., .–). Thersander’s death at the hands of Telephus isnot mentioned, but this would have been irrelevant when the story servedas an example.

I think this fragment is important for the history of epic in general:. The fragment confirms that the motif of the Teuthranian expedition

was not an invention of the poet of the Cypria, because it antedates itscomposition considerably. Certainly, from a narratological point of viewand in relation to the whole story of the Trojan war, it creates suspense,but it was not a mere poetic invention.

. Its date is very close to the date of the Iliad as suggested by WalterBurkert, Martin West and myself (though, of course, this is a much-disputed question (Kullmann b: –)). But it is likely that thestory of Telephus’ heroism is older than the Iliad.

. The conformity of the version of Archilochus with the version of theCypria is astonishing. In a sense this supports the hypothesis that thereexisted a canonical sequence of events in oral poetry, a ‘Faktenkanon’ (asI have called it in Quellen der Ilias), to which the Teuthranian expeditionbelonged (Kullmann : –, –).

. The confirmation of my assumption that this is an old story leads tothe question of its historical background. It obviously mirrors colonialexperiences. The story suggests that in myth the colonization was notone-sided. The figure of Telephus, son of Heracles and of Arcadianorigin and heir to the Mysian King, seems to show that ‘fraternization’also existed and that in this part of the myth aggressive and peacefultendencies were combined.

. The question arises why the whole sequence of the second departurefrom Aulis was concealed by the poet of the Iliad. When I wrote Quellender Ilias I thought this may have been because Homer did not acceptthe mysterious sacrifice and removal of Iphigenia from the altar and herreplacement by a hind. But now I think that the main reason is that thisepisode would have impaired Homer’s conception of the Trojan War,



wolfgang kullmann

which knows about, but deliberately excludes, the Aeolian colonizationof the coast of Asia Minor.

. The story of Telephus, son of Heracles, who was wounded by Achillesbut later healed by him after he had promised to show the Achaeans theright way to Troy, is a link between the myth of the first destruction ofTroy by Heracles and that of the second by the warriors of Agamemnon.The second occurred with the aid of earlier Greek settlers in Asia Minor.

. One may ask whether it is possible that the story of Heracles’ journeyto Troy via Mysia which ended with the first capture of Troy, as partlytold in the Iliad, is the original Trojan myth, whereas the story of itsconquest in order to regain Helen is secondary.

As for the Catalogue of Ships, the new papyrus supports the claim thatThersander originally belonged to it and was the main representative ofBoeotia. Apparently the Iliad has abbreviated the Trojan legend by omittingthe Teuthranian expedition, the death of Thersander and the sacrificeof Iphigenia. Moreover, the old puzzle of why there are five leaders ofthe Boeotians is solved. The original catalogue apparently only knew ofthree leaders, a number which also occurs in other entries. As a result weare able to conclude (with much more confidence than before) that theCatalogue of Ships antedates the Iliad and belongs to a more completeview of the Trojan war myth.

I proceed to the list of heroes. They come from every region of mainlandGreece as well as the southern islands of the Aegean. They are the comman-ders of the ships and warriors from the regions and the cities mentioned ineach entry. But they are not the kings of these regions, similar to the kingsof the Mycenaean territories, i.e. of Thebes, Pylos or Mycenae, accordingto the picture presented by archaeology. Some mythological characters arenot mentioned in the Catalogue, like Nauplius, the eponym of Nauplia,and his son Palamedes. Although they play an important role in the Tro-jan legend, they are suppressed in the Iliad, like Thersander, but unlikehim they cannot be detected in the Catalogue. Perhaps a whole entry wasomitted. Moreover, Messenia is not mentioned. Apparently the Catalogueassumes the situation after the first Messenian War, as does the rest ofthe Iliad in which Agamemnon offers seven Messenian cities to Achillesin .ff. (Kullmann b: , –). Certainly this makes sense onlyif we also suppose that originally Agamemnon came from Sparta wherehe shared a dual kingship with his brother Menelaus. But the Messenian

Cf. Kullmann (: –, –, –, b: , ), Schlange-Schoningen (: ff.). Cf. Kullmann (: –, b: ff.).




fighters from Pherae, mentioned in Iliad .ff. as being under the com-mand of Ortilochus, seem originally to have constituted an autonomouscontingent. Presumably this passage is a mythological relic.

Sometimes the Catalogue provides additional brief information aboutthe heroes. In part this is genealogical in nature: so Ascalaphus and Ialmenuscoming from Orchomenus are called sons of Ares and Astyoche, whosecretly had an affair with the god (.–); Polypoetes is presented as theson of Pirithous, the Lapith, and Hippodamia in .–, and there is anallusion to the battle of the Lapiths and Centaurs on their wedding day(.–).

Some heroes are given a personal characterization. The Locrian Ajax,son of Oileus, is compared, to his disadvantage, with the great Ajax (.–). Menestheus, we are told, is good at deploying chariots and troops(.–). Meges is said to be the son of Phyleus, who was angry with hisfather (Augeas) and emigrated to Doulichion (.–). With respect toThoas we hear that the sons of Oeneus, including Meleager (and implicitlyTydeus), were no longer alive. Tlepolemus’ reason for emigrating to Rhodesis explained by his need to flee because of the murder of Licymnius (.–). The people of the Euboean Elephenor, i.e. the Abantes, wear their hairlong at the back and are good warriors (.–). It is Agamemnon whogives the Arcadians the ships they need (.–).

In general the narrative exhibits features which also occur in othergenealogical literature, as, for example, in the Ps.-Hesiodic Catalogue ofWomen. There we find a similar mixture of genealogy and hints at themythological background. This catalogue poetry appears to have had along oral tradition.

Some of the leaders mentioned in the Catalogue do not occur againin the rest of the Iliad. Apart from Philoctetes, who was left at Lemnos,they are the Phocian Epistrophus (.), brother of Schedius, Agapenor(.), representative of the Arcadians, Nireus of Syme (.), Phidippusand Antiphus (.), sons of the Heraclid Thessalus, leading a contingentfrom the Dodecanese, Guneus from the unknown Cyphos (.), whomthe Enienes and the Peraebi followed, and Prothous (.), leader ofthe Magnetes. And, as Kirk (: ) has stressed, the leadership of theEpeians as described in the Catalogue is not reflected in the rest of the Iliad.Thalpius and Polyxenus (. and ) are completely absent, whereasOtus from Cyllene is called a leader of the Epeians in . alongside Megeswho in the Catalogue represents the people coming from Doulichion. In

Cf. Kullmann (: –).



wolfgang kullmann

sum: nine of forty-six leaders mentioned in the Catalogue (Philoctetes notincluded) do not occur in the rest of the epic; ten will be killed later in thepoem, but some of them probably in conflict with other traditions. Othersare mentioned in an unimportant context.

A certain tension between the Catalogue and the rest of the Iliad cannotbe denied. It seems that many of the leaders owe their inclusion in theCatalogue to the fact that their fathers or grandfathers were prominentcharacters in older legend. This is of course the case with Diomedes,Sthenelus and Euryalus, who belong to the Epigonoi in Theban epics,although they are prominent in the Iliad too (the same holds true forThersander with respect to the original Catalogue). And it is certainly trueof Polypoetes, son of the Lapith Pirithous, and Leonteus, grandson of theLapith Caeneus, who recall the famous fight of the Lapiths and Centaursalluded to in Iliad .ff.

Many leaders in the Catalogue of Ships descend from the Argonauts.This is probably the case with Schedius and Epistrophus whose fatherIphitus is mentioned in Apollodorus’ catalogue of the Argonauts (Bibl...; cf. Ap. Rhod. Argon. .ff.). Eumelus is also connected to thelegend of the Argonauts, firstly through his mother Alcestis, the mostbeautiful of the daughters of Pelias, as mentioned in the Catalogue (.–) – for according to the legend it was Pelias who gave the decisive orderto Jason to bring back the Golden Fleece from Colchis before takingover his realm; and secondly through his father Admetus, who occurs inApollodorus’ list too. The father of Philoctetes was apparently anotherfamous Argonaut, Poeas, mentioned only in Od. ., who killed thegiant Talos, a bronze monster at Crete which possessed only one full bloodvessel, with an arrow in the ankle (Apollod. Bibl. ..). Carl Robertcomments: ‘Because Poias played a decisive role in the oldest form ofthe Talos adventure he must have belonged to the Argonauts already veryearly.’ The reliability of the list of Argonauts in Apollodorus and its relativeage are confirmed by the entry for Philoctetes in the Catalogue of Ships.According to it a city called Thaumakie (Thaumakoi in historical times)belongs to Philoctetes’ region. And Apollodorus mentions that Poeas wasthe son of Thaumacus, apparently the eponymous founder of Thaumakie.This connection is not mentioned in the Iliad at all and is evidentlyunknown to it, so that Apollodorus is not dependent upon the Iliad in thisrespect. Another famous Argonaut was Ancaeus, father of Agapenor, leader

Cf. Robert (: ): ‘Da aber . . . Poias in der altesten Form des Talosabenteuers eine entschei-dende Rolle spielt, muß er schon fruher zu den Argonauten gehort haben.’




of the Arcadians, possibly identical with a Pleuronian Ancaeus, mentionedin Iliad . as the opponent of Nestor in a boxing match. The Iliadmainly presupposes an epic version of the Argonaut saga, as is clear fromIliad .ff., where it is said that the Jasonid Euneos, son of Hypsipyle,named, according to his father, ‘the man with the good ship’, i.e. theArgo, sent wine to the Achaeans. It seems that it was particularly (butnot exclusively) the fathers of the Thessalian leaders in the Catalogue ofShips who participated in the expedition of the Argonauts. Unfortunately,the sources of the catalogue of Argonauts in Apollod. Bibl. .. areunknown, although they may ultimately go back to oral traditions. Elevenor twelve leaders mentioned in the Homeric Catalogue of Ships havefathers or (in one case) another relative among the forty-four participantsin the expedition of the Argonauts enumerated by Apollodorus (Schedius,Epistrophus, Ajax, Diomedes, Agapenor, Odysseus, Tlepolemus, Achilles,Eumelus, Philoctetes, Leonteus and perhaps Protesilaus (?)), and fiveleaders who do not play an important role in the Iliad occur in his listof Argonauts themselves (the Boeotians Peneleos and Leitus, the MinyansAscalaphus and Ialmenus, and Euryalus)! Perhaps the poet of the Catalogueof Ships had no names of their sons at his disposal and so adopted theArgonautic heroes into his own catalogue.

Another legend, ‘The Calydonian Boar Hunt’, is indirectly referred toin the Catalogue through the reference to Oeneus’ son Meleager (andimplicitly Tydeus) when the Aetolian contingent led by Thoas is described(.). Nestor represents sagas of Pylos and Elis, whereas Agamemnonand Menelaus are Pelopids who are in a way colonial intruders into themythology of the Greek mainland in so far as they originally come fromthe Aeolian colonies on Lesbos and the coasts of northern Asia Minor.

The leaders of the Dodecanese are Heraclids of the first generation likeTlepolemus (as is Telephus in Mysia), or of the second (Phidippus andAntiphus, sons of Thessalus). It should be added that Heracles is a mytho-logical character who is ideal as the father of founders of cities in foundation

Some suggestions are made by Drager (: , : –), Scherer (: –). According to Ps.-Hesiod fr. . West (= F . H) Protesilaus was a son of Actor (who was

a son of the Aeolid Deıon). This seems to be the original version. In Iliad .ff. it is said thatProtesilaus is a brother of Podarces, son of the Phylakid Iphiclus (i.e. grandson of Deıon), andthe successor in the leadership of his contingent; but this is perhaps an innovation of the poet ofthe Iliad. Cf. Kullmann (: ). Another Actor, son of Hippasus, and another Iphiclus, son ofThestius, were Argonauts according to Apollodorus and older sources, as is stressed by Grossardt(: –). On the other hand, the extraordinary faculty of running over the corn fields withouttouching them ascribed to the Phylakid Iphiklos (Ps.-Hesiod fr. West; Iliad .) is a featurewhich is characteristic of Argonautic adventures. Cf. West (: ).

Cf. West (: –), Kullmann (b: , –).



wolfgang kullmann

legends because in myth his strength is combined with considerable sexualactivity. Just look at the list of sons attributed to him in Apollod. ..

To sum up, the list of the Achaean leaders is a panhellenic groupingof descendants of famous heroes known to the audience from older epicpoems and local legends. By far the most important epic tradition whichantedates the Trojan cycle of legends is the saga of the ‘Expedition of theArgonauts’. This discovery has considerable consequences. Probably a lotof the heroes of the Catalogue of Ships and the Iliad were either borrowedfrom earlier oral epic legends, especially from an oral Argonaut epic, or ayounger generation of sons, whose fathers were famous in these legends,was invented. It cannot of course be ruled out that through poetic inventionthe fathers of the Trojan War heroes sometimes later became Argonauts.But in the development of Greek mythology the existence of a group offorty-six Achaean leaders as described in the Catalogue of Ships is a verylate phenomenon.

There is significant disagreement, however, between the catalogue ofthe leaders and that of the cities. The Thessalian heroes are not as firmlyrooted in their regions and cities as most of the heroes of other parts of theCatalogue seem to be. This fact is generally accepted even though thereare highly contradictory interpretations. If we assume that the Cataloguemirrors the specific view of a poet of the beginning of the seventh centurythen we have to admit, as I do, the following result. The regions of Achillesand Protesilaus overlap (Halos lies on the Pagasitic Gulf in the region ofProtesilaus, not on the Maliacic Gulf ). In the same way the region ofEurypylus is confused: the Hypereian spring is found in the city of Pheraebelonging to the region of Eumelus, and Ormenion is situated at the footof Mt Pelion on the Pagasitic Gulf belonging to the region of Prothous(neither belonging to the region of Eurypylus). The region of the LapithsPolypoetes and Leonteus extends to the place given to Guneus (the city ofGonnoi is situated in their region, i.e. the main city of the Perrhaebians,a people allocated to the command of Guneus). On the other hand theregion of Guneus surprisingly reaches to the River Titaressos (most likelythe Europos, a tributary of the River Peneios in the middle of Thessaly)and Dodone in the far west. The places attributed to Philoctetes are totallyconfused. Meliboea, Methone and Olizon belong to the region given toProthous (‘about the river Peneios and the mount Pelion’), whereas Thau-makie (the native town of his father) is to be identified with Thaumakoi,which is situated in the south of Thessaly on the border with Phthiotis(Strabo ..), although it does not directly impair the territory of the




other heroes. Perhaps the other cities are only attributed to Philoctetes inorder to give him a region.

I think we cannot avoid following Niese (: ff.) and Giovannini(: ff.) who claim that the poet of the Catalogue has distributedidentifiable cities from a list perhaps of the beginning of the seventh centuryamong well-known characters of myth in order to give each of them akingdom without always knowing precisely the geographical situation. Atthe same time he may have been influenced by mythological associations.It runs against the character of early Greek myth to ascribe to certain heroeslike the Phthian Achilles the rule over a number of definable autonomouscities. This sort of combination is an artificial construct.

We proceed to the list of cities. Just as with the heroes, so the citiessometimes receive an additional remark. The people of Dorion belongto the contingent of Nestor and mention of this place induces the poetto tell the story of the meeting with the Muses of the Thracian singerThamyris of Oichalia, and how Thamyris boasted of being better thanthem (.–). It does not matter whether the poet of the Cataloguehas confused Dorion with Dotion in Thessaly where the Ps.-HesiodicCatalogue of Women places the story (fr. and . West) or viceversa.

Unlike the list of heroes the list of cities seems to be almost contemporarywith the poet’s lifetime. It is remarkable that the distribution of the cites ofThessaly among the heroes is confused whereas, according to Giovannini(: ), every Thessalian city named in the Catalogue can be securelyidentified with the exception of only three: Oichalia, Asterion and Elone.Moreover, Oichalia can probably be deleted from this list. According toVisser’s interpretation of Strabo .. and .. (: ff.) it seems clearthat the historical Oichalia is situated in the Hestiaeotis near Trikka as theCatalogue says (.). The same confusion occurs with respect to Douli-chion and the Echinads. According to Strabo .. an island of Dolichaexisted. But it is impossible that Dolicha and the Echinads ever existedas a region capable of sending forty ships under the leaderhip of Meges,grandson of Augeas and a fugitive from Elis. Apparently the geograph-ical names are correct, but the region has been invented by the poet ofthe Catalogue, probably on mythological grounds. The alleged emigrationof Phyleus contradicts the rest of the Iliad in which, as we mentioned,Meges leads the Epeians. The story of Phyleus’ anger at his father seems

See above p. . Cf. Hirschberger (: F ). See Giovannini (: ).



wolfgang kullmann

to have been of peculiar interest to the poet of the Catalogue. Of coursethe geographical description of the regions contributing to Agamemnonand to Diomedes can only be explained as a contamination of differentmythological traditions.

These are not the only geographical contradictions between the Cata-logue and the rest of the Iliad. We find another inconsistency of this kindin Iliad .. There the Boeotian locality of % �� is mentioned. Strabo(..) identified it with the village � �� which he found near the ForestLake (��V ��'��). (The coexistence of plural and singular forms is com-mon, e.g. in-#:�� and-#�� ore:>� ande�>�.) This contradictsanother occurrence of % �� at Iliad .ff. There it is said that Oresbius,who was killed by Hector, lived at Hyle on the so-called Cephisian Lake,normally identified with the Copais basin. Strabo comments that Homer‘means the Hulike limne ’. But this is of course special pleading. The expla-nation of Wallace (: ) is no better. He suggests that the village wassituated ‘between’ Lake Copais and the Forest Lake. Apparently this is aninconsistency between the Catalogue and the rest of the Iliad.

So far our inquiry suggests that we should analyse the list of citiesindependently from the list of heroes. The sheer number of about citiesraises the question of its probable historical background. As I have pointedout elsewhere (Kullmann b: ff., : ff.), I am convinced thatthe list has something to do with administrative work conducted in thefirst half of the seventh century bc in connection with the organization ofthe Olympic Games.

