The Continuity of Classical Literature Through Fragmentary Traditions 9783110712223, 9783110700374, 9783110712292, 2020950007

Citation preview

The Continuity of Classical Literature through Fragmentary Traditions

Trends in Classics – Supplementary Volumes

Edited by Franco Montanari and Antonios Rengakos Associate Editors Stavros Frangoulidis · Fausto Montana · Lara Pagani Serena Perrone · Evina Sistakou · Christos Tsagalis Scientific Committee Alberto Bernabé · Margarethe Billerbeck Claude Calame · Jonas Grethlein · Philip R. Hardie Stephen J. Harrison · Stephen Hinds · Richard Hunter Christina Kraus · Giuseppe Mastromarco Gregory Nagy · Theodore D. Papanghelis Giusto Picone · Tim Whitmarsh Bernhard Zimmermann

Volume 105

The Continuity of Classical Literature through Fragmentary Traditions Edited by Francesco Ginelli and Francesco Lupi

ISBN 978-3-11-070037-4 e-ISBN (PDF) 978-3-11-071222-3 e-ISBN (EPUB) 978-3-11-071229-2 ISSN 1868-4785 Library of Congress Control Number: 2020950007 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2021 Walter de Gruyter GmbH, Berlin/Boston Editorial Office: Alessia Ferreccio and Katerina Zianna Logo: Christopher Schneider, Laufen Printing and binding: CPI books GmbH, Leck www.degruyter.com

Preface The chapters collected in this volume originate from papers that were originally presented at the panel ‘From Sources to Editions and back again. The Continuity of Classical Literature through Fragmentary Traditionsʼ. The panel was part of the 10th Celtic Conference in Classics (19‒22 July 2017, McGill University and Université de Montréal) and saw a total of fifteen papers delivered over three days. In sending this book to the press, we would like to express our gratitude to the organisers of the Montreal Conference, Elsa Bouchard (Université de Montréal) and William Gladhill (McGill University), who gave us the opportunity to set up our panel and wonderfully hosted us at both universities. We are also grateful to all the colleagues who attended the panel and contributed to it, either by presenting a paper or by joining the fruitful discussions that followed each presentation. We believe that the panel proved a success in promoting the academic cross-fertilisation policy that drives the Celtic Conference in Classics and wish to acknowledge the collaborative and friendly spirit that we experienced in Montreal. Gratitude is also due to Raymond L. Capra, Eva Falaschi, Patrick J. Finglass, Kyriaki Ioannidou, Chiara Monaco, Enrico Emanuele Prodi, and Effie Zagari. We are very grateful to Suzanne Sharland, for her thorough linguistic revision of the entire volume and for further valuable comments and suggestions on its contents. We also wish to credit the Yale Papyrus Collection, Beinecke Rare Book and Manuscript Library, as the source of the images reproduced in Giulio Iovine’s chapter (8), and express our gratitude to Ellen Doon, the Library’s Head of the Manuscript Unit. We also wish to thank Serena Pirrotta, Marco Acquafredda, and Anne Hiller for their interest in the project and thorough assistance throughout the editorial process; Katerina Zianna did a wonderful job in preparing the manuscript for publication and taking care of its editing. Lastly, we would like to acknowledge the support provided by the Department of Cultures and Civilizations of the University of Verona. Francesco Ginelli Francesco Lupi

https://doi.org/10.1515/9783110712223-202

Contents Preface  V List of Figures  IX List of Tables  XI Francesco Ginelli and Francesco Lupi Introduction  1 Stefano Vecchiato Marginalia to Hesiodic Fragments: A Possible Dis-attribution (Fr. 41 M.–W.), a Possible Attribution (Fr. 327 M.–W.), and Some Recently (Re-)discovered Fragments  19 Francesco Lupi To Belong or not to Belong: A Few Remarks on the Lyric Fragments of Sophocles’ Tereus  37 Chiara Meccariello ‘Well Begun is Half Done’? Uses and Misuses of Incipits in Greek Antiquity and Beyond  57 Roberta Berardi Collecting Fragments for a Fragmentary Literary Genre: The Case of Greek Hellenistic Oratory  79 Francesco Ginelli The New Nepos: Prolegomena Toward a Renumbering of Cornelius Nepos’ Fragments  103 Jarrett T. Welsh The Fifth Glossary of Nonius Marcellus  121 Nereida Villagra Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers: A Case Study on Phineus and the Argonauts  145

VIII  Contents Giulio Iovine The Unruly Fragments: Old Problems and New Perspectives in Latin Military Papyri from Dura-Europos (P. Dura 56, 64, 72, 74, 76, 89, 113)  165 List of Contributors  193 Index of Names  195 Index Locorum  205

List of Figures Fig. 1: P. Dura 56: Docket in fr. A, above l. 1.  169 Fig. 2: P. Dura 71: Docket in the upper margin.  170 Fig. 3: P. Dura 66 f: Docket in the upper margin.  170 Fig. 4: P. Dura 63: Docket in the upper margin.  170 Fig. 5: Fr. A.  172 Fig. 6: Fr. B.  173 Fig. 7: Fr. C.  173 Fig. 8: Fr. B with smaller scraps.  174 Fig. 9: Fr. h (m in Marichal’s edition).  174 Fig. 10a: Current status: ] ḳal ̣ [ .  175 Fig. 10b: Marichal’s edition (upside down): e]x̣ Kạl[ .  175 Fig. 11: Letter A: The smaller scraps.  176 Fig. 12: Letter B: The smaller scrap.  176 Fig. 13: Letter A, verso: The addressee’s name.  177 Fig. 14: P. Dura 72.  179 Fig. 15: P. Dura 72: ạṇ[, or ṃạ[?  180 Fig. 16: P. Dura 74: -anae, not -ante.  181 Fig. 17: P. Dura 76: Similar damages.  184 Fig. 18: P. Dura 74: A possible subscription.  185 Fig. 19: P. Dura 89 (= DP 9 recto), col. II, ll. 5–6 (the scrap provides coh and traces of the figure XX, directly to be linked to Pal of Palmyrenorum).  187 Fig. 20: P. Dura 107 (= DP 9 verso), col. II, l. 12.  187 Fig. 21: The papyrus in its current position.  190 Fig. 22: The papyrus turned 180°.  190

https://doi.org/10.1515/9783110712223-204

List of Tables Tab. 1: Tab. 2: Tab. 3: Tab. 4: Tab. 5: Tab. 6:

Soph. frr. 592.4–6+593 R.2: Metrical description (i).  47 Soph. frr. 592.4–6+593 R.2: Metrical description (ii).  48 P. Dura 56: Ordering the fragments.  171 P. Dura 89: Ordering the fragments.  186 Fr. b+c: A reassessment.  189 Fr. b+c: A reassessment.  189

https://doi.org/10.1515/9783110712223-205

Francesco Ginelli and Francesco Lupi

Introduction

 The study of fragments and the search for a method Fragmentary texts play a central role in the field of Classical literature. Modern understanding of ancient cultures is inevitably characterised by loss, since the greater part of what ancient Greece and Rome produced has not survived. With regard to written texts, such loss has come about through a variety of factors: from the weak physical composition of the materials employed to record texts in antiquity, to their destruction by human agency and the ravages of time, to deliberate selection (and concomitant exclusion) in the formation of literary canons.1 The sense of loss that accompanies the study of Classical literature, however frustrating it may be, poses a stimulating challenge to scholars and readers. In fact, the desire to restore what has been lost in the process of textual transmission acts as a sort of Aristotelian primus motus for the study of fragments. It is this desire that keeps motivating scholars in spite of the meagre state of the evidence at hand, with the result that fragmentary literary works are still nowadays collected, studied, and edited, possibly to an even greater extent than in the past.2 The paucity of textual material represents the most obvious and most challenging difficulty for those wishing to conduct research on fragmentary texts. In fact, the smaller the amount of available text, the higher becomes the risk for the intended research to lead to biased results, affecting, amongst other things, aims such as textual restoration, interpretation, and the tentative reconstruction of partially preserved works. Consequently, methodological concerns play a major  In the following pages the pronoun we and the possessive adjective our are generally used to convey both editors' views on the topics treated. While the content of almost all the paragraphs is the outcome of prolonged reflection and mutual discussion between both editors, paragraphs 1 and 2 were authored specifically by Francesco Lupi, and paragraphs 3 and 4 were individually written by Francesco Ginelli.  1 On the relationship between fragments and canon formation, see Most (1997) vi, paragraph ‘1. Fragments and Canons’. 2 The healthy state of fragmentology in contemporary Classical scholarship is demonstrated by the current wealth of conferences and multi-authored volumes addressing the topic of fragments from a variety of perspectives, including the theoretical: see, most recently and most notably, Derda/Hilder/Kwapisz (2017). https://doi.org/10.1515/9783110712223-001

  Francesco Ginelli and Francesco Lupi role in the study of fragmentary literature. Just as in any other field of Classics, however, methods and principles should not be regarded as rigidly immutable. Constantly evolving factors such as scholarly trends and technical-scientific innovations have a considerable impact upon these methods and principles. Moreover, methodological principles are also influenced by practical considerations, and by issues arising from and sometimes peculiar to the texts under scrutiny. These aspects range from the physical condition of the evidence (e.g. in the case of a papyrus text, the degree of fragmentation of the papyrus itself), to the literary rules and conventions to which the fragment(s) may be subject. On the last point, fragments of a prose-text will necessarily entail different issues from fragments of poetry; such issues will both influence the questions raised by the fragments and limit the ways in which scholars are able to answer them. Likewise, the level of literariness of a fragmentary text will play, at least to a degree, a similar role. In other words, if different questions apply to different kinds of texts, it follows that the methods and principles employed in the study of fragmentary literature should also be tailored around the specific features of the texts to which those very methods and principles are applied.3 Therefore there is a need for a constant comparison between the methods, approaches, and goals with which textual fragments — be they Greek or Latin, poetry or prose, literary or para-literary — are investigated in current scholarship. The continuously changing nature of the study of fragmentary literature provided the stimulus for us to bring together scholars variously engaged with textual fragments, and to collect contributions across literary genres, on a wide range of aspects relating to textual fragmentation. The ultimate goal of this book has been to convene a discussion group on some of the most relevant issues that fragmentologists have to face. More often than not, such issues are of a methodological nature. It is our conviction, then, that bringing to the forefront the methodological implications of the study of fragmentary literature is the most viable way to promote debate in the field and to invite critical reflection on its common specificities and limitations. The methodologies employed by this volume’s contributors, however influenced by the factors outlined above, all show common concerns. The purpose of this introduction, therefore, is to highlight some of the methodological concerns and principles that underlie the study of fragmentary literature, broadly defined, and which are fundamental to the field. Despite their differences in topic and  3 Most (1997) vii (point 4) lucidly warns against treating poetic, philosophical, and historical fragments as if they were “all fragments in the same way”: in fact, different kinds of fragments lend themselves to different levels of accuracy in quotation.

Introduction  

aims, the contributions collected in the present volume provide illuminating examples of the application of such principles and methods. To account for those principles, the following pages offer a checklist of the steps needed to ensure that the study and editing of fragmentary texts rest on a methodologically sound base. We find the image of the checklist to be particularly vivid, as it is comparable to the tasks performed by airline pilots prior to taking off, comprising a set of complex and consequential operations aimed at ensuring the success of the flight. Likewise, in the study of textual fragments, preliminary steps need to be taken in order to increase the chances of ‘success’ of any research undertaken in the field. Without pushing the analogy too far, and in order to lay down, at the outset, a basic epistemological premise of fragmentology, it is crucial to point out that with textual fragments ‘success’ can only be measured in terms of likelihood of the results of one’s research being correct.4 This is due to the fragments’ status as remnants of a whole that is unable to be restored to its original completeness. What strategies, then, should one adopt in studying what is no longer fully extant? How should one set about scrutinising texts affected by fragmentation? One possible route, we contend, is to ensure that any hypothesis concerning a fragmentary text is at least compatible with a set of conditions; failure to satisfy any such conditions will result in the hypothesis being heavily undermined. Establishing a checklist of tasks to be undertaken in the initial stages of one’s research will then be helpful in tracing those conditions and granting a methodologically sound approach to the study of textual fragments.

 What is a fragment and how do fragments come about? 2.1 Direct and indirect tradition. Fragments and testimonies The concept of a fragment implies that an entity, which was once whole, is broken and in pieces, indicating that something may be lacking or missing from what

 4 This consideration obviously does not apply to flights: a flight is successful as long as the aircraft lands safely at the intended destination, a result which can be objectively ascertained.

  Francesco Ginelli and Francesco Lupi remains. The etymology itself of the English word fragment (compare Latin fragmentum)5 suggests the idea of a ‘fracture’ befalling an originally homogenous body or object, whatever its physical constitution, nature, and function may be.6 If we limit this observation to textual fragments, the subject of this volume, it becomes evident that a fragment exists so long as the work it originally belonged to does not survive in its entirety. To put it in other words, a portion (p) of a text (T) surviving ‘outside’ of T (e.g. if it is preserved as a quotation by a later author, or anyhow detached from the original text conceived as a whole) does not qualify as a fragment (f) of T until T is lost7. As long as T is extant, in fact, p does not enjoy the status of fragment, not, at least, in the sense in which the term is generally used by Classicists; moreover, any observation on p can be verified against T — or, conversely, any observation on T can be verified against p, at least for the overlapping portion. For instance, a quotation from Sophocles’ Antigone by a later author could be checked and verified against the text of that particular passage as preserved in the manuscripts of Sophocles’ Antigone. Any discrepancies between p (in this case, the Ant. quotation, an instance of indirect tradition of an ancient text) and T (the

 5 On the etymology of the English word fragment, “a sixteenth-century borrowing” from Latin, and the cognates Eng. break and Lat. frango, both going back to the Proto-Indo-European root *bhreg-, ‘break’, see Katz (2017) 27–30. After all, the idea of material ‘brokenness’ implicit in the very etymology of fragment is still clearly felt by English speakers today; cf. the first definition given in the OED Online, s.v. ‘fragment, n.’: “A part broken off or otherwise detached from a whole; a broken piece; a (comparatively) small detached portion of anything”; with regard to works of art, under point 2b, the following definition is given: “An extant portion of a writing or composition which as a whole is lost; also, a portion of a work left uncompleted by its author; hence, a part of any unfinished whole or uncompleted design.” (accessed January 28 2020, https://www.oed.com/view/Entry/74114?rskey=Mgb0Dn&result=1#eid). 6 Fragments have stimulated scholars’ imagination through the centuries, eliciting various metaphorical (or analogical) descriptions to account both for their peculiar status as inherently ‘incomplete’ texts and for the no less peculiar approach required for such texts (Wright 2016, xxiv– xxv lists at least seven different images). The most widespread of such images probably is — as Rudolph Kassel named it — the Schiffbruchmetapher, which likens literary fragments to the disjointed timbers of a wrecked ship; on the image, see esp. Kassel (1991) 243 and n. 2, Dionisotti (1997) 26–27. Building on this image — and shifting the focus from the fragments themselves to the scholarship on fragments — we suggest conceiving collections and critical editions of fragmentary texts as a sort of castaway depot, meant to provide shelter to the ‘survivors’ (namely fragments) of the literary shipwreck that befell countless Graeco-Roman works. 7 The adjective lost is used here with the particular sense of ‘lost as a whole’, ‘not fully extant’, implying that the lost text actually survives in one or more discrete portions. Thus, unless further specified, lost is here virtually synonymous with fragmentary. Similarly, extant is here used with the marked sense of ‘fully extant’, ‘non fragmentary’ (see, however, n. 17 below).

Introduction  

play Ant. as transmitted by the direct tradition of Sophocles’ extant oeuvre) can be rather readily ascertained by means of a close comparison between the two. For f to exist, as we indicated above, the condition instead is that T is lost. When this condition is fulfilled, in the simplest scenario what is known of T ultimately derives from p (or p1, p2, p3 …8), that is, from the sole surviving portion(s) of an irretrievable whole.9 Under these circumstances, such portions can be viewed as fragments in their own right (then p1, p2, p3 … = f1, f2, f3 …). Fragments, unfortunately, cannot be verified against the whole of which they were originally a part; to resort to yet another analogy, fragments are like glittering gems of a lost literary treasure: they shine, but just not enough for us to appreciate the full picture. The implications of this state of affairs are manifold and highlight the defective state of fragments: a) Fragments are both detached (irreparably) from the original text,10 and are disjointed from one another;11 b) Their textual restoration and interpretation is hampered by the lack of the (original) context; c) Various degrees of uncertainty apply, from case to case, to such issues as the fragments’ exact boundaries, authorship, and original placement both within the work (if known) to which they belonged,12 and, if only authorship is known, within their author’s oeuvre;13 d) Fragments are not representative of the lost whole.14  8 This differentiation may be used where more than one portion of a lost text survives. 9 For the sake of clarity, this account is intentionally kept as simple as possible. In practice, however, our knowledge of a no longer extant text is in most cases advanced by a variety of sources, in addition to its surviving portions. At times such knowledge depends exclusively on what, for the moment, we may call external testimonies—that is, texts that tell us something about a lost work without purporting to be actual (i.e. literal) fragments of it. On the concepts of fragment and testimony and their shifting boundaries, see further below. 10 The concept of original is a contentious one: see e.g. Canfora (20192) 11–19. The term original is used in these pages with reference to quantitative aspects, the original amounting to a (virtually) complete text, as opposed to its fragments. 11 The relative ‘disjointedness’ of fragments of the same work is not always irreparable: the tentative joining together of fragments is indeed a well-known practice in fragmentology. For an instance of this see, in this volume, Lupi (ch. 2). 12 On the task of placing fragments within the original structure of a lost work, see Vecchiato (ch. 1) and, again, Lupi (ch. 2, section 4). 13 This last issue is particularly relevant to both Vecchiato (ch. 1) and Ginelli (ch. 5). 14 The point is made most lucidly by Olson in Baltussen/Olson (2017) 396–397. See also Most (1997) vi (paragraph ‘2. Fragments and Synecdoche’), remarking on the “tendency to regard fragments as partes pro toto, as though they contained locked within their narrow compass the secrets of the author’s work as a whole …”.

  Francesco Ginelli and Francesco Lupi In view of these deficiencies, it is necessary to examine fragments in the most accurate way possible and to devote a great deal of attention to every single facet,15 including the ways in which they have come down to us. Complying with this need for all-round accuracy will serve at least two purposes in the study of fragments: 1) To make as much sense as possible of the fragments themselves: What do they mean, deal with, or refer to? Are they textually sound? 2) To recover as much information as possible about the lost whole from which they come: How do the fragments relate to one another and to the original work? How do they influence our knowledge of that work and our understanding of its author?

2.2 Direct and indirect tradition The foregoing observations bring us to the first point of our checklist, which is: (i) to distinguish between directly and indirectly transmitted fragments. Making such a distinction is arguably the first task one has to undertake when approaching the study of a fragmentary text. The concept of directly transmitted fragments applies to fragments of an otherwise lost text,16 which are preserved without any intentional (textual) mediation by a later author.17 In this case, a fragment will have come down to us as part of a written medium (of varying material, format, and function) expressly meant to preserve the text of which the fragment is an extant portion (or, in some cases, the only one). In this sense, therefore, fragments are the remnants of a materially fragmented manuscript item (e.g. a papyrus roll, a parchment codex, an inscription) that suffered physical damage over

 15 Along similar lines, Wright (2016) xxiv–xxv, under point (iii) advocates, in the study of fragmentary tragedy, “a sort of micro-reading, in an attempt to push the fragments to their limits and squeeze out every drop of meaning and nuance”. This point applies, in principle, to all literary genres. 16 The definition is independent of the original nature and function of the lost work and applies equally to prose-texts, poems, official or private documents, and so on. Documentary evidence is at the core of Iovine’s chapter (8) on Latin military papyri from Dura-Europos. 17 However, the possibility is not ruled out that the quoting author and the one who is quoted are in fact coeval: e.g. a Euripidean fragment retrieved from Aristophanes’ Acharnians (425 BCE) — that is, from a play produced during Euripides’ lifetime — is an obvious instance of this possibility.

Introduction  

the course of time.18 Conceptualised in these terms, a textual fragment comes closest to the idea of physical fracture implicit in the word and its root. On the other hand, a fragment belongs to the indirect tradition when it survives thanks to the mediation of an author who intentionally quoted or excerpted a portion of a work which was subsequently lost. Translations (e.g. Latin, Arabic, or Syriac translations of Greek texts) enjoy a sort of ‘in-between’ status: in one (admittedly oversimplified) sense they re-produce a text originally composed in a different language. By adopting a different linguistic code, however, translations inevitably produce new texts, which are different, in various aspects and to various degrees, from their source texts.19

2.3 Fragments and testimonies A second task required of fragmentologists — and the second point in our checklist — is (ii) to make as clear as possible a distinction between the concepts of fragment and testimony. Textual portions such as indirect quotations, summaries, or paraphrases in principle should be regarded as testimonies of a lost text, rather than fragments stricto sensu;20 in fact, they do not, nor do they aim to, reproduce the supposedly exact wording of a text, but consciously deviate from it, delivering, as André Laks puts it, an “élaboration secondaire” of that text.21 However, we should not conceive of this distinction too rigidly, nor should it permit us to devalue testimonies as less worthy of scrutiny than fragments. As will be seen, both categories play a relevant role in a number of chapters of this book, and on occasion the distinction between the two is conveniently re-shaped according to the researcher’s aims and editorial principles, taking into consideration also the specific literary features of the texts under discussion.22 Generally, something that at face value would qualify as a testimony of a lost work (such as  18 One may argue that even texts usually regarded as ‘extant’ are not entirely so. In fact, lacunae (evident or supposed) may affect single portions of a text, to the detriment of its completeness; nevertheless, while to some degree lacunose, such a text will be altogether intelligible and our overall appreciation of it will not be drastically undermined. From a different yet equally important perspective, Most (2010) 371 argues that complete works may be regarded as fragments “inasmuch as the larger cultural context that produced and enjoyed them has been lost”. 19 In this volume the question of whether translations should be considered proper fragments is lucidly tackled by Berardi (ch. 4): see esp. p. 95 n. 54. 20 Testimonies on broader issues such the author’s life and overall literary activity are not considered here. 21 Laks (1997) 238. On fragments and testimonies, see also Balbo (20072) xiii–xvi. 22 On this, see Ginelli (ch. 5), pp. 104–111.

  Francesco Ginelli and Francesco Lupi a paraphrase or a summary) is often no less informative than a fragment in the strict sense.23 If we adopt a content-based perspective, moreover, a text providing information on a lost text without quoting verbatim from it is in principle no different from a proper fragment, as both categories, in a way, allow us to recover ‘parts’ of a lost whole.24 It will be the task of fragmentologists, especially in their capacity as editors of fragments, to formulate explicitly and rigorously the criteria followed in arranging and allocating their material. Clarity and openness about the editorial or working criteria adopted constitutes yet another essential task for the fragmentologist.25 Criteria will vary greatly among editions of fragmentary texts and so will the extent and boundaries of the categories of fragment and testimony. Independently of the fact that the criteria followed in collecting, arranging, and distinguishing between fragments and testimonies may vary according to the aims and methods employed, one must in any case take into careful consideration the manner in which particular fragments have been transmitted to us — whether, that is, they have been directly or indirectly transmitted. With regard, in particular, to the indirect tradition and the issues it poses to scholars, a few observations are appropriate: (a) Indirectly transmitted fragments often reveal marked textual deviations and alterations compared to the supposed original text; (b) The contexts in which these fragments are embedded may bias our interpretation; (c) Both (a) and (b) ultimately originate from the fact that fragments belonging to the indirect tradition are the result of deliberate selection, a process which often entails refunctionalisation (the fragment is made to mean something different from what it meant in its original context, either through (a) or (b), or both) and re-wording (intentional or unintentional, as in the case of quotations from memory). One may further note that (d) indirectly transmitted fragments also show different degrees of faithfulness to the original text. For instance, in the case of allegedly verbatim quotations of a lost work, the practice of quoting from memory, particularly common in antiquity, is likely to hinder our chances of retrieving the

 23 See West (1973) 94–95; most recently and with specific regard to the Presocratics and their doctrines, the point is made by Baltussen (2017) esp. 76–77. 24 On this point, see Laks (1997) 239. 25 See, along similar lines, Wright’s (2016) xxvi point (x) and, in this volume, Ginelli’s (ch. 5) considerations at pp. 108–109.

Introduction  

ipsissima verba of the quoted author.26 One should therefore exercise considerable caution in those cases where such practice is either hinted at or suggested by the author who quotes it, or is in some way detected by scholars. When analysing fragments that have come down to us in the form of quotations, as a general rule one should keep in mind that what we read today is the product of several layers of mediation. Of such layers, memory is but one. In fact, quotations may also be characterised to a degree by covert parodic intervention on the part of the quoter (e.g. the quoting author makes a pun on the prior text for his own purposes), thus carrying textual alterations liable to go unnoticed, at least in some cases, by today’s readers. Another, more obvious layer of mediation resides in the textual transmission of the author who quotes the portion of text. One example will suffice: the poetic quotations preserved in Plutarch’s How to Study Poetry, for instance, are inextricably entwined in the textual history of Plutarch’s treatise; in other words, they are at the intersection of indirect and direct tradition. Fragments, and the contexts in which they are embedded, share the same manuscript fortunes, as it were. A further layer of mediation applies to indirect (that is, not word-for-word) quotations conveying other people’s thoughts, as in cases of indirect speech (oratio obliqua), paraphrases, summaries, allusions or generic references. All such texts will normally be less faithful to the original text that is summarised, paraphrased, or alluded to than verbatim quotations in fragments deriving from direct tradition. It will be the researcher’s task to isolate any interpolations that may have crept into these ‘second-degree fragments’, as we could label them, or to detect any instances of ideological manipulation of the original text (a phenomenon known to be highly relevant, for instance, for philosophical texts).27

 Philological analysis and textual history of fragments As we saw in the previous paragraph, the methodological reflection on such concepts as the distinction between directly and indirectly transmitted fragments (i) and that between testimonies and fragments (ii) should constitute the preliminary two-step phase of any research on fragmentary texts. By duly pondering the  26 On quotations from memory and their inaccuracy, see West (1973) 17–18. 27 On Heraclitus and the Presocratics, see Fronterotta (2013) iii–x.

  Francesco Ginelli and Francesco Lupi specificities of the fragments at hand, researchers will be in a better position to determine the exact nature of the evidence dealt with and, consequently, their research objectives. It will only be when these steps have been performed that the research should proceed to a proper philological analysis of the fragments’ text (iii) — that is, to the third point of our checklist. In principle, this stage will entail the same techniques employed in the study of complete texts, including: 1) Collection and collation of all available manuscripts of the fragments’ source and of all types of available relevant evidence (also of a non-literary nature); 2) Analysis of such features as language, style, and, in the case of fragments of poetry, metre; 3) Comparisons with other works, either extant or fragmentary, of the same author and/or with comparable works (i.e. belonging to the same literary genre and period or relating to the same topic). The researcher’s goals, as well as the attainability of these goals, will differ according to the specific features and state of the fragments: (a) Constitution and interpretation of the text will be influenced by such factors as the actual amount of preserved text, its intelligibility, and soundness; (b) Retrieval of the ‘original’ words of a lost work — should this be the objective aimed at — will be bound by a high degree of uncertainty when dealing with non-verbatim quotations (though, as we saw, caution must be exercised in this respect also with purportedly literal quotations); (c) Reconstruction of a lost work, an inherently speculative task, will depend on how revealing and extensive the available evidence is.28 What is additionally required of fragmentologists is that they also try every possible way to extract useful information — or, conversely, to detect and isolate elements with a potentially biasing effect — from the fragments’ source, which brings us to the fourth point of our checklist (iv). In fact, it is crucial to stress that one should not scrutinise fragments as absolute entities. Especially with fragments belonging to the indirect tradition, whether they are actual or ‘second-degree’ fragments as defined in the previous paragraph, the ‘context of their survival’29 will not only elicit as many questions as the fragments themselves, but will on occasion yield relevant (and often overlooked) evidence for a better understanding of them. One of our contributors, Jarrett T. Welsh aptly describes this  28 Many gaps will inevitably be left unfilled in every attempted reconstruction of a lost work. On the issue of reconstruction (and its limits), see now Baltussen/Olson (2017). 29 The phrase is borrowed from Welsh (ch. 6), p. 124.

Introduction  

as “new information [providing] a better guide to our speculations about [them]” (p. 124). More generally, it follows that reassessment of the available evidence as well as its framing from novel perspectives are crucial tasks in the study of fragments, on the assumption that editions of any text can be improved upon. In fact, contemporary scholarship on fragments often relies on outdated editions, on which our interpretation of a fragmentary text ultimately depends. Many of the chapters collected in this volume stress the fact that new results in the study of Greek and Latin fragments can be obtained not only by scrutinising newly found texts, but also — and far more often — by painstakingly reassessing long-known evidence.30 There are a number of factors that invite caution in the handling of fragments, and several of these originate with the fragments’ sources and their manner of transmission. There may be, for example, conscious or unconscious textual manipulations (parodic and ideological alterations; linguistic adaptations affecting the dialectal facies of a text in the course of manuscript transmission), involuntary alterations (e.g. those originating from the practice of quoting by memory), and the implications of the specific format of the source (e.g. anthology, epitome, cento). All such factors can only be fully appreciated by paying close attention to the sources’ inner rules, aims, quotational techniques, and relationship with other works of a similar kind. Ideally, one should also strive to become acquainted with the sources’ textual history. In particular, the study of fragments should be framed within an overtly historical perspective, in a way that promotes the study of the contexts of their transmission as an equally important task of fragmentology.31 In other words, tracing the history of a fragment will also entail, to some extent, doing the same with its source(s). Among possible questions, and with specific regard, again, to indirectly transmitted fragments, it is important to ask what reasons an ancient author had to quote an extract (not yet a ‘fragment’) from a subsequently lost work, and to quote it in that particular way.32 No less important is to ascertain, whenever possible, whether the quoting author had direct access to the quoted text, or whether he relied on intermediary sources (in other words, how far removed is the quoter from the text he purports to be quoting?). In the absence of a fragment’s original  30 This is part of the continuous interest enjoyed by the study of fragmentary literature, either considered per se as a self-sufficient, specialised scholarly domain, or simply employed as source material for broader research in Classics. 31 The importance of the context of transmission is duly stressed by Conte (19922) vi, Most (1997) vi–vii (paragraph ‘3. Sources of Fragments’). 32 On this aspect, see again Most (1997) vi–vii, who draws an important correlation between an author’s reasons for quoting and his way of quoting.

  Francesco Ginelli and Francesco Lupi context, prominence should at least be given to their ‘new’ context. It is by investigating this context that the fragment’s circulation in antiquity, reception, and survival through the ages (as well as those of its author) will be best understood. Therefore cultural and, to some extent, social history are domains with which the study of textual fragments and their transmission inevitably intersects. The last observation allows us to move to one final, broader point. In most cases, studying fragments will not lead us to recover much of the original whole. Nevertheless, to our partial and somewhat unexpected satisfaction, fragmentology will in many cases end up shedding light on a greater number of aspects of the ancient world than originally envisaged.

 Contributors, chapters and content The following pages offer a brief overview of each chapter’s content. The opening chapter, by Stefano Vecchiato (Chapter 1, ‘Marginalia to Hesiodic fragments. A possible dis-attribution (fr. 41 M.–W.), a possible attribution (fr. 327 M.–W.), and some recently (re-)discovered fragments’), tackles some of the complex tasks that scholarship on Hesiodic fragments has to face. Vecchiato moves into the realms of both direct and — primarily — indirect transmission and analyses crucial editorial matters such as assigning fragments to a specific work of the Hesiodic corpus and proposing plausible interpretations thereof. By means of careful analysis of the surviving evidence and a review of the existing scholarship, Vecchiato also methodically examines some recent ‘Hesiodic’ findings in order to determine their status as genuine, dubious, or spurious fragments. More generally, the Author reminds readers that working with textual fragments necessarily involves a high degree of conjecture, but at the same time healthily advocates a cautious approach, which values first and foremost the evidence at hand. Drama is among the literary genres most affected by fragmentation, and it is therefore not surprising that two chapters in this book focus on Greek tragic fragments. Fragmentary plays involve a wide range of issues, ranging from the constitutio textus of badly preserved and scattered bits of text, often irreparably detached from their original dramatic context, to interpretative problems and

Introduction  

tentative plot reconstructions.33 The present volume offers a significant sample of most of these issues and explores them from a variety of perspectives. In the first of our chapters dealing with dramatic fragments (Chapter 2, ‘To belong or not to belong. A few remarks on the lyric fragments of Sophocles’ Tereus’), Francesco Lupi proposes a reassessment of the textual transmission of two of the few lyric fragments of Sophocles preserved in the indirect tradition (592.4– 6, 593 R.2). Lupi’s investigation takes into account the anthological practice of Stobaeus’ Florilegium (the only source for frr. 592.4–6 and 593 R.2), on the methodological premise that fragments should be placed and understood within the broader frame of their source, even when the latter is of a purely anthological nature. Lupi urges scholars and readers engaging with fragmentary literature to rethink the accepted distinction between fragment and source, which he notes are “too often regarded as discrete and non-communicating entities” (p. 53), and he conceives the task of the editor as a sort of ‘mediation’ between them. Chiara Meccariello’s contribution (Chapter 3, ‘‘Well begun is half done’? Uses and misuses of incipits in Greek antiquity and beyond’) again focuses on Greek drama, yet from an entirely different perspective. Meccariello explores the tantalising issue of incipits (archai), the initial words of literary works, and their employment in Greek antiquity to identify works. Meccariello first offers an overview of the different uses of the ‘arche-system’ as a naming and textual identification device from Aristophanes onwards, then tackles three instances of problematic archai attested for the Euripidean corpus. For all three of the dramas which Meccariello investigates (the fragmentary Archelaus and Meleager, and the extant Rhesus), two or three different openings are attested, a situation which demonstrates the occasional failure of the arche-system as an identification device — its failure, that is, “to secure the textual identity of a work” (p. 76). Most noticeably, the three case studies singled out by Meccariello also enable her to highlight the “diagnostic potential” (p. 76) of the arche-system on matters of textual soundness and authenticity. In Chapter 4 (‘Collecting fragments for a fragmentary literary genre: the case of Greek Hellenistic oratory’), Roberta Berardi focuses on Hellenistic oratory, “a genre that is fragmentary at its core” (p. 81), and for which, surprisingly, no collection of the surviving fragments exists. Berardi’s investigation addresses this

 33 Another important issue is the complex relationship between direct and indirect transmission of plays for which papyri and the later medieval manuscript tradition have by chance preserved overlapping textual portions. Examples of this phenomenon are numerous; a few cases relating to Greek tragedy are referred to in Lupi (ch. 2).

  Francesco Ginelli and Francesco Lupi striking lacuna in Classical scholarship and carefully delineates the scope, editorial criteria, and methodological premises which would best characterise a collection of the fragments of Hellenistic oratory, an undertaking on which she is currently engaged. Berardi provides readers with a useful outline of the preliminary tasks that editing a multi-author and inherently diverse corpus of fragments requires. Berardi reflects on practical issues that have confronted her in the process of producing a new edition, with an English translation and commentary, of P. Schub. 32. She examines this as a specimen case for rhetorical papyri, while suggesting that it may have belonged to a book of professional orations. Berardi then goes on to explain the guiding paradigm she has devised in order to organise the vast source material for her future edition. This methodological principle, labelled the ‘Concentric Circles’ model, is based on the fragments’ exactness and ‘distance’ from the speech of which they were originally a part: quotations of an actual speech are grouped in the innermost ‘Circle’, paraphrases in the second ‘Circle’ radiating outwards, and ‘unclear’ fragments in the third, outermost ‘Circle’. In Chapter 5 (‘The New Nepos: Prolegomena toward a renumbering of Cornelius Neposʼ fragments’), Francesco Ginelli sets out a series of preliminary observations relating to his proposed new numbering of Cornelius Neposʼ fragments. Despite the many critical editions of Nepos produced in the twentieth century, his Deperditorum librorum reliquiae have received little attention. Moreover, the methodologies followed by editors are not always convincing; hence the need for a new collection of Nepos’ fragments. Since only a small part of Nepos’ fragmentary texts is ascribable to a specific lost work based on the information handed down by tradition, scholars have assigned most fragments to particular works according to ʻthematicʼ criteria, i.e. criteria based on the fragments’ content. As Ginelli shows, however, the application of such criteria has often resulted in ungrounded allocations of the fragments within Nepos’ oeuvre, instances of which he discusses in the chapter. Since a number of topics dealt with by Nepos could have appeared in more than one of his works, as suggested by the oftenoverlapping content between biographies and other works, editing the fragments of Nepos raises methodological issues which can only be addressed, as Ginelli shows, with a critically aware approach and by paying close attention to the text. Jarrett T. Welsh (Chapter 6, ‘The fifth glossary of Nonius Marcellus’) concentrates on Nonius Marcellus’ De compendiosa doctrina, an important source of literary fragments of Republican literature. The Author specifically focuses on one of Nonius’ own scholarly sources — the ‘fifth glossary’ (‘Gloss. v’) —, its features, and the implications thereof for “the informed speculations of those who seek to

Introduction  

edit or discuss a fragment” (p. 125) deriving from this source. Welsh first highlights some features observable in the Latin literary quotations Nonius took from ‘Gloss. v’ (such as signs of rewriting; syntactical incompleteness; presence of gaps between the lexicographical definitions given by Nonius and the quotations meant to illustrate them); then, the Author turns his attention to the fifth glossary’s recurrent patterns and methods in handling its literary quotations, as well as its prominent — yet neither exclusive nor direct — connection with the lexicographical work of Augustan scholar Verrius Flaccus. Building on the results of his investigation, Welsh finally proposes a number of considerations one should ponder when dealing with literary fragments deriving from ‘Gloss. v’. General ‘laws’ or ‘principles’, as the Author advises, may not be formulated for those fragments; however, as the study shows, careful attention to the features of ‘Gloss. v’ — as well as, ideally, of all sources of book fragments — suggests at least “some more general possibilities and probabilities” (p. 125), which can in turn guide our speculations on the fragments under scrutiny. In Chapter 7 (‘Mythographus Homericus, ἱστορίαι and fragmentary mythographers: a case study on Phineus and the Argonauts’) Nereida Villagra examines the so-called Mythographus Homericus (MH), a designation given to a lost mythographical work, from which several narratives on mythical subjects (ἱστορίαι) are preserved in papyri and scholia. The fragmentary and conjectural nature of the MH, along with the problematic relationship between scholia and papyri relating to its text, raise several methodological questions, which Villagra illustrates by investigating an ἱστορία concerning the Argonauts’ encounter with Phineus, which is preserved both in the D-scholia to the Odyssey (schol. Od. 12.69‒70), and in some papyrus fragments (PSI 10.117). Villagra’s case study is particularly interesting in that it presents all the formal features held to be characteristic of the MH. It is also a revealing one, as it allows her to show that reconstructing the ‘original’ text of the MH, as well as retrieving its ‘original’ features, are operations bound to remain very uncertain. Nevertheless, the very existence of different stages of the MH’s text, as attested in its various traditions, contributes to a deeper understanding of the MH and its nature, leading Villagra to conclude that “already in antiquity this text was perceived as an open text, and not as the work of an authority” (p. 162). The eighth chapter, by Giulio Iovine (Chapter 8, ‘The unruly fragments. Old problems and new perspectives in Latin military papyri from Dura-Europos (P. Dura 56, 64, 72, 74, 76, 89, 113)’), takes readers into the fascinating territory of Latin papyrology and shows, once more, that editions of fragmentary texts require constant updating and re-evaluation. Iovine demonstrates this by meticu-

  Francesco Ginelli and Francesco Lupi lously proposing new readings of several scraps of Latin epistulae and other documentary typologies preserved among the military papyri of Dura-Europos. The chapter deviates from the overall literary focus of the preceding essays but contributes much to the debate on fragmented texts from Graeco-Roman antiquity and the ways in which contemporary philology and Classics in general can address the manifold issues posed by fragments. While readers may be left thinking that the book raises more questions than it answers — a feeling all too familiar to Classicists engaging with fragmentary literature — this is the outcome that the editors of this book have earnestly hoped for in assembling the volume. Verona, January 2020 Francesco Ginelli Francesco Lupi

References Balbo, A. (2004, 20072), I frammenti degli oratori romani dell’età augustea e tiberiana. Parte prima: Età augustea. Seconda edizione riveduta e corretta, Alessandria. Baltussen, H. (2017), ‘Slim Pickings and Russian Dolls? Presocratic Fragments in Peripatetic Sources after Aristotle’, in T. Derda/J. Hilder/J. Kwapisz (eds), Fragments, Holes, and Wholes: Reconstructing the Ancient World in Theory and Practice, JJP Supplements 30, 73– 90. Baltussen, H./Olson, S.D. (2017), ‘Epilogue: a Conversation on Fragments’, in T. Derda/J. Hilder/J. Kwapisz (eds), Fragments, Holes, and Wholes: Reconstructing the Ancient World in Theory and Practice, JJP Supplements 30, 393–406. Canfora, L. (2002, 20192), Il copista come autore, Palermo. Conte, G.B. (1987, 19922), Letteratura latina. Manuale storico dalle origini alla fine dell’impero romano. Nuova edizione, Florence; English transl.: Latin Literature. A History. Translated by J.B. Solodow. Revised by D. Fowler and G.W. Most, Baltimore/London 1999. Derda, T./Hilder, J./Kwapisz, J. (eds) (2017), Fragments, Holes, and Wholes: Reconstructing the Ancient World in Theory and Practice, JJP Supplements 30. Dionisotti, A.C. (1997), ‘On Fragments in Classical Scholarship’, in G.W. Most (ed.), Collecting Fragments.Fragmente sammeln, Göttingen, 1–33. Fronterotta, F. (2013), Eraclito. Frammenti, Milan. Kassel, R. (1991), ‘Fragmente und ihre Sammler’, in H. Hofmann/A. Harder (eds), Fragmenta dramatica. Beiträge zur Interpretation der griechischen Tragikerfragmente und ihrer Wirkungsgeschichte, Göttingen, 243–253 (repr. in R. Kassel, Kleine Schriften, Berlin/New York 1991, 88–98); English transl. by H. Harvey and D. Harvey, ‘Fragments and their Collectors’, in F. McHardy et al. (eds), Lost Dramas of Classical Athens. Greek Tragic Fragments, Exeter 2005, 7–20.

Introduction  

Katz, J.T. (2017), ‘Reconstructing the Pre-ancient World in Theory and Practice’, in T. Derda/ J. Hilder/J. Kwapisz (eds), Fragments, Holes, and Wholes: Reconstructing the Ancient World in Theory and Practice, JJP Supplements 30, 23–40. Laks, A. (1997), ‘Du témoignage comme fragment’, in G.W. Most (ed.), Collecting fragments = Fragmente sammeln, Göttingen, 237‒272 (repr. in A. Laks, Histoire, doxographie, vérité. Études sur Aristote, Théophraste et la philosophie présocratique, Louvain-la-Neuve/Paris/ Dudley (Mass.) 2007, 27–55). Most, G.W. (1997), ‘Some Reflections on Fragments’, in G.W. Most (ed.), Collecting Fragments = Fragmente sammeln, Göttingen, vi–viii. Most, G.W. (2010), ‘Fragments’, in A. Grafton/G.W. Most/S. Settis (eds), The Classical Tradition, Cambridge (Mass.)/London, 371–377. West, M.L. (1973), Textual Criticism and Editorial Technique Applicable to Greek and Latin Texts, Stuttgart. Wright, M. (2016), The Lost Plays of Greek Tragedy. Volume 1: Neglected Authors, New York/ London.

Stefano Vecchiato

Marginalia to Hesiodic Fragments A Possible Dis-attribution (Fr. 41 M.–W.), a Possible Attribution (Fr. 327 M.–W.), and Some Recently (Re-)discovered Fragments As the title suggests, my aim in this chapter is limited: first, I shall focus on problems surrounding the attribution of two ‘Hesiodic’ fragments to a work attributed to ‘Hesiod’ (the first fragment will be used as a ‘negative’ case, the second as a more-or-less ‘positive’ one); second, I shall survey recent minor findings pertaining to the corpus of ‘Hesiodic’ fragments, and I will concentrate especially on a tantalising new quotation attested in P. Gen. inv. 161 fr. 5 = col. iii. By analysing particular problems, I hope to convey broader reflections on the issues that Fragmentology poses to scholars and students, principally with respect to establishing the most appropriate placement of a given fragment within a critical edition.

 On placing indirectly transmitted fragments of Hesiod: two case-studies Let us consider first Hesiod, fr. 41 M.–W. = 183 M. = *23 H., which is included among the fragments of the Catalogue in the editions of Merkelbach/West (1967) and of Most (2007), (2018), and along with the fragments of the Catalogue and/or the Megalai Ehoiai in the commented edition of Hirschberger (2004): Ps.-Herodianus, Philet. 242, p. 66 Dain: τὰ δὲ ἐκ τόπου ἰσοδυναμεῖ τῇ “ἐξ” προθέσει μετὰ γενικῆς πτώσεως, οἷον “Θήβηθεν”, ἐκ Θηβῶν, “Ἀθήνηθεν παραγίνομαι”, ἀντὶ τοῦ “ἐξ Ἀθηνῶν”. ὅθεν οἱ λέγοντες “ἐξ οἴκοθεν παραγίνομαι” ἁμαρτάνουσι· δὶς γὰρ τὴν “ἐξ” πρόθεσιν παραλαμβάνουσιν. σημειωτέον οὖν τὸ ὁμηρικὸν “ἐξ οὐρανόθεν” (Il. 8.19, 21; 17.548) καὶ τὸ παρὰ τῷ Ἡσιόδῳ· ἐγὼ δ᾽ ἐξ ἀγρόθεν ἥκω

 I thank the organisers of the wonderful panel in Montréal and editors of the present volume, Francesco Ginelli and Francesco Lupi. I am most grateful for helpful advice to Patrick J. Finglass, Francesco Lupi, Enrico Emanuele Prodi, Luca Ruggeri, Christos Tsagalis, and the anonymous referees. Any remaining shortcomings, infelicities, and errors are mine alone. Unless otherwise noted, the translations are mine. https://doi.org/10.1515/9783110712223-002

  Stefano Vecchiato (The -θεν suffix showing) movement from a place has the same value as the preposition ἐξ plus genitive, as (in the following examples): Θήβηθεν, which means ‘from (ἐκ) Thebes’, Ἀθήνηθεν παραγίνομαι, which stands for ‘[I come] from (ἐξ) Athens’. Therefore, those who say ἐξ οἴκοθεν παραγίνομαι [‘I come from the house’] are wrong, for they use twice the preposition ἐξ (in the same syntagm). Exceptions to this rule are the Homeric (formula) ἐξ οὐρανόθεν [‘from the sky’] and the Hesiodic line: ‘I have come from the countryside’.

This citation, first published as a ‘Hesiodic’ fragment in Merkelbach and West’s edition, comes from the abridgement of the Atticist glossary entitled Φιλέταιρος dating in its full form to some time between the second and fourth centuries CE and wrongly assigned to Herodian.1 As Richard Reitzenstein has shown,2 this treatise may contain loans from Herodian’s Συμπόσιον, which probably included not only etymologies, but also Atticist doctrines, like the one attested in the citation preceding our fragment.3 Attic authors avoid using both the preposition ἐξ and the suffix -θεν in the same syntagm. The compiler of the glossary informs his readers that Homer and Hesiod furnished exceptions (σημειωτέον) to this practice, the former by employing the phrase ἐξ οὐρανόθεν, the latter by using the expression ἐγὼ δ᾽ ἐξ ἀγρόθεν ἥκω, which happened to be similar to the standard expression ἀγρόθεν ἥκειν that, as we know from a passage of Pollux’s Onomasticon, was an ‘Attic formation’ (Ἀττικὸς σχηματισμός).4 The fragment under scrutiny derives from direct speech, as is clear from the use of ἐγώ and the first-person verb ἥκω. Τhe entire expression ἀγρόθεν ἥκειν is a hapax in Hesiodic poetry. Neither ἀγρόθεν nor the first-person verb ἥκω are ever attested elsewhere, even separately, in the entire Hesiodic corpus. The word ἀγρόθεν is employed twice in Homer together with a verb of movement (Od. 13.267–268: τὸν μὲν ἐγὼ κατιόντα βάλον χαλκήρεϊ δουρὶ/ἀγρόθεν ‘I struck him with my bronze-tipped spear as he returned from the field’; Od. 15.428: ἀγρόθεν ἐρχομένην ‘as I was coming from the fields’).5 It is difficult to divine the exact metrical position this Hesiodic poetic fragment occupied within the hexameter:

 1 See Dyck (1993) 791–792, Dickey (2014) 341 (no. 56). 2 Reitzenstein (1897) 299–312. 3 On Herodian’s Συμπόσιον, see Dyck (1993) 790, Dickey (2014) 314 (no. 54). 4 Poll. 9.12.3 Bethe: οἱ μὲν Ἀττικοὶ σχηματισμοὶ ἀγρόνδε ἐλθεῖν καὶ ἀγρόθι οἰκεῖν καὶ ἀγρόθεν ἥκειν, καὶ εὐθὺ τῶν ἀγρῶν ἀπελθεῖν, καὶ ἔστε πρὸς τὸν ἀγρόν ‘The Attic formations [are] ‘to go to the country’ and ‘to live in the country’ and ‘to come from the countryside’, and ‘to come straight from the fields/countryside’ and ‘to go towards the field/in the direction of the countryside’. 5 For a relevant example in which ἐγώ, an adverb ending in -θεν, and ἥκω appear in the same verse, see Hom. Il. 5.478: καὶ γὰρ ἐγὼν ἐπίκουρος ἐὼν μάλα τηλόθεν ἥκω ‘for I who am an ally have come from very far’.

Marginalia to Hesiodic Fragments  

the only other occurrence of the first-person verb ἥκω in archaic epic poetry is at line-end (Hom. Il. 5.478, quoted above n. 5). Accordingly, one may suggest that the preserved expression occurred after the third-foot trochaic caesura, but there is no certainty. Scholars have assigned this fragment, which is cited without an ascription, to different works belonging to the Hesiodic corpus. Let us review, one by one, the suggestions made by various scholars to date. Jacques Schwartz argued that the fragment under examination comes from the Cheironos Hypothekai.6 He based his argument on probability: since in Schwartz’s time traces of dialogue were attested only in ‘Hesiodic’ fragments belonging to the Wedding of Ceyx, the Melampodia, the Cheironos Hypothekai, and the Aigimios, he disregarded the Catalogue of Women as a possible candidate. Instead, Schwartz opted for the Cheironos Hypothekai because he thought that it contained the most likely context for someone’s return from the fields. Schwartz was probably thinking that Cheiron was the speaker, and he postulated a scene in which the Centaur was instructing the young Achilles on hunting. The publication of many more papyrus fragments after 1960, and their inclusion in the edition of Hesiod’s fragments by Merkelbach and West, gave scholars a better knowledge of the style of ‘Hesiodic’ poetry and particularly the ‘Hesiodic’ Catalogue of Women. Given that dialogue is also attested in this epic,7 the Catalogue of Women was added to the list of potential ‘candidates’ with respect to the provenance of the fragment under examination. Merkelbach and West opted for placing this fragment in the Catalogue of Women (fr. 41 M.–W.) on the basis of Pi. P. 4.102 ἀντρόθε γὰρ νέομαι (‘I am returning from the cave’), a sentence uttered by Jason to Pelias after his return from Cheiron’s cave, where the hero had spent twenty years being educated by the Centaur and his family.8 West even entertained the thought that the reading ἀγρόθεν (‘from the countryside’), reported by ‘Herodian’, could be changed to ἀντρόθεν (‘from the cave’), as in the Pindaric line just quoted.9 Since we know that ‘Hesiod’ mentioned the episode of Jason’s upbringing by Cheiron (in all probability in the Catalogue of Women, fr. 40 M.–W.: Αἴσων, ὃς τέκεθ’ υἱὸν Ἰήσονα ποιμένα λαῶν,/ὃν Χείρων ἔθρεψ’ ἐνὶ Πηλίῳ ὑλήεντι  6 See Schwartz (1960) 242. 7 See e.g. Cat. fr. 43a.41–43 M.–W. = 69 M. = 37 H., fr. 75.13–25 M.–W. = 48 M. = *3 H., fr. 76.9–23 M.–W. = 48.34–48 M. = *4 H., fr. 165.3 M.–W. = 117 M. = 72 H. (clearly the end of a direct speech; see West’s secure supplement in 165.4: ἦ ῥ᾽· ὃ δὲ ‘so he spoke’), fr. 211.7–13 M.–W. = 152.14–20 M. = 100 H., etc. 8 On the Pindaric passage and on the myth of Jason’s rearing by Cheiron, see Giannini (1995) 457–459. 9 See Merkelbach/West (1967) ad loc.: ‘porro ἀντρόθεν conicere possis’.

  Stefano Vecchiato ‘Aeson, who begot Jason, the shepherd (leader) of the people, as his son, whom Cheiron raised on wooded Pelion’, transl. G.W. Most), it is possible that the Hesiodic citation we know via ‘Herodian’ comes from the Catalogue of Women too. In her commentary on the Catalogue of Women and the Megalai Ehoiai, Martina Hirschberger (2004) 483 placed fr. 41 M.–W. (= *23 Hirschberger) in the class ἐξ ἀδήλων ἐπῶν (incerti operis), ‘of uncertain work’. This seems a more cautious placement, though Hirschberger seems to be convinced that the fragment under examination comes either from the Catalogue of Women or the Megalai Ehoiai. Finally, Most (2007) 260 assigned the fragment to the Catalogue of Women, like Merkelbach and West, but changed its placement. Instead of postulating a context involving Jason, Most, confessing his uncertainty, placed it among ‘fragments about unidentified myths’ in the Catalogue of Women (fr. 183 Most).10 In sum, assigning this fragment to the ‘Hesiodic’ genealogic-catalogic poems (either the Catalogue of Women and/or the Megalai Ehoiai) seems to have become common practice. At this point, it seems appropriate to question whether West’s proposal is actually well-grounded. After all, West’s proposal is the only one for which arguments were provided and it has been basically accepted — albeit with slight modifications — by the editors of the two major ‘Hesiodic’ fragments published after the epoch-making edition of Merkelbach and West. In my view, as I will try to make clear, it is open to serious objections. West’s arguments supporting the attribution of this fragment to the Catalogue, as opposed to any of the other works assigned to Hesiod, is based on the alleged similarity between the ‘Hesiodic’ half-line and the aforementioned Pindaric expression employed by Jason. Hence the collocation of this fragment within the Jason-Cheiron story narrated in the Catalogue.11 To be fair, ‘returning/  10 The placement has not changed in Most’s second edition (Most 2018). 11 As far as this section of the Catalogue is concerned, one should note that the fragment following fr. 41 in M.–W.’s edition, fr. 42 M.–W. (on Cheiron’s wife, a Naiad nymph or a nymph called Nais), is only conjecturally ascribed not only to the Catalogue of Women, but also to this particular myth. Most (2007) places fr. 42 M.–W. amid the ‘Catalogi fragmenta incertae sedis’ (fr. 163 Most) on the basis of the inclusion in his edition of P. Oxy. 2509 (fr. 162 Most = 103 Hirschberger) on Actaeon’s dogs. The papyrus preserves part of a dialogue between a goddess (Artemis, or, less likely, Athena) and Cheiron. Before the dialogue, the goddess arrives at Cheiron’s μέγα σπέος (v. 2) ‘great cave’, where the Centaur ἔνα[ιε/… νηΐδ’ ἔχων νύμφην θυμαρέ’ ἄκ[οιτιν (2–3), ‘(there) dwelled (Cheiron), who had a Naead nymph as his well-pleasing wife’ (transl. G.W. Most). Is this verse that fr. 42 M.–W. = 163 M. should actually be referred to? For an in-depth discussion on P. Oxy. 2509 and its possible pertinence to the ‘Hesiodic’ Catalogue, see Janko (1984), Hirschberger (2004) 393–397, with further references. Finally, an attribution of fr. 42 M.– W. to the Cheironos Hypothekai cannot a priori be excluded.