Most probably we find an allusion to the Olympic Games in Iliad.ff. There Nestor says that the Pylians had sent a four-horse chariotto ‘divine Elis’ in order to win a tripod in an interregional agon. But Augeasconfiscated the horses and the chariot and sent back the driver. Of courserevenge followed later.

It is noteworthy that a four-horse chariot is mentioned. We know ofmany votive offerings of the eighth century bc from Olympia which rep-resent chariots with horses. But these are only two-horse chariots. Thesame holds true for the funeral games in the myth. Only two-horse char-iots occur. We conclude that the passage must have been influenced bythe introduction of four-horse races in Olympia. According to the list ofOlympionikai edited by Moretti (: ff.) this introduction took placein bc. It does not matter whether this date is absolutely correct or not,

Cf. Laser (: ), Dickie (: –), Crielaard (: ff.), Kullmann (b: , :ff.).




but it confirms the allusion. Excavations at Olympia now show that thegames became an interregional institution after bc, as Alfred Mallwitzand Ulrich Sinn have proved in some detailed publications (Sinn ;Mallwitz ). From that date many fountains are found in Olympiawhich were built by groups of participants in the Olympic Games to pro-vide themselves with water. Apparently each city had its own fountain.This implies that the games required a strong organization. The cities hadto be invited to the games and this presupposes the existence of lists. Weknow of such lists from later periods. A long list from Delphi (producedabout bc) contains the names of consuls in different cities who receivedthe theoroi (called theorodokoi) when they brought the invitations (Plassart). Giovannini has pointed to the strong resemblance of this list to theorder of the cities in the Catalogue of Ships and has claimed that sucha list underlies it. The routes of the Delphian list are very similar to thegeographical enumeration of the cities in the Iliad (Giovannini : –, (comparative map)).

If we look at the geographical principle used in the Catalogue, we findthat it is hodological as is usual among singers, and not only in Greece(Gehrke : ff.). But Danek (: –) has convincingly shownfrom his own comparative perspective that the Homeric Catalogue differsfrom all other comparable catalogues of oral poetry in that it provides, as heargues, an almost two-dimensional picture of the geography of mainlandGreece which can only be explained by its panhellenic tendency. Whetherthis is a deviation from the hodological principle or not depends on thedefinition and is of minor importance.

Without doubt, the strategy of the poet of the Catalogue had a strongpolitical dimension. He belonged to a Greek world which was character-ized by intensive communication between cities, probably along the routesaccording to which the contingents are enumerated. The organization ofan event like the Olympic Games would have been much more difficult in aworld of or more autonomous cities than in an extended kingdom likethose of the Asiatic monarchies of the ancient world which was ordered ina strictly hierarchical manner. This historically attested intensive commu-nication is mirrored not only in the Catalogue but also in the stories of therecruiting of the Achaean contingents by Agamemnon and Menelaus (andperhaps Palamedes and Nestor) which we find in the Homeric epics andthe Cypria. Of course, this picture is deceptive. It is one thing to persuadesomeone to participate in the Olympic Games by means of a personal visit,and another to recruit an army of , warriors against Troy in thisway. This is possible only in poetry.



wolfgang kullmann

Some smaller problems remain. Did the poet of the Catalogue followa contemporary geographical list exclusively or are some names of citiesinserted which were known to him only from the mythical tradition? Theproblem of the site of Oichalia in the Catalogue seems to have been solved.However, the situation is unclear with respect to Eutresis. Allegedly thecity of Eutresis was not settled between and bc (Goldman ).This seems incompatible with the myth that Amphion and Zethus, sonsof Antiope, the builders of the walls of Thebes and first founders of thatcity, were born at Eutresis (Strabo ..). They are firmly incorporatedinto the genealogical systems and cannot owe their existence in mythto a Mycenaean tradition. Eutresis should have existed as a geographicalnotion as early as the first half of the seventh century. But there is noevidence.

Let me summarize. The rigid formal structure of the Catalogue of Shipssuggests an origin independent of the Iliad, to which it was adapted byits poet. Whether it was nevertheless composed by him for an earlierperformance of another epic cannot be excluded, but seems improbable.Additional arguments for independence are inconsistencies between theCatalogue and the rest of the Iliad. Special attention is to be paid to theapparent substitution of the main leader of the Boeotians Thersander withthree fillers. This hypothesis presupposes that the Teuthranian expeditionwas a pre-Iliadic part of the Trojan legend, which has been confirmed bythe new Archilochus papyrus published in . Moreover, the enormousmass of mythological and geographical detail of the Catalogue cannot besufficiently explained when it is attributed to the poet of the Iliad himself.A great deal of the list of heroes consists of the names of the descendants ofthe heroes of older legends and oral epics. This list seems to be older thanthe detailed list of cities which apparently reflects the contemporary worldof the poet, although it is older than the Iliad as well. The inconsistenciesbetween both lists and between the list of cities and the rest of the Iliadcorroborate this supposition.

For prosopographical reasons it is obvious that the epic tradition con-cerning the Argonauts antedates the tradition concerning the Trojan War.

Our inquiry results in the following picture of the relative chronologyof the Homeric Catalogue of Ships and its constituent parts:

For a more detailed account of the prosopography of the Homeric Catalogue of Ships, cf. Kullmann(: ff., esp. ff.): () The list of the fathers of the Achaean leaders mentioned in the Catalogueof Ships; () The list of Argonauts mentioned by Apollod. Bibl. .. which recur among theAchaean leaders of the Catalogue of Ships and their fathers and grandfathers.




. Iliad including Catalogue of Ships.. Original Catalogue of Ships without allusions to the situation on the

battlefield, but including the Boeotian leader Thersander. The authortook account of the Teuthranian expedition, mentioned in the Cypriaaccording to Proclus and in a poem of Archilochus.

. The author of the original Catalogue of Ships used a near-contemporarygeographical list of cities composed for bureaucratic purposes, which isnot in full agreement with the rest of the Iliad.

. The author of the original Catalogue of Ships used oral epics which wereolder than the Iliad, concerning the ‘Expedition of the Argonauts’, the‘Seven against Thebes’ and other legends, in order to find mythologicalcharacters who could help the Atreidae in the Trojan War. He tookover either some of the participants in these events themselves or theirdescendants whom he found among the myths or invented himself.



chapter 13


Martin West

The occasion of the Oslo conference prompted me to do something Ihad long had in mind: to make a synthesis of the views I have formed in thecourse of over forty years on the datings of various early Greek hexameterpoems. The dating of the Iliad alone, or the Odyssey, or the Cyclic epics,or the Hymns, or the poems of Hesiod, would of course have affordedmaterial enough for a conference paper. But there is much to be said,especially perhaps in the context of this collective enterprise, for attemptingan overview and for proposing a comprehensive chronological framework.Naturally I cannot, within the set limits, review the history of scholarshipon the question or engage at length with the work of others who havecome to divergent conclusions. I cannot do much more than explain whatI regard as the best criteria in these matters, and what conclusions I derivefrom them. If I make more reference to my own previous publicationsthan to those of anyone else, it is not because I wish to pose as thesupreme fountain-head of truth but because I am ex proposito drawingtogether and building on (or in some cases modifying) my earlier work andthoughts.

In attempting to establish a chronology for these poems we have to usea combination of relative and absolute criteria. We start with an overalltime-frame. A hundred years ago scholars of repute were happy to put theIliad and Odyssey in the tenth or ninth century. Nowadays I think almosteveryone would want to put them somewhere between and (orat the very latest ). We see the matter in sharper focus. We no longerthink of Greek history as beginning with the first Olympiad in , withonly a misty mythical era before it. Archaeology has given us a picture ofthe Mycenaean world and a general framework for the centuries betweenthe collapse of the palaces and the flowering of the archaic polis. We havelearned how to date some elements of the material culture described inthe Homeric poems. We have built up a model of the stages by which




the Greeks gained knowledge of foreign peoples and places, and we canmeasure the Homeric picture against it. We no longer imagine that thepoems as we have them could have been preserved unchanged over timewithout being fixed in writing, and we accordingly rule out a date ofcomposition earlier than we believe a written text was likely to exist. Weno longer take it as an axiom that epic poems must be earlier than lyric oriambic or elegiac poems.

Within our time-frame there are few anchor-points for absolute dating.Hence relative chronology has a major part to play. If we can establish asequence for at least some of these poems, that should help us to find the bestplace for them within our framework of absolute chronology. Sometimesit is a question not of relative dating among these poems themselves butof dating relative to a lyric poet such as Alcaeus. Relative and absolutechronology thus become entangled.

By what criteria may we judge one poem to have been composed beforeor after another? I will suggest three: () verbal echoes, adaptations, orimitations; () difference in the degree of linguistic development; and ()thematic dependence.

1 verbal echoes, adaptations, imitations

In earlier times this criterion was employed with greater confidence thannow seems appropriate. It was assumed that Homer was the oldest poet,and verbal parallels between the Homeric poems and Hesiod or other poetswere automatically taken as echoes of Homer. The presence of all theseapparent echoes of Homer then seemed to confirm his priority.

It is different now. We can often say that an elegiac or melic poet isusing epic diction, but we can no longer treat ‘epic’ as being necessarilysynonymous with the Iliad and Odyssey. We have become much more awarethan some of our predecessors that there existed a broad current of oralepic, and that phraseology known to us from the two epics we have wasnot necessarily unique to them.

Still, there are a few cases where a passage of individual character in oneearly hexameter poem does plausibly look like the model for a passage inanother poet. Several such instances point to the currency of the Worksand Days from the mid-seventh century. C. A. Trypanis made a persuasivecase for the influence of the Hesiodic poems on the Hymn to Demeter.

Semon. < Op. –; Alc. < Op. –; Sapph. < Op. – (and Theog. –). Trypanis (); cf. Richardson (: –).



martin west

I have elsewhere listed half a dozen instances of possible Hesiodic influenceon the Iliad. I have reservations now about one of them, the catalogueof Nereids in .– (< Hes. Theog. –), which Zenodotus andAristarchus athetized and which may indeed be an interpolation. Theothers I stand by, and I think they can be added to. There are no placeswhere I see evidence of the converse relationship, the Iliad influencingHesiod.

There are many passages of the Odyssey that are evidently modelled onpassages in the Iliad. The material, with the analysis it requires, is sufficientto fill a book, and just such a book has been written by Knut Usener. Iam not aware of any instance of the converse relationship, though I wouldnot wish to deny that the poet of the Iliad knew an Odyssey, not necessarilyidentical with the Odyssey that we have.

There are two passages of seventh-century elegy that have often beentaken to be inspired by the Iliad. One is Tyrtaeus .–, about what adisgraceful sight it is to see an old man killed in the front line insteadof a young one; this certainly stands in a close relationship with Priam’sappeal to Hector in Il. .–, and equally certainly is not derived fromit. The other is Mimnermus’ comparison of human lives to leaves, withits parallel in Glaucus’ speech to Diomedes in Il. .–. Neither of theIliadic passages fits its context as well as does the elegiac parallel, and theycannot be used as evidence of priority.

Among the Hymns, the one to Aphrodite has always been consideredone of the earliest, and certain intertextual relationships are consistent with

West (: –); on the Nereid catalogue, Nickau (: –), West (a: ). The otherpassages I cited were: Il. . < Theog. ; Il. .– < Theog. –; Il. . < Op. ; Il..– < Theog. –; Il. . < Op. –. The following may be added: Il. . < Theog.f.; Il. .– < Op. f., ; Il. .– < Theog. , ; Il. . and ��'�� < Theog.; Il. . < Theog. ; Il. . < Theog. ; Il. .– < Theog. , –, –; Il. .f.< Theog. f.; Il. .– < Theog. –; Il. ., . < Theog. , ; Il. .f., f.,.f. < Theog. ; Il. .– < Op. –; Il. .f. < Op. f., ff.; Il. .– < Theog., f., = , ; Il. .f. < Theog. –. Hesiod’s Titanomachy and Typhonomachyseem to have made a particular impression.

Usener (). He concludes () that the Iliad must already have existed in fixed form when ourOdyssey was conceived and that the poet of the Odyssey must often have heard it recited in this fixedform.

This is suggested particularly by the two references to Odysseus as the father of Telemachus, Il. .

and .. See West (: ); on the Tyrtaeus passage also Von der Muhll (: ), who cites earlier

discussions. There are many other Iliadic passages that recall the manner of martial elegy and nodoubt reflect current poetry of this type: . (cf. Callin. .f.), .– ≈ .– (cf. Tyrt..–), .– (cf. Callin. .–), . (cf. Tyrt. .f.), .–, .–, –, .– (cf.Tyrt. .–).




this. The description of the goddess’s toilet at – looks like the modelfor the closely similar one in Od. .–, where it no longer has its propercontext of preparation for a seduction (cf. Il. .–). The hymn appearsalso to have exercised a certain influence on the one to Demeter (Heitsch: –; Richardson : –; and now Faulkner (: –)).The Delian part of the Apollo hymn in my opinion owes a certain amountto the Pythian, though I know that many have taken the opposite view(West : –).

2 difference in degree of linguistic development

The epic language was a traditional and conservative idiom whichresponded only gradually to changes in the spoken dialects of the poets.But neologisms, newer words and newer grammatical forms, made theirway into it over time, and their presence, or the extent to which theyestablished themselves as regular features, provides possible indicators ofrelative chronology. If we had the text of an epic poem composed in thetwelfth or eleventh century bc and could set it beside the Iliad, there is nodoubt that we would easily be able to establish on purely linguistic groundsthat the Iliad was the younger of the two. There would be morphologicaland phonological differences much greater than any to be found withinthe extant corpus.

As things are, we have a quantity of texts and fragments composedwithin a period of perhaps or years in a rather ossified language,and linguistic differences with a potential chronological significance donot leap to the eye. They exist, but they disclose themselves only throughcareful analysis and measurement. Richard Janko in his remarkable firstbook, Homer, Hesiod and the Hymns (), gave us the most sophisticatedstudy of such phenomena. The trouble is, as he recognized, that a numberof imponderable factors are involved. It is not just a matter of tracing a slowbut steady trickle of changes. Individual poets will have differed in theirpropensity to archaism and resistance to newer forms. Dialect differencesin their vernaculars will have affected both the choices available to themand their perceptions of what was the poetic (non-vernacular) alternative.Differences of genre and subject matter will also have affected the outcome.Janko hoped to overcome these difficulties by means of a broad-basedstatistical inquiry using a variety of linguistic criteria. He certainly madegreat efforts to take all relevant considerations into account and sparedhimself no pains. He made himself a computer concordance of Homer(possibly the first ever) and deployed sophisticated statistical methods. He



martin west

arrived at a sequence of texts ordered according to the frequency of newerforms, and made suggestions for converting his relative chronology intopossible absolute chronologies by making certain assumptions. It seems,however, that few scholars have felt able to embrace his conclusions asdependable. When we are having to compare texts composed in differentareas by poets schooled in different local traditions – one of them perhaps aman aged thirty, another a septuagenarian – it is far from clear that Janko’smethod will tell us reliably in what order they were produced. There area few texts showing linguistic features (especially lexical and stylistic) thatseem obviously late, and we can agree that these are poems of relatively latecomposition: the Cypria, the Hymn to Hermes, the pair of hymns to Heliosand Selene, the Batrachomyomachia. But for the rest it seems to me that thelinguistic criterion is less useful and trustworthy than other indicators, andthat where it offers a result that conflicts with arguments based on otherkinds of evidence, it should be treated with great caution.

3 thematic dependence

The most helpful criterion in my view is thematic dependence. Like theother criteria, it has to be used with circumspection. A recently publishedpapyrus (P.Oxy. ) has given us part of an elegy of Archilochus in whichhe recalled an episode from heroic legend: the repulse of the Achaeans byTelephus when, having landed in Mysia under the misapprehension that itwas the Troad, they attacked his city supposing it to be Troy. Archilochusmust have had the story from hexameter epic current in his time; hisadaptation of it into elegiacs is a reflection of his enthusiasm for thispoetry. We know that the story was told in the Cypria. But we cannot inferthat the Cypria itself existed in Archilochus’ time, because it is clear thatmuch of the material in it did not first appear there but came from oldertradition. We can only say that this particular episode was being related inepic in Archilochus’ vicinity by the mid part of the seventh century.

It is a different matter when Alcaeus (fr. ) writes of Achilles calling onhis mother, the sea-nymph, who then went and supplicated at the kneesof Zeus to persuade him to do something about her dear son’s '��.Alcaeus is clearly echoing, not just some unspecified epic that containedan episode similar to the one we know from Iliad , but one in which Zeusmust have responded as Thetis wished, that is, by allowing the Trojans togain the upper hand, to be driven back only when Achilles was inducedto forget his wrath against Agamemnon: an epic, then, with the samefundamental architecture as our Iliad. It would be an excess of caution todoubt that Alcaeus is alluding to the Iliad itself – possibly an Iliad with




some recensional differences from ours, but still the Iliad. In another place(fr. ) he wrote of how Xanthus (Scamander) had its flow obstructedby corpses, a hyperbolic motif special to Achilles’ battle with the river inIl. .–; here again we can say that he is following the Iliad. The‘Ammonius’ commentary draws attention to the imitation.

What we have to look for in establishing dependence is thematic materialthat we have reason to think was new at the time or specific to a particularpoem. I argued in a recent article (West c) that the Aethiopis containedthree major motifs that were innovations with respect to the Iliad: the arrivalat Troy of an army of Amazons under Penthesilea, the further arrival ofan army of Aithiopes under Memnon, and the posthumous translation ofAchilles to the White Island in the Black Sea. Memnon and the Aithiopeswere introduced – I would claim invented – as the immediate response tothe problem created when the Iliad poet postponed Achilles’ death fromshortly after Hector’s death to an unspecified later time: a new antagonistwas needed to hold back the Achaeans’ advance as Hector had done, untilAchilles could overcome him and pursue his followers to the Scaean Gates.The Amazon episode, which preceded the arrival of Memnon, had nosuch organic role, and I suggested that it was introduced at a later stage. Idistinguish accordingly between a Memnonis (without the Amazons) andthe later Aethiopis (which included them). Now, the poet of the Odysseyknows the Memnon episode. He mentions Memnon as the handsomestman that Odysseus saw at Troy (.), and he mentions that the bonesof Antilochus, whom Memnon killed (.–), are with those of Achillesand Patroclus (.); the Memnon poet had used Antilochus as Patroclushad been used in the Iliad, as a special friend of Achilles whose fall inbattle provoked Achilles to go after the enemy hero. On the other hand theAmazon episode and Achilles’ translation to the White Island find no echoin the Odyssey or are contradicted in it. This suggests the chronologicalsequence Iliad–Memnonis–Odyssey, Aethiopis. For indications of absolutedating we can turn to the evidence of art. Vase painters begin to depictrecognizably Iliadic scenes from c. . Memnon and Penthesilea firstappear around the end of the century. There is another early literaryreference to Memnon in a fragment of Alcman (PMGF ) – again aroundthe end of the seventh century, or early in the sixth.