Marginalia to Hesiodic Fragments  

coming from the countryside’ is different from ‘returning/coming from the cave’, as it is the word ἄντρον that specifically evokes Cheiron’s abode. So, West’s suggestion seems to work only if one accepts his aforementioned conjecture ἐγὼ δ᾽ ἐξ ἀντρόθεν ἥκω based on the Pindaric passage in P. 4; if not, the reference is deprived of this specific attribution. To emend the transmitted text in this case is, in my view, arbitrary; but it should be noted that if, following West, one decides to emend the text, the conjecture ΑΝΤΡΟΘΕΝ for ΑΓΡΟΘΕΝ is clearly neither the sole option, nor the most economical. A more defensible emendation would be, for example, ΑΓΡΟΘΕΝ into ΑΡΓΟΘΕΝ ‘from the city of Argos’/‘from the Argolis’.12 This exempli gratia alternative would radically change our view on the fragment and the implications envisaged by West with regard to its placement. In any case, the main point which should be stressed again is that there are no cogent reasons to change the transmitted text, which is perfectly sound both in terms of metre and in terms of grammar and provides a plausible meaning. Therefore, when all is said and done, I think that the most reliable placement of this fragment should be tout court among the fragments of uncertain placement (fragmenta incertae sedis, frr. 303–342 M.–W.). My verdict is based on a balanced consideration of the relevant data: to put it briefly, there is simply neither a decisive factor for the placement of this fragment in any given ‘Hesiodic’ poem, nor the slightest contextual hint suggesting the relevance of the fragment to the Catalogue, much less to the Jason-Cheiron story narrated therein.13 After this ‘deconstructive’ section, let me briefly suggest a more (albeit speculative) ‘constructive’ hypothesis about the possible placement(s) of a Hesiodic incertae sedis paraphrase fragment (fr. 327 M.–W. = 277 M.) which, as far as I know, has not been commented upon by scholars so far.  12 If we follow this suggestion, it should be noted that before citing the half-verse from Hesiod, ‘Herodian’ mentions examples such as ἐκ Θηβῶν/Θήβηθεν, ἐξ Ἀθηνῶν/Ἀθήνηθεν and ἐξ οἴκοθεν (παραγίνομαι) as if they belonged to the same class. It may well be that he also quoted Homer with a ‘non-city example’ (ἐξ οὐρανόθεν) and then Hesiod with a ‘city example’ (Ἀργόθεν instead of ἀγρόθεν). The list below furnishes two examples from Homer providing a parallel for the structure ‘preposition showing movement from a place + name of a city + -θεν suffix + verb of movement’, and is indicative of how Ἀργόθεν could fit in a relevant context: 1. Il. 24.492: ὄψεσθαι φίλον υἱὸν ἀπὸ Τροίηθεν ἰόντα; 2. Od. 9.38: ὅν μοι Ζεὺς ἐφέηκεν ἀπὸ Τροίηθεν ἰόντι. I would like to stress that the point here is not to propose an alternative to the conjecture proposed by West, but simply to show that the latter is not the only possibility and, above all, that emending the paradosis would be motiveless, in this case. See further below. 13 Dräger (1993) 99–101 argues on mythographical basis against the attribution of fr. 41 M.–W. to the Jason–Cheiron episode narrated in the Catalogue, but he does not seem to be willing to propose a complete dismissal of the fragment from that poem.

  Stefano Vecchiato Audacis Excerpta, gramm. VII p. 332.5–7 Keil: qui primum his observationibus in componendis carminibus usi sunt? Phemonoe dicitur Apollinis vates prima per insaniam ita locuta, cuius Hesiodus14 meminit. Who were the first ones who made use of these observations [scil. with respect to the hexameter] in composing their poems? Phemonoe, the priestess of Apollo, is said to have been the first to speak in this way, because of insanity; Hesiod mentions her. (Transl. G.W. Most)

This citation comes from the Excerpts from the Books on Grammar of Scaurus and Palladius by the Latin grammarian Audax.15 The context deals with a general discussion on the metrical form of the hexameter, and Phemonoe, the priestess of Apollo, is said to be the first to use it, presumably in order to utter prophecies; Audax then adds that Hesiod mentioned her. As far as this point is concerned, I am not sure whether or not the text allows us to go so far as to argue that Hesiod declared Phemonoe to have been the first person to speak in hexameters due to the Apollonian possession; the only certain fact that we can gather from this excerpt seems to be that he mentioned her somewhere. Who was this Phemonoe and how could she be most plausibly placed in the poems attributed to Hesiod? Various sources concur with Audax’s notice in making Phemonoe the first priestess of Apollo and thus the first Pythia to utter prophecies in hexameters.16 A scholium to Euripides’ Orestes is relevant for the genealogy of Phemonoe (schol. MTAB Eur. Or. 1094 Schwartz): Ὕαμος ὁ Λυκώρου μετὰ τὸν κατακλυσμὸν βασιλεύων τῶν περὶ τὸν Παρνασὸν Ὕαν πόλιν ἔκτισε καὶ γήμας Μελάνθειαν τὴν Δευκαλίωνος ἐποίησε Μελαινίδα· ἐκ δὲ Μελαινίδος Δελφὸς, ὃς τὴν ἀρχὴν τοῦ μητροπάτορος παραλαβὼν ἔγημε Κασταλίαν, ἀφ’ ἧς ἡ κρήνη τοὔνομα ἔχει, καὶ ἔσχε Καστάλιον καὶ Φημονόην, ἥν φασι πρώτην ἑξαμέτρῳ χρῆσαι. πόλιν δὲ κτίσας ὁ Δελφὸς Δελφίδα ὠνόμασε. Hyamos of Lycoros, who reigned over the region of Parnassus after the Flood, established the city of Hya and having married Melantheia, the daughter of Deucalion, begot Melainis. Delphos was the son of Melainis who took over the kingdom of his mother’s father and married Castalia, from whom the spring gets its name, and he begot Castalios and Phemonoe, whom they say was the first to make utterances in hexameters. Having founded a city, Delphos named it Delphi.

 14 Sittl (1890) 56–57 proposed, unconvincingly in my opinion, to change the reported reading Hesiodus of all the mss. into Heliodorus, believing that the authority meant here was the ancient grammarian and metrician Heliodorus (first century CE). Against this emendation see also the objections raised by Schwartz (1960) 74–75. 15 On this grammatical work, see Goetz (1896). 16 See Strab. 9.3.5, Paus. 10.5.7, 10.6.7; cf. also Paus. 10.12.10.

Marginalia to Hesiodic Fragments  

According to this source, Phemonoe was the daughter of Delphos, future founder of the town of Delphi, and of Castalia, after whom the famous spring of Mt Parnassus was named. Delphos was in turn the son of Melainis, daughter of Melantheia who was the daughter of Deucalion.17 Phemonoe therefore would be Deucalion’s great-great-granddaughter. It should be noted that the information provided by the Orestes scholium that Melantheia (the great-grandmother of Phemonoe) was Deucalion’s daughter seems to match the evidence coming from an anonymous source quoted shortly after the citation of Hesiod’s authority (fr. 4 M.–W. = 5 M.) for the detail of the name of Deucalion’s mother (schol. Hom. Od. 10.2 Dindorf, cf. Hes. fr. 4 M.–W. = 5 M.): Δευκαλίων, ἐφ’ οὗ ὁ κατακλυσμὸς γέγονε, Προμηθέως μὲν ἦν υἱὸς, μητρὸς δέ, ὡς οἱ πλεῖστοι λέγουσι, Κλυμένης, ὡς δὲ Ἡσίοδος, †Πρυνείης (v.l. Πρυνόης, Προνόης Dindorf),18 ὡς δὲ Ἀκουσίλαος (fr. 34 Fowler), Ἡσιόνης τῆς Ὠκεανοῦ [τοῦ Προμηθέως]. ἔγημε δὲ Πύρραν τὴν Ἐπιμηθέως καὶ Πανδώρας … γίνονται δὲ τῷ Δευκαλίωνι θυγατέρες μὲν δύο Πρωτογένεια καὶ Μελάνθεια, υἱοὶ δὲ Ἀμφικτύων καὶ Ἕλλην. τοῦ Προμηθέως del. West : καὶ Προμηθέως Sturz Deucalion, in whose time the Flood happened, was the son of Prometheus, and his mother, according to most of the sources, was Clymene, but according to Hesiod was †Pryneie (v.l. Prynoe, Pronoe), and according to Acusilaus was Hesione the daughter of Ocean [the son (?) of Prometheus]. He married Pyrrha who was the daughter of Epimetheus and Pandora … Deucalion had two daughters who were called Protogeneia and Melantheia, and two sons called Amphictyon and Hellen.

It is interesting that shortly after the start of the modern reconstruction of the ‘Hesiodic’ Catalogue of Women, within the section concerning Deucalion and his children, the offspring of a daughter of Deucalion, Thyia, who is not mentioned in the account preserved in the Odyssey scholium quoted above, is discussed. Thyia slept with Zeus and gave birth to Macedon, eponymous hero of the Greek region, and to Magnes (Hes. fr. 7 M.–W. = 7 M. = 3 H.). We are further informed (Hes. fr. 8 M.–W. = 8 M. = 42 H.) that Magnes begot Dictys and Polydectes: therefore, in the poem the record of Thyia’s descendants goes at least as far as Deucalion’s great-grandchildren (Dictys and Polydectes). Then, in Hes. fr. 9 M.–W. (= 9 M. = 4 H.), the account of the offspring of Hellen, Deucalion’s son, begins. Analogously, it may well be that Hesiod also briefly treated Melantheia’s lineage, as

 17 For a survey of the various traditions about Delphos, see Waser (1901). 18 On the textual problem regarding the name of Deucalion’s mother in the Catalogue, see West (1985) 50–51.

  Stefano Vecchiato he did with Thyia’s offspring in frr. 7–8 M.–W., and that he mentioned Melantheia’s great-granddaughter, Phemonoe. Unfortunately, the only securely attested daughters of Deucalion in the Catalogue are Pandora, Thyia, and Protogeneia.19 There can be no certainty regarding Melantheia,20 who, as we have seen, appears in a genealogy not explicitly attributed to ‘Hesiod’, and so may well come from a different source. Although we are on thin ice, in the light of the above data one can speculate, in my view, on two possibilities: (a) Phemonoe may not have been mentioned in the Catalogue of Women, but in the other genealogical poem attributed to ‘Hesiod’, the Megalai Ehoiai. Admittedly, although no sources confirm that Deucalion’s offspring were discussed in this poem as well, at the same time no single element excludes this possiblity. Angelo Casanova noticed that, according to the surviving fragments, the ME often adds ‘new’ members to the offspring of characters already mentioned in the Catalogue.21 This observation would apply perfectly to the present case; however, the exact structure of the ME, as opposed to the Catalogue, is hotly debated, and scholarly consensus on the very nature of this poem has not yet been reached.22 (b) Taking into account Phemonoe’s close relationship with Delphi, Apollo, and the prophetic sphere (as noted above, her father was Delphos, founder of Delphi, and Apollo made her first priestess and seeress of the Delphic shrine), one may suggest that Phemonoe occurred in the only poem which treats both prophecy and genealogy attributed to ‘Hesiod’ — the Melampodia. This poem featured not only the deeds of Melampus, but also those of his descendants and of the competing branch of seers — Teiresias and his offspring.23 Apollo seems to have played a considerable role in the poem: the Melampodia certainly mentioned Claros (fr. 278 M.–W. = 214 M.), a place near Colophon where a prophetic shrine strongly influenced by Delphi and Apollonian elements was founded by

 19 See West (1985) 50–56; for an in-depth analysis of Deucalion’s family, with a specific focus on Pandora, see Casanova (1979a) 135–187. 20 Cf. Most’s cautious rubric to Hes. frr. 2–6 M. (2007, 43): “Deucalion’s children: Pandora, Thyia, Hellen; Graecus, Protogeneia, Melantheia?” 21 See Casanova (1979b) 236. He detected eight secure cases out of seventeen fragments. 22 An excellent survey of the ME fragments and relevant problems thereof can be found in Hirschberger (2004) 81–86, and especially D’Alessio (2005). 23 On the Melampodia and its structure, see Schwartz (1960) 210–228, Löffler (1963), Cingano (2009) 121–123, more specifically Cozzoli (2016) 153: “l’opera si articola soprattutto sotto forma di una contesa tra due rami genealogici d’indovini quello di Melampo con Calcante e Anfiloco e quello di Tiresia con Manto e Mopso, uno minio (-pilio-tessalico) e uno tebano e si conclude nella definitiva scomparsa del primo ad opera del secondo”.

Marginalia to Hesiodic Fragments  

Teiresias’ daughter Manto.24 Apollo himself killed Amphilocus, the last descendant of the Melampodids (fr. 279 M.–W. = 215 M.), and this fact seems to contextualise the Melampodia as the story of the foundations, by families of seers, of shrines and sanctuaries which were established in competition with Delphi under divine protection.25 In a narrative pattern of this sort, the mention of Phemonoe, devoted priestess of Apollo, would not be surprising. After all, in the Greek tradition Phemonoe soon became an exemplary ancestral and mythical seeress, like the protagonists of the Melampodia. According to Clement of Alexandria (Strom. 1.21.107), Phemonoe lived before legendary seers and singers of hexameters such as Musaeus, Orpheus, and Linus.26 In another passage of Clemens’ Stromateis (1.12.134), Phemonoe is presented as the Delphic counterpart to the most renowned seers of the Greek mythical past who are connected to different geographical areas, such as Idmon, the seer of the Argonauts, Mopsus, the son of Teiresias’ daughter Manto, Amphilochus, and Alcmaeon (note that Teiresias, Manto, Mopsus, and Amphilochus all featured in the Melampodia). All this is, of course, speculative, and a good deal of caution should be exercised: the safest decision is to keep the fragment under examination within the fragmenta incertae sedis (fragments of uncertain placement) class, perhaps signalling these possibilities in the apparatus.

 ‘Better than nothing’: recent Hesiodic trouvailles The time of revolutionary papyrus discoveries of the last century seems over as far as the works of ‘Hesiod’ are concerned, and no major papyrus has been unearthed from the sands of Egypt for decades. Yet, in the last years, the patience and precision of various scholars have allowed for minor findings that have enlarged the Hesiodic corpus, albeit slightly. I summarise them below with a brief discussion focusing mainly on their nature and placing:  24 See Momigliano (1934), Cozzoli (2016) 151–153. 25 Cozzoli (2016) 150 notes that it “sembra collocare l’opera [i.e. the Melampodia] nell’ambito della storia sacra di un santuario come quello di Claro che nasce in competizione con Delfi sotto la protezione del dio”. 26 Note that in an uncertain poem ‘Hesiod’ touched upon the figure of the seer Linus; he is said to be the son of the Muse Ourania and the object of laments of bards and lyre-players at every banquet and dance (Hes. fr. inc. sed. 305 M.–W. = 255 M. = *11 H.); he is further credited with being ‘knowing in all kinds of expertise’ (Hes. fr. inc. sed. 306 M.–W. = 256 M. = *12 H.).

  Stefano Vecchiato (a) A new papyrus fragment of Book 1 of the Catalogue of Women has appeared. In 2013 Danbeck managed to combine two tiny papyrus scraps, P. Oxy. 2075 fr. 5 + P. Oxy. 2481 fr. 12 (already classed respectively as fr. 117 M.–W. = 8 H. and fr. 85 M.–W. = 120 H.) from rolls containing fragments of Book 1 of the Catalogue in a single fragment, ‘by virtue of their transmitting the ends and beginnings, respectively, of three verses nearly identical to the description of Trojan Erichthonios’ wealth at Iliad 20.220–2’.27 The resulting fragment, however, is not particularly revealing, and it is difficult to ascertain its exact topic and the character(s) it treats. Nonetheless, given the position of the fragments in their respective rolls, it seems fair to agree with Danbeck that this new fragment ‘belongs either closely preceding fr. 23a [scil. M.–W.] or following fr. 26 [scil. M.–W.]: the extended family being treated is that of a female Aiolid, either a Thessalian or Aitolian branch’.28 (b) There has also been the publication of a new one-word fragment, i.e. πορθμός, ‘strait’ (the exact inflection is not determinable), preserved in a passage of Pollux’s Onomasticon (9.18 Bethe). This fragment, first discovered in the ms. Laurentianus plut. 56.1 (= L) by Maass in 1880, and duly included and signalled by Bethe (with the bracketed indication ‘fr. nov.’) in his edition of Pollux, was surprisingly neglected by the subsequent editors of Hesiodic fragments, until Matijašić brought it back to the attention of the scholarly community.29 Matijašić cautiously speculates about the possible work in which this word was used by ‘Hesiod’, and identifies two fragments (possibly from the Catalogue) in which the word may have appeared.30 In my view, given the lack of conclusive data, the safest decision will be to place it amongst the fragmenta incertae sedis.31 (c) Also significant is a passage of Plutarch’s De Herodoti malignitate (14.857f), omitted by all the editions of Hesiodic testimonia and/or fragments, which relates that Hesiod was in agreement with many other poets, such as Homer, Pisander, Archilochus, and Pindar, in ignoring an Egyptian and Phoeni-

 27 Danbeck (2013) 15. 28 Danbeck (2013) 30. This is now Hes. fr. 307 Most2 in the Addenda fragmentis section of Most (2018) 402–403. 29 See the detailed discussion of the unfortunate history of this fragment in Matijašić (2017) 340–342. 30 Both are paraphrase, not verbatim, fragments: Hes. fr. 149 M.–W. = fr. 245 M. apud D.S. 4.85.4–5 (where the strait of Messina could have been mentioned); fr. 188A M.–W. = fr. 130 M. apud P. Mich. inv. 1447, col. ii, ll. 7–9 (where the strait between Boeotia and Euboea is mentioned): see Matijašić (2017) 343–344. 31 Most (2018) 404 now prints it as fr. 309 Most2 in the Addenda fragmentis section of his edition.

Marginalia to Hesiodic Fragments  

cian version of Heracles, and in depicting the hero only with Boeotian and/or Argive traits. Given how general the content is, I proposed to classify Plutarch’s passage amid the “Miscellaneous Judgements” testimonia on Hesiod (cf. Hes. TT 152– 157 Most).32 Most (2018) 404 now prints it as a ‘proper’ fragmentum (fr. 308 Most2), unconvincingly in my opinion, in the Addenda fragmentis section. (d) There have also been two neglected fragments erroneously attributed by the sources to Hesiod — two new spurious fragments,33 preserved by two entries of Byzantine lexicographical works (EtGen AB s.v. ἱμερτῆς = Dion. Per. 537; EtGud s.v. ἀφωσιωμένος De Stefani, which I have tentatively attributed to the shadowy tragic poet Isidorus, TrGF 1 211).34 To these findings one may now add the following: (1) Ps.-Ammon., Diff. 340, p. 88.11–13 Nickau: ξυστὸν τό τε ἀκόντιον ὡς παρ’ †ἡσιόδω† (codd., Ἡροδότῳ Vulcanius) ἐν τῷ πρώτῳ (cf. Hdt. 1.52), καὶ τὸ οἰκοδόμημα ὡς παρὰ Ξενοφῶντι ἐν Οἰκονομικῷ (X. Oec. 11.15). As I have shown elsewhere,35 the reading ἡσιόδω (sic, read Ἡσιόδῳ), shared by all the manuscripts of Ps.-Ammonius’ De adfinium vocabulorum differentia, can be defended on good grounds; if we accept the paradosis, we can lay hands on a new one-word fragment (ξυστόν, ‘javelin’) from Book 1 of the Hesiodic Catalogue of Women. Bonaventura Vulcanius’ emendation of Ἡσιόδῳ to Ἡροδότῳ cannot, however, be ruled out, given that also the word ξυστόν (albeit with a slightly different meaning from the one attested in Ps.-Ammonius’ entry) does, in fact, appear in Hdt. 1.52, and that the confusion between the names of Hesiod and Herodotus is found in various textual traditions.36 Therefore, perhaps, the best placement of Ps.-Ammonius’ entry is amid the fragmenta dubia. (2) A further and ‘virtual’, so to speak, one-word quotation fragment (this time embedded in a paraphrase fragment) can perhaps be detected in Hes. fr. 263 M.–W. = 202 M., from The Wedding of Ceyx (apud schol. A.R. 1.1289, p. 116 Wen-

 32 See Vecchiato (2017) 120. 33 The recovery of spurious fragments of a given author may seem at first sight to be a frivolous exercise. However, in my view such fragments testify to an active yet indirect, perhaps even subconscious reception of that author through the ages. Therefore, collecting and studying them may be useful both from a historical and a philological point of view. 34 See the discussion and analysis of these fragments in Vecchiato (2019a). 35 See Vecchiato (2019b). 36 Cf. for instance the examples quoted in Vecchiato (2019b) 648 n. 7, reported here in a slightly modified form: Plin. nat. 15.3 = Hes. fr. 347 M.–W. (perhaps Hdt. 5.82.2 is meant, but the issue is very uncertain), Arist. HA 601b = Hes. fr. 364 M.–W. (cf. Hdt. 1.106.2, but see Olson 1987), Strab. 1.3.18 = Hes. fr. 368 M.–W. (cf. Hdt. 2.10.3).

  Stefano Vecchiato del): Ἡσίοδος ἐν τῷ Κήυκος γάμῳ ἐκβάντα φησὶν αὐτὸν [scil. Heracles] ἐφ’ ὕδατος ζήτησιν τῆς Μαγνησίας περὶ τὰς ἀπὸ τῆς ἀφέσεως αὐτοῦ [scil. Heracles] Ἀφετὰς καλουμένας ἀπολειφθῆναι.37 One can conjecture that the etymology of Aphetae (which will have been present in the poem, and possibly elucidated by the poet himself),38 a place on the Pagasaean Gulf named after Heracles’ abandonment (ἄφεσις) by the Argonauts, was explained in the poem by the use of a form of the verb ἀφίημι, attested in the epic diction with the meaning of ‘send away, let go, loose’.39 In fact, both the derivative (cf. Chantraine, DELG, and Frisk, GEW, s.v. ἵημι) noun ἄφεσις and the adjective ἄφετος appear to be later and nonepic forms: the former is first attested in Isocr. Or. 17.29, while the latter in ‘Aesch.’ PV 666. For an analogous case, one may compare the explanation of the name of the island Anaphe from the verb ἀναφαίνω in A.R. 4.1717: … Ἀνάφην δέ τε λισσάδα νῆσον/ἴσκον, ὃ δὴ Φοῖβός μιν ἀτυζομένοις ἀνέφηνεν. If this line of reasoning is right, this would be the first instance of the verb ἀφίημι in the Hesiodic corpus (the exact inflection, however, would not be determinable), and at the same time a ‘new’ one-word fragment of the Wedding of Ceyx. (3) Orion, Etym. s.v. ἀήτης, p. 19 Sturz: ἀήτης, ὁ ἄνεμος, παρὰ τὸ ἀεῖν καὶ πνεῖν. ἀῶ, ἀήσω, ἀήτης. οὕτως Ἡσίοδος φησίν· διασκιδνᾶσιν ἀέντος. I take the quotation διασκιδνᾶσιν ἀέντος to be actually a corruption of διασκιδνᾶσιν ἀέντες which occurs in Hom. Il. 5.526, an expression which will have been wrongly attributed by Orion (or one of his sources) to Hesiod possibly by means of the reminiscence of Hes. Th. 875 ἄλλοτε δ’ ἄλλαι ἄεισι διασκιδνᾶσί τε νῆας.40 Accordingly, Or. Etym. s.v. ἀήτης should be placed amongst the fragmenta spuria (spurious fragments).41

 37 For an analysis of Hes. fr. 263 M.–W, see Merkelbach/West (1965) 301–302; on Aphetae’s etymology, cf. also Most (2007) 279 n. 9. 38 So Merkelbach/West (1965) 302; the taste for etymological plays displayed through the whole ‘Hesiodic’ corpus is of course well-known. 39 See LSJ9 s.v., II 1a–b; cf. esp. Hom. Il. 1.25 = 379 ἀλλὰ κακῶς ἀφίει. 40 The possibility that one of the two aforementioned verses should be read in Orion’s entry was first envisaged by Larcher (apud Sturz 1820, 19). For an analogous case (a Homeric line wrongly attributed to Hesiod due to reminiscence in wording of a Hesiodic line), one may compare Hes. fr. sp. 406 M.–W. apud EtGen AIAIIB s.v. εἰλυφάζων; as far as this fragment is concerned, I do not share the invasive proposals of Casadio (1990/1993), whose collation does not take into account the textual contribution of AII, which invalidates his emendations: see Baldi (2013) 184 ad EtSym ε 153. For another example, cf. also Hes. fr. sp. 403 M.–W. apud EM s.v. τῶ p. 773.22 Gaisford, with Sittl (1890) 49. 41 See also above n. 33. This source is at the same time a ‘new’ testimonium Homericum for Hom. Il. 5.526: it does not appear in the massive apparatus of West’s edition of the Iliad (1998, vol. 1)

Marginalia to Hesiodic Fragments  

(4) EtSym α 1074/75 Lasserre–Livadaras: ἀντιφερίζω. ἐναντιοῦμαι, ἐξισοῦμαι. παρὰ τὸ φέρω, φερίζω, ὡς θέρω, θερίζω. Ἡσίοδος. καὶ ἀντιφέρεσθαι τὸ ἐναντιοῦσθαι λέγει.42 The verb λέγει seems to have as implicit subject Ἡσίοδος, mentioned shortly before in the same entry of the Etymologicum Symeonis as the authority of the lemma ἀντιφερίζω, i.e. Hes. Op. 210 (see below). The source for this passage of the EtSym is EtGen AB α 926 Lasserre–Livadaras: ἀντιφερίζω (Hes. Op. 210)· τὸ ἐναντιοῦμαι ἢ ἐξισοῦμαι· παρὰ τὸ φέρω φερίζω, ὡς θέρω θερίζω, οἷον (l.c.)· ‘ἄφρων δ’, ὅς κ’ ἐθέλῃ πρὸς κρείσσονας ἀντιφερίζειν’· ἀμφότερον (op.cit. 211)· ‘νίκης τε στέρεται πρός τ’ αἴσχεσιν ἄλγεα πάσχει’· καὶ τῆς νίκης στερεῖται καὶ πρὸς τῇ αἰσχύνῃ ἄλγεα πάσχει. Ἡσίοδος. καὶ ἀντιφέρεσθαι (Hom. Il. 1.589 etc.), ἐναν-τιοῦσθαι.43 No instances of ἀντιφέρεσθαι occur in the Hesiodic corpus. The fact that the mention of ἀντιφέρεσθαι, a synonym of ἀντιφερίζειν appended to the end of the entry, should actually be referred to Hom. Il. 1.589 χραισμεῖν· ἀργαλέος γὰρ Ὀλύμπιος ἀντιφέρεσθαι is made clear by schol. ex. T Hom. Il. 1.589 Erbse: ἀντιφέρεσθαι: ἐναντιοῦσθαι ἢ ἐξισοῦσθαι, which presents the same interpretamentum as the Etymologica. Entia non sunt multiplicanda praeter necessitatem. The compiler or compilers of the EtSym entry seem to have wrongly concluded that the sentence regarding ἀντιφέρεσθαι in the EtGen should be referred to Hesiod, and to have added the verb λέγει of their own accord. So, another spurious fragment.44 (5) There has also been a new, albeit very lacunose, verbatim citation of a ‘Hesiodic’ verse (or verses?). This has resurfaced, after the recent new edition of Furley/Gysembergh (2017), in fr. 5 = col. iii of P. Gen. inv. 161 (second century CE), a papyrus containing scraps of an anonymous treatise on various practices of divination (such as extispicy and pyromancy). I print below fr. 5 of the papyrus according to the text established by Furley/Gysembergh, with the poetic quotation in bold type:

 ad loc., nor in his additions ad testimonia (West 2000, vol. 2, vi–vii), nor in the excellent complementary article of Kassel (2002). 42 This fragment was first pointed out by Sittl (1890) 48. He was not yet aware of the EtSym entry, but only the derivative passage of Ps.-Zonaras’ Lexicon (p. 216 Tittmann), which repeatedly draws on the EtSym: see Alpers (1972) 741–743, Baldi (2013) xli–xlii. Thus, Ps.-Zonaras’ entry does not have independent value, because it is copied from the EtSym. 43 See Lasserre/Livadaras (1992) 78 ad loc. On the EtGen as fundamental source of the EtSym, see Baldi (2013) xxvi–xxviii with further references. 44 See also above n. 33.

  Stefano Vecchiato .

. . [......].ων δ[ ].ν[.]ερων διεπ.[ ]ω̣ν χ̣αλεπῶν δ.[ ἐ]ν̣ τ̣ έ̣ λειαν ἔϲεϲθ[αι, ὥϲ φηϲι]ν̣ Ἡϲ̣ίο̣ δ̣οϲ· ἐϲθλ̣[ ]. ̣ἀ̣[έ]θλοιϲι χερειον[ ] ἱ̣ ε̣ροῖϲ θεῶν ὄντων .[ [....]αι ἐπεπτυγμέ[ναι [ἔϲο]ν̣ται ἐνκλημ̣ατ..[ . . .

5

The fact that the ‘Hesiodic’ quotation should be identified with the damaged or uncertain words at ll. 5–6, and not with what comes before and/or after, is guaranteed by two data: (i) after Ἡϲ̣ίο̣ δ̣οϲ in the papyrus there is a high stop; its function will have been to introduce the citation, which of course will have followed Hesiod’s mention at l. 5;45 (ii) the form ἀέθλοισι and the comparative χερείων (whose ending is irretrievable to us) at l. 6 are typical of poetry and in particular of epic diction, while the words in l. 7 are not (also, they do not fit in a hexameter, unless one assumes synizesis in θεῶν; more importantly, the form ὄντων of the participle of εἰμι is not attested in epic diction — it should be ἐόντων46). Accordingly, we obtain the remains of a previously unknown verse, or more, attributed to ‘Hesiod’: ἐϲθλ̣[ … ]. ̣ἀ̣[έ]θλοιϲι χερειον[… noble(s?) … (with?) contests (or prizes?47) … inferior(s? bad?) …

It is impossible to tell if the quotation extended up to the beginning of line 7. The purely exempli gratia reconstruction proposed by Furley and Gysembergh ἐϲθλοὶ δ᾿ οὐ χαλεποῖϲ ἀέθλοιϲι χερείονέϲ εἰϲιν is ruled out by the length of the lines (not to mention the violation of Hermann’s Bridge): as the editors themselves state, every line of the columns of this papyrus must contain approximately twenty letters.48 Even if we assume that the quotation reaches the beginning of line 7, the

 45 The reading of Hesiod’s name in the papyrus, despite the objections raised by Costanza (2018) 11, is almost certain: see Furley/Gysembergh (2018). 46 Cf., however, Hom. Od. 7.94 ἀθανάτους ὄντας καὶ ἀγήρως ἤματα πάντα. 47 The possibility that ἆθλα could here mean ‘prizes’, and not ‘contests’, as Furley/Gysembergh (2017) implicitly assume, cannot a priori be ruled out, given the absence of context. 48 Furley/Gysembergh (2017) 1.

Marginalia to Hesiodic Fragments  

proposed reconstruction is too long to be accommodated between the end of line 5 and the beginning of line 7. I do not want to propose an alternative reconstruction here, because I agree with the editors that “the Hesiodic quote cannot be definitely restored”,49 given the lack of contextual hints that could help in divining the function of a Hesiodic quotation within this context, and its proper meaning.50 Rather, I would like to propose a solution for an alleged metrical problem pointed out by A. Vergados. With respect to the sequence ] ̣ ἀ̣[έ]θλοιϲι χερειον[ at l. 6, Vergados observes “that ‐θλ‐ universally make position in Hesiod, creating a metrical problem. This might be an indication that the ‘quote’ is late and spurious”.51 One may add that there is not any single instance of -θλ- not making position not only in the extant Hesiodic corpus, but also in the extant hexametric poetry; thus, apparently, the aforementioned sequence could not run in a hexameter, the resulting scansion being ⏑––⏑⏑–⏑. However, the issue can be framed differently: as far as the opposition ἄεθλος vs ἆθλος is concerned, M.L. West noticed that “[i]t may be thought that this [scil. ἀεθλ-] is the lectio difficilior and ἀθλ- a modernisation, as in 435 [scil. Hes. Th. 435 ἐσθλὴ δ’ αὖθ’ ὁπότ’ ἄνδρες ἀεθλεύωσ’ (ἀθλεύωσιν a) ἐν ἀγῶνι]. But the introduction of epic and hyper-epic forms against the meter is not uncommon in the MSS. of Hesiod … It is likely enough that ἄεθλος belongs in this class”.52 This statement seems to be confirmed also by the following instances, not quoted by Vergados and Furley/Gysembergh: (1) at Hes. Th. 800 ἄλλος δ’ ἐξ ἄλλου δέχεται χαλεπώτερος ἆθλος, the reading of b is ἄεθλος;53 (2) at Hes. Op. 656 the reading of φ11 (= Vat. gr. 1332 [fourteenth century]) and ψ21 in marg. (= Vindob. phil. gr. 242 [a. 1452]) is ἄεθλ’ ἔθεσαν παῖδες μεγαλήτορες· ἔνθα μέ φημι, while all the other mss. have ἆθλ’.54 Only Rzach (19133) and Solmsen (19903) print ἄεθλ’, assuming synizesis.

 49 Furley/Gysembergh (2017) 13. 50 Furley/Gysembergh (2017) 12 suggested that “[t]he quote might contain a thought such as ‘the Good are not worse through trials’”, while Vergados, apud Furley/Gysembergh (2017) 12–13, “a sense such as ‘it is through (hard) contests that good men prove themselves superior to the χερείονεϲ’”. Actually, there are no elements in the remains of the text that quotes the verse(s) of ‘Hesiod’ that could allow for accepting either suggestion. 51 Vergados apud Furley/Gysembergh (2017) 12. 52 West (1966) 376. 53 b is a lost and important manuscript of the recensio of the Theogony; it can be reconstructed from five extant mss. See West (1964) 172–174, from whom I draw the siglum. 54 See Rzach (19133) ad loc. For these mss. of the Works and Days, I use the sigla employed by West (1974) 177, 183.

  Stefano Vecchiato Since synizesis in the initial ἀε- of the form ἀεθλ-, while theoretically possible,55 is not definitely attested in early Greek hexameter, it would prima facie seem best not to introduce it by conjecture. Therefore, one may think that also the present case was affected by the introduction of an ‘epic’ form against the meter, and thus that the original reading in the ‘new’ fragment was ἄθλοισι. This, of course, would not imply that the papyrus should be emended: the anonymous author of the treatise could well have found the ‘Hesiodic’ verse(s) with the form ἀέθλοιϲι already in his source. However, a form like ἄεθλος is a word with an ‘open’ spelling concealing established contraction; therefore, whenever it occurs in metrically and prosodically unambiguous positions, the issue at stake is basically orthographical and phonetical:56 synizesis in forms of ἄεθλος is attested in poets as early as Pindar and (possibly) Archilochus.57 My own editorial preference would be to assume synizesis also in the form attested in this new fragment. Whatever the decision, the main point is that there is no need to state, as Vergados does, that “the ‘quote’ is late and spurious”, nor that the prosodical difficulty is indeed problematic, given that it can be easily overcome, as we have seen: this fragment, in my view, will then have to be printed among the genuine Hesiodic fragmenta incertae sedis. Hopefully, therefore, this new fragment can perhaps provoke further reflections by the scholars that will also engage with the variant readings of the two Hesiodic cases cited above (Th. 800; Op. 656).

References Alpers, K. (1972), ‘‘Zonarae’ lexicon’, RE 10A, cols 732–763. Baldi, D. (2013), Etymologicum Symeonis Γ–Ε, Turnhout. Casadio, V. (1990/1993), ‘[Hes.] fr. 406 M.–W.’, MCr 25/28, 21–22. Casanova, A. (1979a), La famiglia di Pandora. Analisi filologica dei miti di Pandora e Prometeo nella tradizione esiodea, Florence. Casanova, A. (1979b), ‘Catalogo, Eèe e Grandi Eèe nella tradizione ellenistica’, Prometheus 5, 217–240.

 55 See below. 56 On this, see West (1982) 12–13 and the examples provided there. 57 Pindar presents eight secure cases of ἄεθλος with synizesis: see Pi. O. 3.15, 9.108, P. 1.99, N. 9.9, 10.32, 11.23, I. 3.9, 6.48. At Archil. fr. 182.1 W.2 (apud Hephaest. De poem. 7.2) εὖτε πρὸς ἆθλα δῆμος ἡθροΐζετο the reading of all the mss. is ἄεθλα; ἆθλα is a conjecture by Fick, but the paradosis was approved by West himself (1982) 13.

Marginalia to Hesiodic Fragments  

Cingano, E. (2009), ‘The Hesiodic Corpus’, in F. Montanari/A. Rengakos/C. Tsagalis (eds), Brill’s Companion to Hesiod, Leiden/Boston, 91–130. Costanza, S. (2018), ‘La tipologia divinatoria di P. Gen. inv. 161 inerente alla ieroscopia’, ZPE 207, 5–13. Cozzoli, A.T. (2016), ‘Un poema mantico: la Melampodia pseudoesiodea’, SemRom, n.s., 5, 145–162. D’Alessio, G.B. (2005), ‘The Megalai Ehoiai: a Survey of the Fragments’, in R. Hunter (ed.), The Hesiodic Catalogue of Women. Constructions and Reconstructions, Cambridge, 176–216. Danbeck, D. (2013), ‘Hesiod, Catalogue of Women 85+117’, ZPE 187, 15–30. Dickey, E. (2014), ‘A Catalogue of Works Attributed to the Grammarian Herodian’, CPh 109, 325–345. Dräger, P. (1993), Argo Pasimelousa. Der Argonautenmythos in der griechischen und römischen Literatur, Stuttgart. Dyck, A.R. (1993), ‘Aelius Herodian: Recent Studies and Prospects for Future Research’, ANRW 2.34 (1), 772–794. Furley, W./Gysembergh, V. (2017), ‘Divination, Pyromancy, Hesiod: P. Gen. inv. 161 has more to offer’, ZPE 203, 1–23. Furley, W./Gysembergh, V. (2018), ‘A Note on the Reading ‘Hesiod’ (ZPE 203 (2017) 1–23)’, ZPE 207, 14. Giannini, P. (1995), ‘Commento alla Pitica Quarta’, in B. Gentili/P. Angeli Bernardini/E. Cingano/ P. Giannini (eds), Pindaro. Le Pitiche, Milan, 426–510. Goetz, G. (1896), ‘Audax (2)’, RE 2.2, col. 2278. Hirschberger, M. (2004), Gynaikōn Katalogos und Megalai Ēhoiai. Ein Kommentar zu den Fragmenten zweier hesiodeischer Epen, Munich/Leipzig. Janko, R. (1984), ‘P.Oxy. 2509: Hesiod’s Catalogue on the Death of Actaeon’, Phoenix 38, 299– 307. Kassel, R. (2002), ‘Testimonia Homerica’, RhM 145, 241–251. Lasserre, F./Livadaras, N. (1992), Etymologicum Magnum Genuinum, Symeonis Etymologicum una cum Magna Grammatica, Etymologicum Magnum Auctum, vol. 2, Athens. Löffler, I. (1963), Die Melampodie: Versuch einer Rekonstruktion des Inhalts, Meisenheim am Glan. Matijašić, I. (2017), ‘A Neglected Hesiodic Fragment in Pollux’, Mnemosyne 70, 340–346. Merkelbach, R./West, M.L. (1965), ‘The Wedding of Ceyx’, RhM 108, 300–317. Merkelbach, R./West, M.L. (1967), Fragmenta Hesiodea, Oxford. Momigliano, A. (1934), ‘Manto e l’Apollo di Claro’, RFIC 12, 313–321. Most, G.W. (ed.) (2007), Hesiod. The Shield, Catalogue of Women. Other Fragments, Cambridge (Mass.)/London. Most, G.W. (20182), Hesiod. The Shield, Catalogue of Women. Other Fragments, Cambridge (Mass.)/London. Olson, S.D. (1987), ‘Aristotle Hist. Anim. 8.18.3 and the Ἀσσύϱιοι λόγοι of Herodotus’, Historia 36, 495–496. Reitzenstein, R. (1897), Geschichte der griechischen Etymologika. Ein Beitrag zur Geschichte der Philologie in Alexandria und Byzanz, Leipzig. Rzach, A. (1902, 19133), Hesiodi Carmina, Leipzig. Schwartz, J. (1960), Pseudo-Hesiodeia. Recherches sur la composition, la diffusion et la disparition ancienne d’œuvres attribuées à Hésiode, Leiden. Sittl, K. (1890), ‘Die Glaubwürdigkeit der Hesiodfragmente’, WS 12, 38–65.

  Stefano Vecchiato Solmsen, F. (1970, 19903), Hesiodi Theogonia, Opera et Dies, Scutum, Oxford. Sturz, F.W. (1820), Orionis Thebani Etymologicon, Leipzig. Vecchiato, S. (2017), ‘Nota editoriale a tre frammenti ‘esiodei’ (‘Hes.’ FF 340–42 M.–W. = 287– 89 M.)’, Hermes 145, 118–120. Vecchiato, S. (2019a), ‘Spuria Hesiodea Neglecta’, MH 76.1, 22–29. Vecchiato, S. (2019b), ‘A New Fragment of the Hesiodic Catalogue of Women?’, Mnemosyne 72, 647–651. Waser, O. (1901), ‘Delphos’, RE 4.2, col. 2700. West, M.L. (1964), ‘The Medieval and Renaissance Manuscripts of Hesiod’s Theogony’, CQ 14, 165–189. West, M.L. (1966), Hesiod. Theogony, Oxford. West, M.L. (1974), ‘The Medieval Manuscripts of the Works and Days’, CQ 24, 161–185. West, M.L. (1982), Greek Metre, Oxford. West, M.L. (1985), The Hesiodic Catalogue of Women. Its Nature, Structure, and Origins, Oxford. West, M.L. (1998, 2000), Homeri Ilias. Vol. 1, Stuttgart/Leipzig; vol. 2, Munich/Leipzig.

Francesco Lupi

To Belong or not to Belong A Few Remarks on the Lyric Fragments of Sophocles’ Tereus

 Introduction Among the fragmentary plays of Sophocles, Tereus has in recent years enjoyed considerable critical attention. Most noticeably, in 2016 scholarship on the play benefited from the publication of a new papyrus fragment by Samuel Slattery in volume 82 of The Oxyrhynchus Papyri series. 1 Tereus told the story of Procne, wife of the Thracian king Tereus, and Procne’s sister Philomela. In order to avenge Tereus’ rape of Philomela, the two sisters murdered Itys, the only son of Procne and Tereus, and set up a cannibalistic banquet, during which an unsuspecting Tereus was served the flesh of his son. 2 In the present paper, my aim is neither to discuss at length matters of plot nor to propose a new, overarching interpretation of the tragedy. Rather, I intend to focus on two lyric fragments attributed to the play by late antique compiler Johannes Stobaeus (fifth century CE) in the Anthology. In so doing, I will assess what contribution to establishing the text can be deduced from the transmission of the fragments and the manner in which they are quoted by their source. In particular, I will argue that lines 4–6 of fr. 592 R.2 should be printed together with fr. 593 R.2 as a continuous text, with the former preceding the latter. I shall also review the colometric arrangement of the text(s) with reference to the manuscript

 I wish to thank Anna Maganuco for her insightful comments on the paper and for generously sharing in advance of publication some data on the transmitted colometry of Sophocles’ Trachiniae (see n. 41 below); for further comments and observations, gratitude is due to Guido Avezzù, John Hilton, Andrea Rodighiero, and to both referees. I also wish to thank Suzanne Sharland for carefully revising and improving the English of this paper. For ease of reference, tragic fragments are referred to by number followed by abbreviation of their editor’s name.  1 Following its publication, the new papyrus fragment (P. Oxy. 5292) has been extensively investigated by Finglass (2016). For further recent contributions on the text of Tereus’ select fragments, see De Stefani (1998); Casanova (2003); Tammaro (2015) 83–86; Librán Moreno (2015) and (2016). 2 On the myth, see Monella (2005) esp. 79–125 (on Sophocles’ version); Fitzpatrick/Sommerstein (2006) 141–149 (addenda in Sommerstein/Talboy 2012, 261–265); Casanova (2003) 66–78; Milo (2008) 7–13. https://doi.org/10.1515/9783110712223-003

  Francesco Lupi evidence. Thereafter, I will discuss the identity of the persona canens, and show that, contrary to what is generally assumed by modern editors, the attribution of these lines to the Chorus is not the only possibility and is certainly not confirmed by the manuscript evidence. Lastly, I will address, albeit very briefly, the problematic placement of these fragments within the plot of the play. More generally, it is also my intention to raise a few methodological points that are relevant to the editing of indirectly transmitted fragments.

 The text Let us first examine frr. 592 and 593 R.2 as printed in the fourth volume of the TrGF. 3 With regard to the former, Radt’s arrangement of the text is based on the underlying assumption that lines 1–3 and 4–6, which show exact metrical correspondence, 4 in the original choral context belonged to the same strophic pair and stood in metrical responsion: 5 (a) Soph. fr. 592 R.2 ΧΟ. ἀλλὰ τῶν πολλῶν καλῶν τίς χάρις, εἰ κακόβουλος 3 φροντὶς ἐκτρίψει τὸν εὐαίωνα πλοῦτον; *** τὰν γὰρ ἀνθρώπου ζόαν ποικιλομήτιδες ἆται 3 πημάτων πάσαις μεταλλάσσουσιν ὥραις

4

metrum: E |2 D– |3 E–e– |

 3 Text and apparatuses (slightly modified and reduced) are based on the edition by Radt (19992) 437–445 (frr. 581–595b R.2); for further details here omitted, see ibid., 443–444, app. The metrical descriptions marked as ‘metrum’ are Radt’s own. Translators of each quoted passage are indicated within round brackets. More recently, Tereus’ fragments (or selections thereof) have also been edited by Lloyd-Jones (20032) 290–301; Fitzpatrick/Sommerstein (2006) 162–173; Milo (2008) 27–112, 115–124 (frr. inc.); Diggle (1998) 72–74. 4 Obtained through minor textual changes; see below p. 39, app. ad lines 3 and 4. 5 Radt’s arrangement follows that of Pearson (1917) vol. 2, 235; the first to assume metrical responsion between fr. 592.1–3 and 592.4–6 R.2 was Bergk (1833) 29–30, who, however, placed lines 4–6 (joined directly to fr. 591 R.2) in the strophe and lines 1–3 in the “extrema pars” of the corresponding antistrophe (i.e. 591 + 592.4–6 ~ * * * + 592.1–3 R.2 See also Radt (1983) 229. On these points, however, see below pp. 40–41.

To Belong or not to Belong  

1–3 Plut. De aud. poet. 21b–c 6 πρὸς δ’ ἐκεῖνα τὰ περὶ τοῦ πλούτου ‘δεινὸς–ἰδεῖν’ (Soph. fr. 88 R.2) ἀντιπαραθήσει πολλὰ τῶν Σοφοκλέους, ὧν καὶ ταῦτ’ ἐστὶ ‘γένοιτο–ἀνήρ’ (fr. 835 R.2) καὶ ‘οὐδὲν–φρονεῖ’ (fr. 836 R.2) καὶ ‘ἀλλὰ–πλοῦτον;’ (fr. 592.1-3 R.2) | 4–6 Stob. Flor. 4.34 (περὶ τοῦ βίου, ὅτι βραχὺς καὶ εὐτελὴς καὶ φροντίδων ἀνάμεστος).39 (SMA) Σοφοκλέους Τηρεῖ· ‘τὰν–ὥραις’ chori notam ante lemma habet Stob. S || 3 φρ. ἐκτρίψει van Herwerden, Tucker : φρ. ἐκτρέφει codd. (defendit Wilamowitz [‘‘semper tenet’ (ἐκ- abundat)’ (teste Radt)], ‘sed τὸν εὐαίωνα πλοῦτον ad τῶν πολλῶν καλῶν spectare mihi quidem certum videtur esse’ Radt); φρ. ἐκτρίβει Reiske; ἐκτρέφει φρ. Bergk (postea φρ. ἐκστρέψει maluit) metri gratia (‘sed anceps in dactylo-epitritis interdum breve esse potest’ monet Radt); φρ. ἀντρέψει van Herwerden; φρ. ἐκτρέψει von Blumenthal; obelum adhibuit Buchwald || 4 γὰρ Bergk metri gratia : δ᾽ codd. | ζόαν Dindorf : ζωὰν (sine acc. S) codd. || 5 ἅτ- Μ, ἄτ- A || 6 μεταλλάσουσιν S | ὥρας A, αὔραις Burges, (μεταλλάσσουσι) μορφαῖς Lehrs; obelum adhibuit Buchwald But what pleasure comes from the many splendid things, if thought that gives bad counsel is to destroy the wealth that makes life happy? … For the life of men is transformed by the cunning wiles of ruinous error that bring calamities at all seasons. (Transl. Lloyd-Jones 20032, 299) (b)

Soph. fr. 593 R.2 ζώοι τις ἀνθρώπων τὸ κατ᾽ ἦμαρ ὅπως ἥδιστα πορσύνων· τὸ δ᾽ ἐς αὔριον αἰεὶ 3 τυφλὸν ἕρπει

metrum: –e–D |2 –e–D– |3 e– Stob. Flor. 4.34 (περὶ τοῦ βίου, ὅτι βραχὺς καὶ εὐτελὴς καὶ φροντίδων ἀνάμεστος).40 (S) ἐν ταὐτῷ· 7 ‘ζώοι– ἕρπει’ 1 ζῴη Wagner, ζήτω Blaydes, ζώει ? Buchwald | ἆμαρ Gleditsch, Blaydes || 2 ἅδιστα Wolff | γὰρ (pro δ᾽ ἐς) Gleditsch, Blaydes, δ᾽ ἐπ᾽ (i.e. ἐφέρπει) West (teste Radt) | ἀεὶ Dindorf || 3 ‘τυφλὸς Friedländer’ Wilamowitz (teste Radt) Let any man procure as much pleasure as he can as he lives his daily life; but the morrow comes ever blind. (Transl. Lloyd-Jones 20032, 301)

 6 For the sake of space, the first three quotations in Plutarch’s passage — one from Sophocles’ Sons of Aleus (fr. 88 R.2), the other two from unknown play — are omitted. 7 On this lemma, see below pp. 43–44.

  Francesco Lupi These fragments are two of only three lyric passages which the manuscript tradition explicitly attributes to Tereus, with the third one being fr. 591 R.2. 8 In terms of metre, all these fragments are kat’enóplion-epitrites. 9 Our evidence for these excerpts rests entirely on the indirect tradition, as they are all preserved either as excerpts in Johannes Stobaeus’ Anthology, 10 or quoted within another author’s text (as is the case with fr. 591 R.2, a few lines of which fragment are also transmitted by the second-century rhetor Favorinus, and with fr. 592.1–3 R.2, exclusively preserved by Plutarch). 11 On account of metre and content, the fragments have been variously joined together in order to reconstruct their original choral context. However, we do not know whether they both come from the same ode (or a portion thereof) or whether any other kat’enóplion epitrite fragment of Sophocles’ should be assigned to the same lyric context. Scholars have suggested various configurations, though not conclusively. 12 Before proceeding any further, a preliminary remark on fr. 592 R.2 is in order. Among recent editors of Tereus, David Fitzpatrick and Alan Sommerstein have cautiously, and, in my opinion, rightly, objected to the attribution of lines 1–3 of fr. 592 R.2 to that play and have done so on several grounds: 1. There is no positive evidence for such an attribution, as the source for lines 1–3 (Plutarch) merely attributes them to Sophocles, without mentioning any play; 13 2. Lines 1–3 and 4–6 make different points, as they ascribe the “reversals of fortune to the ‘ill-counselled thinking’ of humans [κακόβουλος φροντὶς, lines 2– 3]”, and to the “‘cunning wiles’ [ποικιλομήτιδες ἆται, line 5] of, one must assume, a malevolent deity” respectively; 14

 8 Fr. 591 R.2: Stob. Flor. 4.29a (περὶ εὐγενείας. ὅτι εὐγενεῖς οἱ κατ᾽ ἀρετὴν ζῶντες, κἂν μὴ λαμπρῶν ὦσι γεγονότες πατέρων [ἀγαθοὶ τὸν τρόπον]).12 (SMA); lines 1 and (part of) 2 are also transmitted, without mention of either author or play, by Favorinus (Exil. 11.1 [col. 9, ll. 25–26]). Sommerstein (2010) has cautiously suggested, on several grounds (among which metre), that PMG 960, a lyric fragment preserved by Clement of Alexandria, may come from Sophocles’ Tereus (see esp. pp. 657–658; see further Sommerstein/Talboy 2012, 263–264). On the other hand, see Milo (2011), who proposes identifying Clement’s quotation as a Hellenistic lyric fragment. 9 On this class of metres, see Gentili/Lomiento (2003) 204–219. 10 On Stobaeus’ textual tradition, see below nn. 23, 27. 11 See app. test. above (p. 39) and n. 8. 12 See Radt (19992) 443, app. font. ad frr. 591–592 R.2. On the structure of kat’enóplion metres in Sophocles, see Gentili/Lomiento (2003) 210. 13 Milo (2008) 88 wrongly states that Plutarch “omette autore e indicazione della tragedia cui [lines 1–3 of fr. 592 R2] appartengono”. 14 Text quoted is Fitzpatrick/Sommerstein (2006) 186, ad fr. 592.4–6 (= J).

To Belong or not to Belong  

3.

The metrical correspondence between lines 1–3 and lines 4–6 may be due to mere coincidence.

For the above reasons, it is certainly safer to accept the editorial caution of Fitzpatrick and Sommerstein, and not assume that these lines certainly belong to Tereus. One should always bear in mind that the proportion of Sophocles’ lost plays amounts to an overwhelming 94 to 95% of the entire corpus Sophocleum; 15 therefore, any assumption based on metrical similarity is subject to a high degree of uncertainty. Sheer metrical coincidence has been argued to exist also between frr. 593 and 879a (inc. fab.) R.2, which Bergk (1833) 34 thought might have been in metrical responsion, a hypothesis rejected both by Radt and Fitzpatrick and Sommerstein. 16 Radt’s note on fr. 593 R.2, in particular, is worth citing in full, as it provides a commendable piece of editorial caution: … tam frequens in dactylo-epitritis similitudo versuum et tam ingens deperditorum Sophoclis versuum numerus est ut similitudinem inter F 879a et hoc fr. [= fr. 593 R.2] intercedentem fortunae tribuere malim; accedit quod responsio illa coniecturis non levibus extorquenda est.