The last third of the seventh century was, I believe, a crucial period forthe crystallization of the epic tradition. Oral epic no doubt continued toexist, but some poems were being written up and so taking on a more stableidentity and a more or less fixed form. We know that this happened sooner

Cf. Meyerhoff (: –), Fowler (: ); and now Faulkner (: –). Cf. Eisenberger (: ). On the date of Alcman cf. West (: –, a).



martin west

or later, and the hypothesis that it was at this period allows us to constructa model that works out very nicely. By the end of the century it appearsfrom various indicators that certain of these epics, identifiable with onescurrent in the classical period, were widely familiar. The Odyssey looks asif it comes quite late in the series. Its poet knows not only the Iliad andMemnonis but other poems about the Trojan War resembling those latercurrent in the Epic Cycle.

Firstly, he shows an extensive acquaintance with the subject matter of theLittle Iliad, and must have known, if not that very text, something prettysimilar. He refers to Ajax’ defeat over the armour of Achilles (.ff.);Deiphobus as Helen’s last husband (cf. ., .); the fetching of Neop-tolemus from Scyros and his defeat of Telephus’ son Eurypylus (.ff.);Odysseus’ entry into Troy disguised as a beggar (.ff.); Epeius’ buildingof the horse (.ff., .ff.). In other words he is familiar with a wholeseries of episodes that came in the Little Iliad.

Secondly, there is a striking link between the Odyssey and the Iliou Persis.According to Proclus’ summary the Iliou Persis began with the Trojanswondering what to do with the Wooden Horse after the apparent departureof the Achaeans. Some wanted to push it over a cliff, some wanted to setfire to it, and some considered that it was a sacred object and ought to bededicated to Athena. This third view prevailed: the horse was taken intothe city, the warriors hidden inside it emerged in the night and opened thegates, and Troy was sacked. Menelaus found Helen at Deiphobus’ house,slew him, and took her back to the ships, thus accomplishing the purposefor which the war had been fought in the first place. A good ending forthis epic; but the debate over the horse has been thought an unlikely pointat which to begin it. Yet the compass of the poem exactly matches the songthat is put in Demodocus’ mouth at Od. .–. Demodocus divergesin one detail from the Iliou Persis as represented in Proclus: of the threeoptions considered by the Trojans for dealing with the Wooden Horse, twoare the same, but the third is different, cutting it open rather than burningit. Otherwise everything tallies; if Odysseus accompanies Menelaus to thehouse of Deiphobus where Helen is to be found, it is because the Odysseypoet wants Odysseus’ deeds to be given emphasis in this song which issung in his presence among the Phaeacians. No one nowadays, I suppose,would argue that this short passage of the Odyssey provided the basis onwhich the Iliou Persis was constructed. On the contrary, the Odyssey poet

As did Blass (: f.), ‘Nun ist das, was Demodokos in 3 von der Einnahme Trojas singt,gerade in dem gewahlten Anfang so merkwurdig ahnlich mit dem Anfang von Arktinos’ �� , wiedemselben Proklos gibt, daß ich wenigstens uberzeugt bin: der Verfasser hat dieses Lied der Odysseezum Muster gehabt.’




must have known a poem similar to the Iliou Persis in scope and content.It is hardly going too far to say that he represents Demodocus as singingthe Iliou Persis.

Thirdly, he was familiar with the �� , the Return of theAchaeans, as a subject of epic. This is the theme of the song with whichPhemius entertains the suitors in ., and it is one of the topics onwhich Aeolus questions Odysseus in .; the others are ‘Ilios’ and ‘theships of the Achaeans’, which sounds like a version of the Catalogue ofShips. The Return of the Achaeans is the background against which thepoet composed his own epic of Odysseus’ Return. His references to theother heroes’ returns are in fair agreement with the content of the CyclicNostoi. Nestor in book (–, –) relates to Telemachus howAgamemnon and Menelaus quarrelled over the return: Menelaus wantedto set sail straight away, but Agamemnon insisted on waiting and makingsacrifices to appease the anger of Athena. So half the army set sail, includingNestor and Diomedes, who both got safely home. Menelaus followed, buthe was delayed at Sounion by the death of his helmsman and then caughtby a storm which carried him to Crete and from there with five of his shipsto Egypt, where he remained for seven years. Meanwhile Agamemnonhad got home and been murdered by Aegisthus with Clytaemestra. Sevenyears later Orestes appeared and avenged his father, and Menelaus returnedfrom Egypt immediately after that. Telemachus learns more from Menelaushimself in book , including how the Locrian Ajax perished at the �#�� from the anger of Athena and Poseidon (–). All of thisagrees with the Nostoi as summarized by Proclus, except that in ProclusAjax’ death is associated with the U1�� in Euboea insteadof the �#�� at Tenos. In the parallel account in Apollodorus,however, the storm is located near Tenos and Ajax’ body is washed up onMyconos. It may be that this was in fact the Nostoi version and that Proclushas been modified to fit the later vulgate version where it was Euboea.In any case the Odyssey poet knows an epic account of all these eventsthat was closely related to the Nostoi current in classical times. The Nostoi,so far as we can see, made only one brief allusion to Odysseus’ return:Neoptolemus, making his way home overland, encountered Odysseus atMaroneia. Clearly the poem did not contain the main story of Odysseus’return, and the reason is no doubt that there was already a separate epicdealing with that, some precursor of our Odyssey.

It has long been recognized that the Odyssey also drew upon an Argonau-tic epic. Perhaps the same poem lies behind Mimnermus’ lines (–a)about Jason’s successful recovery of the Golden Fleece from the ends ofthe earth. The epic did not survive into later times, but we can reconstruct



martin west

something of its narrative. In a recent reexamination of the question (West) I have shown that this Argonautica could not have been composedmuch before . This is one among other reasons for bringing the date ofthe Odyssey down into the latter part of the seventh century. More of thislater.

The Odyssey in its turn served as the basis for new invention. Proclus’Trojan Cycle ended with the Telegony, very plausibly ascribed to a Cyre-naean poet Eugammon active in the s. As summarized by Proclus, itappears as a ragbag of legends about the end of Odysseus’ life, in which thenumber of his sons was increased from one to four (or possibly five), bornof three different mothers. The first part of the poem covered the journeyinland into Thesprotia that Teiresias foretold in the Nekyia (.–).The second part dealt with the arrival in Ithaca of Telegonus, a son thatOdysseus had fathered during his stay with Circe. The two parts no doubthad separate origins: the first was designed to glorify a local dynasty inThesprotia, while the final product, in which Odysseus had a late son byPenelope called Arcesilas, honoured the Battiads of Eugammon’s own cityCyrene.

We can thus construct a chronological sequence Argonautica > Odyssey> Thesprotis > Telegony. Further reflexes of the Odyssey can be seen in the‘Hesiodic’ Catalogue of Women, where Nestor’s daughter Polycaste, whobathes Telemachus on his visit to Pylos in the Odyssey (.), became hiswife and bore him a son (fr. ), and where the Cephallenians were saidto be the progeny of Hermes and Calypso, a union evidently suggested byHermes’ visits to Calypso’s island in the Odyssey.

In another paper (West b) I have examined reflexes of Ionian epicin the Lesbian poets, who give us a useful chronological peg. By their time,c. –, it seems clear that certain celebrated hexameter poems werecurrent in something like a fixed form, no doubt due to the existence ofwritten texts – not that many people read written texts, of course. Wehave seen that the Lesbians knew the Works and Days and the Iliad. Theycertainly knew a poem or poems about the beginnings of the Trojan War,and about its end, though they were very selective in what they chose tomention. They were fascinated above all by the two most glamorous char-acters in the story: Helen, who caused the war, and Achilles, the supremehero who fought in it. Something of the same perspective characterizedthe Cypria, whose poet gave elaborate accounts of the births of Helen andAchilles, both resulting from initiatives of Zeus to lay the basis for thegreat war he planned. Both birth legends are referred to by the Lesbianpoets: Helen’s birth from an egg found and fostered by Leda, and the




wedding of Peleus and Thetis in Cheiron’s house, attended by all the gods,which resulted in the birth of Achilles. There is also reference to Helen’sabandonment of Menelaus and of her daughter Hermione.

All this agrees well enough with the Cypria for that to be a possiblesource, except that the Cypria seems to have been a rather late compilationwith no organic unity, a long concatenation of episodes, most of whichfind no echo in the remains of the Lesbians. The language of its fragments,especially fr. , makes a date before seem unlikely (Wackernagel :–). In any case it is not likely that any epic singer in the time of Sapphoand Alcaeus was so ‘cyclic’ in his outlook that he set out to cover in onepoem the whole story of the Trojan War from the wedding of Peleus to thepoint where the Iliad begins. What is more easily conceivable is a poemabout how the war began, from the births of Achilles and Helen to theJudgement of Paris, his abduction of Helen, and her reception at Troy.Such a poem might well have continued to be transmitted after the timeof the Lesbians and been absorbed in the eventual encyclopaedic Cypria.

As regards the end of the war, the Lesbians clearly know the main eventsand what happened in the Returns from Troy. One of Alcaeus’ politicalsongs (, cf. A h) contained a substantial digression on the LocrianAjax’ assault on Cassandra in the shrine of Athena, the goddess’s angerat the act, and the storm that she roused against the homeward-boundAchaeans because they had neglected to stone Ajax to death. This is ingood agreement with the Iliou Persis so far as we can tell from Proclus’summary. Sappho () retails a cult legend that the Atreidai were delayedat Lesbos on their way home from Troy and were unable to continue untilthey had prayed to Zeus, Hera, and Dionysus. This connects with theNostoi, though it diverges in detail. Proclus does not mention Lesbos, butthat may be the result of excessive compression. The Odyssey version, whichas we have seen is closely related to the Nostoi, gives more detail. It tellsthat Nestor and Diomedes stopped at Lesbos, and Menelaus came soonafterwards and found them there (.–). Sappho diverges from bothepics in having Agamemnon and Menelaus at Lesbos together; in the epicsthe brothers had parted at Troy and never saw each other again after that.So there is a variant version current at Lesbos, but the overall frameworkis the same. We cannot say positively that Sappho and Alcaeus knew‘the’ Iliou Persis or ‘the’ Nostoi, but they certainly knew epic poetry – orepics – of similar content. The fact that the Odyssey poet knew poems closelyrelated to the canonical Aethiopis, Little Iliad, Iliou Persis, and Nostoi makes

Sapph. .ff., ; Alc. .ff.



martin west

it all the more likely that these poems enjoyed some general currency bythe end of the seventh century.

There is no reliable evidence that the Lesbians knew the Odyssey itself.To suggest that they might not have known it may seem preposterous tothose who take it for granted that the two Homeric epics were composedgenerations earlier. But this is just the sort of inherited prejudice that wemust take care to exclude from our reckoning. Our aim is to build a soundedifice from a fresh appraisal of the evidence, taking nothing for granted.

Besides narrative epic, the Lesbians will have known hymns of theHomeric type, and some of Alcaeus’ hymns show connections withsome of the extant Homeric Hymns. The hymn to Dionysus in which hetold the story of Hera’s entrapment in a special chair made by Hephaestus(a–e Voigt) had the same subject matter as a major Homeric hymnof which we only have fragments and which I will discuss later. In hishymn to Hermes ( + SLG S) Alcaeus told how the infant god stoleApollo’s cattle and then his bow, and this naturally provokes comparisonwith the Homeric hymn. Thirdly there is his hymn to the Dioscuri (,with which PMG perhaps belongs), in which he describes them asriding over land and sea on their swift horses, and as saving sailors fromdeath, leaping in brightness from afar onto their ships’ extremities. Therd Homeric Hymn has very similar content. Fourthly there is a fragmentof a hymn to Artemis, ascribed by some editors to Alcaeus but by othersto Sappho, in which the goddess swears an oath of eternal virginity whiletouching the head of her father Zeus, and he assents and grants her herspecial role as a huntress. In the Homeric Hymn to Aphrodite (.ff.) thereis a similar description, but applied to Hestia instead of Artemis.

It is clear that the lyric and the hexameter hymns stand in a commontradition, and it is quite conceivable that there are more direct relationships.In the cases of Alcaeus’ Dionysus and Dioscuri hymns there seems no reasonwhy he should not have known the poems later transmitted in the Homericcollection. The Homeric Hymn to Hermes, on the other hand, can hardlybe dated as early as : it contains too many words and expressions notparalleled before the fifth century, and I would prefer to regard it, withWilamowitz and others, as a later descendant of the hymn that Alcaeusknew. The Hymn to Aphrodite is almost certainly early enough to have

See West (b: n. ). Alc. L–P = Sapph. A Voigt. Cf. Lobel and Page (: ), Page (b: –), Treu (:

–), Liberman (: xciv). Wilamowitz-Moellendorff (: , : n. ); cf. Liberman (: ). For a clear and cautious

comparison of Alcaeus’ poem with the Homeric hymn see Page (b: –).




been known to the Lesbian poets. But one may think it more likely thatHestia’s oath in the hexameter poem is adapted from a similar passage inan Artemis hymn, and that such a hymn served as the direct source for theLesbian poet, than that the Lesbian personally transferred the motif fromHestia to Artemis.

So much for my three criteria and what they yield. Let me now beginto synthesize. I have constructed a number of sequences: Hesiod beforethe Iliad, Semonides and the Lesbians; the Iliad before the Memnonis, theOdyssey and the Lesbians (and of course before the Doloneia, which wasnever an independent poem but composed to stand in the Iliad in itspresent place); the Memnonis before the Odyssey and Alcman; the LittleIliad, Iliou Persis and Nostoi before the Odyssey and no doubt before theLesbians; and so on. The linguistic criterion puts the Cypria and certain ofthe Homeric Hymns well after all these.

To progress towards an absolute chronology we need to find fixed pointsto which to attach some of the drifting threads. We can roughly datethe non-hexameter poets who are involved. Alcman and the two Lesbiansflourished in the late seventh and/or the early sixth century. Semonides ofAmorgos is less easy to fix. The ancient synchronism with Archilochus

may be based on nothing but the fact that these were two of the threeprincipal iambographers. However, if it is true that Semonides was one ofthose who led the Samian colony to Amorgos, that would put him in themid-seventh century and agree with the Archilochus synchronism.

There is no good reason to believe that the Iliad is earlier than themid-seventh century, than the generation of Archilochus. When scholarssuch as Joachim Latacz continue to assert an eighth-century dating, it isdue to a phenomenon that may be called dogmatic drag: obsolete viewsgo on exercising a pull even after being discredited, and put a brake onprogress to a new position. When scholars were forced to give up theold datings to the eleventh or tenth or ninth century, they continued tocling to the highest dating that seemed admissible. The fact is that thereis no clear reflex of the Iliad in art before c. , or in literature beforeAlcaeus. The weaponry and combat tactics described in it have a seriesof features characteristic of the first half of the seventh century (van Wees: esp. –; id. in eod. (: –)). The design of Achilles’ shield,with its scenes of battles, cities, and so on, has its closest archaeological

Euseb. Chron. Ol. ; cf. Rohde (: I. –). Suda s.v. Q�''��, obviously by confusion with Q�'��. Cf. Kullmann (: ): ‘Latacz’ Ansatz der Ilias im Jh. . . . beruht auf reiner Konvention und

kann nicht richtig sein.’



martin west

parallels in Cypro-Phoenician metal dishes and bowls of the period –. Walter Burkert has famously connected the passage about the wealthof Egyptian Thebes (.–) with that city’s prosperity in the twenty-fifth dynasty (–) and with the display of booty taken from it byAssurbanipal in (: –). I have proposed another linkage witha historical event in the Near East, connecting the flood which washedaway the Achaean fortifications after the war (.–) with Sennacherib’slevelling of Babylon by diverted river waters in . The reference in .f.to the !x��'��4�� 4��1�4�� implies knowledge of the Scythians, andthus of the Black Sea, that cannot antedate the seventh century. The sameis true of the reference to Apollo’s shrine at Delphi as a paradigm of riches(.f.): it was a functioning oracular centre in the eighth century, but ofonly local importance and at that time far surpassed by Perachora in thewealth of its dedications. Similarly, when Nestor in .– speaks of aprize-winning, four-horse chariot team that his father had sent to competeat games in Elis, the poet must have the Olympic games in view, which didnot begin to attract more than a local crowd until around and did nothave chariot racing till . A further clearly seventh-century element isthe Gorgon blazon on Athena’s aegis in .f. and on Agamemnon’s shieldin .f. Taking all the evidence together, I concluded that the likeliesttime for the composition of the Iliad was between and (West :–). I remain convinced that this dating cannot be far wrong.

With regard to Hesiod I am more inclined to modify the opinions thatI first put forward in my edition of the Theogony over forty years ago (West: –). I then dated the Theogony between and ; as for theWorks and Days, which we know is the later poem of the two, I thoughtit was not likely to be later than , with as the latest possible date.I repeated these views unchanged in my edition of the Works and Days(West : n. ). If we are dating Semonides rightly, remains anapproximate terminus ante quem for Hesiod. But I am now more reluctantto go up into the mists of the eighth century. I originally did so on thegrounds that the Amphidamas at whose funeral games Hesiod won a prizein Chalcis (Op. –) is connected by Plutarch with the Lelantine War,and that that shadowy conflict perhaps belongs in the last third of theeighth century. But those are very insecure data. Two other considerations

Cf. Morgan (: –), who finds no convincing evidence of a temple in the sanctuary areabefore the mid-seventh century; Dickie (: ).

Paus. ..; Dickie (: f.); Crielaard (: f.); Kullmann (b: –). Burkert (: n. ) = (: ). Gorgons are first attested in this function around and

do not appear in art at all before .




suggest a somewhat later dating. One is that if we are dating the Iliad andeverything else to after , an eighth-century dating for Hesiod wouldleave him isolated up there, a whole generation or two generations earlierthan anything else that survived in written transmission. This seems apriori unlikely. The other point is that the catalogue of rivers in Theogony– includes the Nile and the Danube, besides some lesser rivers of thesouthern and western shores of the Black Sea. One cannot say that thiswould have been impossible in the late eighth century, but it would be lesssurprising a generation later. So I am now content to contemplate a datefor Hesiod c. –.

The river list also includes the Phasis, at that time a rather mythical water-course associated with the Argonautic expedition, not yet firmly located.

Hesiod doubtless knew the Argonaut legend, which is also implied inthe Iliad (.–). But the particular Argonautica from which the poet ofthe Odyssey borrowed extensively can hardly have been composed before themiddle of the seventh century, because it appears to reflect knowledge ofa sector of the northern Pontos, from the Crimea to the Straits of Kerch,which so far as we can judge from the archaeological material was firstbeing explored at that period (West : –, ).

This gives us a terminus post quem for the Odyssey itself. What is the termi-nus ante quem? It is difficult to find one earlier than Eugammon’s Telegony,which we place in the s, and Stesichorus (PMGF ). However, thepicture of the oriental world in the Odyssey would appear to suit the seventhcentury better than the sixth. Sidon, destroyed in , is treated as a flour-ishing city (cf. especially .). This does not mean that the poet had notheard of the destruction – he is telling of a much earlier time – but we maysay that Sidon’s prosperity is still a living memory. The picture of Egyptseems to reflect the period of friendly Graeco-Egyptian relations in thetime of Psammetichus (Dickie : –). Then there is the passage whereMenelaus recalls his visit to Libya, where the sheep quickly grow horns,

for they give birth three times a year:there neither master nor shepherd goes lackingin cheese and meat, nor in sweet milk,but they ever provide abundant milk.

(.–)

So by the time the Odyssey is composed Libya is known and famed for itsflocks, and this will surely be at about the time of the foundation of Cyreneor not long before (Dickie : –).

See now M. L. West ().



martin west

I observed earlier that the Memnonis existed before the Odyssey, andthat the same is true of three other epics identifiable with, or at leastmuch resembling, the Little Iliad, Iliou Persis and Nostoi current in theclassical period. The Memnonis at least is post-Iliadic, and there is no signof Memnon in literature or art before the end of the seventh century. TheseCyclic poems all seem to belong sometime in the second half of the seventhcentury, with the Memnonis in the last quarter.

Memnon is one of a series of heroes who arrive unheralded at Troytowards the end of the war bringing foreign armies to help the Trojans. Theothers are the Amazon Penthesilea, whose story was prefixed to Memnon’sin the Aethiopis, the Thracian Rhesus who appears in the Doloneia, andthe Mysian Eurypylus, the son of Telephus, who featured in the Little Iliadand was killed by Achilles’ son Neoptolemus. This motif of the relief forcefrom foreign parts is evidently a novel development serving the purposeof filling out and prolonging the tale of Troy’s defeat. The story of Neop-tolemus and Eurypylus is known to the Odyssey poet, and incidentally toArchilochus (fr. ), whereas the two allusions to Neoptolemus in the Iliad(.–, .–) seem to be interpolated. The Doloneia, like theMemnonis, appears to be not later than , as the evidence of artindicates that it was already part of the Iliad as known in the Pelopon-nese by that time or soon after (Friis Johansen : ; Snodgrass :, ).

Turning to the Homeric Hymns, I begin with the long hymn to Diony-sus which has come down to us only in fragments and which is nowadaysnumbered as Hymn . It may plausibly be identified with the hymn whoseexistence was long ago inferred by Bergk and Wilamowitz from poetic andartistic references to a singular story about Dionysus and Hephaestus.

The story is reconstructed as follows. When Hera gave birth to the crip-ple Hephaestus she was disgusted with him and cast him down fromheaven into the sea. He grew up there among the Nereids, developing hisengineering skills. Then he sent his mother a fine throne he had made,which was fitted with a secret mechanism so that when she sat down shefound herself trapped. None of the gods was able to free her, and evi-dently Hephaestus would have to be induced to come back and undowhat he had done. Ares went and tried to capture him by force, butfailed because he defended himself with fire, which Ares could not face.Then Dionysus went equipped with wine and authorized to say that ifhe consented to come he could have Aphrodite for his wife. The wine

Bergk (: n. ), Wilamowitz-Moellendorff (: –); cf. West (b).




made Hephaestus drunk, and he was escorted back to Olympus in a merrystate, riding on a donkey. He released Hera and duly received Aphrodite(somewhat to her distaste). Dionysus’ reward was to be admitted as anOlympian.

I have mentioned that Alcaeus retold the story in a hymn of his own, andWilamowitz inferred a seventh-century date for the hexameter model. Infact we already find a series of echoes of it in the Iliad: .ff. (Hephaestusthrown out of heaven by Zeus to Lemnos); .– (a fine throne tobe made by Hephaestus), – (Zeus throwing gods about, would havethrown Hypnos down into the sea), – (promise of Charis as wife);.ff. (Hera bound fast by Zeus, who is throwing gods out of heaven;none of them can free her); .–, ff. (Hephaestus has Charis aswife; was once thrown out by Hera because a cripple; his sojourn with theNereids and practice of crafts). Charis in these passages is a substitute forAphrodite in the hymn. The Dionysus hymn, then, or whatever form ofit was known to the Iliad poet, is to be dated not later than the middlethird of the seventh century.

The Hymn to Aphrodite may be put twenty or thirty years later. It isthe closest in style of the Hymns to the Iliad, and the prophecy which itcontains about Aeneas’ descendants ruling among the Trojans in the futureis related to the similar prophecy in Iliad .–. Reinhardt even arguedthat the hymn was by the poet of the Iliad himself; but later scholars havegiven good grounds for seeing its poet as an epigone with his own narrativestyle, a slightly more developed vocabulary, and a derivative use of theIliad and Hesiod. Still, both of the prophecy passages must have beencomposed to gratify the same aristocratic family in the Troad, and therecannot be a big time-gap, probably a decade or two at most, between thesetwo celebrations of it. I mentioned earlier that the hymn’s description ofAphrodite’s toilet looks like the model for a passage in the Odyssey; thatagain would suit a dating to (say) the s or s.

The Aphrodite hymn was also drawn upon by the poet of the Hymn toDemeter, which is therefore later. As it contains no mention of Athens, itis assumed to date from before the time of Pisistratus, when the Mysteriescame under Athenian control. It can be dated with some probability tothe first half of the sixth century, but not more narrowly within thistime-span.

Demodocus’ song in Od. .ff. likewise reflects the tale: there is Aphrodite as Hephaestus’(unfaithful) wife, Ares is again involved, and the motif of the trap-furniture appears in a differentapplication.

To the references in West (b: ) add now Faulkner (: –).



martin west

Hes. (Th., Op.)

680

670

660

650

640

630

620

610

600

590

580

570

560

550

540

530

520

510

500

490

480

470

460

450

440

430

420

Relative sequences Time scale Other poems: suggested datings

H.Dion. (1)

SemonidesIliad

Argon.

Memnonis

Aethiopis

Alcman

Doloneia

H.Ap.Pyth.

Thesprotis

H.Dem.

Telegony

H.Ap.D+P (523)

H.Ap.Del.

Lesbians

H.Aphr.

Odyssey

Il.parva. Il.Pers. Nostoi

H.Art. (9)

Thebais (t.p.q.). Aspis

Corinthiaca (t.p.q.)

Cypria

Margites. Pythagorean Orphica

H.Dion (7). Orphic Theogony

H.Pan (19) Parmenides

H.Herm.

H.Heph. (20)

Pisander (t.p.q.)

‘Hes.’ Cat.

Panyassis

Empedocles

‘Epimenides’ Chresmoi

H.Helios/Selene (31–2)

Fig. . A chronology of early Greek epic.




The Pythian part of the Hymn to Apollo was apparently composed shortlyafter the First Sacred War of /, when the Delphic sanctuary waswrested from the control of Crisa by an alliance of Phocians and othersand Crisa itself was destroyed. The warning to the Crisaeans at the end(–) is a post eventum prophecy referring to the change of governance. Itis an attractive conjecture that the hymn was composed for the first PythianGames in . If so, that would be the first exact date that we could attachto a Greek hexameter poem.

The next exact date is , the year when Polycrates celebrated a com-bined Delian and Pythian festival on Delos in honour of Apollo. As Burkertand Janko have seen, this is the likely occasion of the combined Delian andPythian Hymn that has come down to us (Burkert : –; Janko :–). It is a deliberate conflation of the older Pythian Hymn with a morerecent Delian Hymn that had also had an independent existence. As to thedate of the Delian Hymn, Burkert noted that it presupposes a temple ofApollo on Delos, something that did not exist until about / (:).

Figure . sums up the conclusions of the discussion. For the sake of amore complete picture I have included in the right-hand column suggesteddatings for various poets and poems not considered above, based on a rangeof considerations that could not be discussed within the limits of the presentessay; I have argued for some of them in previous publications. I do notpretend that these results are definitive. But they represent an attemptat a new overall chronology of early Greek epic founded on reason andobservation rather than tradition and convention, and as such I hope theywill be welcomed.

For the Homeric Hymns listed cf. West (b: –); for the Margites, ibid. (); for theThebais, West (a: ); for the Corinthiaca ascribed to Eumelus, West (a: –); forthe pseudo-Hesiodic Catalogue of Women, West (: –); for early Pythagorean Orphica and thefirst Orphic Theogony, West (: –, –); for ‘Epimenides’, ibid. (–).



Bibliography

Alden, M. () ‘Lions in paradise: lion-similes in the Iliad and the lion-cubs ofIl. .–’, CQ 55: –.

Alexanderson, B. () ‘Homeric formulae for ships’, Eranos 68: –.Allan, W. () ‘Arms and the man: Euphorbus, Hector, and the death of Patro-

clus’, CQ 55: –.() ‘Performing the will of Zeus: the A�.� >�#�: and the scope of early

Greek epic’, in Performance, Iconography, Reception: Studies in Honour ofOliver Taplin, ed. M. Revermann and P. Wilson. Oxford: –.

Allen, T. W., W. R. Halliday and E. E. Sikes (eds.) () The Homeric Hymns.Oxford.

Andersen, Ø. () ‘The making of the past in the Iliad ’, HSPh 93: –.() ‘Allusion and the audience of Homer’, in Paısi-Apostolopoulou ():

–.Anderson, M. J. () The Fall of Troy in Early Greek Poetry and Art. Oxford.Arrighetti, G. (ed.) () Esiodo: Opere, with introduction, translation and com-

mentary. Turin.Athanassakis, A. N. () The Orphic Hymns. Missoula, Mont.Auerbach, E. () Mimesis: Dargestellte Wirklichkeit in der abendlandischen Lit-

eratur. Bern.Ballabriga, A. () Les fictions d’Homere. Paris.Bartonek, A. and G. Buchner () ‘Die altesten griechischen Inschriften von

Pithekoussai (. Halfte des VIII. bis . Halfte des VI. Jh.)’, Sprache 37.2:–.

Beck, W. () ‘t��dxv=" in the Hesiodic Catalogue and Antimachus’, ZPE73: –.

Berg, N. and D. Haug () ‘Dividing Homer (continued): innovation vs.tradition in Homer – an overlooked piece of evidence’, SO 75: –.

Bergk, T. () Griechische Literaturgeschichte, vol. I. Berlin.Bernabe, A. (ed.) () Poetarum Epicorum Graecorum Testimonia et Fragmenta,

vol. I. Leipzig, Berlin and Stuttgart. nd edn. .(ed.) () Poetarum Epicorum Graecorum Testimonia et Fragmenta, vol. II.

Leipzig, Berlin and Munich.



Bibliography

Betegh, G. () The Derveni Papyrus: Cosmology, Theology and Interpretation.Cambridge.

Bittlestone, R., J. Diggle and J. Underhill () Odysseus Unbound: the Search forHomer’s Ithaca. Cambridge.

Blass, Friedrich () Die Interpolationen in der Odyssee: Eine Untersuchung.Halle.

Blumel, W. () Die aiolischen Dialekte: Phonologie und Morphologie derinschriftlichen Texte aus generativer Sicht. Gottingen.

Blumer, W. () Interpretation archaischer Dichtung: die mythologischen Partiender Erga Hesiods ( vols.). Munster.

Bohme, R. () Orpheus: das Alter des Kitharoden. Berlin.() Orpheus: der Sanger und seine Zeit. Bern.() Peisistratos und sein homerischer Dichter. Bern.() ‘Homer oder Orpheus?’, ZPE 71: –.() Der Lykomide: Tradition und Wandel zwischen Orpheus und Homer. Bern.

Bolte, F. () ‘Ein pylisches Epos?’, RhM 83: –.Bowra, C. M. () ‘Homeric words in Cyprus’, CQ 20: –.

() ‘Homeric words in Arcadian inscriptions’, JHS 54: –.Bravo, B. (a) ‘Luoghi di culto nella chora di Olbia Pontica’, in Problemi della

chora coloniale dall’ Occidente al Mar Nero: atti del quarantesimo convegno distudi sulla Magna Grecia. Taranto: –.

(b) ‘Un frammento della Piccola Iliade (P. Oxy. ), lo stile narrativotardo-arcaico, I racconti su Achille immortale’, QUCC n.s. 67: –.

Brillante, C. () ‘Nestore gerenio: le origini di un epiteto’, in Atti e memoriedel secondo congresso internazionale di micenologia 1991, vol. I, ed. E. DeMiro:–.

Brown, C. O. () ‘Anactoria and the �� '�+4'� at Sappho fr., Voigt’, QUCC 61: –.

Buchholz, H.-G. and F. Matz (eds.) (–) Archaeologia Homerica: die Denkmalerund das fruhgriechische Epos. Gottingen.

Buck, C. D. () The Greek Dialects, nd edn. Chicago.Buecheler, F. () ‘Paraphrase eines Gedichtes uber den Raub der Persephone’,

in Berliner Klassikertexte, vol. V., eds. F. Buecheler, W. Schubart andH. Diels. Berlin: –.

Buiskikh, S. () ‘Das Achilleus-Heiligtum in der Chora von Olbia Pontica ausarchaischer Zeit’, in Problemi della chora coloniale dall’ Occidente al Mar Nero:atti del quarantesimo convegno di studi sulla Magna Grecia. Taranto: –.

Burgess, J. S. () The Tradition of the Trojan War in Homer and the Epic Cycle.Baltimore and London.

() ‘The epic cycle and fragments’, in The Blackwell Companion to AncientEpic, ed. J. M. Foley. Oxford: –.

() ‘Neoanalysis, orality, and intertextuality: an examination of Homericmotif transference’, Oral Tradition 21: –.

() The Death and Afterlife of Achilles. Baltimore.() ‘The hypertext of Astyanax’, Trends in Classics 2: –.



Bibliography

Burkert, W. () ‘Die Leistung eines Kreophylos: Kreophyleer, Homeriden unddie archaische Heraklesepik’, MH 29: –, reprinted in Burkert ():–.

() ‘Das hunderttorige Theben und die Datierung der Ilias’, WS 89: –,reprinted in Burkert (), –.

() ‘Kynaithos, Polycrates and the Homeric hymn to Apollo’, in Arktouros:Hellenic Studies Presented to Bernard M. W. Knox, ed. G. W. Bowersock,W. Burkert and M. C. J. Putnam. Berlin and New York: –. Reprinted inBurkert (), –.

() Homo Necans: the Anthropology of Ancient Greek Sacrificial Ritual andMyth, tr. P. Bing. Berkeley and Los Angeles.

() ‘The making of Homer in the sixth century B.C.: rhapsodes versus Stesi-choros’, in Papers on the Amasis Painter and his World: Colloquium sponsoredby the Getty Center for the History of Art and Humanities and Symposiumsponsored by the J. Paul Getty Museum. Malibu, Calif.: –. Reprinted inBurkert (), –.

() ‘“Irrevocabile verbum”: Spuren mundlichen Erzahlens in der Odyssee’,in Horen – Sagen – Lesen – Lernen: Bausteine zu einer Geschichte der kom-munikativen Kultur. Festschrift fur Rudolf Schenda zum 65. Geburtstag, ed.U. Brunold-Bigler and H. Bausinger. Berlin: –. Reprinted in Burkert(), –.

() The Creation of the Sacred: Tracks of Biology in Early Religions. Cambridge,Mass. and London.

() Kleine Schriften I: Homerica, ed. C. Riedweg. Gottingen.Cairns, D. L. (a) ‘Introduction’, in Cairns (b): –.

(ed.) (b) Oxford Readings in Homer’s Iliad. Oxford.Cantilena, M. () Omero: Odissea, vol. VI, introduzione e commento a cura

di J. Russo e A. Heubeck con aggiornamenti di Mario Cantilena, settimaedizione completamente rinnovata. Milan.

Cassola, F. (ed.) () Inni omerici. Milan.Chadwick, J. () ‘The descent of the Greek epic’, JHS 110: –.Chantraine, P. () ‘Le role et la valeur de 7�- dans la composition’, RPh 16:

–.() ‘Le divin et les dieux chez Homere’, in La notion du divin depuis Homere

jusqu’a Platon. Entretiens sur l’Antiquite classique . Geneva: –.() Grammaire homerique, vol. II. Paris.() Grammaire homerique, vol. I, troisieme tirage avec une conclusion nouvelle.

Paris. st edn. .() Dictionnaire etymologique de la langue grecque. Paris.

Cingano, E. () ‘Tradizioni su Tebe nell’epica e nella lirica greca arcaica’, inPresenza e funzione della citta di Tebe nella cultura greca: Atti del ConvegnoInternazionale (Urbino 7–9 luglio 1997), ed. P. Angeli Bernardini. Pisa andRoma: –.