The first point made by Radt can also be applied to fr. 592.1–3 R2, which Radt instead prints together with — yet detached from — lines 4–6. 17 More generally, a point of method and editorial caution ought to be raised here: that one (or more) fragments from an unknown play may be combined together with another fragment from a known play should not be promoted beyond the apparatus, unless, in addition to the mere acknowledgment of a metrical similarity, compelling arguments supporting the attribution can be produced. A simpler, perhaps more convincing textual arrangement can be obtained by joining together frr. 592.4–6 18 and 593 R.2 in the form of a continuous text (592.4– 6+593 R.2), thus positing an even closer relationship between them than hitherto assumed. 19 To the best of my knowledge, the case for these two groups of lines

 15 There are only seven extant plays out of a number that in the sources oscillates between 123 and 113 genuine plays: for details, see Avezzù (2012) 43–44. 16 See Fitzpatrick/Sommerstein (2006) 185–186, ad fr. 593 (= H); fr. 879a R.2 does not feature in Diggle’s (1998) edition, while Milo (2008) places it among the incerta (120–124). 17 This is the case too in Lloyd-Jones (20032) 298; Diggle (1998) 74 (= fr. IV a–b); Milo (2008) 87. 18 For the sake of greater clarity, I should stress that lines 1–3 of fr. 592 R2 are not going to be dealt with in the present paper, as it cannot be proven that they belong to Tereus. 19 A close relationship between the fragments has been persuasively argued to exist by Fitzpatrick and Sommerstein (2006) 186–187, ad fr. 592.4–6 R.2 (J): “In the present fragment the chorus

  Francesco Lupi being printed together as a consecutive text has not yet been made. 20 This arrangement, as I shall endeavour to show, is more convincing than other options: the textual arrangement I am proposing is not only satisfactory in terms of content, language, and metre, but can also be supported if one takes into account the manner in which Stobaeus, the only source of the fragments, tends to quote texts. As I will show, separate lemmatization of an originally continuous text is compatible with Stobaeus’ anthological practice. In particular, we can argue the following for frr. 592.4–6 and 593 R.2: 1. Both fragments certainly come from Tereus, as explicitly attested by Stobaeus; 2. Both fragments make a similar point and overlap considerably in terms of lexis. Scholars have already remarked upon the content overlap, but they have not specifically remarked on the language employed. Lexical overlap strengthens the possibility that frr. 592.4–6 and 593 R.2 were a continuous text; 3.

When the fragments are joined together, the resulting textual unit makes perfect sense and produces a thought progression that is in itself satisfying, as is the resulting metrical arrangement.

Before discussing the points just made, I provide the following provisional text and translation (based on that of Lloyd-Jones 20032, with changes):

 state, by way of explaining or justifying some earlier utterance, that human life is full of calamities caused by ‘ruinous error’; and fr. H (593), recommending a hedonistic approach that takes no thought for the morrow, is just the sort of utterance that might be justified in this way … It is therefore quite possible that these two fragments come from the same ode”. On the argument of Fitzpatrick and Sommerstein, see below. 20 In 1859 Theodor Bergk proposed the sequence frr. 879a *** + 592.4–6 ~ fr. 593 *** + 592.1–3 R.2 (I take this piece of information from Radt 19992, 443, app. test. ad fr. 592 R.2, as I could not inspect Bergk’s article myself). In the larger reconstruction proposed by Bergk, the two fragments are placed one after the other, but they are thought of as belonging to different strophes (in other words, they are not consecutive in the sense that I am suggesting here). A few years earlier, Rossbach/Westphal (1856) 448 had joined the fragments together as part of the same strophe, yet in the reverse order (593 + 592.4–6 R.2), and so later did Gleditsch (1868) 32 (879a + 592.1–3 ~ 593 + 592.4–6 R.2), albeit with minor differences.

To Belong or not to Belong  

21 … (?) τὰν γὰρ ἀνθρώπου ζόαν ποικιλομήτιδες ἆται πημάτων πάσαις μεταλλάσσουσιν ὥραις· ζώοι τις ἀνθρώπων τὸ κατ᾽ ἦμαρ ὅπως ἥδιστα πορσύνων· τὸ δ᾽ ἐς αὔριον αἰεὶ τυφλὸν ἕρπει

4

1 γὰρ Bergk : δ᾽ codd. | ζόαν Dindorf : ζωὰν (sine acc. S) codd. || 3 post ὥραις colon posui … since the life of men is transformed by the cunning wiles of ruinous error that bring calamities at all seasons, let any man procure as much pleasure as he can as he lives his daily life; the morrow comes ever blind. 22

(i) With regard to the first point — that both fragments certainly come from Tereus — scholars have not considered the import of the fact that frr. 592(.4–6) and 593 R.2 are placed one after the other in their only source (Stobaeus’ manuscript S). In this manuscript, dated to the end of the tenth century, 23 they are introduced by different lemmata, Σοφοκλέους Τηρεῖ and ἐν ταὐτῷ respectively. 24 Despite the separate lemmatization, it remains possible that the two excerpts originally formed a continuous text. This possibility should be considered within an anthological framework. The extraction of different ἐκλογαί — the chosen bits of text deemed worthy of anthologization — from an originally continuous text is not at all unprecedented in Stobaeus. Overall, it is the practice of anthologists to separate and lemmatise in smaller textual units lines that originally stood together as a continuous text: at least two cases can be found among Stobaeus’ tragic quotations that  21 Radt prints the chori nota preceding fr. 592 R.2 without angular brackets (see above, section 2), as do Lloyd-Jones (20032) and Fitzpatrick/Sommerstein (2006); round brackets are added instead by Diggle (1998); angular brackets, both for fr. 592 R.2 and for fr. 593 R.2, by Milo (2008) and, for the latter alone, by Lloyd-Jones. As I will argue in section 3 below (pp. 50–51), the editorial choices of Radt, Lloyd-Jones, and Fitzpatrick and Sommerstein should be avoided. 22 Points of syntax and translation are discussed below p. 46. 23 On the dating of S (Vindobonensis Philologicus Graecus 67), the oldest manuscript of Stobaeus’ Florilegium (books 3–4 of the Anthology), see Piccione (1994b) 189, with n. 4. On the manuscript, see also Hense (1894) vii–xxii; Di Lello-Finuoli (1977/1979) 364–367, (1999) 17–22; on its history during Renaissance humanism, see Speranzi (2010), (2013) 127, 285–286 (no. 95); Ferreri (2014) 546–548. Further information on the manuscript tradition of the Florilegium in Ranocchia (2011) esp. 348–352. 24 In the Anthology, the occurrence of ἐν ταὐτῷ indicates that the eclogue it introduces and the preceding one belong to the same literary work: see Piccione (1994a) 296, (1999) 143.

  Francesco Lupi clearly show such an anthological practice. The case of lines 229–236 from Euripides’ Orestes, which in one branch of the tradition of Stobaeus are lemmatised in three successive entries, 25 is especially remarkable. Otto Hense, editor of the latter half of Stobaeus’ Anthology (the so-called Florilegium) in the app. ad loc. summarises this practice admirably: “discerpere malunt gnomologi contextum fabularum quam exhibere continuum” (‘The collectors of pithy idioms prefer to dismember the context of the stories rather than to present them as a continuous narrative’). 26 This typically anthological habit is fully compatible with the hypothesis that frr. 592.4–6 and 593 R.2 were originally consecutive textual portions. It should then be noted that, in the Orestes’ sequence referred to above, Stobaeus’ manuscripts M and A 27 place the lemma ἐν ταὐτῷ before each subsequent quotation from the same text. The same lemma is found before fr. 593 R.2 in Stobaeus’ manuscript S (int. marg.). A warning must be issued before I proceed: nothing certain about the original structure of a fragmentary work can be inferred from the type of lemma(ta) employed by Stobaeus, as Rosa Maria Piccione has convincingly shown in an extensive paper on the topic. 28 However, if we look at Stobaeus’ quotations from extant tragedies as well as from Hesiod’s Works and Days, we notice that the vast majority of quotations introduced with the lemma ἐν ταὐτῷ (in at least one of the three main manuscripts of the Florilegium, SMA) come from a passage further into the original text than the one that precedes them in the anthology. In an overwhelming number of cases (ca. 87%), quotations introduced by ἐν ταὐτῷ contain lines that in the original structure both of the plays and of Hesiod’s

 25 Stob. Flor. 4.36.1 (Eur. Or. 211–212), 4.36.2 (Eur. Or. 229–230), 4.36.3 (Eur. Or. 231–232), 4.36.4 (Eur. Or. 233–236 [line 234 is omitted]). On a smaller scale, the same arrangement is found at Stob. Flor. 4.19.1 (Eur. Hel. 726–727), 4.19.2 (Eur. Hel. 728–733). For details, see Piccione (1999) 147. 26 Lemmatization of successive excerpts from the same play is not unparalleled in anthologies on papyrus either: Eur. Hipp. 403, 404, 406, 407, 407a, 408, 409, 410, 413–414, 415–423 are quoted successively in P. Berol. inv. 9772, fr. 2, cols III, l. 6–IV, l. 7, on which see Piccione (2017b) esp. 72–74; on the papyrus, see also Carrara (2009) 182–187. On gnomic anthologies and gnomologia, see now Piccione (2017a). 27 The Escurialensis Σ II 14 (M) and the Parisinus Graecus 1984 (A) belong to the same branch of the Florilegium’s bipartite manuscript tradition, while the other branch has its main representative in S. 28 Piccione (1999). See esp. p. 169, where the scholar argues that “la formulazione di conclusioni basate sull’osservazione dei lemmi [in Stobaeus] non può costituire un criterio oggettivo di giudizio e rischia di cadere in errore”.

To Belong or not to Belong  

poem follow, at varying distances, the lines quoted by Stobaeus in the preceding eclogue. 29 While not overwhelming, this provides us with a significant body of evidence on which to base the joining together of frr. 592.4–6 and 593 R.2; at the very least, it shows that the proposed textual arrangement is perfectly compatible with Stobaeus’ anthological practice. If we return to the case of the successive quotations from Euripides’ Orestes in particular, there is nothing that prevents us from positing a similar situation for the Tereus’ quotations. In Tereus’ case, moreover, the exceptional metrical nature of the quotations (kat’enóplion-epitrites) and their being placed one after the other in the paradosis reduces by a wide margin the likelihood that Stobaeus was drawing on different sources. 30 Moreover, the separate lemmatization itself is by no means an indication that for frr. 592.4–6 and 593 R.2 Stobaeus was drawing on two different sources (source ‘A’ for the first quotation, source ‘B’ for the second). Should this have been the case, the likelihood of frr. 592.4–6 and 593 R.2 originally being detached from one another would perhaps be greater. (ii) I shall now focus on a few details of language and style. By joining frr. 592.4–6 and 593 R.2 together, the text receives further stylistic nuances. First, the connection between the fragments is strengthened by the chiasmus linking lines 1 and 4: ἀνθρώπου ζοὰν ~ ζώοι … ἀνθρώπων. Second, the succession of what might be labelled as ‘units of time’ — ὥραις (line 3), ἦμαρ (line 4), and αὔριον (line 6) — also appears significant, and allows the poet to move from the perennial gnomic perspective of the πᾶσαι ὧραι (line 3) to men’s daily striving for a happy life (lines 4–5), only to eventually resort to another — and conclusive — gnomic statement: ‘The morrow (αὔριον) comes ever blind’. These three units of time may also be taken to represent the three temporal dimensions of past (first gnome), present (paraenetic statement), and future (second gnome).

 29 Poetic quotations introduced by ἐν ταὐτῷ (in bold): (a) in progressive line order, Stob. Flor. 3.2.11–12 (Hes. Op. 270–272, 346–348), 3.10.11–12 (Hes. Op. 40–41, 352), 3.20.34–35 (Eur. Med. 446–447, 520–521), 3.22.20–21 (Soph. Aj. 125–130, 758–761), 3.29.2–3–4 (Hes. Op. 21–24, 308– 313, 317–320), 4.8.10–11 (Soph. OT 380–382, 873–874), 4.22g.169–170 (Hes. Op. 56–57, 67–68), 4.22g.194–195 (Eur. Med. 408–409, 568–570), 4.36.1–2–3–4 (Eur. Or. 211–212, 229–230, 231– 232, 233–236), 4.52a.7–8 (Eur. Or. 1034, 1509); (b) not in progressive line order, Stob. Flor. 4.29b.36–37 (Eur. El. 550–551, 369–370), 4.52a.6–7 (Eur. Or. 1523, 1034). 30 Regarding tragic drama in particular, Stobaeus’ dependence on different sources is shown by the presence of multiple quotations of the same excerpt (or part thereof) in different chapters (or different subsections of the same chapter): see e.g., on Euripidean dittographies in the Anthology, Di Gregorio (1980) esp. 53, 55–56. On the rarity of lyric excerpts from tragedy in the Anthology, see Piccione (1994a) 295, with n. 74.

  Francesco Lupi (iii) As far as content is concerned, it is significant that the succession general statement (lines 1–3) followed by an injunction to conduct one’s life in a certain way on account of the preceding statement (lines 4–5) is found, for instance, at lines 780–789 of Euripides’ Alcestis. The Euripidean passage thus provides a significant parallel to the restored textual unit of Tereus. 31 In these lines from the Alcestis, Herakles reflects on human life and its uncertainty. The general, gnomic truth that the hero states at lines 782–786 prompts Herakles’ recommendation that one should live without thinking too much of the future, since it cannot be known in advance. Exactly the same sequence — general truth followed by hedonistic appeal — is found in the restored Tereus’ fragment. In Sophocles, however, the hedonistic appeal is followed by a further gnome, which may act as an apt conclusion to a lyric stanza or lyric period. One may also compare a passage in Sophocles’ Oedipus rex (lines 977–979), 32 where we find a similar sequence of general truth followed by an appeal to live in a certain way on account of the general truth just stated (Jocasta’s words in the Oedipus rex passage, however, do not qualify as a properly hedonistic appeal). From a syntactic point of view, in the proposed enlarged text the particle γάρ (1), which is due to a conjecture first proposed metri gratia by Bergk, would qualify as an instance of an “anticipatory γάρ”. We can apply to our passage Denniston’s (19542, 69) words on the use of γάρ as “anticipatory in the strict sense”: “[t]he sentence opens with the γάρ clause, the whole of the main clause being postponed. γάρ would here” — and in the Tereus passage as well — “naturally be rendered ‘since’ or ‘as’”. 33

 31 The Alcestis passage (lines 783–786) is referred to by Fitzpatrick/Sommerstein (2006) 186 in their comment ad 593 (= H).2–3 R.2; it is also quoted more fully (lines 780–789) by Milo (2008) 92–93. 32 (ΙΟ.) τί δ’ ἂν φοβοῖτ’ ἄνθρωπος ᾧ τὰ τῆς τύχης/κρατεῖ, πρόνοια δ’ ἐστὶν οὐδενὸς σαφής;/εἰκῆ κράτιστον ζῆν, ὅπως δύναιτό τις. (“But what should a man be afraid of when for him it is the event that rules, and there is no certain foreknowledge of anything? It is best to live anyhow, as one may.”; transl. Lloyd-Jones 19972). 33 I do not deem the resulting asyndeton between the γάρ clause in lines 1–3 (= 592.4–6 R.2) and the following precept in lines 4–5 (= 593.1–2 R.2) as a fatal objection to the proposed syntactical reading: one may argue that the lexical overlap analysed above (p. 45) lends further cohesion to the enlarged fragment, thus smoothing the harshness of the asyndeton. However, an alternative reading is also possible, as one may not rule out the possibility that the γάρ clause refers back to a now-lost portion of the ode (in that case, a full-stop after ὥραις in line 3 would suit the syntax better). On this reading, in lines 1–3 the Chorus would provide some sort of explanation of a previous statement (this is basically the view of Fitzpatrick and Sommerstein quoted above [n. 19], with the obvious difference that they propose identifying such statement with fr. 593 R.2);

To Belong or not to Belong  

I shall now focus on some metrical aspects raised by the restored textual unit and provide a tentative metrical description: 34 Tab. 1: Soph. frr. 592.4–6+593 R2: Metrical description (i). τὰν γὰρ ἀνθρώπου ζόαν ποικιλομήτιδες ἆται () πημάτων πάσαις μεταλλάσσουσιν ὥραις· () ζώοι τις ἀνθρώπων τὸ κατ᾽ ἦμαρ ὅπως () ἥδιστα πορσύνων· τὸ δ᾽ ἐς αὔριον αἰεὶ () τυφλὸν ἕρπει () ()

–⏑–––⏑– –⏑⏑–⏑⏑–– –⏑–––⏑–––⏑–– ––⏑–– –⏑⏑–⏑⏑– ––⏑–– –⏑⏑–⏑⏑–– –⏑––

epitrtr‸ hemf epitrtr (stesich) reiza (penthemia) hemm reiza (penthemia) hemf epitrtr

It cannot be determined whether this section amounted to an entire strophe or, at least, to a self-contained section. Metrical signals supporting the hypothesis of an enclosed stanza may not be claimed to exist for certain. However, if we suppose verbal synapheia between 5 and 6 and thus split αἰεί (αἰ-/ εὶ), as some scholars do, 35 the entire sequence is rounded off by a metrum reizianum (here in the form reiza), a colon typically used as strophic clausula. 36 The possibility that ἕρπει at line 6 was followed by a hiatus, moreover, cannot be ruled out either. However, the colometry proposed above is not the only possible one as, once again, the manuscript has something to tell us. It is interesting to note that Stobaeus’ manuscript S seemingly preserves traces of a colometric arrangement for both frr. 592(.4–6) and 593 R.2. 37 It is therefore worthwhile to focus, if briefly, on this aspect  lines 4–5 would then amount to an ‘isolated’ precept, delivered in asyndeton (compare Iocaste’s maxim at Soph. OT 979 [above n. 32], with Finglass 2018, 467, ad loc., referring to Eur. Alc. 782). 34 For metrical abbreviations, see Gentili/Lomiento (2003) xiii–xiv. 35 Thus e.g. Diggle (1998); Milo (2008); with verbal synapheia lines 5–6 scan as reiza hemm (= line 4) | reiza; cf. e.g. Ar. Av. 452–453. 36 Gentili/Lomiento (2003) 204, who refer, as an example, to Eur. Alc. 908–910 = 932–934 (three reiziana). In Sophocles, reizc (2ionma‸‸) is used as a clausula at Aj. 181 = 191 (ibid., 214); two reiziana (reizc and reizb: 614–614a = 625–625a) round off the second strophic pair of the Antigone’s second stasimon (see Giannachi 2011, 49–50). 37 The excerpts (fol. 163r) do not present a properly colometric layout, i.e. with different cola arranged on different lines. However, spacings are clearly visible that mark the end of wellattested cola. Compare the spacing between anapaestic dipodies at (e.g.) Stob. 4.26.6 (Eur. fr. 103 Kn.) in S (fol. 140v), noted by Hense (1909) 651, app. ad loc. On the presence of spacing as a possible indication of “embryonic” colometric practice already in fourth- to second-century BCE papyruses lacking colometric arrangement, see Pordomingo (2005) passim. Here and below I borrow the term embryonic from Pordomingo (ibid., 186, remarking on the presence of “une division colométrique «embrionnaire»” in P. Stras. W.G. 304–307).

  Francesco Lupi too. It is indeed rather striking that the transmitted ‘colometry’ has been, at least to my knowledge, completely disregarded by Sophoclean scholars. The only remark I have come across is, perhaps not surprisingly, that by the last editor of Stobaeus’ Florilegium, Otto Hense. In the app. ad loc., Hense has a passing remark on the cola’s disposition found for fr. 592(.4–6) R.2, but no further consideration is given to such evidence. 38 It is significant to note that S’s ‘embryonic’ colometric division of fr. 592(.4–6) 2 R. did not escape the attention of the great sixteenth-century Hellenist and Greek scholar Marcus Musurus. In the Parisinus Graecus 2130 (fol. 102v), 39 a by-product of the so-called ‘classe Trincavelliana’ of Stobaeus’ manuscript tradition in the hand of Musurus himself, 40 the scholar reproduces the ‘colometric’ arrangement of S, for which I propose the following metrical description: Tab. 2: Soph. frr. 592.4–6+593 R.2: Metrical description (ii). τὰν δ᾽ ἀνθρώπου ζωὰν ποικιλομήτιδες ἆται () πημάτων πάσαις () μεταλλάσσουσιν ὥραις () ()

() –––––– –⏑⏑–⏑⏑–– –⏑––– ⏑–––⏑––

an‸‸ (?) 41 hemf penthemtr (hypodo) end (ia‸ átakton)

 38 Hense (1912) 837 carefully reports that “a v.[erbo] ποικιλομ. novum colon incipit in S, dein a πημάτων et a μεταλλ.”. As for fr. 593 R.2, for which there seems to be a similar arrangement in the manuscript, Hense’s apparatus is silent instead. However, slightly nuanced spacings are visible both after ἀνθρώπων (593.1 R.2) and, this time also with a dot within it, after πορσύνων (593.2 R.2). There also seem to be similar spacings, with a dot in both cases, both after αἰεὶ (593.2 R.2) and after ἦμαρ (593.1 R.2): below (Tab. 2), I do not consider this last instance, thus joining into a single colon τὸ κατ᾽ ἦμαρ and ὅπως ἥδιστα πορσύνων. On S’s use in arranging the poetic quotations on the page, see, in general, Hense (1894) xviiiff. (the colometry of lyric excerpts from tragedy, however, is not discussed). 39 On the manuscript, see Speranzi (2009) esp. 261–263, (2010) esp. 339–348, (2013) 127–130, 219–220; Ferreri (2012) 105–107, (2014) 487–488. 40 Speranzi (2010) 342 states that “il copista — o, meglio, il compilatore — del Par. Gr. 2130 è, senza alcun dubbio Marco Musuro” (my emphasis). 41 It is preferable to accept, as all modern editors do, Dindorf’s slight alteration ζωὰν > ζόαν (Doric form of ζόη), thus restoring the metrical sequence – – – – ⏑ –, i.e. prosdo (2ia‸‸ átakton). Maganuco (2018) 103 identifies this colon in the colometry attested by most Sophoclean manuscripts (LΛKAYTTaTeZoZn) at Soph. Tr. 1023 (such form of the dochmiac prosodiac is also found at Aesch. Suppl. 890). In iambic context, the erroneous substitution of ζωή for ζόη in the manuscript tradition of the Florilegium is attested at Stob. 4.50.63 (Soph. fr. 556 R.2), where Porson’s emendation ζόη is metrically guaranteed (⏑–6). On ζόα, “die einzige form welche die tragödie kennt”, see further Wilamowitz-Moellendorff (18952) vol. 1, 154, ad Eur. HF 664.

To Belong or not to Belong  

ζώοι τις ἀνθρώπων () τὸ κατ᾽ ἦμαρ ὅπως ἥδιστα πορσύνων· () τὸ δ᾽ ἐς αὔριον αἰεὶ () τυφλὸν ἕρπει () ()

––⏑––– ⏑⏑–⏑⏑–– –⏑––– ⏑⏑–⏑⏑–– –⏑––

prosdo (ia‸‸) reizd (pentheman) penthemtr (hypodo) reizd (pentheman) epitrtr

Do the colometric clues provided by S suggest any solution to the metrical interpretation of these lines? Perhaps they do not solve the matter, but nevertheless they demonstrate Stobaeus’ (or his source’s) awareness of a specific metrical arrangement of the text. 42 Unfortunately, we lack any papyrus evidence for these lines, but the colometric traces discussed above make it possible that Stobaeus’ source for these lyric fragments may ultimately have been a Tragödienliederpapyrus, an anthology of choral odes from tragedy, of which we have instances dating from the Ptolemaic era. 43 At any rate, the evidence of manuscript S should be mentioned in future editions of Sophocles’ fragments, or at least taken into account in establishing the colometry of frr. 592.4–6+593 R.2. 44

 42 It may be noted in passing that the ‘embryonic’ colometric arrangement provided by S gives emphasis to a series of assonant word-endings (A~A1, B~B1, C~C1): see cola 3–8, … πάσαις / … ὥραις / … ἀνθρώπων / … πορσύνων· / … αἰεὶ / … ἕρπει. 43 See BKT 5.2, pp. 79–84 (P. Berol. inv. 9771) (third cent. BCE), containing almost the entire parodos of Euripides’ Phaethon; P. Stras. W.G. 304–307 recto (third cent. BCE), containing lyric excerpts from three Euripidean plays (Phoenician Women, Medea, and an as-of-yet unidentified play [fr. **953m Kn.]). On both manuscripts, see Carrara (2009) 103–113; Pordomingo (2005) 186, 197–199, (2013) 59–64, 80–93. Neither papyrus presents a colometric arrangement; however, according to Pordomingo (2005), traces of an ‘embryonic’ colometric division may in fact be present. Still in the third century BCE, anapaestic lines (either lyric or non-lyric) without colometry are also attested: see Fassino (1999) 4 n. 25. On musical papyruses, “per i quali in tutte le età è normale una scrittura senza separazione di cola o di versi”, see ibid., n. 27. 44 The case of Eur. fr. 453 Kn., quoted at Stob. 4.14.1, is highly relevant for the present discussion. Also for this lyric fragment (lines 1–9: str.; 10–12: ant.; 1–8 are transmitted by Polyb. 12.26.5 too), S (fol. 104v) preserves traces of a colometric division, as the presence of spacings (not remarked by Hense) clearly attests. Almost half of the resulting cola match the colometric layout adopted by Kannicht (2004) 489–490; where they do not, this is due either to textual corruption (as in line 6; however, the corrupt paradosis χαρίεσσαν προειδεώραν presupposes the metrically plausible χαρίεσσαν προσιδεῖν ὥραν [Polyb.]: see Martinelli 1987, 172–174; Kannicht accepts Austin’s transposition metri gratia προσιδεῖν χαρίεσσαν ὥραν) or to different, yet (again) metrically plausible, readings (line 4: the paradosis is defended by Martinelli 1987, 165–172). Stobaeus’ colometric division is anyway imperfect at lines 1, 9, 10. Fr. 453 Kn. has also come down to us in a second-century BCE papyrus without colometry (P. Köln 398, fr. a, col. I; see Kannicht 2004, add., 1162–1163), which partly preserves a few more lines of the antistrophe as well as fourteen iambic trimeters preceding the strophe. The papyrus, although damaged, confirms Polybius’

  Francesco Lupi

 Tracing the persona canens: on editorial caution again I now turn to the more provocative part of my paper, in which I wish to challenge the general assumption that frr. 592.4–6 and 593 R.2 are voiced — actually sung — by the Chorus. The very lyrical nature of these lines may be taken as an indication that the speaker — actually singer — is the Chorus. Moreover, all modern critical editions of Sophocles and Stobaeus alike remark on the fact that in manuscript S fr. 592.4–6 R.2 is lemmatised with a chori nota. My point here is that the latter fact — namely, the presence of the chori nota in Stobaeus’ paradosis — does not prove anything with certainty with regard to attributing the fragments to a specific character. In fact, there is no proof that the compendium for χοροῦ in S actually identifies the speaker as the Chorus. 45 It is not, in other words, a sure pointer to the identity of the speaker. In all its occurrences in the manuscript (twenty-two), in fact, the compendium seemingly suggests that the copyist was exclusively aware of the non-iambic nature of the passage quoted. 46 With one exception, 47 the chori nota only appears in the case of lyric or anapaestic fragments, and if we take it as a nota personae proper, such explanation is contradicted in three cases out of five (60%) by the evidence of directly transmitted texts. 48 Therefore, we should reject the practice of printing the text with the nota personae ‘XO’ preceding the fragment, as its restoration in the text can only be conjectural. We may be inclined to believe that the Tereus lines under scrutiny were

 reading at line 6 and is consistent with Polybius’ and Stobaeus’ text at line 4 (see Kannicht 2004, app. ad loc.). 45 The points made here are based on what I have argued in Lupi (2017), which presupposes and confirms Meridor (1974). 46 The claim seems to receive further support from Meridor’s (1974, 134) observation that iambic (i.e. non-lyric) eclogues known from the direct tradition to be choral utterances lack the chori nota: see e.g. the cases listed ibid., n. 11. 47 Stob. 4.19.13 (Eur. fr. 261 Kn.) (3ia); to this eclogue, one may add Stob. 3.8.2 (Soph. fr. 61 R.2), whose iambic nature, though long disputed, is accepted by Radt (19992) 137 and further argued for by Lupi (2017) esp. 159–161. At any rate, there are reasons to believe that both fragments were not spoken by the Chorus: see Lupi (2017) 161–162, 164–166 (with previous bibliography). 48 Stob. 4.41.6 (Eur. IA 28–33: Old Man); 4.44.15 (Eur. Tro. 101–102: Hecuba); 4.44.34 (Eur. Hipp. 203–207: Nurse); the other two cases are Stob. 3.37.4 (Soph. El. 1082–1083) and 4.16.11 (Eur. Ba. 389–394); see Lupi (2017) 163; Meridor (1974) 132–134. With regard to fragments, an attribution to the Chorus is conjecturally, yet almost unanimously rejected by scholars for Stob. 4.48b.17 (Eur. fr. 119 Kn.); for an overview of recent positions on the issue, see Lupi (2017) 163, with nn. 29–30.

To Belong or not to Belong  

in fact sung by the Chorus; however, given that the compendium for χοροῦ in S arguably has no bearing on the attribution of the fragment, it is safer editorial practice to at least enclose the chori nota within angular brackets or follow it with a question mark. Hesitation must, therefore, guide us when dealing with tragic fragments for which Stobaeus’ manuscript S adds the chori nota: it may be worthwhile undertaking a thorough reassessment of the problem.

 Placing the fragment: the search for a possible context and its limits Before I conclude, I will briefly touch upon the possible context of the choral ode to which our fragment belongs. While there is no scholarly consensus on the placement of the lyric fragments of Tereus, in what follows I will focus only on the restored enlarged text (frr. 592.4–6+593 R.2). The textual sequence here conjecturally reconstructed cannot be securely placed in the plot of the play. 49 As Fitzpatrick and Sommerstein (2006) 186 remark: The fragment [scil. fr. 593 R.2] is evidently from a choral song. A recommendation to live for the pleasure of the moment might in principle be either (i) a short-sighted utterance like that of Iocaste in OT 977–9, made just before a disaster which an audience can foresee but the speaker cannot, or (ii) a reaction to a disaster that seems to make nonsense of all imaginable ethical principles … It is hard to find a plausible context in Tereus for an utterance of type (i) by the chorus; (ii) is more likely, and the chorus may well be reacting to the terrible fate of the innocent Philomela.

 49 On the placing of frr. 592–593 R.2 as an admittedly difficult task, see Kiso (1984) 73; Dobrov (1993) 208. Hypotheses on the fragments’ original placement in the plot of the play vary greatly among scholars. Webster (19692) 4 acknowledged the existence of a “close correspondence both of metre and thought between the choric fragments of the Tereus and the parodos of the Trachiniae”; this observation alone may suggest an entirely different context than the one proposed in the present section of the paper. Moving towards the other end of the play’s dramatic action, Kiso (1984) 73 tentatively assigns both fragments to the exodus (“[f]rr. 592R and 593R cannot be securely placed, but the sentiment is quite suitably understood as remarks commenting on a human level on what happened in the drama”); Dobrov (1993) 208 more firmly states that the fragments’ (592–593 R.2) “commentary on the dangers of presumption and the mutability of human fortune … can be applied only to the final crisis of the play” (my emphasis). No hypothesis is put forth by Sutton (1984) 129; Hourmouziades (1986) 137 only remarks that frr. 591–593 R.2, “commenting on the mutability of human life, probably belong to choral odes”.

  Francesco Lupi In what follows, I intend to challenge the contention of Fitzpatrick and Sommerstein that fr. 593 R.2 should not be understood as a “short-sighted utterance … made just before a disaster which an audience can foresee but the speaker cannot (i)”. A possible context — one may suggest — could be a choral ode preceding the cannibalistic banquet of Tereus. 50 This hypothesis has not, to my knowledge, been proposed so far. According to the myth, in order to avenge the rape and glossectomy of her sister Philomela by her husband Tereus, Procne killed their only son, Itys, and served his flesh to his unsuspecting father. Could our fragment, with its hedonistic appeal, function as a superbly ironic introduction to the disaster? The ode, with its recommendation that one procure as much daily pleasure as one can, would be particularly well placed prior to a banquet, and would contrast sharply the anticipated enjoyment of the feast with the ominous tension felt by the knowing Athenian audience. In fact, the choral ode may even have been performed at the same time as the banquet. The dramatic effect would somehow be similar to the one that is prepared in the second stasimon of Sophocles’ Ajax: the Chorus, unaware of Ajax’s real intention to take his own life, sings a joyful ode to celebrate the hero’s apparent change of mind. Ajax has left the stage (at line 692) only to reappear in the third episode (at line 815), where he will eventually commit suicide. In a similar way, by placing the hedonistic appeal of frr. 592.4–6+593 R.2 just before (or during) Tereus’ nefarious banquet, Sophocles would masterfully build up to the final catastrophe, by contrasting the apparent innocence of those who sing our fragment (whose identity remains, as we have seen in section 3, an open issue) with the audience’s awful foreknowledge. The proposed context would obviously favour identifying the persona canens with the Chorus; a stasimon performed while the killing of Itys and Tereus’ banquet take place (off-stage) 51 would in turn present itself as a compelling option. This placement, however, turns out to be problematic as soon as we consider that the Chorus in Tereus was female. 52 As Procne’s confidants throughout the play, the women of the Chorus must have been aware of the heroine’s revenge plot ever since its inception. 53 Thus, the scenario here suggested would rule out the Chorus immediately, as they cannot be unaware of Procne’s murderous plan. The Chorus cannot sing the ode to which frr. 592.4–6+593 R.2 belonged just before (or during)

 50 I wish to thank Dr Gary Vos, to whom I owe the suggestion. 51 See Monella (2005) 118–119; Fitzpatrick/Sommerstein (2006) 152. 52 Finglass (2016) 66, with n. 32. 53 See e.g. Fitzpatrick/Sommerstein (2006) 178.

To Belong or not to Belong  

the sisters’ serving of their revenge and be unaware of it. 54 Thus, interpreting fr. 593 R.2 as a reaction to a disaster, as Fitzpatrick and Sommerstein do, seems to be the better option. What disaster, though, was being reacted to remains an open issue. I suggest that frr. 592.4–6+593 R.2 could be a reaction to the sisters’ revenge after its accomplishment, a scenario — I contend — that is as equally possible as that proposed by Fitzpatrick and Sommerstein (see above). However, in light of the markedly sententious tone of frr. 592.4–6+593 R.,2 any attempt to locate them within the drama is bound to remain mere speculation.

 Conclusion In analysing these editorial minutiae, I hope to have shown that, if there is room for improvement in editing an indirectly transmitted text, this may still be found in long-known manuscripts. Understanding the manner in which the source quotes texts, its use of paratextual elements (such as lemmata), and a thorough inspection of the manuscripts themselves, is imperative for experts in fragmentary literature as well as for any scholars who engage with fragmentary texts. In other words, a return to the sources is probably a task worth undertaking. ‘The mediated tradition’ (tradizione mediata) is a term with which the Italian scholar Michele Curnis, 55 in an essay on the Platonic quotations preserved by Stobaeus,  54 On Philomela’s role, see Monella (2005) 116–118; on the revenge theme in the play, see Fitzpatrick/Sommerstein (2006) 153–157; Scattolin (2013) 131. 55 See Curnis (2011) 71–76. Curnis stresses the importance of the ‘indirect tradition’, a definition which he considers “obsoleta e ambigua” (p. 76). Most importantly, on Stobaeus’ role in the transmission of works attested in the direct tradition (such as Homer, Euripides, Isocrates, Plato), he observes that “la trascrizione dello Stobeo, e dunque la tradizione manoscritta della sua raccolta, non è che una mediazione tra i versanti testuali circolanti nella tarda antichità (IV–V secolo) — da cui i singoli estratti si sono separati per avere vita autonoma all’interno dell’Anthologion stesso — e i testimoni medioevali dei rispettivi autori, inevitabilmente più affascinanti, più ricchi, completi, ma non per questo necessariamente più esatti” (ibid.). Curnis further argues (pp. 76–77 n. 11) that the preference for the ‘direct tradition’ over “altre forme di attestazione” of a text is due to a (pre)judice, according to which “la documentazione indiretta, nelle sue svariate forme, è percepita come frammento, partizione, [p. 77] segmento non autonomo e di per sé insufficiente alla comprensione dell’opera in questione. Tale giudizio è lecito per quanto concerne l’intelligenza complessiva, ma cessa di esserlo quando si applica alla recensio di tutte le testimonianze utili a ricostruire criticamente il testo”. For the Tereus fragments discussed in this paper, we can only rely on Stobaeus’ testimony; however, Curnis’ lucid reflection can be adduced to pre-empt any a priori belittling of the manuscript evidence — textual and otherwise — found in Stobaeus.

  Francesco Lupi has recently rechristened the indirect tradition. In keeping with this brilliant renaming, I wish to stress that the editor’s task must also be a mediation between the inner rules of two different domains, too often regarded as discrete and non-communicating entities, that of the author who quotes and the one who is quoted.

References Avezzù, G. (2012), ‘Text and Transmission’, in A. Markantonatos (ed.), Brill’s Companion to Sophocles, Leiden/Boston, 39–57. Bergk, T. (1833), Commentatio de fragmentis Sophoclis, in Viro summo Godofredo Hermanno praesidi suo diem natalem a.d. IV Cal. Decembres MDCCCXXXIII gratulatur Societas Graeca interprete T. Bergk, Leipzig. Carrara, P. (2009), Il testo di Euripide nell’antichità. Ricerche sulla tradizione testuale euripidea antica (sec. IV a.C. – sec. VIII d.C.), Florence. Casanova, A. (2003), ‘Osservazioni sui frammenti del Tereo’, in G. Avezzù (ed.), Il dramma sofocleo. Testo, lingua, interpretazione. Atti del Seminario Internazionale, Verona, 24–26 gennaio 2002, Stuttgart/Weimar, 59–68. Curnis, M. (2011), ‘Plato Stobaeensis. Citazioni ed estratti platonici nell’Anthologion’, in G. Reydams-Schils (ed.), Thinking through Excerpts. Studies on Stobaeus, Turnhout, 71–123. Denniston, J.D. (1934, 19542), The Greek Particles, Oxford. De Stefani, C. (1998), ‘Una nota a Sofocle’, Maia 50.3, 427–431. Diggle, J. (1998), Tragicorum Graecorum fragmenta selecta, Oxford. Di Gregorio, L. (1980), ‘Lettura diretta e utilizzazione di fonti intermedie nelle citazioni plutarchee dei tre grandi tragici: II’, Aevum 54, 46–79. Di Lello-Finuoli, A.L. (1977/1979), ‘A proposito di alcuni codici Trincavelliani’, RSBN n.s., 14–16, 349–376. Dobrov, G. (1993), ‘The Tragic and the Comic Tereus’, AJPh 114.2, 189–234. Fassino, M. (1999), ‘Revisione di P.Stras. W.G. 304–307: nuovi frammenti della Medea e di un’altra tragedia di Euripide’, ZPE 127, 1–46. Ferreri, L. (2012), ‘Le vicende umanistiche dello Stobeo di Vienna e l’ingrata fatica di rintracciarne la progenie’, Schede umanistiche, n.s., 26, 67–109. Ferreri, L. (2014), L’Italia degli umanisti. Marco Musuro, Turnhout. Finglass, P.J. (2016), ‘A New Fragment of Sophocles’ Tereus’, ZPE 200, 61–65. Finglass, P.J. (2018), Sophocles. Oedipus the King, edited with Introduction, Translation, and Commentary, Cambridge. Fitzpatrick, D./Sommerstein, A.H. (2006), ‘Tereus’, in A.H. Sommerstein/D. Fitzpatrick/T. Talboy (eds), Sophocles. Selected Fragmentary Plays. Volume I: Hermione, Polyxene, The Diners, Tereus, Troilus, Phaedra, Oxford, 141–195. Gentili, B./Lomiento, L. (2003), Metrica e ritmica. Storia delle forme poetiche nella Grecia antica, Milan. Giannachi, F.G. (2011), Sofocle, Antigone. I canti, Pisa/Rome. Gleditsch, H. (1868), Die Sophokleischen Strophen metrisch erklärt. II Theil, Berlin.

To Belong or not to Belong  

Hense, O. (1894), Ioannis Stobaei Anthologii libri duo posteriores, vol. 1, Berlin. Hense, O. (1909), Ioannis Stobaei Anthologii libri duo posteriores, vol. 2, Berlin. Hense, O. (1912), Ioannis Stobaei Anthologii libri duo posteriores, vol. 3, Berlin. Hourmouziades, N.C. (1986), ‘Sophocles’ Tereus’, in J.H. Betts/J.T. Hooker/J.R. Green (eds), Studies in Honour of T.B.L. Webster, vol. 1, Bristol. Kannicht, R. (2004), Tragicorum Graecorum Fragmenta, vol. 5, Euripides, Göttingen. Kiso, A. (1984), The Lost Sophocles, New York. Librán Moreno, M. (2015), ‘Abubillas, cucos y aves rapaces: la autoría de Sófocles, Fr. 581 R. (= Arist., HA 633a 17–28) (Tereo)’, Emerita 83, 247–263. Librán Moreno, M. (2016), ‘A Conjecture on S. fr. 581.9 Radt’, Hermes 143, 491–497. Lloyd-Jones, H. (1994, 19972), Sophocles. Ajax. Electra. Oedipus Tyrannus. Repr. with corrections, Cambridge (Mass.)/London. Lloyd-Jones, H. (1996, 20032), Sophocles. Fragments. Repr. with corrections and additions, Cambridge (Mass.)/London. Lupi, F. (2017), ‘In margine a Soph. fr. 61 R.2: tra ecdotica sofoclea e tradizione stobeana’, MEG 17, 157–170. Maganuco, A. (2018), ‘Esametri dattilici in Sofocle. I casi di Soph. Phil. 839-42; Soph. Tr. 1010– 4, 1018–22, 1031–40’, Lexis 36, 92‒110. Martinelli, M.C. (1987), ‘Osservazioni metrico-testuali sul fr. 453 N.2 (= 71 Austin) del Cresfonte di Euripide’, SCO 37, 165–175. Meridor, R. (1974), ‘A Propos the New Edition of Euripides’ Troades’, SCI 1, 132–136. Milo, D. (2008), Il Tereo di Sofocle, Naples. Milo, D. (2011), ‘Clemente Alessandrino, Stromata V 5, 27’, RAAN, n.s., 75, 565–573. Monella, P. (2005), Procne e Filomela: dal mito al simbolo letterario, Bologna. Pearson, A.C. (1917), The Fragments of Sophocles, edited with additional Notes from the Papers of Sir R.C. Jebb and Dr W.G. Headlam, 3 vols, Cambridge. Piccione, R.M. (1994a), ‘Sulle fonti e le metodologie compilative di Stobeo’, Eikasmos 5, 281– 317. Piccione, R.M. (1994b), ‘Sulle citazioni euripidee in Stobeo e sulla struttura dell’Anthologion’, RFIC 122.2, 175–218. Piccione, R.M. (1999), ‘Caratterizzazione di lemmi nell’Anthologion di Giovanni Stobeo. Questioni di metodo’, RFIC 127.2, 139–175. Piccione, R.M. (2017a), ‘Sentenze, antologie gnomiche e gnomologi’, in Corpus dei papiri filosofici greci e latini (CPF). Testi e lessico nei papiri di cultura greca e latina, Parte II.3: Gnomica, Florence, 3–24. Piccione, R.M. (2017b), ‘GNOM 3 (P.Berol. 9772), Antologia περὶ γάμου’, in Corpus dei papiri filosofici greci e latini (CPF). Testi e lessico nei papiri di cultura greca e latina, Parte II.3: Gnomica, Florence, 54–76. Pordomingo, F. (2005), ‘La colométrie dans les papyrus ptolémaïques’, Aevum(ant), n.s., 5, 179–202. Pordomingo, F. (2013), Antologías de época helenística en papiro, Florence. Radt, S. (1983), ‘Sophokles in seinen Fragmenten’, in J. de Romilly (ed.), Sophocle. Sept exposés suivis de discussions (Entretiens sur l’Antiquité classique, 29), Geneva, 185–222 (repr. without discussion in H. Hofmann/A. Harder (eds), Fragmenta dramatica. Beiträge zur Interpretation der griechischen Tragikerfragmente und ihrer Wirkungsgeschichte, Göttingen 1991, 79–110, and A. Harder/R. Regtuit/P. Stork/G. Wakker (eds), Noch einmal zu... Kleine Schriften von Stefan Radt zu seinem 75. Geburtstag, Leiden/Boston/Köln 2002, 263–292).

  Francesco Lupi Radt, S. (1977, 19992), Tragicorum Graecorum Fragmenta, vol. 4, Sophocles, Göttingen. Ranocchia, G. (2011), ‘Aristone di Chio in Stobeo e nella letteratura gnomologica’, in G. Reydams-Schils (ed.), Thinking through Excerpts. Studies on Stobaeus, Turnhout, 339–386. Rossbach, A./Westphal, R. (1856), Metrik der griechischen Dramatiker und Lyriker nebst den begleitenden musischen Künsten. Dritter Theil. Griechische Metrik, Leipzig. Scattolin, P. (2013), ‘Le notizie sul Tereo di Sofocle nei papiri’, in G. Bastianini/A. Casanova (eds), I papiri di Eschilo e di Sofocle, Atti del convegno internazionale di studi, Firenze 14– 15 giugno 2012, Florence, 119–141. Slattery, S. (2016), ‘5292. Sophocles, Tereus’, in The Oxyrhynchus Papyri 82, 8–14. Sommerstein, A.H. (2010), ‘An Overlooked Tragic Fragment: PMG 960’, in F. Cortés Gabaudan/ J.V. Méndez Dosuna (eds), Dic mihi, Musa, virum. Homenaje al profesor Antonio López Eire, Salamanca, 653–658. Sommerstein, A.H./Talboy, T. (eds) (2012), Sophocles. Selected Fragmentary Plays. Volume II: The Epigoni, Oenomaus, Palamedes, The Arrival of Nauplius, Nauplius and the Beacon, The Shepherds, Triptolemus, Oxford. Speranzi, D. (2009), ‘Un «libellus» del Florilegio di Stobeo e la scrittura dell’anziano Giano Lascaris’, MEG 9, 253–265. Speranzi, D. (2010), ‘Vicende umanistiche di un antico codice. Marco Musuro e il Florilegio di Stobeo’, Segno e Testo 8, 313–350. Speranzi, D. (2013), Marco Musuro. Libri e scrittura, Rome. Sutton, D.F. (1984), The Lost Sophocles, Lanham/London. Tammaro, V. (2015), ‘Su alcuni frammenti tragici’, Eikasmos 26, 83–92. Webster, T.B.L. (1936, 19692), An introduction to Sophocles, London. Wilamowitz-Moellendorff, U. von (1889, 18952), Euripides, Herakles, 2 vols, Berlin.

Chiara Meccariello

‘Well Begun is Half Done’? Uses and Misuses of Incipits in Greek Antiquity and Beyond

 Introduction Honoured and noble warrior Where are the sheep Where are the wild oxen And with you I did not In our city In former times

These lines are part of a longer text inscribed on a Mesopotamian cuneiform tablet dated to the first quarter of the second millennium BCE.1 What at first sight might seem to be a cryptic poem involving a warrior and some cattle, instead turns out to be not a unitary composition, but a sequence of incipits of hymns. In this and several other cuneiform tablets from the Sumerian, Old Babylonian and Kassite Periods (2000–1155 BCE), lists of incipits, like modern lists of titles, are used to record the overall contents of an archive, allowing users to browse through a collection without having to consult all of its items directly.2 In recording the contents of sets of written works for practical purposes, the cuneiform incipit tablets offer a parallel, and an antecedent, for one of the fields in which incipits, or archai, were used in Greek antiquity — that of cataloguing. But the extant Greek sources point to a broader employment of the arche-system, and preserve not only rich evidence of various uses, but also information on problematic cases. The lack of correspondence between a fragmentary arche and the actual incipit of the work preserved as a whole, or the parallel circulation of different fragmentary archai of works no longer extant, present the ancient and modern scholar with a philological challenge. Do different archai indicate the existence of a spurious and a genuine text, or do they point to the circulation of different authorial versions of the same work? Alternatively, do they stem from

 1 Jena, HS 1504. Translation by Dalby (1986) 478. 2 A comprehensive list and full discussion of these catalogues can be found in Dalby (1986); see also Hallo (1963). https://doi.org/10.1515/9783110712223-004

  Chiara Meccariello other phenomena such as interpolations, editors’ excisions, accidental corruption, or intentional manipulation? Or do they simply derive from mistakes in attribution? In this chapter I shall first offer an overview of the different uses of incipits in Greek antiquity, focusing on literary and paraliterary texts, and thereafter discuss a few controversial examples which show archai ‘at work’ in ancient and modern philological discussions.3 The cases under consideration here raise issues that stem not only (and not always) from the fragmentary state of the relevant works, but also from the fragmented nature of our knowledge of ancient authors’ compositional practices, on the one hand, and of the dynamics of textual transmission, especially before the Alexandrian period, on the other. Ancient scholars were often in a better position than we are in both respects. Thus, in approaching ancient and modern philological discussions and their intertwining, I assume by default the bona fides of ancient critics and their ability to access more information than we have today, albeit with less-refined methodologies. Accordingly, I base my own discussion of controversial archai on an attempt to reconstruct the thought and resources of ancient scholars, as refracted by our intermittent documentation.

 Uses of the arche in Greek antiquity: an overview The earliest extant example of use of the arche as an identification device is probably found in Aristophanes’ Clouds, first performed in 423 BCE. In a passage of the agon between the Kreitton Logos and the Hetton Logos, the former, in describing old education, refers to the practice of having young boys learn songs by heart, songs “such as ‘Pallas the terrible, sacker of cities’ (Παλλάδα περσέπολιν δεινάν), or ‘A strain that sounds afar’ (τηλέπορόν τι βόαμα), singing it in the mode their fathers handed down”.4

 3 I will not examine here the case of scriptural incipits written on amulets, a common apotropaic practice in Late Antiquity: for a full study of this phenomenon, see Sanzo (2014). 4 Ar. Nub. 961–968, translation by Sommerstein (1982) 103. A word-play on the opening of a scolion by Harmodius can already be found at Acharn. 1093: see Olson (2002) 336. For other quotations of incipits in Aristophanes see Pax 1270 and the “battle of the prologues” in Frogs (below, p. 65).

‘Well Begun is Half Done’?  

In this comic passage, brief textual sequences, syntactically incomplete, are used to recall two perhaps familiar compositions, ᾄσματα whose titles are not indicated, and perhaps never existed. The scholia label both fragments as archai, and try to identify the authors of their respective songs.5 The jumble of hypotheses suggests that it was by no means an easy task. The former fragment is ascribed to Lamprocles or Stesichorus,6 but of the longer openings quoted from their poems, only one is compatible with the comic sequence.7 The other arche is said to be untraceable, as already Aristophanes (presumably the grammarian) found it ἐν ἀποσπάσματι (‘in fragmentary form’). But certain unidentified people (τινες), the notes continue, ascribed it to a Cydidas of Hermione and were able to quote a slightly longer version of the incipit, with λύρας after βόαμα.8 While the difficulty points to an early loss of the relevant compositions, their oral character as seems to emerge from the Aristophanic description might have played a role in the scholia’s failure to provide a secure identification: their exact and complete text, after all, might have never been part of a widely circulated written edition. This, of

 5 Schol. Ar. Nub. 967 Holwerda. 6 The ascription to Stesichorus is traced to Eratosthenes (fr. 101 Strecker), while Chamaeleon (fr. 31B Martano) is said to have remained uncertain (ἀπορεῖ in schol. 967b.α). A Phrynichus (PCG 78) is quoted for the ascription to Lamprocles: this may be the comic poet or the grammarian, but a papyrus from the third century CE, P. Oxy. XIII 1611 (frr. 5 + 6 + 43), which refers to the same issue (Stes. fr. 322a Finglass, Chamaeleon fr. 31A Martano), seems to suggest that Phrynichus parodied this poem of Lamprocles (l. 165 παρα[ποιεῖ], ‘took over’), which would obviously point to the playwright. 7 It is the sequence Παλλάδα περσέπτολιν δεινὰν/θεὸν ἐγρεκύδοιμον, which Phrynichus ascribed to Lamprocles (schol. 967b.α). Another incipit, Παλλάδα περσέπτολιν κλῄζω πολεμοδόκον ἁγνάν, is reported in schol. 967a.α, 967b.α (where the quotation of this fragment is apparently ascribed to Chamaeleon), 967b.β, Tzetz. Chil. 1.25, l. 686 Leone and P. Oxy. 1611. A scholium on Aristides (schol. Aristid. Or. 46.162) accounts for the difference between this incipit and the one quoted by Aristophanes by pointing out that δεινὰν is added by the playwright γελοίως. Note that in these sources the quotation continues for a few more words after ἁγνάν, and there are discrepancies in the wording of the last part between the Aristophanic scholia on the one hand and the Aristides scholium and Tzetzes on the other. The matter remains obscure: see esp. Dover (1968) 215 and Davies/Finglass (2014) 595–596. 8 Schol. 967a.α μὴ εὑρίσκεσθαι, ὅτου ποτ’ ἐστίν· ἐν γὰρ ἀποσπάσματι εὑρεῖν Ἀριστοφάνη ἐν τῇ βιβλιοθήκῃ (Ar. Byz. fr. 379 Slater). τινὲς δέ φασιν αὐτὸ Κυδίδου Ἑρμιονέως. A slightly different version of the same sentence is found in 967a.β, while the ascription to Cydidas and the fuller incipit, but not the reference to Aristophanes, are also in 967b.α. On the possibility that Cydidas is a corruption for Cydias, a poet known from other sources, see Dover (1968) 215; for the hypothesis that the original name was Ceceidas, an early dithyrambic poet mentioned in Nub. 985 (but several variants are attested: cf. the critical apparatus in Dover 1968, 60), see Guidorizzi/Del Corno (1996) 300.

  Chiara Meccariello course, also makes the arche status of these fragments in principle more uncertain, as it is only witnessed by the scholia themselves. However, the label is likely to be right, as the quotation of the arche of a poetic composition in lieu of title was an established identification device already by the fourth century. It is with this purpose that Aristotle quotes the arche of an Archilochean iambus and of an elegy by Solon, and similar cases are attested in several authors of different periods and provenance, who use the arche to indicate a work of which no title is given, as probably no title was available.9 In addition, the quotation of the opening words, and sometimes also the final words of a piece, with or without other designations, is used to refer to subsections of literary works: Xenophon, Diodorus Siculus, and Plutarch, for example, use this technique to indicate the song of the Sirens in the Odyssey, a brief section of Euripides’ Phoenissae, and the parodos of Euripides’ Electra respectively.10 A handful of papyri provide further evidence for both practices. The earliest, the so-called Vienna epigrams papyrus (P. Vindob. G 40611), dates to the third century BCE and contains a list of incipits of epigrams and stichometric markers. The general heading τὰ ἐπιζητούμενα τῶν ἐπιγραμμάτων (‘the epigrams that are sought out’) suggests that these archai are a list of desiderata, epigrams selected to be either recopied (e.g. to form an anthology) or investigated (e.g. in order to establish their authorship).11 In the absence of a similar indication, it is even harder to identify the purpose of a second-century BCE papyrus, P. Mich. inv. 3498 + 3250b recto, 3250a and c recto, which contains lists of archai of lyric poems and tragic songs.12 The headings παρόδων ἀρχαί, to be understood as ‘beginnings  9 Arist. Rhet. 3.1418b29–30, Ath. 5.2.4–7 Kenyon, Strab. 10.3.13, D.H. Comp. 22.58, Aristid. HL 1, p. 280.20–21 Jebb, Heph. 15.10, p. 50.21–22 Consbruch etc. The Lives of Diogenes Laertius also show a wide, if not regular, use of the arche as a naming device in the absence of a title for several typologies of writings, both in prose and verse: see e.g. 1.119, 2.42, 5.27, 5.60. Other examples can be found in Nachmanson (1941) 37–49. Since Aristotle, the arche is often part of a relative sentence of the type οὗ/ἧς/ὧν (ἡ) ἀρχή, which follows an appropriate generic designation such as ἴαμβος, ἐλεγεία and several others: on this structure, see also below (p. 62). 10 X. Mem. 2.6.11 ἃ μὲν αἱ Σειρῆνες ἐπῇδον τῷ Ὀδυσσεῖ ἤκουσας Ὁμήρου, ὧν ἐστιν ἀρχὴ τοιάδε τις· ‘Δεῦρ’ ἄγε δή, πολύαιν’ Ὀδυσεῦ, μέγα κῦδος Ἀχαιῶν’ (Od. 12.184); D.S. 10.9.8 καὶ τοῦτο γνοίη ἄν τις ἐπιστήσας τοῖς ἐν ταῖς Εὐριπίδου Φοινίσσαις στίχοις, ἐν οἷς οἱ περὶ τὸν Πολυνείκην εὔχονται τοῖς θεοῖς, ὧν ἡ ἀρχὴ ‘βλέψας ἐς Ἄργος’ (Eur. Phoen. 1364) ἕως ‘εἰς στέρν’ ἀδελφοῦ τῆσδ’ ἀπ’ ὠλένης βαλεῖν’ (1375); Plut. Lys. 15.3 τινος Φωκέως ᾄσαντος ἐκ τῆς Εὐριπίδου Ἠλέκτρας τὴν πάροδον ἧς ἡ ἀρχή ‘Ἀγαμέμνονος ὦ κόρα, ἤλυθον, Ἠλέκτρα, ποτὶ σὰν ἀγρότειραν αὐλάν’ (Eur. El. 167–168), πάντας ἐπικλασθῆναι. 11 Parsons/Maehler/Maltomini (2015) 10–11. 12 Borges/Sampson (2014) 9–35. For a recent re-edition with detailed commentary and further bibliography see Borrelli et al. (2019).