() ‘A catalogue within a catalogue: Helen’s suitors in the Hesiodic Catalogueof Women (frr. –)’, in Hunter (): –.



Bibliography

Cirio, A. M. () Lettura di Omero: Canto X dell’ Iliade. Palermo.Clark, M. () ‘Formulas, metre and type-scenes’, in Fowler (a): –.Clay, J. S. () The Wrath of Athena. Princeton. nd edn. , Lanham.

() The Politics of Olympus: Form and Meaning in the Major Homeric Hymns.Princeton.

() ‘Odyssean animadversions’, in Montanari (): –.Colakovic, Z. () ‘Homer: Cor Huso ili Avdo, rapsod ili aed?’ [Homer: Cor

Huso or Avdo, rhapsodos or aoidos?], Almanah 27–8: –.(a) The Epics of Avdo Mededovic: A Critical Edition, with introduction by

R. L. Fowler ( vols.). Podgorica.(b) ‘The singer above tales: Homer, Mededovic, and traditional epics’, in

Colakovic (a), vol. II: –.Colakovic, Z. and M. Rojc-Colakovic () Mrtva glava jezik progovara [The

Dead Head Speaks in Tongues]. Podgorica.Colli, G. () La sapienza greca, vol. I. Milan.Crespo, E. (a) Elementos antiguos y modernos en la prosodia homerica. Minos

Suppl. . Salamanca.(b) ‘La cronologıa relativa de la metatesis de cantidad en jonico-atico’,

CFC(G) 12: –.Crielaard, J. P. () ‘Homer, history and archeology: some remarks on the date

of the Homeric world’, in Homeric Questions: Essays in Philology, AncientHistory and Archaeology, Including the Papers of a Conference Organized by theNetherlands Institute at Athens (15 may 1993), ed. J. P. Crielaard. Publicationsof the Netherlands Institute at Athens . Amsterdam: –.

Crissy, K. () ‘Herakles, Odysseus, and the bow: Odyssey .–’, CJ 93: –.Currie, B. G. F. () ‘Homer and the early epic tradition’, in Epic Interactions:

Perspectives on Homer, Virgil and the Epic Tradition Presented to Jasper Griffinby Former Pupils, ed. M. J. Clarke, B. G. F. Currie and R. O. A. M. Lyne.Oxford: –.

Curti, M. () ‘L’officina dei cicli’, SCO 43: –.Dale, A. M. (ed.) () Euripides: Helen. Oxford.D’Alessio, G. B. () ‘The Megalai Ehoiai: a survey of the fragments’, in Hunter

(): –.Danek, G. () Studien zur Dolonie. Vienna.

(a) ‘Darstellung verdeckter Handlung bei Homer und in der sudslawischenHeldenlied-Tradition’, WS 111: –.

(b) Epos und Zitat: Studien zu den Quellen der Odyssee. Vienna.(c) ‘Odysseus and the bow’, in Paısi-Apostolopoulou (): –.() ‘Achilles and the Iliad’, in Eranos: Proceedings of the 9th International

Symposium on the Odyssey (2–7 September 2000), ed. M. Paısi-Apostolopoulou.Ithaca, Greece: –.

(a) Bosnische Heldenepen. Klagenfurt.(b) ‘Traditional referentiality and Homeric intertextuality’, in Montanari

(): –.(), review of Tsagarakis (), Gnomon 75: –.



Bibliography

() ‘Der Schiffskatalog der Ilias: Form und Funktion’, in Ad Fontes! Festschriftfur Gerhard Dobesch zum funfundsechzigsten Geburtstag am 15. September 2004,ed. H. Heftner and K. Tomaschitz. Vienna: –.

() ‘Developments of traditional narrative techniques: flashback accountsin Homer and Avdo Mededovic’, in Colakovic (a) vol. II: –.

Davies, M. (ed.) () Epicorum Graecorum Fragmenta. Gottingen.(a) ‘The date of the epic cycle’, Glotta 67: –.(b) The Epic Cycle. Bristol.(ed.) () Sophocles: Trachiniae, with introduction and commentary. Oxford.

Debiasi, A. () L’epica perduta: Eumelo, il Ciclo, l’occidente. Rome.Dickie, M. () ‘The geography of Homer’s world’, in Homer’s World: Fiction,

Tradition, Reality, ed. Ø. Andersen and M. Dickie. Papers of the NorwegianInstitute at Athens . Bergen: –.

Dihle, A. () Homer-Probleme. Opladen.Dirlmeier, F. () ‘Die Giftpfeile des Odysseus’, in Ausgewahlte Schriften zu

Dichtung und Philosophie der Griechen, ed. H. Gorgemanns. Heidelberg:–. First published in SHAW ().

Dodds, E. R. () The Greeks and the Irrational. Berkeley.Dougherty, C. () The Poetics of Colonization: from City to Text in Archaic

Greece. New York and Oxford.Dover, K. () Lysias and the Corpus Lysiacum. Berkeley and London.Dowden, K. () ‘Homer’s sense of text’, JHS 116: –.

() ‘The epic tradition in Greece’, in Fowler (a): –.Drager, P. () Argo pasimelousa. Stuttgart.

() Untersuchungen zu den Frauenkatalogen Hesiods. Stuttgart.() Die Argonautika des Apollonios Rhodios: das zweite Zornepos der griechis-

chen Literatur. Munich and Leipzig.(ed.) () Apollodor: Bibliotheke. Gotter- und Heldensagen, with translation

and commentary. Dusseldorf and Zurich.Easterling, P. E. () ‘Gods on stage in Greek tragedy’, in Religio Graeco-Romana:

Festschrift fur Walter Potscher, ed. J. Dalfen, G. Petersman and F. F. Schwarz.Grazer Beitrage Suppl. . Graz and Horn: –.

Edwards, G. P. () The Language of Hesiod in its Traditional Context. Oxford.Edwards, M. W. () ‘Homeric speech introductions’, HSPh 74: –.

() ‘Topos and transformation in Homer’, in Homer: Beyond Oral Poetry, ed.J. M. Bremer, I. J. F. De Jong and J. Kalff. Amsterdam: –.

() ‘Neoanalysis and beyond’, ClAnt 9: –.() The Iliad: a Commentary, vol. V. Cambridge.

Eisenberger, H. () ‘Der Mythos in der aolischen Lyrik’, dissertation, Frankfurt.Emlyn-Jones, C. () ‘The reunion of Penelope and Odysseus’, in Homer, eds.

I. McAuslan and P. Walcot. Oxford: –, –. First published in G&R (): –.

Erbse, H. () Beitrage zum Verstandnis der Odyssee. Berlin.Evelyn-White, H. G. () Hesiod, The Homeric Hymns and the Homerica. Cam-

bridge, Mass.



Bibliography

Faulkner, Andrew () The Homeric Hymn to Aphrodite: Introduction, Text, andCommentary, Oxford.

Fenik, B. () Typical Battle Scenes in the Iliad: Studies in the Narrative Techniquesof Homeric Battle Description. Wiesbaden.

Fernandez-Galiano, M. () ‘Books XXI–XXII’, in Heubeck et al. (–),vol. III: –.

Fick, A. () Die homerische Ilias, nach ihrer Entstehung betrachtet und in derursprunglichen Sprachform wiederhergestellt. Gottingen.

Finkelberg, M. () ‘Ajax’s entry in the Hesiodic Catalogue of Women’, CQ 38:–.

() ‘Formulaic and nonformulaic elements in Homer’, CPh 84: –.() ‘Odysseus and the genus “hero”’, G&R 42: –.(a) The Birth of Literary Fiction in Ancient Greece. Oxford.(b) ‘Time and Arete in Homer’, CQ 48: –.() ‘The Cypria, the Iliad, and the problem of multiformity in oral and

written tradition’, CPh 95: –.() ‘Homer as a foundation text’, in Finkelberg and Stroumsa ():

–.() ‘Oral theory and the limits of formulaic diction’, Oral Tradition 19:

–.() Greeks and Pre-Greeks: Aegean Prehistory and Greek Heroic Tradition.

Cambridge.Finkelberg, M. and G. G. Stroumsa (eds.) () Homer, the Bible, and Beyond:

Literary and Religious Canons in the Ancient World. Leiden.Fittschen, K. () Untersuchungen zum Beginn der Sagendarstellung bei den

Griechen. Berlin.Focke, F. () Die Odyssee. Stuttgart and Berlin.Foley, H. P. (ed.) () The Homeric Hymn to Demeter. Princeton.Foley, J. M. () Traditional Oral Epic: The Odyssey, Beowulf, and the Serbo-

Croatian Return Song. Berkeley.() Immanent Art: From Structure to Meaning in Traditional Oral Epic.

Bloomington, Ind.() ‘Oral tradition and its implications’, in Morris and Powell ():

–.() Homer’s Traditional Art. University Park, Pa.

Forssman, B. () ‘Zur Lautform der lesbischen Lyrik’, MSS 33: –.Foster, B. R. () Before the Muses: An Anthology of Akkadian Literature, rd edn.

Bethesda, Md.Fowler, D. P. () ‘On the shoulders of giants: intertextuality and classical

studies’, MD 39: –, reprinted in D. P. Fowler (): –.() Roman Constructions: Readings in Postmodern Latin. Oxford.

Fowler, R. L. () The Nature of Early Greek Lyric: Three Preliminary Studies.Toronto.

() ‘Genealogical thinking, Hesiod’s Catalogue, and the creation of the Hel-lenes’, PCPhS 44: –.



Bibliography

() Early Greek Mythography. Oxford.() ‘Hesiod’s myths’, review of Blumer (), CR n.s. 53: –.(ed.) (a) The Cambridge Companion to Homer. Cambridge.(b) ‘The Homeric question’, in Fowler (a): –.

Fowler, R. W. (), review of Janko (), Phoenix 37: –.Friedrich, R. () Stilwandel im homerischen Epos. Heidelberg.Friis Johansen, K. () The Iliad in Early Greek Art. Copenhagen.Frankel, H. () Dichtung und Philosophie des fruhen Griechentums, nd edn.

Munich. st edn. , New York.Funghi, M. S. () ‘The Derveni papyrus’, in Laks and Most (): –.Galinsky, G. K. () The Herakles Theme. Oxford.Gantz, T. () Early Greek Myth. Baltimore and London. Reprinted in

( vols.).Garbrah, K. A. () ‘A linguistic analysis of selected portions of the Homeric

Odyssey’, Glotta 47: –.() ‘The scholia on Odyssey . –’, Eranos 76: –.

Garner, R. () From Homer to Tragedy. London and New York.() ‘Achilles in Locri’, ZPE 96: –.

Garvie, A. F. (ed.) () Homer: Odyssey, Books VI–VIII, with introduction andcommentary. Cambridge Greek and Latin classics. Cambridge.

Gehrke, H.-J. () ‘Die Geburt der Erdkunde aus dem Geiste der Geometrie:Uberlegungen zur Entstehung und zur Fruhgeschichte der wissenschaftlichenGeographie bei den Griechen’, in Gattungen der wissenschaftlichen Literaturbei den Griechen, ed. W. Kullmann, J. Althoff and M. Asper. ScriptOralia .Tubingen: –.

Gercke, A. () ‘Die Einnahme von Oichalia’, Neue Jahrbucher 8: –.Giovannini, A. () Etude historique sur les origines du Catalogue des vaisseaux.

Bern.Goldman, H. () Excavations at Eutresis in Boeotia. Cambridge, Mass.Goold, G. P. () ‘The nature of Homeric composition’, ICS 2: –.Graf, F. () Eleusis und die orphische Dichtung Athens in vorhellenistischer Zeit.

Berlin and New York.Graziosi, B. () Inventing Homer: the Early Reception of Epic. Cambridge.Graziosi, B. and J. Haubold () Homer: the Resonance of Epic. London.Grethlein, J. () Das Geschichtsbild der Ilias: eine Untersuchung aus

phanomenologischer und narratologischer Perspektive. Gottingen.Griffin, J. () ‘The epic cycle and the uniqueness of Homer’, JHS 97: –,

reprinted in Cairns (ed.) (b): –.() Homer on Life and Death. Oxford.() ‘Homeric words and speakers’, JHS 106: –.(ed.) () Homer: Iliad IX. Oxford.

Grober, C. () Mehmeds Brautfahrt (Smailagic Meho): Volksepos der sudslavischenMohammedaner. Vienna.

Groningen, B. A. van () La composition litteraire archaıque grecque: procedes etrealisations, nd edn. Amsterdam. st edn. .



Bibliography

Grossardt, P. () Die Erzahlung von Meleagros: zur literarischen Entwicklung derkalydonischen Kultlegende. Leiden, Boston and Cologna.

Hackstein, O. () Die Sprachform der homerischen Epen: Faktoren morpholog-ischer Variabilitat in literarischen Fruhformen. Serta Graeca . Wiesbaden.

Hainsworth, J. B. () The Flexibility of the Homeric Formula. Oxford.() Homer. Oxford.(a) ‘Books V–VIII’, in Heubeck et al. (–), vol. I: –.(b) ‘The epic dialect’, in Heubeck et al. (–), vol. I: –.() The Iliad: a Commentary vol. III. Cambridge.

Hajnal, I. () Troia aus sprachwissenschaftlicher Sicht: die Struktur eine Argu-mentation. Innsbruck.

Hall, E. () Inventing the Barbarian: Greek Self-Definition through Tragedy.Oxford.

Hall, J. M. () Ethnic Identity in Greek Identity. Cambridge.() Hellenicity: Between Ethnicity and Culture. Chicago.

Hammond, N. G. () Epirus: the Geography, the Ancient Remains, the Historyand the Topography of Epirus and Adjacent Areas. Oxford.

Hansen, W. () Ariadne’s Thread: a Guide to International Tales Found inClassical Literature. Ithaca, NY and London.

Hardie, C. G. () ‘Two descents into the underworld’, in Evolution of Con-sciousness: Studies in Polarity for Owen Barfield, ed. S. Sugerman. Middletown,Conn.: –.

Haslam, M. () ‘Homeric papyri and the transmission of the text’, in Morrisand Powell (): –.

Haubold, J. (), review of Tsagarakis (), JHS 122: –.() ‘Heracles in the Hesiodic Catalogue of Women’, in Hunter ():

–.Haug, D. () Les phases de l’evolution de la langue epique: trois etudes de linguis-

tique homerique. Hypomnemata . Gottingen.() ‘Does Homeric Greek have prepositions? Or local adverbs? And what’s

the difference anyway?’, in Grammatical Change in Indo-European Lan-guages: Papers Presented at the Workshop on Indo-European Linguistics at theXVIII International Conference on Historical Linguistics, Montreal, 2007, ed.V. Bubenık, J. Hewson and S. Rose. Amsterdam: –.

Heath, J. () ‘The legacy of Peleus: death and divine gifts in the Iliad ’, Hermes120: –.

Heath, M. () ‘Greek literature’, subject review, G&R 55: –.Hedreen, G. () ‘The cult of Achilles in the Euxine’, Hesperia 60: –.Heitsch, E. () Aphroditehymnus, Aeneas und Homer. Gottingen.

() Altes und Neues zur Ilias: Uberlegungen zur Genese des Werkes. Mainz andStuttgart.

() ‘Neoanalytische Antikritik: zum Verhaltnis Ilias–Aithiopis’, RhM 151:–.

Heubeck, A. () ‘Books IX–XII’, in Heubeck et al. (–), vol. II: –.() ‘Books XXIII–XXIV’, in Heubeck et al. (–), vol. III: –.



Bibliography

Heubeck, A., S. West, J. B. Hainsworth, A. Hoekstra, J. Russo and M. Fernandez-Galiano (–) A Commentary on Homer’s Odyssey ( vols.). Oxford.

Hinrichs, G. () ‘De Homericae elocutionis vestigiis Aeolicis’, dissertation,Jena.

Hirschberger, M. () Gynaikon Katalogos und Megalai Ehoiai: ein Kommentarzu den Fragmenten zweier hesiodeischer Epen. Munich and Leipzig.

Hock, H. () ‘The so-called Aeolic inflection of the Greek contract verbs’,Ph.D. dissertation, Yale University.

Hoekstra, A. () Homeric Modifications of Formulaic Prototypes: Studies in theDevelopment of Greek Epic Diction. Amsterdam.

() The Sub-Epic Stage of the Formulaic Tradition: Studies in the HomericHymns to Apollo, to Aphrodite, and to Demeter. Amsterdam and London.

Holoka, J. P. (ed.) () Simone Weil’s The Iliad or the Poem of Force: A Criti-cal Edition, with introduction, translation and commentary. New York andFrankfurt am Main.

Holscher, U. (), review of Kullmann (), Gnomon 38: –.() Die Odyssee: Epos zwischen Marchen und Roman. Munich.

Hommel, H. () Der Gott Achilleus. Heidelberg.Hooke, S. H. () Middle Eastern Mythology from the Assyrians to the Hebrews.

London.Hooker, J. () ‘The cult of Achilles’, RhM 131: –.Horrocks, G. C. () ‘The antiquity of the Greek epic tradition: some new

evidence’, PCPhS 26: –.() Space and Time in Homer: Prepositional and Adverbial Particles in the

Greek Epic. New York.() ‘The Ionian epic tradition: Was there an Aeolic phase in its development?’,

Minos 20–2: –.() ‘Homer’s dialect’, in Morris and Powell (): –.

Hunter, R. L. (ed.) () The Hesiodic Catalogue of Women: Constructions andReconstructions. Cambridge.

Hupe, J. (ed.) () Der Achilleus-Kult im nordlichen Schwarzmeerraum vomBeginn der griechischen Kolonisation bis in die romische Kaiserzeit. Leidorf.

Huxley, G. L. () Greek Epic Poetry. London.Jacobsen, T. () The Harps that Once . . . : Sumerian Poetry in Translation. New

Haven and London.Jacoby, F. (ed.) (–) Die Fragmente der griechischen Historiker ( vols.). Berlin

and Leiden.Jameson, M. H. and I. Malkin () ‘Latinos and the Greeks’, Athenaeum 86.2:

–.Janko, R. () ‘The use of ��, �� and �� in Homer’, Glotta 57: –.

() Homer, Hesiod, and the Hymns: Diachronic Development in Epic Diction.Cambridge and New York.