‘Well Begun is Half Done’?  

of choral songs’, and μερῶν [ἀρχ]|αί, perhaps ‘beginnings of (theatrical) acts’,13 indicate subsections of the roll and assure us that the arche-system was already used both for entire works and for sections of works. A less ambitious example of the practice is attested in O. Wilck. II 1488 (second century BCE), where we find a brief list of epigram openings which might stem from a school context.14 Similar lists of epigram incipits and Sapphic incipits, the latter incorporated into a bibliographical work which also contained stichometric information, have been unearthed from Roman Oxyrhynchus; and a list of titles, incipits and stichometry of Hyperides’ speeches, seemingly arranged according to the alphabetical order of the incipits, is preserved in a second- or third-century roll of the same provenance.15 Archai are also widely used in papyri preserving readers’ aids or digests collectively known as paraliterary works. One example are the Callimachean diegeseis, brief expositions of the contents of Callimachus’ poems.16 In the richest extant witness to this textual typology, P. Mil. Vogl. I 18 (first century CE), the general title ‘diegeseis of Callimachus’ four books of Aetia’ (τῶν δ� Αἰτίων Καλλιμάχου διηγήσεις) is preserved in the margin above col. vi, while in the body of the columns each diegesis is preceded by the incipit of the relevant elegy.17 The archai here clearly function as titles, since they are the only available element that identifies the single poem to which each diegesis pertains. However, in the case of the epyllion Hecale, both the genitive of the title (col. x, l. 362 Ἑκάλης: diegesis is  13 P. Mich. inv. 3250c recto, col. ii, ll. 4–5 and col. iii, ll. 5–6. See also the heading τωνσιμω | ἀρχαί at 3250a recto, col. ii, ll. 4–5, hesitantly interpreted as τῶν Σιμω(νίδου) ἀρχαί in Borges/ Sampson (2012) 32–33, but perhaps more likely to be a corruption of στασίμων | ἀρχαί (σιµ̣ω̣ M.L. West apud Bernsdorff 2014; cf. Geißler 2014, 19). On the problematic syllable division required by the supplement [ἀρχ]|αί at 3250c recto, col. iii, ll. 5–6 and on the uncertain meaning of μερῶν, perhaps a corruption of μελῶν, see Borges/Sampson (2012) 30–31. According to Borrelli et al. (2019) 49, the incorrect syllabification is best explained if we suppose that another word followed μερῶν and further specified this rather generic noun, otherwise the scribe could have written ἀρχαί fully on l. 5. However, the layout of the other two headings, with their generous blank space on both sides, provides sufficient grounds to explain this division. 14 Cribiore (1996) 230 no. 243, SH 976. 15 Epigram incipits: P. Oxy. LIV 3724 (late first century CE), containing 175 incipits of epigrams on cols ii–viii; see Maltomini (2003). Sapphic incipits: P. Oxy. XXI 2294 (second century CE), on which see Puglia (2008). List of Hyperides’ speeches: P. Oxy. XLVII 3360, on which see Otranto (2000) 69–72. Note that there is evidence of an ancient editorial arrangement of both Sappho’s poems and Hyperides’ speeches by alphabetical order of the first lines. An example of the latter is provided by the famous Hyperides roll P. Lond. inv. 108+115 etc. (LDAB 2423). As to the order of Sappho’s poems in the Alexandrian edition, see most recently Prodi (2017), esp. 573–574. 16 Van Rossum-Steenbeek (1998) 74–81 and 259–278. 17 For a rich discussion of this papyrus see Falivene (2011).

  Chiara Meccariello clearly to be understood) and the first line are quoted. Similarly, in the papyrus collections of Homeric hypotheses, summaries of single books of the Iliad and Odyssey, the book number is usually indicated, followed by its first line; sometimes the last line is also set out after the summary.18 Coexistence of two identification devices, title and arche, is found almost consistently in papyri containing dramatic hypotheses. For example, in most of the papyri which preserve alphabetical collections of the plot-summaries of Euripidean plays (the so-called ‘narrative hypotheses’), each summary is preceded by a heading which includes the following:19 1) The title of the play, followed by the relative clause οὗ/ἧς/ὧν (ἡ) ἀρχή; 2) The first line of the play, quoted in its entirety regardless of its syntactical structure; 3) The formula ἡ δ(ὲ) ὑπόθεσις, which marks the beginning of the summary.20 These paraliterary papyri show that the arche is used both in the absence of titles and alongside them. In addition, from an anonymous commentary preserved in a second-century CE papyrus we can glean what seems to be a disambiguating function of the arche between two works with the same title: here the commentator states that Stesichorus wrote two Palinodes, and quotes their respective incipits as recorded by the Peripatetic Chamaeleon.21 While this function of the arche could have been useful in the case of dramatic corpora as well — Euripides, for example, wrote several pairs of homonymous plays — disambiguating specifications appended to the main title are used to this effect in the extant collections of Euripidean hypotheses, and the consistent employment of the arche also

 18 Van Rossum-Steenbeek (1998) 53–74 and 246–259. 19 For a concise overview of the narrative hypotheses of Euripides’ plays, see van RossumSteenbeek (1998) 1–32 and 185–228. 20 For a recent discussion of this formula and its implications, see Meccariello (2016). 21 P. Oxy. XXIX 2506 fr. 26 col. i = PMG 193, Stes. fr. 90 Finglass, Chamael. fr. 32 Martano [μέμ]|φεται τὸν Ὅμηρο[ν ὅτι Ἑ|λέ]νην ἐποίησεν ἐν Τ[ροίαι] | καὶ οὐ τὸ εἴδωλον αὐτῆ[ς, ἔν] | τε τ[ῆι] ἑτέραι τὸν Ἡσίοδ[ον] | μέμ[φετ]αι· διτταὶ γάρ εἰσι πα|λινωδ[ίαι ]λλάττουσαι, καὶ ἔ|στιν ῆ μὲν ἀρχή· ‘δεῦρ’ αὖ|τε θεὰ φιλόμολπε’, τῆς δέ· | ‘χρυσόπτερε παρθένε’, ⟦ερ⟧ὡς | ἀνέγραψε Χαμαιλέων (text quoted according to Finglass’ edition). Several scholars have doubted the testimony of the papyrus, especially in light of earlier passages (Pl. Phaedr. 243b, Isocr. Or. 10.64) which clearly refer to only one Palinode. We cannot exclude the possibility that the commentator is just identifying two distinct parts of a single work, as perhaps betrayed by the original asymmetry of the text before emendation (ἡ μὲν ἀρχή… τῆς δέ). For concise summaries of the long scholarly debate and further references, see Montanari (1989), Martano (2012) 241 n. 2, and Davies/Finglass (2014) 309, with nn. 57–59.

‘Well Begun is Half Done’?  

alongside unambiguous titles suggests this was not the primary function of dramatic incipits.22 In the case of the Euripidean hypotheses, an indication of the purpose of the arche comes, by contrast, from the medieval manuscripts. Here several of the same narrative hypotheses are each prefixed to the text of the relevant play and no archai are quoted. In fact, the dramatic text does not need to be recalled or securely identified, because it is copied on the same physical object. It is only when text and paratext exist as separate entities that stronger identification devices are required. The archai included in papyrus collections of hypotheses thus seem to function as lemmata in self-standing commentaries or hypomnemata: they recall each of the summarised dramas unambiguously, in the absence of physical contiguity, and provide a means to identify not simply the relevant work, but rather the text, the actual words that the work comprises. Albeit infrequent in ancient literature, quotation of both the title and the first line, as found in paraliterary works, is common in literary passages with markedly bio-bibliographical contents. The Contest of Homer and Hesiod, for example, mentions both the title and the arche of Thebais, Epigoni and Hymn to Apollo, as well as indicating the number of lines of the first two texts.23 Another example is the so-called Vita Herodotea of Homer, which records the title and incipit of the Ilias minor.24 Despite the paucity of evidence, there are good reasons to trace this practice to Callimachus, especially on the grounds of passages of Athenaeus in which the quotation of the arche, side by side with other identifiers and stichometric information, is explicitly ascribed to his Pinakes.25 An interesting formulation of the  22 Titles of homonymous plays with disambiguating specifications in Euripidean hypotheses: Phrixus I and II in P. Oxy. XXVII 2455, the lost Hippolytus in P. Oxy. LXVIII 4640, Autolycus I in P. Vindob. G 19766. On dramatic titles, see Sommerstein (2002). 23 Certamen ll. 255–259 and 317–318 Allen. Bassino (2019) 188 tentatively suggests that the quotation of these incipits might have “didactic purposes”. 24 [Hdt.] Vita Hom. 16, ll. 203–205 Allen. In Plut. Sol. 8.2–3, the first two lines of Solon’s Salamis are quoted, and then the title of the elegy and its length are added. While the presence of title, incipit and stichometry might well betray the bibliographical origin of the information (cf. the wording τοῦτο τὸ ποίημα Σαλαμὶς ἐπιγέγραπται καὶ στίχων ἑκατόν ἐστι in 8.3), in this case the quotation of the opening lines has a clear narrative function in the context of a description of Solon’s performance of the elegy in the agora. Nachmanson (1941) 39–40, on the other hand, believes that Plutarch quotes the first distich because the title “sich noch nicht allgemein eingebürgert hat”. 25 See esp. Ath. 6.244a = Callim. fr. 434 Pfeiffer τοῦ Χαιρεφῶντος καὶ σύγγραμμα ἀναγράφει Καλλίμαχος ἐν τῷ τῶν παντοδαπῶν πίνακι γράφων οὕτως ‘δεῖπνα ὅσοι ἔγραψαν· Χαιρεφῶν Κυρηβίωνι.’ εἶθ’ ἑξῆς τὴν ἀρχὴν ὑπέθηκεν· … στίχων τριακοσίων ἐβδομήκοντα πέντε’; Ath. 13.585b– c = Callim. fr. 433 Pfeiffer. ἥτις (scil. Γνάθαινα) καὶ νόμον συσσιτικὸν συνέγραψεν … ἀνέγραψε δ’

  Chiara Meccariello usefulness of archai in a bibliographical context can be found, if much later, in Porphyry’s introduction to his own edition of Plotinus’ writings. This includes a proper catalogue, which indicates, for each treatise, the number, title, and arche. The latter item, he writes, is meant to facilitate the identification of each work.26 To sum up, the extant evidence shows a nuanced application of the archesystem as both a naming device, in the absence of a title, and as a means of secondary, textual identification. The system is thus capable of securing the correct identification of a work against the intrusion of spurious material or ambiguities stemming from the existence of homonymous works and shifting titles. In the next section, I shall discuss a few concrete examples of this practice, focusing on cases where the system fails to work smoothly. As we shall see, several problematic archai are attested and, as I shall suggest in the end, there seem to be specific reasons why the arche could turn into an inefficient or insufficient device.

 Failures of the system The existence of a single title and two or more different archai is a crucial problem we encounter in our sources. The problem can stem from different situations: for example, a given author may have written two or more works with the same title, or two or more versions of the same work; or a single work may have undergone interpolation or athetesis, which affected its opening; or else a work with a certain title may have been entirely replaced by another, which either had or assumed the same title. A clear case of double archai stemming from athetesis is attested for Hesiod’s poems. The Pergamene Crates of Mallus is said to have athetised the proems of both the Theogony and the Works and Days because they were not particularly relevant to their respective works.27 From the Hesiodic scholia we also learn that the Peripatetic Praxiphanes and the Alexandrian Aristarchus deleted the proem of the Works and Days; the former even claimed to have found a copy of the poem

 αὐτὸν Καλλίμαχος ἐν τῷ τρίτῳ πίνακι τῶν Νόμων καὶ ἀρχὴν αὐτοῦ τήνδε παρέθετο· … στίχων τριακοσίων εἴκοσι τριῶν. Cf. Pfeiffer (1968) 129–130 and Blum (1991) 157. 26 Porph. Vita Plot. 4 ἦν δὲ καὶ τὰ γεγραμμένα ταῦτα ἃ διὰ τὸ μὴ αὐτὸν ἐπιγράφειν ἄλλος ἄλλο ἑκάστῳ τοὐπίγραμμα ἐτίθει. αἱ δ᾿ οὖν κρατήσασαι ἐπιγραφαί εἰσιν αἵδε· θήσω δὲ καὶ τὰς ἀρχὰς τῶν βιβλίων, εἰς τὸ εὐεπίγνωστον εἶναι ἀπὸ τῶν ἀρχῶν ἕκαστον τῶν δηλουμένων βιβλίων. Porphyry’s edition of the Enneads was published in the early fourth century: see Armstrong (1969) ix. 27 Hes. T 50 Most, Crat. fr. 78 Broggiato.

‘Well Begun is Half Done’?  

without the first ten lines.28 Similarly, Pausanias states that the Boeotians living around the Helicon had removed the prefatory invocation of the Muses from the text of the Works and Days, and had shown him a copy of the poem engraved on lead — presumably one without the first ten lines.29 In practice, athetesis must have produced two quantitatively different versions of the first part of the poem, and obviously two different archai. The existence of different incipits of the same work may also be due to the circulation of a spurious prooimion meant to replace the genuine one. This is the case with Plato’s Theaetetus, as we learn from a commentary preserved on a second-century CE papyrus, according to which there were two prooimia to that work, of equal length, one spurious and one genuine. Both are quoted by the arche.30 Similar dynamics may underlie the examples of problematic archai attested for the Euripidean corpus.31 In what follows I shall discuss three examples — Archelaus, Meleager, and Rhesus — in which the variation of the arche involves early sources, including the playwright Aristophanes, Dicearchus, and the Alexandrian scholar Aristarchus. Our earliest source for the first two cases is Aristophanes’ Frogs. In the famous contest-scene of the play, Aeschylus and Euripides examine and criticise each other’s plays, with Dionysus acting as judge. A substantial portion of this examination (lines 1119–1248) is devoted to prologues, with Aeschylus and Euripides quoting opening passages of their own plays and exposing them to the other’s criticism. Notably, the reference is explicitly to ‘prologues’, and nowhere in the play are the quoted passages labelled as archai, although most of them demonstrably are the very first lines of their respective dramas.32 From line 1198 on, Aeschylus reproaches Euripides for the structural and metrical predictability of his verse: the older tragedian regularly interrupts the rival’s recitation of his openings after the fifth element of the first, second or third line, and completes the trimeter with the sequence ληκύθιον ἀπώλεσεν (‘… lost his little flask of oil’). While Aristophanes does not specify from which play each passage derives, the scholia normally provide relevant information. As to the Euripidean open-

 28 Hes. T 49 Most, Praxiph. frr. 28A and 28B Matelli. See Matelli (2009) 32–42. 29 Paus. 9.31.4 = Hes. T 42 Most. See Cingano (2009) 102–103 and Montanari (2009) 316–320. 30 P. Berol. inv. 9782, col. iii, ll. 28–36 φέρ[ε]|[τ]α̣ι δ̣ὲ̣ καὶ ἄλλο προοί|μ̣ιον ὑπόψ[υ]χρον | σχεδὸν τῶν ἴσων | στίχων, οὗ ἀρχή· | ‘ἆρά γε, ὦ παῖ, φέρεις τὸν | [π]ε[ρὶ Θε]αιτήτου λόγον;’ | τὸ δὲ γνήσιόν ἐστιν, | ο̣ ὗ̣ ἀρχή· ‘ἄρτι, ὦ Τερψί|ων’. 31 Several cases are discussed in Haslam (1975). 32 But see the discussion of the Meleager below.

  Chiara Meccariello ings, in most cases the play’s identification offered in the scholia is straightforward and confirmed by other sources. Two quotations, however, are problematic, and one of the two already puzzled ancient scholars.

. Archelaus The first incipit to receive the ληκύθιον treatment is the following:33 Αἴγυπτος, ὡς ὁ πλεῖστος ἔσπαρται λόγος, ξὺν παισὶ πεντήκοντα ναυτίλῳ πλάτῃ Ἄργος κατασχών

The scholia on this passage are worth quoting in full:34 (1206a) Αἴγυπτος, ὡς ὁ πλεῖστος ἔσπαρται λόγος: Ἀρχελάου αὕτη ἐστὶν ἡ ἀρχὴ (1206b) ὥς τινες ψευδῶς φασιν. οὐ γὰρ φέρεται νῦν Εὐριπίδου λόγος οὐδεὶς τοιοῦτος. (1206c) οὐ γάρ ἐστι, φησὶν Ἀρίσταρχος, τοῦ Ἀρχελάου, εἰ μὴ αὐτὸς μετέθηκεν ὕστερον, ὁ δὲ Ἀριστοφάνης τὸ ἐξ ἀρχῆς κείμενον εἶπεν. ὥς τινες· ψευδῶς (φασιν om.) interpunxit Strecker : Ἀρχελάου αὕτη … ὥς τινες ψευδῶς Schnee : Ἀρχελάου αὕτη … ὥς τινες ψευδῶς Roemer || φασιν M : om. VEΘBarb

Despite some textual and punctuation uncertainties, we can be reasonably certain that according to these notes some unnamed scholars (τινες) identified the lines as the beginning of the Archelaus, but others said that this identification was incorrect (ψευδῶς), as no such passage by Euripides was preserved. The statement of Aristarchus is also reported, according to which this was not the beginning of the Archelaus, unless Euripides reworked the play’s original beginning. Evidently, Aristarchus’ Archelaus had a different arche, and indeed the following opening of the play is widely attested in Hellenistic and later authors:35 Δαναὸς ὁ πεντήκοντα θυγατέρων πατὴρ Νείλου λιπὼν κάλλιστον †ἐκ γαίας† ὕδωρ, ὃς ἐκ μελαμβρότοιο πληροῦται ῥοὰς  33 Ar. Ran. 1206–1208 = Eur. TrGF F 846. 34 Schol. Ar. Ran. 1206; numbering and text as in Chantry (1999). 35 Eur. TrGF F 228. The quotation of the fragment in D.S. 1.38.4 belongs to a section derived from Agatharchides of Cnidus, a younger contemporary of Aristarchus (FGrHist 86 F 19). The date of the anonymous On the Nile, another witness to the Euripidean fragment (FGrHist 647), is unknown, but the contents of his work are likely to reach back to the Hellenistic period as well: see Gambetti (2012) for a discussion of this author.

‘Well Begun is Half Done’?  

Αἰθιοπίδος γῆς, ἡνίκ’ ἂν τακῇ χιὼν †τέθριππεύοντος† ἡλίου κατ’ αἰθέρα, ἐλθὼν ἐς Ἄργος ᾤκισ’ Ἰνάχου πόλιν· Πελασγιώτας δ’ ὠνομασμένους τὸ πρὶν Δαναοὺς καλεῖσθαι νόμον ἔθηκ’ ἀν’ Ἑλλάδα.

5

For Aristarchus the Aristophanic arche, Αἴγυπτος, ὡς ὁ πλεῖστος ἔσπαρται λόγος, was already a fragment, but he clearly was able to access the whole text of the Archelaus, in which he found a different incipit, conceivably Δαναὸς ὁ πεντήκοντα θυγατέρων πατήρ. Since he was unable to find the Aristophanic opening anywhere in the Euripidean production, it is safe to conclude that this incipit was lost by the beginning of the second century BCE, and probably never made it to Alexandria. Two main scenarios may explain this fact: (a) the Aristophanic lines are from a play that, as a whole, never made it to Alexandria; (b) the Aristophanic lines are an incipit that, as such, never made it to Alexandria as part of a work. The first scenario is, in principle, not implausible. We know that an early loss affected the Euripidean production, so that the Alexandrian corpus of Euripides only included about the 85% of his original production.36 Alexandrian scholars were indeed aware of this fact,37 and even if the scholium does not state it explicitly, Aristarchus’ take on the Αἴγυπτος-arche quoted in Aristophanes might implicitly be that it stemmed from a lost play. In fact, the hypothesis of a reworking of the Archelaus seems just a compromise (εἰ μὴ) between the ascription of the lines to this play, which conceivably Aristarchus found in a previous commentator, and their absence from the Alexandrian text, but is not presented as Aristarchus’ preferred hypothesis. However, “it is prudent to posit unknown Euripidean tragedies with the greatest reluctance”, since most of the lost plays must have been satyr dramas;38 and it seems unlikely that Aristophanes would here quote an isolated satyric incipit.39 As to (b), we can infer from the scholium that Aristarchus considered authorial reworking a possible scenario. Was he aware of such a scenario in other cases? While the biographical tradition on Euripides might have preserved and at times produced stories of authorial reworking already in the early Hellenistic  36 See the numerical distinction between the entire Euripidean production and the extant plays in the Lives of Euripides, TrGF T 1 IA.9, 28–29 and T 1 IB.5, 57–59. 37 This is indicated by the phrase οὐ σῴζεται appended to the title of the satyr drama Theristai in the didascalic section included in Aristophanes of Byzantium’s hypothesis of the Medea (Hyp. (a) Eur. Med., l. 43 Diggle = TrGF vol. 5.1, p. 425, test. i). 38 Scullion (2006) 187 (my italics). On the numerical consistency of the Euripidean σῳζόμενα and οὐ σῳζόμενα, see Kannicht (1996) and Meccariello (forthcoming). 39 See in particular Ran. 860–864 and 1119–1121, where the reference is to “tragedies”.

  Chiara Meccariello period,40 in the specific case of the Archelaus, authorial reworking might have seemed a concrete possibility in light of the notion, well-established in antiquity, that Euripides wrote the play for King Archelaus, while spending the last part of his life at his court in Macedon.41 If we accept that the Aristophanic opening is from Euripides’ Archelaus, we need to assume that the play was known to the Athenian audience by 405, so that it would be recognisable when recited during the comic agon of the Frogs.42 Thus, if the play was first produced in Macedon, we should suppose that it was also performed in Athens shortly afterwards. Conversely, Euripides may have produced the play in Athens before leaving for Macedon, and then performed the play there, too. In either case, he may well have written two different versions of the opening, only one of which — the Macedonian version and not the Athenian one quoted in the Frogs — was extant in the second century BCE, at least in the manuscripts to which Aristarchus had access. In this scenario, the Macedonian version could well have found favour in Alexandria for several reasons. The extensive Nile reference, with its icastic emphasis on the Egyptian provenance of Danaus, could have been an important factor. Likewise, the genealogy of Archelaus outlined in the rest of that prologue (TrGF F 228a), which connects the Macedonian royal house to Zeus through Heracles, might have been of particular interest to the Ptolemaic family.43 An ideological preference for this prologue (or part thereof) might explain why the other had become unknown by the time of Aristarchus. This explanation remains plausible even if the revision of the opening was not Euripidean, as some textual elements seem to suggest. In particular, two lines of the non-Aristophanic incipit were already considered redundant in antiquity, and their athetesis has been supported by modern scholars.44 Other oddities and

 40 Later evidence on this phenomenon: Plut. Am. 756b–c (on the reworking of the opening of the Melanippe Sophe) and P. Oxy. LXXVI 5093 (on an alleged overall revision of the Medea). On the complicated relationship between anecdotes and text in the case of the beginning of Euripides’ Phoenissae, see Meccariello (2014a). 41 TrGF vol. 5.1 (13) testt. ib–iia2. Euripides’ journey to Archelaus’ court is already presupposed in Arist. Pol. 5.1311b30–34, but this passage shows that the tradition on the matter had an anecdotal nature already in Aristotle’s time. For a sceptical approach to this tradition, see Lefkowitz (20122) 91 and Scullion (2003). 42 See esp. Scullion (2003) 394–395. 43 Cf. Theocritus’ Idyll 17 (encomium of Ptolemy II), esp. 13–33. 44 Tib. De fig. Demosth. 48 οἱ γὰρ τρεῖς περὶ τοῦ ποταμοῦ στίχοι περιττοί (i.e. ll. 3–5), Diggle (1998) 94, Collard/Cropp/Gibert (2004) 351.

‘Well Begun is Half Done’?  

textual problems affect the fragment.45 However, if this was the text available to Aristarchus, these details (as far as we can tell) were still not enough to instill suspicion, and to make the Aristophanic arche perform a diagnostic function on the soundness and authenticity of the opening of the Archelaus ascribed to Euripides in the Hellenistic period. And similarly, most of modern scholars’ cautious lack of trust in the τινες of the scholium, who are suspected to have been misled by the similarity of the two incipits,46 has resulted in including the fragment quoted in Frogs among those incertae sedis while retaining the other opening among the genuine Archelaus fragments.47

. Meleager My second Euripidean example stems from the same comic scene. The tragic lines recited by Euripides this time are the following:48 Οἰνεύς ποτ᾿ ἐκ γῆς πολύμετρον λαβὼν στάχυν θύων ἀπαρχάς

The scholium ascribes the lines to the Meleager, and adds that they are not strictly opening lines, but found μετὰ ἱκανὰ τῆς ἀρχῆς. The proper ἀρχὴ of the play is also adduced: it runs Καλυδὼν μὲν ἥδε γαῖα Πελοπίας χθονός, in agreement with several other sources.49 Unlike the case of the Archelaus, the scholiast here does not show doubt nor does he preserve different hypotheses. His statement, however, collides with two elements: first, the structure of the Aristophanic fragment, most notably the lack of connective particles and the use of ποτε, which can be compared to incipits such as TrGF F 819.1 Σιδώνιόν ποτ’ ἄστυ Κάδμος ἐκλιπών (Phrixus II) or TrGF F

 45 A striking feature of l. 3 is the genitive in -οιο, unparalleled in iambic trimeters. For this and other issues, see esp. Harder (1985) 182–188. 46 For example, the start with a proper name, the reference to fifty children, the participial clause conveying movement, the mention of the arrival in Argos: see Harder (1985) 181. 47 Notable exceptions are Haslam (1975) 171, Dover (1993) ad 1206–1208, who favours the hypothesis of a prologue reworking, either by Euripides in Macedon or, preferably, by others in the fourth century, Sommerstein (1996) ad 1206–1208, Scullion (2006). 48 Ar. Ran. 1240–1241 = Eur. TrGF F 516. 49 Schol. Ar. Ran. 1238 Chantry Οἰνεύς ποτ’ ἐκ γῆς: ἔστι μὲν ἐκ Μελεάγρου μετὰ ἱκανὰ τῆς ἀρχῆς. ἡ δὲ ἀρχὴ τοῦ δράματος ‘Καλυδὼν μὲν ἥδε γαῖα Πελοπίας χθονός’ (Eur. TrGF F 515.1). The latter fragment is quoted by several authors, including Aristotle (Rhet. 3.1409b8), who, however, gives no indication of its opening position, and ascribes it to Sophocles.