() ‘P. Oxy. : Hesiod’s Catalogue on the death of Actaeon’, Phoenix 38:–.

(a), review of West (), CPh 81: –.



Bibliography

(b) ‘The Shield of Heracles and the legend of Cycnus’, CQ 36: –.() The Iliad: a Commentary, vol. IV. Cambridge.() ‘The Homeric poems as oral dictated texts’, CQ 48: –.(), review of West (), CR n.s. 50: –.() ‘The Derveni papyrus (Diagoras of Melos, Apopyrgizontes Logoi?): a new

translation’, CPh 96: –.(), review of West (), CW 97: –.

Jens, W. () ‘Die Dolonie und ihr Dichter’, StudGen 8: –.Jones, P. V. () ‘Introduction’, in Homer: German Scholarship in Translation,

ed. G. M. Wright and P. V. Jones. Oxford: –.Jong, I. J. F. de () ‘Homeric words and speakers: an addendum’, JHS 108:

–.() A Narratological Commentary on the Odyssey. Cambridge.() Narrators and Focalizers: the Presentation of the Story in the Iliad. London.

st edn. , Amsterdam.() ‘Homer’, in Time in Ancient Greek Literature, ed. I. J. F. de Jong and

R. Nunlist. Studies in Ancient Greek Narrative . Leiden: –.Jorgensen, O. () ‘Das Auftreten der Gotter in den Buchern �–' der Odyssee’,

Hermes 39: –.Kahane, A. and M. Mueller (eds.) The Chicago Homer. www.library.northwestern.

edu/homer.Kakridis, J. T. () Homeric Researches. Lund. First published in Greek,

.() ‘Probleme der griechischen Heldensage’, Poetica 5: –.

Kannicht, R. (ed.) () Euripides: Helena, vol. II. Heidelberg.Kelly, A. () ‘Neoanalysis and the Nestorbedrangnis: a test case’, Hermes 134:

–.() A Referential Commentary and Lexicon to Homer, Iliad VIII. Oxford.

Kennedy, D. () ‘Virgilian epic’, in The Cambridge Companion to Virgil, ed.C. Martindale. Cambridge: –.

Kern, O. (ed.) () Orphicorum fragmenta. Berlin.Kirk, G. S. () The Songs of Homer. Cambridge.

() The Iliad: a Commentary, vol. I. Cambridge.() The Iliad: a Commentary, vol. II. Cambridge.

Kirscher, T. () ‘Die Bogenprobe’, Hermes 120: –.Kirsten, E. (), review of Welter () and Winterscheidt (), Gnomon 18:

–.Koenen, L. () ‘Greece, the Near East and Egypt: cyclic destruction in Hesiod

and the Catalogue of Women’, TAPhA 124: –.Kohnken, A. () ‘“Meilichos orga”: Liebesthematik und aktueller Sieg in der

Neunten Pythischen Ode Pindars’, in Pindare: huit exposes suivis de discussions,ed. A. Hurst. Entretiens pour l’etude de l’antiquite classique . Geneva: –.

Kolb, F. () ‘Die Bau-, Religions- und Kulturpolitik der Peisitratiden’, JDAI92: –.



Bibliography

Korenjak, M. () ‘Homerische Intertextualitat ohne Formeln? Zweiphorminxspielende Helden in Ilias und Odyssee’, MD 40: –.

Kouremenos, T., G. M. Parassoglou and K. Tsantsanoglou (eds.) () TheDerveni Papyrus. Florence.

Krarup, P. () ‘Verwendung von Abstracta in der direkten Rede bei Homer’,C&M 10: –.

Krischer, T. () Formale Konventionen der homerischen Epik. Munich.Kruger, A. () ‘Die orphische �3�� ’, Hermes 73: –.Kullmann, W. () Das Wirken der Gotter in der Ilias. Berlin.

() Die Quellen der Ilias (Troischer Sagenkreis). Hermes Einzelschriften .Wiesbaden.

(), review of Dihle (), Gnomon 49: –.() ‘Oral poetry theory and neoanalysis in Homeric research’, GRBS 25:

–, reprinted in Kullmann (): –.() ‘Ergebnisse der motivgeschichtlichen Forschung zu Homer (Neoanal-

yse)’, in Zweihundert Jahre Homer-Forschung, ed. J. Latacz. Stuttgart: –.Reprinted in Kullmann (), –.

() Homerische Motive, ed. R. J. Muller. Stuttgart.() ‘Festgehaltene Kenntnisse im Schiffskatalog und im Troerkatalog der

Ilias’, in Vermittlung und Tradierung von Wissen in der griechischen Kultur, ed.W. Kullmann and J. Althoff. ScriptOralia . Tubingen: –. Reprintedin Kullmann (b): –.

(), review of Latacz (), Gnomon 73: –.(a) ‘Homer und Kleinasien’, in Kullmann (b): –.(b) Realitat, Imagination und Theorie: Kleine Schriften zu Epos und Tragodie

in der Antike, ed. A. Rengakos. Stuttgart.(c) ‘The two Nekyiai of the Odyssey and their oral sources’, in Kullmann

(b): –.() ‘Ilias und Aithiopis’, Hermes 133: –.() ‘&�#'�� t4�� x�� (* –)’, in Contests and

Rewards in the Homeric Epics: Proceedings of the 10th International Symposiumof the Odyssey, ed. M. Paısi-Apostolopoulou, A. Rengakos and C. Tsagalis.Ithaca, Greece: –.

() ‘Poesie, Mythos und Realitat im Schiffskatalog der Ilias’, Hermes 137:–.

Laks, A. and G. W. Most (eds.) () Studies on the Derveni Papyrus. Oxford.Larson, S. () ‘Boiotia, Athens, the Peisistratids and the Odyssey’s catalogue of

heroines’, GRBS 41: –.Laser, S. () Sport und Spiel. Archaeologia Homerica T. Gottingen.Latacz, J. (ed.) () Homer: Tradition und Neuerung. Darmstadt.

() Troia und Homer: der Weg zur Losung eines alten Ratsels. Munich.Liberman, G. (ed.) () Alcee: Fragments, with translation ( vols.). Collection

des universites de France. Paris.Linforth, I. M. () The Arts of Orpheus. Berkeley and Los Angeles.Lloyd-Jones, H. () The Justice of Zeus, nd edn. Berkeley. st edn. .



Bibliography

Lobel, E. and D. L. Page () ‘A new fragment of Aeolic verse’, CQ 2: –.Lohmann, D. () Die Komposition der Reden in der Ilias. Berlin.Lord, A. B. () ‘Homer’s originality: oral dictated texts’, TAPhA 94: –,

reprinted in Latacz (): –.() The Singer of Tales. Cambridge, Mass.() Epic Singers and Oral Tradition. Ithaca, NY.() The Singer Resumes the Tale, ed. M. L. Lord. Ithaca, NY.

Lorimer, H. L. () Homer and the Monuments. London.Lowenstam, S. () ‘Talking vases: the relationship between the Homeric poems

and archaic representations of epic myth’, TAPhA 127: –.Macleod, C. W. (ed.) () Homer: Iliad, Book XXIV, with introduction and

commentary. Cambridge Greek and Latin Classics. Cambridge.Malkin, I. () The Returns of Odysseus: Colonization and Ethnicity. Berkeley.Mallwitz, A. () Bericht uber die Ausgrabungen in Olympia, vol. XI. Berlin and

New York.Malten, L. (a) ‘Altorphische Demetersage’, ARW 12: –.

(b) ‘Der Raub der Kore’, ARW 12: –.Marjanovic, L. (ed.) (–) Hrvatske narodne pjesme, vols III and IV. Zagreb.Martin, R. () The Language of Heroes: Speech and Performance in the Iliad.

Ithaca, NY.() ‘Hesiod’s metanastic poetics’, Ramus 21: –.() ‘Pulp epic: the Catalogue and the Shield ’, in Hunter (): –.

Martindale, C. () Redeeming the Text: Latin Poetry and the Hermeneutics ofReception. Cambridge.

Mastronarde, D. J. (ed.) () Euripides: Medea, with introduction and com-mentary. Cambridge Greek and Latin Classics. Cambridge.

Mededovic, A. (a) Serbo-Croatian Heroic Songs, vol. III, coll. M. Parry, ed.D. E. Bynum, with translation, introduction, notes and commentary byA. B. Lord. Cambridge, Mass.

(b) Serbo-Croatian Heroic Songs, vol. IV, coll. M. Parry, eds. D. E. Bynumand A. B. Lord, with translation, notes and commentary. Cambridge,Mass.

() Serbo-Croatian Heroic Songs, vol. VI, coll. M. Parry, ed. D. E. Bynum,with introduction and notes. Cambridge, Mass.

Meier, W. D. () Die epische Formel im pseudohesiodeischen Frauenkatalog: eineUntersuchung zum nachhomerischen Formelgebrauch. Zurich.

Meillet, A. () Apercu d’une histoire de la langue greque. Paris.Meister, K. () Die homerische Kunstsprache. Leipzig. Reprinted in Stuttgart,

.Meliado, C. () ‘Un nuovo frammento esiodeo in uno scolio a Teocrito’, ZPE

145: –.Mendez Dosuna, J. () ‘Metatesis de cantidad en jonico-atico y heracleota’,

Emerita 61: –.Meyerhoff, D. () Traditioneller Stoff und individuelle Gestaltung: Untersuchun-

gen zu Alkaios und Sappho. Hildesheim, Zurich and New York.



Bibliography

Monro, D. B. () A Grammar of the Homeric Dialect. Oxford.Montanari, F. (ed.) () Omero tremila anni dopo. Rome.

() ‘The case of Book Ten and the unity of the Iliad plot in ancient schol-arship’, Trends in Classics 2: –.

Moretti, L. () Olympionikai: i vincitori negli antichi agoni olimpici. Rome.Morgan, Catherine () Athletes and Oracles: The Transformation of Olympian

and Delphi in the Eighth Century B.C. Cambridge.Morpurgo Davies, A. () ‘Mycenaean and Greek language’, in Linear B: A 1984

Survey, ed. A. Morpurgo Davies and Y. Duhoux. Louvain and La-Neuve:–.

Morris, I. () ‘The use and abuse of Homer’, ClAnt 5: –, reprinted inCairns (b), –.

() ‘Homer and the Iron Age’, in Morris and Powell (): –.Morris, I. and B. B. Powell (eds.) () A New Companion to Homer. Leiden.Most, G. W. (ed.) () Hesiod: Theogony; Works and Days; Testimonia, with

introduction and translation. Loeb Classical Library. Cambridge, Mass. andLondon.

Nagy, G. () The Best of the Achaeans. Baltimore.() Pindar’s Homer. Baltimore.(a) Homeric Questions. Austin, Texas.(b) Poetry as Performance: Homer and Beyond. Cambridge.() Homeric Responses. Austin, Texas.() Homer’s Text and Language. Champaign, Ill.

Nelis, D. () ‘The reading of Orpheus: the Orphic Argonautica and the epictradition’, in Roman and Greek Imperial Epic, ed. M. Paschalis. Herakleion:–.

Nickau, K. () Untersuchungen zur textkritischen Methode des Zenodotos vonEphesos. Berlin and New York.

Niemeier, W.-D. () ‘Aegina – first Aegean “state” outside of Crete?’, in Politeia:Society and State in the Aegean Bronze Age, ed. R. Laffineur and W.-D.Niemeier. Aegeum . Liege: –.

Niese, B. () Der homerische Schiffskatalog als historische Quelle betrachtet. Kiel.Nilsson, M. P. () ‘Gotter und Psychologie bei Homer’, ARW 22: –.Notopoulos, J. A. () ‘Studies in early greek oral poetry’, HSPh 68: –.Nunlist, R. () ‘Some clarifying remarks on “focalization”’, in Montanari

(): –.() ‘The Homeric scholia on focalization’, Mnemosyne 56: –.

Obbink, D. () ‘Cosmology as initiation vs. the critique of the orphic mysteries’,in Laks and Most (): –.

() ‘. Archilochus: Elegies (more of VI and XXX )’, TheOxyrhynchus Papyri 69: –.

Ogden, D. () Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds:a Sourcebook. Oxford.

Okhotnikov, S. B. and A. S. Ostroverkhov () Svyatilishche Akhilla na OstroveLevke (Zmeinom). Kiev. Russian with brief English summary.



Bibliography

Olson, S. D. () Blood and Iron: Stories and Storytelling in Homer’s Odyssey.Leiden.

Osborne, R. () Greece in the Making, 1200–479 BC. London.() ‘Early Greek colonization? The nature of Greek settlement on the West’,

in Archaic Greece: New Approaches and New Evidence, ed. N. Fisher andH. van Wees. London: –.

() ‘Ordering women in Hesiod’s Catalogue’, in Hunter (): –.Oswald, R. () Das Ende der Odyssee: Studien zu Strukturen epischen Gestaltens.

Graz.Page, D. L. (a) The Homeric Odyssey: the Mary Flexner Lectures Delivered at

Bryn Mawr College, Pennsylvania. Oxford.(b) Sappho and Alcaeus. Oxford.() ‘The Sources of the Iliad ’, review of Kullmann (), CR n.s. 11:

–.Paısi-Apostolopoulou, M. (ed.) () Homerica: Proceedings of the 8th International

Symposium on the Odyssey (1–5 September 1996). Ithaca, Greece.Parker, R. C. T. () ‘The Hymn to Demeter and the Homeric Hymns’, G&R 38:

–.() Athenian Religion: A History. Oxford.

Parry, A. A. () Blameless Aegisthus: A Study of �� and Other HomericEpithets. Mnemosyne Suppl. . Leiden.

Parry, A. M. () ‘Have we Homer’s Iliad?’, YClS 20: –, reprinted in Parry(), –.

(a) ‘Introduction’, in A. M. Parry (b): ix–lxii.(ed.) (b) The Making of Homeric Verse: the Collected Papers of Milman Parry.

Oxford.() The Language of Achilles and Other Papers. Oxford.

Parry, M. () ‘Studies in the epic technique of oral verse-making: II. TheHomeric language as the language of poetry’, HSPh 43: –.

Pavese, C. O. () Tradizione e generi poetici della Grecia arcaica. Rome.() Studi sulla tradizione epica rapsodica. Rome.

Peersman, J. () ‘Plisthene et sa famille dans les genealogies de la traditionhesiodique’, AC 62: –.

Pelliccia, H. () ‘Two points about rhapsodes’, in Finkelberg and Stroumsa(): –.

Penglase, R. () Greek Myths and Mesopotamia: Parallels and Influence in theHomeric Hymns and Hesiod. London and New York.

Pestalozzi, H. () Die Achilleis als Quelle der Ilias. Zurich.Peters, M. () Untersuchungen zur Vertretung der indogermanischen Laryngale

im Griechischen. Vienna.() ‘Zur Frage einer “achaischen” Phase des griechischen Epos’, in O-o-pe-

ro-si: Festschrift fur Ernst Risch zum 75. Geburtstag, ed. A. Etter. Berlin andNew York: –.

() ‘Beitrage zur Griechischen Etymologie’, in Miscellanea Linguistica Graeco-Latina, ed. L. Isebaert. Collection d’etudes classiques . Namur: –.



Bibliography

Pierson, W. () ‘Uber die Tmesis der Praposition vom Verbum bei den griechis-chen Dichtern, insbesondere bei Dramatikern und Lyrikern’, RhM 11: –,–, –.

Plassart, M. A. () ‘Inscriptions de Delphes: la liste des Theodoriques’, BCH45: –.

Powell, B. B. () review of Nagy (b), BMCRev ... bmcr.brynmawr.edu//...html [accessed September ].

Powell, J. U. (ed.) () Collectanea Alexandrina. Oxford.Prinz, F. () ‘Herakles’, in Wissowa et al. (–), suppl. : –.Privitera, G. A. () Il ritorno del guerriero. Turin.Pucci, P. () Odysseus Polytropos: Intertextual Readings in the Odyssey and the

Iliad. Ithaca, NY and London.Reece, S. () ‘The Cretan Odyssey: a lie truer than the truth’, AJPh 115: –.Reichel, M. and A. Rengakos (eds.) () Epea pteroenta: Beitrage zur Homer-

forschung. Festschrift fur Wolfgang Kullmann zum 75. Geburtstag. Stuttgart.Reinhardt, K. () Die Ilias und ihr Dichter. Gottingen.Rengakos, A. () ‘Zeit und Gleichzeitigkeit in den homerischen Epen’, A&A

40: –.() ‘Zur Zeitstruktur der Odyssee’, WS 111: –.() ‘Narrativitat, Intertextualitat, Selbstreferentialitat: die neue Deutung der

Odyssee’, in Reichel and Rengakos (): –.Richardson, N. J. () The Homeric Hymn to Demeter. Oxford.

() The Iliad: a Commentary vol. VI. Cambridge.Risch, E. () Wortbildung der homerischen Sprache, nd edn. Berlin and New

York.Rix, H. et al. () Lexikon der indogermanischen Verben, nd edn. Wiesbaden.Robert, C. () Studien zur Ilias. Berlin.

() Die griechische Heldensage, vol. III.. Griechische Mythologie , th edn.Berlin.

Rohde, E. () Kleine Schriften ( vols.). Tubingen and Leipzig.Romilly, J. de () Perspectives actuelles sur l’epopee homerique. Paris.Ruijgh, C. J. () L’element acheen de la langue epique. Assen.

() ‘D’Homere aux origines proto-myceniennes de la tradition epique’, inHomeric Questions, ed. J. P. Crielaard. Amsterdam: –.

() ‘La date de la creation de l’alphabet grec et celle de l’epopee homerique’,BO 54: –.

Russell, D. A. F. () ‘De imitatione’, in Creative Imitation and Latin Literature,ed. D. West and T. Woodman. Cambridge: –.

Russo, J. A. () ‘Is “oral” or “aural” composition the cause of Homer’s formulaicstyle?’, in Oral Literature and the Formula, ed. B. A. Stolz and R. S. Shannon.Ann Arbor: –.

Rutherford, I. C. () ‘Catalogues of women: formulas, voice and death inEhoie-poetry, the Hesiodic Gunaikon Katalogos and the Odysseian Nekuia’,in Matrices of Genre: Authors, Canons, and Society, ed. M. Depew and D.Obbink. Center for Hellenic Studies Colloquia . Cambridge, Mass.: –.



Bibliography

() ‘Hesiod and the Literary Traditions of the Near East’, in Brill’s Com-panion to Hesiod, ed. F. Montanari, A. Rengakos and C. Tsagalis. Leiden:–.

Rutherford, R. B. (ed.) () Homer: Odyssey, Books XIX and XX, with introductionand commentary. Cambridge Greek and Latin Classics. Cambridge.

() Homer. Greece and Rome: New Surveys in the Classics . Oxford.() ‘From the Iliad to the Odyssey’, in Cairns (b): –.

Sale, W. M. () ‘The Trojans, statistics, and Milman Parry’, GRBS 30: –.

Sauge, A. () Iliade: langue, recit, ecriture. Bern.Scarpi, P. () Le Religioni dei misteri, vol. I. Milan.Schadewaldt, W. () ‘Einblick in die Erfindung der Ilias: Ilias und Memnonis’,

in Festschrift fur Karl Reinhardt. Weimar: –. Reprinted in Schadewaldt(), –.

() Von Homers Welt und Werk: Aufsatze und Auslegungen zur homerischenFrage, th edn. Stuttgart.

Schein, S. L. () The Mortal Hero. Berkeley.() ‘Homeric intertextuality: two examples’, in Euphrosyne, ed. A. Rengakos

and J. M. Kazazis. Stuttgart: –.() ‘Mythological allusion in the Odyssey’, in Montanari (): –.

Scherer, B. () Mythos, Katalog und Prophezeiung: Studien zu den Argonautikades Apollonios Rhodios. Stuttgart.

Schischwani, S. () ‘Mundliche Eurytossage in der Odyssee’, in �� :�� !� �� , (3–8 ��"��1993), ed. M. Paısi-Apostolopoulou. Ithaca, Greece: –.

Schlange-Schoningen, H. () ‘@E�3�, ��1 ��: Odysseus, Aias undPalamedes’, in Geschichte und Fiktion in der homerischen Odyssee, ed.A. Luther. Zetemata . Munich: –.

Schmitz, T. A. () ‘Ist die Odyssee “spannend”? Anmerkungen zurErzahltechnik des homerischen Epos’, Philologus 138: –.

() Modern Literary Theory and Ancient Texts: an Introduction. Oxford.Schoeck, G. () Ilias und Aithiopis. Zurich.Schroeder, C. M. () ‘A new monograph of Aristarchus?’, JHS 127: –.Scodel, R. () ‘Pseudo-intimacy and the prior knowledge of the Homeric

audience’, Arethusa 30: –.() Listening to Homer: Tradition, Narrative, and Audience. Ann Arbor.

Scott, J. A. () ‘Repeated verses in Homer’, AJPh 32: –.Seaford, R. () Reciprocity and Ritual: Homer and Tragedy in the Developing

City-State. Oxford.Semsovic, S. () ‘Smailagic Meho: usporedna analiza dviju pjesama istog

naslova’ [Smailagic Meho: a comparative analysis of two songs with thesame title], Almanah 31–2: –.

Severyns, A. () Le cycle epique dans l’ecole d’Aristarque. Liege.Shipp, G. P. () Studies in the Language of Homer, nd edn. Cambridge. st

edn. .



Bibliography

Sinn, U. () ‘Die Stellung der Wettkampfe im Kult des Zeus Olympios’,Nikephoros 4: –.

Slatkin, L. M. () The Power of Thetis. Berkeley.Snodgrass, A. () ‘Homer and Greek art’, in Morris and Powell (): –.

() Homer and the Artists. Cambridge.Steinruck, M. () Rede und Kontext. Zum Verhaltnis von Person und Erzahler in

fruhgriechischen Texten. Bonn.() ‘Comment faire l’eloge d’une femme’, Metis 11: –.

Stoessl, F. () Der Tod des Herakles. Zurich.Strunk, K. () ‘Die sogenannten Aolismen der homerischen Sprache’, disserta-

tion, Cologne.Stuber, K. () Zum dialektalen Einheit des Ostionischen. Innsbruck.Szemerenyi, O. () ‘The genitive singular of masculine -a-stem nouns in Greek’,

Glotta 35: –.(), review of Ruijgh () and Strunk (), JHS 79: –.

Taplin, O. () ‘The earliest quotation of the Iliad?’, in Owls to Athens: Essayson Classical Subjects Presented to Sir Kenneth Dover, ed. E. M. Craik. Oxford:–.

() Homeric Soundings: the Shaping of the Iliad. Oxford.Thalmann, W. G. () The Swineherd and the Bow. Ithaca, NY and London.Theiler, W. () ‘Ilias und Odyssee in der Verflechtung ihres Entstehens’, MH

19: –.Thomas, R. F. () ‘Virgil’s Georgics and the art of reference’, HSPh 90: –.Thompson, S. (–) Motif Index of Folk Literature: a Classification of Narrative

Elements in Folktales, Ballads, Myths, Fables, Mediaeval Romances, Exempla,Fabliaux, Jest-Books and Local Legends ( vols.), nd English edn. Copen-hagen. st edn. –, Helsinki.

Thornton, A. () People and Themes in Homer’s Odyssey. Dunedin.Tracy, S. V. () The Story of the Odyssey. Princeton.Treu, M. (ed.) () Sappho: Lieder, with translation, th edn. Munich.Trypanis, C. A. () ‘�� _�� !" �� 7�� 7�� .� �'��

A:'��’, Athena 48: –.Tsagalis, C. () The Oral Palimpsest: Exploring Intertextuality in the Homeric

Epics. Harvard and Cambridge, Mass.Tsagarakis, O. () Studies in Odyssey 11. Hermes Einzelschriften . Stuttgart.Usener, K. () Beobachtungen zum Verhaltnis der Odyssee zur Ilias. Tubingen.Valk, M. H. Van Der () Textual Criticism of the Odyssey. Leiden.Vian, F. () Les Argonautiques Orphiques. Paris.Visser, E. () Homerische Versifikationstechnik: Versuch einer Rekonstruktion.

Frankfurt am Main.() Homers Katalog der Schiffe. Stuttgart and Leipzig.

Von der Muhll, P. () Kritisches Hypomnema zur Ilias. Basel.Wachter, R. () ‘Grammatik der homerischen Sprache’, in Homers Ilias:

Gesamtkommentar. Prolegomena, ed. J. Latacz. Munich and Leipzig: –.() Non-Attic Greek Vase-Inscriptions. Oxford.



Bibliography

() ‘Greek dialects and epic poetry: Did Homer have to be an Ionian?’, in#��$ �� %&��$: Actes du Ve congres internationale de dialectologiegrecque, Athenes, 28–30 septembre 2006, ed. M. Hatzopoulos. Athens and Paris:–.

Wackernagel, J. () Sprachliche Untersuchungen zu Homer. Gottingen.() Vorlesungen uber Syntax, mit besonderer Berucksichtung von Griechisch,

Lateinisch und Deutsch, vol. II, nd edn. Basel.Walcot, P. () Hesiod and the Near East. Cardiff.Wallace, P. W. () Strabo’s Description of Boeotia: a Commentary. Heidelberg.Wathelet, P. () Les traits eoliens dans la langue de l’epopee grecque. Rome.

() ‘Thebes de Boetie, vue par Homere et par Eschyle ou le reflet dedeux sensibilites differentes’, in Serta Leodiensia secunda: melanges publiespar les Classiques de Liege a l’occasion du 175. anniversaire de l’Universite, ed.A. Bodson, P. Wathelet and M. Dubuisson. Liege: –.

() ‘Les phases dialectales de l’epopee grecques et l’apport de l’eolien’, Eikas-mos 14: –.

Webster, T. B. L. () ‘Early and late in Homeric diction’, Eranos 54: –.Wees, H. van () ‘The Homeric way of war: the Iliad and the hoplite phalanx’,

G&R 41: –, –.() War and Violence in Ancient Greece. London and Swansea.

Welter, G. () Aigina. Berlin.West, M. L. () ‘Alcmanica’, CQ 15: –.

(ed.) () Hesiod: Theogony. Oxford.() ‘Cynaethus’ Hymn to Apollo’, CQ 25: –.(ed.) () Hesiod: Works and Days, with introduction and commentary.

Oxford.() The Orphic Poems. Oxford.() The Hesiodic Catalogue of Women: Its Nature, Structure, and Origins.

Oxford.() ‘The rise of Greek epic’, JHS 108: –.(a) ‘Alcman and the Spartan royalty’, ZPE 91: –.(b) ‘The descent of Greek epic: a reply’, JHS 112: –.() ‘The date of the Iliad ’, MH 52: –.(ed.) (–) Homeri Ilias ( vols.). Stuttgart, Leipzig and Munich.() ‘The invention of Homer’, CQ 49: –.(a) Studies in the Text and Transmission of the Iliad. Munich and Leipzig.(b) ‘The fragmentary Homeric Hymn to Dionysus’, ZPE 134: –.(a) ‘“Eumelos”: a Corinthian epic cycle?’, JHS 122: –.(b) ‘The view from Lesbos’, in Reichel and Rengakos (): –.(ed.) (a) Greek Epic Fragments: From the Seventh to the Fifth Centuries BC,

with introduction and translation. Loeb Classical Library. Cambridge, Mass.and London.

(ed.) (b) Homeric Hymns; Homeric Apocrypha; Lives of Homer, with intro-duction and translation. Loeb Classical Library. Cambridge, Mass. andLondon.



Bibliography

(c) ‘Iliad and Aethiopis’, CQ 53: –.() ‘Odyssey and Argonautica’, CQ 55: –.() ‘Phasis and Aia’, MH 64: –.

West, S. (a) ‘An alternative nostos for Odysseus’, LCM 6: –.(ed.) (b) Omero: Odissea, vol. I, with commentary. Milan.() ‘Laertes revisited’, PCPhS 215: –.() ‘Phoenix’s antecedents: a note on Iliad ’, SCI 20: –.() ‘Terminal problems’, in Hesperos: Studies in Ancient Greek Poetry Pre-

sented to M. L. West on his Seventieth Birthday, eds. P. J. Finglass, C. Collardand N. J. Richardson. Oxford: –.

Wilamowitz-Moellendorff, U. von () ‘Hephaistos’, Nachrichten von derKoniglichen Gesellschaft der Wissenschaften zu Gottingen (phil.-hist. Klasse): –, reprinted in Wilamowitz-Moellendorff (), –.

() Sappho und Simonides. Berlin.() Der Glaube der Hellenen, vol. II. Berlin.() Kleine Schriften vol. V.. Berlin.

Willcock, M. M. (ed.) (–) The Iliad of Homer: Books I–XXIV, with intro-duction and commentary ( vols.). Basingstoke.

() ‘Neoanalysis’, in Morris and Powell (): –.Winterscheidt, H. () Aigina: eine Untersuchung uber seine Gesellschaft und

Wirtschaft. Wurzburg.Wissowa, G., W. Kroll, K. Mittelhaus and K. Ziegler (eds.) (–) Paulys Realen-

cyclopadie der classischen Altertumswissenschaft. Stuttgart.Witte, K. (a) ‘Homeros’, in Wissowa et al. (–), vol. : –.

(b) ‘Zur Flexion homerischer Formeln’, Glotta 3: –.Wolf, F. A. () Prolegomena ad Homerum. Halle.

() Prolegomena to Homer, 1795, tr. A. Grafton, G. W. Most and J. E. G.Zetzel, with introduction and notes. Princeton.

Woodhouse, W. J. () The Composition of Homer’s Odyssey. Oxford.Wyatt, N. () Religious Texts from Ugarit: The Words of Ilimilku and his

Colleagues, nd edn. London.Wyatt, W. F. () ‘Short accusative plurals in Greek’, TAPhA 97: –.

() Metrical Lengthening in Homer. Rome.() ‘Homer’s linguistic ancestors’, EEThess 14: –.() ‘Homer’s linguistic forebears’, JHS 112: –.

Zielinski, T. () Die Behandlung gleichzeitiger Ereignisse im antiken Epos.Philologus Suppl. : –.



General index

Abantes

absolute dating , , , –, ,

accusativea-stem plural , ,

o-stem plural , ,

Acesander

Achilleis

Achilles , , –, , , , , ,–, –, , , , , , ,, , , , ,

and horsemanship –

Acrisius

Actaeon

Actor

Admetus

Aeacidae

Aeacus

Aegina

Aegisthus ,

Aeneas ,

Aeolic dialect –, –, –,

Aeolic phase , , , –, , –, ,

Aeolisms –, ,

Aeolus , ,

Aetolians

Agamemnon –, , , , , , ,, , , , , , , , ,,

Agapenor , ,

Aithiopes

Aethiopis , , , , , , , , ,

and Homeric poems

and Iliad

Ajax , –, , , –, , , , ,

Ajax the Locrian , , ,

Alcaeus , , , , , ,

Alcestis

Alcinous

Alcman ,

dating of

Alcmena

Aleus

Alexandrian scholarship –, ,

allusion , , , , , , , ,–, , , , –, –, ,,

alphabet ,

Amazons

[Ammonius]

Amorgos

Amphiaraus

Amphidamas

Amphimedon

Amphion

analysis , , , , , , , ,

Ancaeus

Anderson, M. J.

Andromache , , ,

Anticlea , , ,

Antilochus

Antimachus of Colophon

Antinous

Antiphus , ,

Aphrodite

Apollo , , , , , , , , ,,

Apollodorus , , , , ,

Apollonius of Rhodes

Arcadians , ,

Arcado-Cypriote dialect

Arcesilaus , ,

archaisms , , , , , , , –, , ,, , ,

false –,

Archilochus , , , , , ,

Arctinus of Miletus

Ares ,

Arete

Argia


General index

Argonautica ,

dating of

and Odyssey –

Argonauts –,

Aristaeus

Aristarchus , , , , , ,

Aristophanes (comic playwright) ,

Aristophanes (grammarian) ,

arming scenes ,

Artemis , ,

Ascalaphus ,

Asopids

Assurbanipal

Asterion

Astyanax , –

Astyoche

Atalante

Athena , , , , , , , ,

Athenians

Atlantids

Atreidae ,

Atreus

Attic dialect ,

Attica

audiences , , , , , , , –,, , , , , , ,

narrative/authorial ,

Auge

Augeas ,

Aulis , ,

authorial self-reference

Autolycus

Avdo Mededovic , –

Babylon

Bacchylides

bards , ,

Batrachomyomachia

Battiads

Baubo

Bellerophon ,

Bernabe, A.

Boeotia

Boeotians , , ,

book production –

Burgess, J. , –, ,

Cadmeia

Caeneus

Caicus, River

Calchas ,

Calydonian Boar Hunt –

Calypso

Cassandra ,

Catalogue of Ships ,

Boeotian leaders in –

and Iliad , , –

list of cities in –

list of heroes in –

Catalogue of Women , –, , , ,

authorship of

content of –

dating of –

and Hesiodic poems

and Homeric Hymn to Apollo

and Homeric poems –

and Iliad –

and Odyssey , –,

and other early epic

and Shield of Heracles

and Thebais

and Theogony –

and Trojan Cycle –

writing down of

Cebriones –

Centaurs ,

Cephallenians

Cephisian Lake

Cephisus, River

Cerberus ,

Chalcis

Chamaeleon

Chantraine, P. ,

Charis

Cheiron

Chiron

Chloris , ,

Circe , , , ,

Clay, J. S.

Clement of Alexandria

Clonius ,

Clytaemestra ,

Colakovic, Z.

contraction , –, –,

Copais, Lake ,

Corinth

crasis

Crete , ,

Cretheus ,

Crisa

Crissy, K.

Currie, B.

Cycle, Epic , , , –, , ; see alsoAithiopis, Cypria, Epigonoi, Ilias Mikra,Iliou Persis, Nostoi, Telegony, Thebais

Cyphos

Cypria , , , , , , , , ,, ,

dating of

and Iliad

Cyrene –, ,


General index

D’Alessio, G. B.

Danae

Danek, G. , , , , , , , ,,

��

Danube

Dardanus

dativea-stem plural ,

o-stem plural ,

plural in -� �

Deıon

Deiphobus

Delos

Delphi

Demeter , , , –, , , –

Demetrius of Phalerum

Demodocus , , ,

Demophoon

Deucalion

A�-/;��- , , , –

diektasis –

Dieuchidas

diffusion model , –, , –

digamma , , , , , –, ,

Diomedes , , , –, , , , ,, , , ,

Dionysus , , , , ,

Dioscuri

Dodds, E. R.

Dodecanese ,

Dodone

Dolicha

Dolon , ,

Doloneia , , , , , , , ,

arming scenes in

authorship of , ,

dating of ,

formulas in –

and Iliad , , ,

and Odyssey

speeches in

statistical analysis of

and South Slavic epics –

treatment of time in –

Dorion

Doulichion ,

dual, athematic

Echinads

Edwards, G. P. ,

Edwards, M.

Egypt , ,

Elephenor

Eleusinians , ,

Eleusis

Elis , ,

Elone

Elpenor

Elysium

M'�

Enienes

Enipeus, River ,

Ennius

Epeians , ,

Epeius

Ephorus from Cyme

epic dialect –, –, –, , , , ,

epic poetryauthorship of

Epigoni

Epigonoi

Epistrophus , ,

Eratosthenes , ,

Eriphyle

7+ , ,

�# for ��

Euboea ,

Euchenor

Eugammon ,

Eumaeus , ,

Eumelus , ,

Eumolpus

Euneos

Euripides

Europa ,

Europos

Euryalus ,

Eurylochus

Eurypylus , ,

Eurystheus

Eurytus –

Eutresis

external prolepsis , ,

feminist literary criticism

Festus

Forest Lake

formulas , , , , , , , , , –,, , –, , , , , ,

Fowler, R.

40/4� ,

Garbrah, K. A.

genitive –,

a-stem , , , –

a-stem plural , , ,

a-stem singular , , , , , ,

o-stem singular , ,

genre ,


General index

Gerania

Gerenoi

Glaucus , ,

Gonnoi

Griffin, J. ,

Guneus ,

guslar

Hades , , , , , ,

Hainsworth, B. ,

Haliartus

Hall, J.

Halos

Hebe ,

Hecataeus

Hecate , , , , ,

Hector , , , , , , , ,

Helen , , , , , , , , ,,

Helios , , ,

Hephaestus , , ,

Hera , , , , ,

Heracleis

Heracles , , , , , , , , ,

and Eurytus –

in Iliad

and Iphitus –

in Nekuia –

Heraclides Ponticus

Heraclids

Hermes , , , , ,

phonology of –

Hermione

Herodotus , , , ,

Hesiod , ,

life of –

Hesiodic poems , , ; see also Theogony,Works and Days

dating of –

and Homeric Hymn to Demeter

and Iliad

Hinrichs, G.

Hipparchus

Hippodamia

Hippomenes

Hoekstra, A. , , , ,

Homerand Hesiod ,

identity of

life of –

Homeric Hymn to Aphrodite ,

dating of ,

and Homeric Hymn to Demeter

and Odyssey

Homeric Hymn to Apollo , , ,

dating of

Homeric Hymn to Demeter , , , ,

authorship of

dating of ,

and P.Berol. –

self-consciousness in

and Theogony

Homeric Hymn to Dionysusdating of –

Homeric Hymn to Hermes ,

dating of

Homeric Hymns ,

Homeric poems ; see also Iliad, Odysseyauthorship of ,

composition of

and Epic Cycle , , –,

and Hesiodic poems , , , –

moral judgement in

writing down of , –,

Homeric society , ,

Hyle

Hypereian spring

Hypnos

Hypothebai

Hypsipyle

Ialmenus ,

Idomeneus ,

Iliad ; see also Homeric poems, Doloneia,Catalogue of Ships

authorship of

book

dating of –

and early elegy

and Odyssey , , , , , –, , ,, –, ,

Nestorbedrangnis

Teichoskopia

Ilias Mikra , , , ,

dating of

and Homeric poems –

and Odyssey

Iliou Persis , , ,

dating of

and Odyssey –

imitation –,

imperfect tense –

Inachids

innovations , , , , –, , , , ,–, –, , , , ,

interpolations , , , , , ,

intertextuality , , , , –, , –,, , , , ,

in Catalogue of Women –


General index

Iole ,

Ionian Isles

Ionic dialect , –, , , ,

Ionic gap ,

Iphiclus ,

Iphigenia , ,

Iphimede

Iphitus , , , –,

Isles of the Blessed

Isocrates

Janko, R. , –, , , , , , , ,–

Jason ,

Jones, B. –

[Justin Martyr]

��

Kirk, G. S. ,

Kolakovic, M.

Krischer, T.

Kullmann, W. , ,

Laodamas

Laomedon

Lapiths , ,

Leda

Leitus ,

Lelantine War , ,

Lemnos ,

Leonteus , ,

Lesbian poetsdating of

and Ionian epic –

Lesbos

Leuce

Leucon

Libya

Licymnius

Lilaia

Linear B , ,

linguistic flexibility –

literacy

Lloyd-Jones, H.

[Longinus] ,

Lord, A. B. , ,

Lucian

Lycia

Lycurgus

Lydians

Magnetes

Malea, Cape

Maroneia

Marsyas

Medon

Megalai Ehoiai , ,

'�4� '�4�� , –, –, ,

Meges , ,

Meıonia

Meister, K. ,

Melampus , ,

Melaneus

Meleager ,

Meliboea

Melissa

Memnon , ,

Memnonis , , , ,

dating of

Menelaus , –, , , , , , ,, , , , ,

Menestheus

Meones

Meriones ,

Messene

Messenia

Methone

metrical constraints , –, ,

Miletus

Mimnermus ,

Minos

Minyans

Minyas

model text , , , , , , ,

Monro’s Law

Moschus

motif alienation , ,

motif transference , , –, , –,, ,

multiformity ,

Musaeus

Mycenae

Mycenaean dialect ,

Myconos

Mysia , , ,

Nagy, G.

narrative layers

Nauplius

�� ,

�� –

Nekuia –, , –, ,

strata in

Neleus , , ,

neoanalysis –, , , , , , ,, , , –

and oral theory –, –

Neoptolemus , , , , , ,

neo-unitarianism

Nereids , , ,


General index

Nestor , , , , , , , , ,,

Nestor’s cup

Nikandre kore

Nile

Nile, River

Niobe

Nireus

n-mobile

Nonnus

Nostoi , , ,

dating of

and Odyssey

Ocalea

Odysseus , –, , , , , ,

as bowman –

and Heracles

Odyssey ; see also Homeric poems, Nekuiabooks – –

content of –

dating of

end of , , ,

influences on

and other early epic –

second Nekuia ,

Oedipus

Oeneus

Oichalia , , , , , ,

Olbia

Olizon

Olympia

Onchestus

oral poetry , , –, , , , , , ,, –, , , , , , ,

and innovation

and writing ,

oral theory –, , , , , , , ,

Orchomenus ,

Oresbius

Orestes ,

Orion

Ormenion

Orpheus , , , ,

Orphic Hymn to Demeter

Ortilochus ,

Otus

Page, D.

Palamedes ,

Pamphos

Pandora

Panopeus

Parian Marble

Paris ,

Parry, A. M.

Parry, M. , , , , , ,

Patroclus , , , , , ,

Pausanias , , , ,

Peleus , ,

Pelias , ,

Pelion, Mount , ,

Pelopidae ,

Peneios, River ,

Peneleos ,

Penelope , , , , , , , ,, , , , ,

Penthesilea ,

Peraebi

Periander

Periclymenus

Perimedes

Pero , ,

Perrhaebians

Persephone , , , , , , , ,

Perseus

��+3�'� –

Phaeacians

Phasis, River

Phemius , , , ,

Pherae ,

Phidippus , ,

Philochorus

Philoctetes , , , , , , ,

Philodemus

Philostratus

Phocians ,

Phoenicia

Phoenix

Phrygia ,

Phyleus ,

Pindar ,

Pirithous ,

Pisistratean redaction ,

Pisistratus ,

Plautus

Pleisthenes

Plutarch

Podarces ,

Poeas

Polycaste ,

Polycrates

Polydamas

Polyidus

Polyphemus ,

Polypoetes , ,

Polyxenus


General index

Pontos

Porthaon

Poseidon , , , , , , ,

phonology of –

��/�� ,

prevocalic vowel-shortening , ,

Priam , ,

Proclus , , , , , , ,

Prometheus

prosodic habit –

Protesilaus , ,

Proteus

Prothoenor ,

Prothous ,

Psammetichus

Pylians

Pylos , , ,

quantitative metathesis , , , –, , ,

Quintus of Smyrna

Rarian plain

relative chronology , , –,

ancient testimony to –

historical approach to –

linguistic approach to , –, , , –

literary approach to , –, , , –,–

pictorial evidence of –

and written texts ,

Rhadamanthus

rhapsodes , ,

Rhea ,

Rhesus

Rhodes

Ruijgh, C. J.

Russo, J. ,

Sack of Oichalia –

Sale, W. M.

Salmoneus , ,

Samothrace

Sappho ,

Sarpedon , –,

Saussure, F.

Scamander

Schedius ,

Schein, S. L. , ,

Schischwani, S.

Scodel, R.

Scyros ,

Semonides ,

dating of

Sennacherib

Shakespeare

Shield of Heracles , , ,

Shipp, G. P. , , , ,

Sidon

Simias

Sisyphus

Snodgrass, A.

Sophocles

Sounion

South Slavic epics , , , –

Sparta

speeches –, –, , –

statistical method , –, –, –, ,–

Stesichorus , , ,

Sthenelus ,

Strabo , ,

stylization , , , ,

synizesis , , , –, , –

Talos

Tegea

Teiresias , , , ,

��/�� ,

Telegonus ,

Telegonydating of

and Odyssey

Telemachus , , , , , , ,

Telephus , , –, ,

Tenos

Teucer

Teuthrania

Teuthranian expedition –,

Teuthras

Thalmann, W. G.

Thalpius

Thamyris ,

Thaumacus

Thaumakie ,

Thebais

Thebes ,

Thebes, (Egypt)

Theoclymenus

Theocritus

Theogony , , ; see also Hesiodic poemsend of ,

and Works and Days ,

Thersander , , , ,

Thesprotia , , ,

Thessalians ,

Thessalus ,

Thetis , , , , , ,

Thoas

Thrasymedes


General index

Thrinacia

Titaressos, River

Tlepolemus , ,

tmesis

in archaic Latin poetry

in classical Greek authors

with 7� or �� –

and free adverbials

in Hellenistic poetry

in Hesiodic poems

in Homeric Hymn to Apollo

in Homeric poems –

in Linear B

in Shield of Heracles

traditional referentiality , –, , , ,,

Triptolemus , ,

Trojans –

Troy

Tydeus ,

type-scenes , , ,

Tyro , , ,

Tyrtaeus

Unbestimmtheitsstellen

unitarianism , , ,

S��1�� –

Usener, K. ,

Visser, E.

vocative as nominative

Wackernagel, J.

West, M. , , , , ,

Works and Days; see also Hesiodic poemsdating of

writing , , –, –, , , , ,

Xanthus

Zenodotus ,

Zethus

Zeus , , , , , , , , ,, , , ,

as Xenios

Zielinski’s Law ,


Index locorum

Acesander (FGrH) F

Aesch.Ag.

–

[Aesch.]TGF III

Alc.fr. a L–P = Voigt

fr. .ff. L–P/Voigt

fr. L–P/Voigt

fr. . L–P/Voigt

fr. L–P/Voigt

fr. L–P = A Voigt

fr. A h Voigt

fr. L–P/Voigt

fr. a L–P = Voigt

fr. a–e Voigt

fr. L–P/Voigt

fr. L–P/Voigt

fr. S SLG

Alcm.fr. PMGF

Alcmaeonisfr. Bernabe/Davies/West

Ap. Rhod.Argon.

.ff.

.

Apollod.Bibl.

.. , ,

..

..–

.

...

..

..

..

..

Epit..

Ar.Nub.

Pax

Archil.fr. IEG

fr. IEG

P.Oxy. ,

Arist.fr. Rose

ArrianPeripl. M. Eux.

Asclepiades (FGrH) F

Bacchyl..–

Callim.Hymn

–

–

Hymn

–

CEG.

ClaudianDe Raptu Proserpinae

.–

Clem. Al.Protr.

..

..

Cypriafr. Bernabe/Davies/West

fr. . Bernabe/Davies/West

fr. Bernabe


Index locorum

Descent of Ishtar–

–

Diod...–

Diog. Laert..

Ephorus (FGrH) F

Eratosth. (FGrH)I B.–

I B

Eur.Hel.

–

–

Tro.–

schol. ad Andr.

Euseb.Chron.

Ol.

Harp.s.v. A# +��

Hdt..

.

.�.

.–

..

.

Hecataeus (FGrH)I F a

Heraclid. Pont.fr. .–

fr.

Hes.Op.

–

–

–

–

ff.

–

–

–

–

–

Theog.–

–

–

–

–

–

–

– ,

–

–

–

–

–

–

–

–

–

– ,

–

– ,

–

–

[Sc.]hypothesis

– ,

fr. M–W = H

fr. M–W

fr. M–W

fr. M–W = H

fr. a. M–W = . H

fr. a. M–W = . H

fr. a M–W = H ,

fr. a. M–W = . H

fr. b M–W

fr. . M–W = . H

fr. .– M–W = .– H


Index locorum

fr. . M–W = . H

fr. .– M–W = .– H

fr. M–W = H ,

fr. a M–W = H

fr. M–W = a H

fr. . M–W = a. H

fr. M–W = H

fr. M–W = H

fr. . M–W = . H

fr. M–W = a+b H

fr. M–W = H

fr. . M–W = . H

fr. H

fr. M–W= ∗ H

fr. M–W = H ,

fr. M–W = H

fr. M–W = H ,

fr. M–W = H

fr. M–W = H

fr. M–W

fr. . M–W = . H

fr. . M–W = . H

fr. .– M–W = .– H

fr. .– M–W= .– H

fr. . M–W = . H

fr. . M–W = . H

fr. M–W = H

fr. a M–W

fr. M–W = H

fr. M–W

fr. M–W = H

fr. A M–W

fr. M–W = ∗ H ,

fr. M–W

fr. M–W

fr. M–W

fr. M–W

Hom.Il.

. ,

.

.

.–

.

.ff.

.

.–

. ,

.

.ff.

.

.–

.–

.

.

.

.

.–

.ff.

.–

.

.

.

.

. ,

.

.

.

.

.–

. , ,

.–

.

.–

.–

.–

.–

.–

.–

.–

.–

. ,

.

.–

.

.–

.

.

.–

.–

.

.

.–

.–

.–

.ff.

.–

.–

.–

. ,

.–

.–

.

.

.

.

.

.

.

.


Index locorum

Hom. (cont.).–

.

.

.

.–

.

.ff.

.

.

.

.

.

.

. ,

.

.

.

.

.

.

.

.

.–

.–

.ff.

.–

.

.–

.ff.

.–

.ff.

.ff.

.

.ff.

.

.– –

.

. ,

.–

.–

.

.

. ,

.–

. ,

.–

.

. ,

.

.

.

.–

.

.ff.

.–

.–

.

.

.

.

.

.

.

.

.–

.

.

.

.–

.–

.–

.

.– ,

.–

.–

.

.

.–

.–

.–

.

. ,

.–

.

.–

.

.–

.

.–

.

.

.ff.

.

.

.–

.–

.ff.

. ,

.–

.–

.–

.

.

.

.

.

.–

.

.


Index locorum

.–

.

.–

.

.– –

.ff.

.

.–

.ff.

.

.–

.

.–

.–

.–

.–

.–

.–

.

.

.

.

.

.

.

.ff.

.–

.

.

.

.

.–

.

.

.

.

.–

.

.–

.–

.–

.–

.–

.ff.

.–

.

.

.

.

.

.–

.

.

.–

.

.–

.

.

.

.–

.

.– ,

.–

.–

.

.

.–

.– –

.–

.–

.ff.

. ,

.–

.

. ,

.–

.

.

.–

.

.–

.

. ,

.

.–

.

.–

.

.

.–

.

.

.

. ,

.

.–

.

.–

.

.

.

.

.

. ,

.

.

.

.

.

.


Index locorum

Hom. (cont.).

.

.–

.

.

.–

.–

.–

.

.–

.–

Od..

.

.–

.–

.

.–

.

.

.–

.–

.

.– ,

.–

.

.

.

.–

.–

. ,

.–

.–

.

.–

.

.–

.

.

.

.–

.

.–

.ff.

.

.–

.–

.–

.–

.

.

.

.

.

.–

.–

.

.–

. ,

.–

.–

.–

.

.–

.

.

.–

.

.– ,

.

.ff.

.

.–

.

.ff.

.–

.

.ff.

.–

.–

. , ,

.

.

.

.–

.–

.–

.–

.–

.–

.

.

.–

.–

.–

.–

.–

.

.–

.

.–

.–

.–

.ff.

.–

.–

.ff.

.–

.

. ,

.–


Index locorum

.–

.–

.–

.ff.

.–

.–

.–

.ff.

.

.ff.

.

.ff.

.–

.–

.–

.–

.– ,

.–

.– ,

.–

.–

.–

.

.–

.

.–

.

.–

.–

.–

.–

.–

.

.

.ff.

.–

.–

.ff.

.–

.

.

.

.–

.–

.ff.

.

.–

.ff.

.ff.

.

.

.

.

.ff.

.

.–

.– ,

.– ,

.–

.–

.

.–

.–

.

.

.–

.

.–

.

. ,

. ,

.

.

.–

.– ,

.

.

.–

.

.–

. ,

.

.

.

.–

.–

.–

.–

.

.

schol. ad Il..

.–

. ,

.a

. ,

.a

.a Erbse

.

.

.b

.

schol. ad Od..

.

.

Hyg.Fab.

Hymn. Hom.

–


Index locorum

Hymn. Hom. Ap.–

–

Hymn. Hom. Cer.– ,

–

–

–

– , , , ,

–

–

–

–

–

–

–

–

–

–

–

–

–

–

, ,

–

–

–

–

Hymn. Hom. Merc.–

Hymn. Hom. Ven.

ff.

–

Hymn. Orph..–

Ilias parvafr. Bernabe = Davies = West ,

fr. .– Bernabe = .– Davies = .–

West

Isoc.Paneg.

,

[Longinus]Subl.

. ,

LucianVer. hist.

.

Lycurg.Leoc.

Marm. Par. (FGrH) A

A , ,

Mimnermus–a

Mosch.Europa

Orphic Argonautica

P.Berol.

–

– ,

–

–

–

–

–

–

– , , ,


Index locorum

– ,

P.Derv.xiii.

xvi.

xxi.

xxii.

xxii.

xxvi.

xxvi.

xxvi.–

Palaephatus (FGrH) F

Paus... ,

..

..

..

..

..

..

..

..

..

..

..

..

..

..

..

.. ,

..

.. ,

Philoch. (FGrH) F

Philostr.Her.

.

.

.–

Pind.schol. ad Ol.

.

.c ,

schol. ad Pyth..a Drachmann

Pl.Leg.

d–e

PMG

Procl.Chrestomathia

– Severyns = –.–

Bernabe

– Severyns = . Bernabe

– Severyns = .– Bernabe

– Severyns = .– Bernabe

Sapphofr. .ff. L–P/Voigt

fr. L–P/Voigt

fr. . L–P/Voigt

fr. . L–P/Voigt

fr. L–P/Voigt

fr. L–P/Voigt

Semon. IEG

Sen.Med.

–

Soph.TGF IV F

Stesichorusfr. PMG

fr. PMGF

fr. PMGF

Strabo..

..

..

..

..

..

..

..

Sudas.v. dZ'��

s.v. Q�''��

Thebaisfr. Bernabe = Davies = West

Thuc..

Tyrtaeus.– IEG

.–

.–

.–

Xen.Cyr.

.


relative_chronology_in_early_greek_epic_poetry_-_edd._o._andersen,_d._t._t._haug_-_2012.pdf

Documents