  Chiara Meccariello 539a Φοίβου ποτ’ οὐκ ἐῶντος ἔσπειρεν τέκνον (Oedipus);50 second, the reasonable certainty that all the other quotations in this portion of the agon are strictly initial. A possible explanation is that the scholiast did not have the full text of the Meleager and thus his statement was nothing more than an educated guess. A scenario in which he only had a different fragmentary arche of the Meleager (presumably Καλυδὼν μὲν ἥδε γαῖα Πελοπίας χθονός) and a general knowledge of the myth of Meleager might indeed explain the conjecture that the Aristophanic lines Οἰνεύς ποτ᾿ ἐκ γῆς πολύμετρον λαβὼν στάχυν/θύων ἀπαρχάς came from the Meleager, but did not constitute its arche.51 While keeping in mind the composite character of our scholia, we can argue against this hypothesis by pointing out that our set of marginal annotations on this scene shows more than a superficial knowledge of some plays’ incipits. The scholiast is able to provide the actual continuation of the lines truncated by Aeschylus in most cases.52 The fact that he does not do so for the Meleager might arouse suspicion. However, a similar note is also missing at the beginning of the second Phrixus, which cannot be due to lack of documentation, since its prologue was well known in antiquity, and its first six lines are even quoted by Tzetzes in the twelfth century.53 Such occasional lack of information is thus more likely to be due to the intermittent character of the extant marginal notes than to lack of data on the scholiast’s part. On these grounds, it is reasonable to assume that the scholiast had access to the entire prologue of the Meleager, where he found the Aristophanic lines not at the beginning but, as he states, μετὰ ἱκανά. There are two main ways to explain this data: either (a) Aristophanes used a non-initial passage; or (b) interpolation and/or reworking affected the beginning of the Meleager, so that the original opening quoted by Aristophanes lost its initial position.

 50 Of course the use of ποτε, while paralleled in Euripidean archai, is not limited to incipits: see e.g. Hipp. 24, HF 10 and 64. 51 Note that Apollodorus’ summary of this myth shows resemblances with the Aristophanic incipit: see in particular Bibl. 1.8.2 τὰς ἀπαρχὰς Οἰνεὺς θεοῖς πᾶσι θύων μόνης Ἀρτέμιδος ἐξελάθετο. 52 Schol. Ar. Ran. 1213b πηδᾷ χορεύων: τὸ δὲ λοιπὸν τοῦ ἰάμβου ‘παρθένοις σὺν Δελφίσιν’, 1219a δυσγενὴς ὤν: τὸ λοιπὸν τοῦ ἰάμβου ‘πλουσίαν ἀροῖ πλάκα’, 1233 θοαῖσιν ἵπποις: τὸ δὲ λοιπὸν τοῦ ἰάμβου ‘Οἰνομάου γαμεῖ κόρην’. 53 Tzetz. schol. Ran. 1225 (IV.3, 1047, 19 Koster). As to the Meleager incipit, the scholia recentiora provide the impossible continuation οὐκ ἔθυσεν Ἀρτέμιδι, but this is “nil nisi αὐτοσχεδίασμα grammaticorum” (Kannicht ad F 516.2).

‘Well Begun is Half Done’?  

In favour of (a), we must stress two facts. First, non-initial lines from prologues, if appropriately cut, may actually read as openings. To take just one example, line 50 of the Phoenissae, μούσας ἐμὸς παῖς Οἰδίπους Σφιγγὸς μαθών, is in syntactic continuity with the last portion of line 49, τυγχάνει δέ πως, but when quoted alone shows the lack of connective particles typical of opening lines. Second, not every Euripidean opening has a metrical and syntactical structure alterable with the lecythion joke. In our specific case, it is worth noting that the fragment receives the lecythion treatment twice, once in the first and once in the second line.54 How many suitable first and second lines could Aristophanes have found in the relatively recent Euripidean production? In the extant nineteen plays, for example, there are none. The thesis of a later interpolation (b), however, is dominant among modern scholars, who prefer to consider the Meleager prologue an example of failure of the arche-system due to post-Aristophanic reworking, rather than a target of comic manipulation.55

. Rhesus My third and last example pertains to the Rhesus. An anonymous introduction prefixed to its text in the medieval manuscripts contains a brief discussion of the authenticity of the play, which, according to some scholars (generically called ἔνιοι), was not Euripidean, but displayed a rather more Sophoclean character, despite being recorded as genuine in the Didascaliae. After pointing out that the interest in sky matters supports Euripidean authorship, the scholiast adds a further remark:56 πρόλογοι δὲ διττοὶ φέρονται. ὁ γοῦν Δικαίαρχος ἐκτιθεὶς τὴν ὑπόθεσιν τοῦ Ῥήσου γράφει κατὰ λέξιν οὕτως· ‘νῦν εὐσέληνον φέγγος ἡ διφρήλατος’. καὶ ἐν ἐνίοις δὲ τῶν ἀντιγράφων ἕτερός τις φέρεται πρόλογος, πεζὸς πάνυ καὶ οὐ πρέπων Εὐριπίδῃ· καὶ τάχα ἄν τινες τῶν

 54 Ar. Ran. 1238–1242. 55 See Kannicht’s apparatus ad TrGF FF 515, 516. Haslam (1975) 171 connects the complicated situation of the Archelaus and Meleager incipits precisely to their being part of the lecythion scene, and in a more prominent way than the other incipits, being, respectively, the first Euripidean opening to receive the lecythion treatment and the only one to receive it twice. This, according to Haslam, would have encouraged changes to the relevant Euripidean texts after the performance of the Frogs, “evidently to save them from a recurrence, in actual performance, of the treatment they had got from Aristophanes”. For sensible objections to this hypothesis, already advanced by Fritzsche (1845) 367 for the Meleager, see Scullion (2006) 190. 56 Hyp. (b) Rh., ll. 26–44 Diggle = TrGF vol. 5.2, p. 642, test. ia (Dicearch. fr. 114 Mirhady).

  Chiara Meccariello ὑποκριτῶν διεσκευακότες εἶεν αὐτόν. ἔχει δὲ οὕτως· ‘ὦ τοῦ μεγίστου Ζηνὸς ἄλκιμον τέκος, / … / Πριάμου, βίᾳ πρόρριζον ἐκτετριμμένην’. 1 Δικαίαρχος Nauck : δικαίαν VL : om. Q | ἐκτιθεὶς V : ἐπιτιθεὶς LQ || 1–2 ὁ γοῦν Δικαίαρχος … γράφει κατὰ λέξιν οὕτως· ‘νῦν–διφρήλατος’ καὶ . ἐν ἐνίοις δὲ …” Schwartz : οὕτως· Luppe

The passage suffered from textual corruption, and what we read is the result of an important correction: the name Δικαίαρχος is Nauck’s emendation of the manuscripts’ δικαίαν. The text with δικαίαν makes no sense, either syntactically or semantically;57 on the other hand, a connection between Dicearchus and the genre of hypotheses is known from other sources, albeit very few, so it seems appropriate that he be described as ἐκτιθεὶς τὴν ὑπόθεσιν τοῦ Ῥήσου.58 Once the emendation is accepted, there is still another problem concerning the boundaries of the Dicearchean quotation: is the iambic trimeter the only part of the quotation, or do the remarks on the actors’ prologue also belong to it?59 As I understand the passage,60 the trimeter is the only part of the Dicearchean quotation. This quotation and the following discussion of the actors’ prologue are the two parts of the whole argument about the existence of two prologues. The mention of ὑπόθεσις and ἀρχή side by side reminds us of the papyrus hypotheses, and leads us to think that the commentator found that arche in a collection of narrative hypotheses.61 But why does the commentator quote the first line, and only that? And why does he include, by contrast, eleven lines when quoting the other prologue? Before addressing these questions, we should note that the Rhesus as we read it, to which this note is prefixed, does not have a iambic prologue, and accordingly a prefatory note preserved in the medieval manuscripts, in pointing out that the chorus of the play is made of φύλακες Τρωϊκοί, adds that they also ‘deliver the prologue’ (προλογίζουσι).62

 57 The adjective is omitted in one late manuscript probably for that reason. For an attempt to save the paradosis, see Carrara (1992). 58 On the Dicearchean authorship, see most recently Verhasselt (2015) and the bibliography therein. 59 Various possibilities are registered in Kannicht’s apparatus, TrGF vol. 5.2, pp. 642–643. See also Merro (2008) ad loc. 60 See Meccariello (2014b) 71–73 and the bibliography therein. 61 Luppe (1990) goes as far as to add after οὕτως as the beginning of the quotation from Dicearchus, reproducing the usual heading of the papyrus hypotheses. 62 Hyp. (c) Rh., ll. 52–53 Diggle.

‘Well Begun is Half Done’?  

Thus, we have evidence of three different beginnings for the play: Our choral incipit, preserved in the medieval manuscripts and also presupposed by the brief note on the prologue speaker: βᾶθι πρὸς εὐνὰς τὰς Ἑκτορέους; 2) The iambic arche quoted in Dicearchus’ hypothesis: νῦν εὐσέληνον φέγγος ἡ διφρήλατος; 3) The iambic arche of the actors’ prologue allegedly preserved in some copies: ὦ τοῦ μεγίστου Ζηνὸς ἄλκιμον τέκος.

1)

These could be the archai of different plays, or they may just reflect a shifting situation limited to the beginning of one single play. In other words, the Rhesus might have originally started with a prologue in iambic trimeters whose arche was that quoted by Dicearchus, and — in some witnesses — with the dull prologue singled out as an actor’s interpolation. But it is equally possible that one or both of the iambic archai pointed to a different Rhesus from what we have — a possibility that, of course, has its bearing on the well-known authenticity problem of this tragedy. Certainly the author of the anonymous introduction knew at least two archai, that of the actors’ prologue and the Dicearchean line; but he might also have read a third arche, if he knew the choral beginning, as in our version of the play. Let us suppose that he had the same Rhesus we have, without an iambic prologue, and, in addition, that he had some copies of the same Rhesus with the actors’ prologue prefixed, and then the Dicearchean hypothesis with a different arche (but not the entire Dicearchean prologue). From the latter arche he did not deduce that Dicearchus had a different play, but only the existence of a different prologue. Why? Perhaps, we can argue, because the summary was clearly that of the same Rhesus. However, judging from the wording of the introduction, and in particular the sequence πρόλογοι δὲ διττοὶ φέρονται, it seems that the scholiast was able to trace the Dicearchean line to a still existing prologue. Indeed, the word φέρονται points to the actual circulation of the two prologues, as for example in the hypothesis to Aristophanes’ Clouds, which refers to the existence of two versions of the play with the words διτταὶ φέρονται Νεφέλαι, and also in the above-mentioned Platonic commentary preserved in P. Berol. inv. 9782.63 And φέρεται is the  63 Hyp. Ar. Nub. A8 Holwerda (on the existence of two versions of the Clouds, both of which survived into the Hellenistic period, see Dover 1968, lxxx–xcviii), P. Berol. inv. 9782, col. iii, ll. 28–30 φέρ[ετ]α̣ι δ̣ὲ̣ καὶ ἄλλο προοίμ̣ιον. See, conversely, οὐ φέρεται in schol. Ar. Ran. 1206 quoted above.

  Chiara Meccariello verb that our scholiast uses for the spurious prologue, which he is able to quote extensively. The parallel circulation of the Dicearchean and the actors’ prologue would also account for the phrasing ἐν ἐνίοις δὲ τῶν ἀντιγράφων, which seems to point to a minority attestation of the actors’ prologue and an implied opposition to other copies bearing the Dicearchean one. Of course, in this case the reference to Dicearchus would only be meant to give authority to the first prologue, as opposed to what the scholiast considers an uninspired interpolation. Also, in the light of our previous overview of the arche-system, the quotation of the incipit only is unsurprising. It does not need to be due to the unavailability of the whole prologue; on the contrary, it is striking that more than the arche has been quoted for the second prologue. After all, of the two prooimia of Plato’s Theaetetus only the beginnings are quoted. In our case, the asymmetry might stem from the scholiast’s desire to provide the reader with textual material to corroborate his severe judgment on that prologue, since he probably expected that it would be rightfully excluded from most copies. However we explain this three-archai situation, we can easily imagine how in similar cases the arche-system could go haywire. In a catalogue of incipits without hypotheses, a reader would have no way of telling whether the arche they found quoted, if different from the one that opened their copy of a given play, pointed to the same play, only with a different beginning, or to a different play at all. In other words, the arche is not a sufficient disambiguation device. It is tempting to suggest that in the collection of hypotheses known from the papyri, title, arche and hypothesis all functioned as identification devices, each in its own way; that the summaries, besides offering a preview and an overview of the contents, occasionally usable as self-standing mythical narrations, were also meant to offer yet another form of identification, thus further securing the textual identity of the plays from possible intrusions of spurious works.

 Conclusions As shown in the previous sections, the quotation of incipits as naming and textidentification devices was widespread in Greek antiquity and was variously employed in literary and paraliterary works. However, we also have evidence of possible or actual failures of the system due to variations in literary openings. It now remains to consider briefly if and how this system and its shortcomings interacted with ancient literary reflection and practice.

‘Well Begun is Half Done’?  

Beginnings seem often to have attracted the attention of Greek literary critics. The scholia on the first line of the Iliad, for example, show traces of critical approaches to the poem’s opening, including a zetema on the reasons why Homer chose to start with the ill-omened word μῆνιν. Zenodotus solved the question by pointing out that a proem ought to be lively and capture the attention of the audience.64 Several ancient critics and rhetors show awareness that attention and goodwill are inevitably secured or compromised from the beginning, and that the opening has particular prominence because it is more likely to stick in the listener’s memory.65 We can expect that, for the same reasons, ancient authors devoted particular attention to the opening words of their works. In this context one may mention the story reported by Dionysius of Halicarnassus, according to which when Plato died a tablet was found on which he had reworked the beginning of the Republic several times, a sign of his literary industry (φιλοπονία).66 It is not unlikely, I believe, that the widespread practice of quoting the arche, with or without a title, to identify a work also increased ancient authors’ attention to this specific portion of the text.67 And, paradoxically, this may have made incipits a preferred target of reworking. The very prominence of the opening seems to have been, potentially, a double-edged sword. The arche-system may have contributed to making works’ beginnings vulnerable, a sort of perennial work-in-progress, liable not only to authorial revision, but also non-authorial alteration, such as athetesis and interpolation. This, in turn, may sometimes have resulted in the arche no longer being a secure identification device. At any rate, while the examples we have seen may be indicative of the occasional failure of the arche-system — since in some cases the arche may have failed to secure the textual identity of a work — they also show the system’s diagnostic potential, to the benefit of both ancient and modern scholars.

 64 Schol. Il. 1.1a and 1b Erbse; cf. Luc. VH 2.20. 65 See for example D.H. Lys. 17, Is. 7. The “battle of the prologues” in Aristophanes’ Frogs may also be adduced in this context. 66 D.H. Comp. 25.209–218; cf. D.L. 3.37. 67 See also Kenney (1970).

  Chiara Meccariello

References Armstrong, A.H. (1969), Plotinus. Ennead, Volume I: Porphyry on the Life of Plotinus. Ennead I, Cambridge (Mass.). Bassino, P. (2019), The ‘Certamen Homeri et Hesiodi’. A Commentary, Berlin/Boston. Bernsdorff, H. (2014), ‘Notes on P.Mich. inv. 3498 + 3250b recto, 3250a and 3250c recto (List of Lyric and Tragic Incipits)’, APF 60, 3–11. Blum, R. (1991), Kallimachos: The Alexandrian Library and the Origins of Bibliography, translated from the German by H.H. Wellisch, Madison (German edition: R. Blum, Kallimachos und die Literaturverzeichnung bei den Griechen, Frankfurt 1977). Borges, C./Sampson, M. (2012), New Literary Papyri from the Michigan Collection: Mythographic Lyric and a Catalogue of Poetic First Lines, Ann Arbor. Borrelli, B./Colella, L.C./D’Angelo, M./Di Tuccio, A./Nicolardi, F./Parisi, A. (2019), ‘On the Michigan Ptolemaic Lyric Papyrus Inv. 3498 + 3250b, 3250c, 3250a: New Readings and Remarks’, ZPE 210, 21–53. Carrara, P. (1992), ‘Dicearco e l’hypothesis del Reso’, ZPE 90, 35–44. Chantry, M. (1999), Scholia in Aristophanem III: Scholia in Thesmophoriazusas; Ranas; Ecclesiazusas et Plutum; Fasc. Ia, Continens Scholia Vetera in Aristophanis Ranas, Groningen. Cingano, E. (2009), ‘The Hesiodic Corpus’, in F. Montanari et. al. (eds), Brill’s Companion to Hesiod, Leiden, 91–130. Collard, C./Cropp, M./Gibert, J. (eds) (2004), Selected Fragmentary Plays, vol. II, Warminster. Cribiore, R. (1996), Writing, Teachers, and Students in Graeco-Roman Egypt, Atlanta. Dalby, A. (1986), ‘The Sumerian Catalogs’, The Journal of Library History 21, 475–487. Davies, M./Finglass, P.J. (2014), Stesichorus: The Poems, Cambridge. Diggle, J. (1998), Tragicorum Graecorum Fragmenta Selecta, Oxford. Dover, K. (1968), Aristophanes: Clouds, Oxford. Dover, K. (1993), Aristophanes: Frogs, Oxford. Falivene, M.R. (2011), ‘The Diegeseis Papyrus: Archaeological Context, Format, and Contents’, in B. Acosta-Hughes/L. Lehnus/S. Stephens (eds), Brill’s Companion to Callimachus, Leiden/Boston, 81–92. Fritzsche, F.V. (1845), Aristophanis Ranae, Zurich. Gambetti, S. (2012), ‘Anonymous, On the Nile (647)’, in I. Worthington (ed.) (2012) Brill’s New Jacoby, consulted online in March 2018 (https://dx.doi.org/10.1163/1873-5363_bnj_a647). Geißler, C.J. (2014), ‘Anmerkungen zu einer Liste mit lyrischen und tragischen Gedichtanfängen (P.Mich. inv. 3498 + 3250 b recto, 3250 c recto und 3250 a recto)’, APF 60, 12–24. Guidorizzi, G./Del Corno, D. (1996), Aristofane. Le Nuvole, Milan. Hallo, W.W. (1963), ‘On the Antiquity of Sumerian Literature’, Journal of the American Oriental Society 83, 167–176 (repr. in The World’s Oldest Literature: Studies in Sumerian BellesLettres, Leiden/Boston 2010, 137–150). Harder, A. (1985), The Kresphontes and Archelaos of Euripides, Leiden. Haslam, M.W. (1975), ‘The Authenticity of Euripides, Phoenissae 1–2, and Sophocles, Electra 1’, GRBS 16, 149–174. Kannicht, R. (1996), ‘Zum Corpus Euripideum’, in C. Mueller-Goldingen/K. Sier (eds), Lenaika: Festschrift für Carl Werner Müller zum 65. Geburtstag, Stuttgart, 21–31. Kenney, E.J. (1970), ‘That Incomparable Poem the ‘Ille ego’?’, CR, n.s., 20, 290. LDAB = Leuven Database of Ancient Books (https://www.trismegistos.org/ldab/index.php).

‘Well Begun is Half Done’?  

Lefkowitz, M. (1981, 20122), The Lives of the Greek Poets, Baltimore. Luppe, W. (1990), ‘Dikaiarchos und der Rhesos-Prolog’, ZPE 84, 11–13. Maltomini, F. (2003), ‘Considerazioni su P. Oxy. LIV 3724: Struttura e finalità di una lista di incipit epigrammatici’, ZPE 144, 67–75. Martano, A. (2012), ‘Chamaeleon of Heraclea: The Sources, Text and Translation’, in E. Matelli/ A. Martano/D. Mirhady (eds), Praxiphanes of Mytilene and Chamaeleon of Heraclea: Text, Translation, and Discussion, New Brunswick, 157–338. Matelli, E. (2009), ‘Questioni di epica arcaica comuni a Prassifane e ad Aristarco’, Aevum 83, 31–60. Meccariello, C. (2014a), ‘The Opening of Euripides’ Phoenissae between Anecdotal and Textual Tradition’, ZPE 190, 49–56. Meccariello, C. (2014b), Le hypotheseis narrative dei drammi euripidei. Testo, contesto, fortuna, Rome. Meccariello, C. (2016), ‘Title, Arche, Hypothesis. Notes on the Heading and Arrangement of the Euripidean Hypotheses on Papyrus’, in T. Derda/A. Łajtar/J. Urbanik (eds), Proceedings of the 27th Congress of Papyrology, Warsaw, July – August 2013, vol. II, Warsaw, 1185–1200. Meccariello, C. (forthcoming), ‘Eight and Counting. New Insights on the Number and Early Transmission of Euripides’ Satyr Dramas’, in A. Antonopoulos/G. Harrison/M. Christopoulos (eds) (forthcoming), Reconstructing a Genre From its Remnants: Studies in Satyr Drama, Berlin/New York. Merro, G. (2008), Gli scoli al Reso euripideo, Messina. Montanari, F. (1989), ‘29. Chamaeleon. 4T: POxy 2506, fr. 26, col. I 1-14: De Stesichoro’, in Corpus dei papiri filosofici greci e latini (CPF), vol. I.1, Florence, 409–413. Montanari, F. (2009), ‘Ancient Scholarship on Hesiod’, in F. Montanari et al. (eds), Brill’s Companion to Hesiod, Leiden, 313–342. Nachmanson, E. (1941), Der griechische Buchtitel, Gothenburg. Olson, S.D. (2002), Aristophanes: Acharnians, Cambridge. Otranto, R. (2000), Antiche liste di libri su papiro, Rome. Parsons, P.J./Maehler, H./Maltomini, F. (2015), The Vienna Epigrams Papyrus (G 40611), Berlin/ Munich/Boston. Pfeiffer, R. (1968), History of Classical Scholarship from the Beginnings to the End of the Hellenistic Age, Oxford. PMG = Page, D.L. (1962), Poetae Melici Graeci, Oxford. Prodi, E.E. (2017), ‘Text as Paratext: Pindar, Sappho, and Alexandrian Editions’, GRBS 57, 547– 582. Puglia, E. (2008), ‘P. Oxy. 2294 e la tradizione delle odi di Saffo’, ZPE 166, 1–8. Sanzo, J.E. (2014), Scriptural Incipits on Amulets from Late Antique Egypt: Text, Typology, and Theory, Tübingen. Scullion, S. (2003), ‘Euripides and Macedon, or the Silence of the Frogs’, CQ 53, 389–400. Scullion, S. (2006), ‘The Opening of Euripides’ Archelaus’, in D. Cairns/V. Liapis (eds), Dionysalexandros: Essays on Aeschylus and His Fellow Tragedians in Honour of Alexander F. Garvie, Swansea, 185–200. SH = Lloyd-Jones, H./Parsons, P. (1983), Supplementum Hellenisticum, Berlin/New York. Sommerstein, A.H. (1982), Aristophanes: Clouds, Warminster. Sommerstein, A.H. (1996), Aristophanes: Frogs, Warminster.

  Chiara Meccariello Sommerstein, A.H. (2002), ‘The Titles of Greek Dramas’, SemRom 5, 1–16 (repr. with addenda in The Tangled Ways of Zeus and other studies in and around Greek Tragedy, Oxford 2010). TrGF = Kannicht, R. (2004), Tragicorum Graecorum Fragmenta, vol. 5, Euripides, Göttingen. Van Rossum-Steenbeek, M. (1998), Greek Readers’ Digests? Studies on a Selection of Subliterary Papyri, Leiden. Verhasselt, G. (2015), ‘The Hypotheses of Euripides and Sophocles by ‘Dicaearchus’’, GRBS 55, 608–636.

Roberta Berardi

Collecting Fragments for a Fragmentary Literary Genre The Case of Greek Hellenistic Oratory

 Introduction and methodological premises Contrary to the idea that most of the fragments of ancient authors have been already collected and edited, a scholar interested in Hellenistic history and literature may find it surprising that a collection of fragments of Hellenistic orators is not something we would be able to find right now on the shelves of any library.1 The reasons are multiple. Oratory in the Hellenistic age has been for a long time seen as a minor aspect of the social, political, and literary scenery of the GraecoRoman world. The myth according to which oratory quite mysteriously dissolved itself after the battle of Chaeronea in 338 BCE, retreating suddenly to the anonymity of rhetorical schools, has been widely accepted and taken for granted for centuries, while remarks were limited to the stylistic evolution of Greek prose.2 This glaring gap in Classical scholarship began to be filled only recently, when the need for a re-evaluation of oratory in the Hellenistic period was at last recognised.3 The focus has however always tended to be mainly on a few isolated and better known personalities such as Demochares of Leuconoe, to whom Gabriele Marasco4 dedicated a monograph, in which he even included and commented on a few fragments. In 1972 Wooten,5 in his unpublished doctoral thesis,

 1 The project I worked on for my DPhil (2016–2020) intended to fill this gap and provided a first commented edition of fragments of Hellenistic orators. 2 Still a fundamental reference is Norden (1898). See also the more recent Porter (1997). 3 Contrary to the widespread belief that the Greek poleis lost their independence and political freedom after the battle of Chaeronea, there is evidence that the Athenian institutions where oratory took place (namely the Council, the Assembly and the courts) continued functioning throughout the Hellenistic era (as also shown by epigraphical evidence), albeit in a slightly different way. On this, see Habicht (1995). 4 Marasco (1984). 5 See Wooten (1972). Understandably for an unpublished thesis, this work contains several flaws and eccentricities, not only in the choice of names and fragments (see below), but also due to some odd choices, such as considering the literary sources on oratory in the period to be comparable to the rhetorical exercises on papyrus. Moreover, the whole discussion is biased by the https://doi.org/10.1515/9783110712223-005

  Roberta Berardi tried to write an account of oratory in the Hellenistic period, although his main argument was that it was still too much dependant on Demosthenes and had no typically new elements. Kremmydas’ and Tempest’s book Hellenistic Oratory: Continuity and Change,6 which was published in 2013, provided a less biased approach to the subject,7 and emphasised for the first time the concept of change alongside continuity, as the title implies. Despite the importance of the reconsideration of Hellenistic oratory that we owe to this book, in its pages no role has been given to the actual personalities involved: the orators — the people who pronounced the speeches whose nature we aim to study — still remain out of the picture. Therefore, a major issue still requires investigation: who were the Hellenistic orators? Do we have any fragments left and are they worthwhile collecting? Or are the traces of names and testimonies just too vague and irretrievable, which is the impression gleaned from reading the secondary literature on the subject? The answer is fortunately positive: we do indeed have names and fragments. However, the need for a theoretical framework and perhaps the still-tacit belief that the anonymity of rhetorical schools won over the role of single personalities, have been clear impediments to the compilation of a first collection of fragments. Moreover, the growing tendency for the figures of orators and historians to be identifiable, from the end of fourth century BCE until the end of the Hellenistic era (and beyond), has resulted in many oratorical fragments being edited in collections of fragmentary historians. Both Müller’s and Jacoby’s collections are indeed filled with testimonia and sometimes fragments of speeches, which are considered historical pieces, just because their authors were both orators and historians, and are mainly known for the latter role.8

 author’s intention to demonstrate that oratory throughout the whole Hellenistic period still mainly depended on the model of Demosthenes. 6 Kremmydas/Tempest (2013). 7 The authors immediately acknowledge (but distance themselves from) the prejudices on which previous scholarship had built its reflection on Hellenistic oratory, when they say (p. 3): “The Athenocentricity of our extant literary sources at one time led to a misplaced assumption which dominated scholarship, namely that oratory died at the battle of Chaeronea alongside Athenian political freedom”. 8 Some were also quite famous, as for instance Neanthes of Cyzicus, FGrHist 84 (Laqueur 1935).

Collecting Fragments for a Fragmentary Literary Genre  

A revealing case is the school of Isocrates, whose last generation of pupils — mostly both orators and historians — can be seen as examples of Hellenistic orators, of whom perhaps the most famous is Theopompus of Chios.9 This model applies equally to other orators outside of the Isocratean school and further afield than Athens. An example for the same period is Anaximenes of Lampsacus (see D.H. Amm. 2 on his vast production),10 one of the main historians at the Macedonian court, who was also an orator. This is true not only for the early Hellenistic period, as my survey of the sources for the Hellenistic orators will hopefully show, but remains valid until the end of the period. One of the last examples is probably Apollonius Molon, one of Cicero’s teachers of rhetoric but also an eminent historian.11 Before having a detailed look at the different sources, at the nature of the material we can collect, and at the methodology that is safe to adopt in compiling an edition of fragments for a genre that is fragmentary at its core, I shall briefly consider two preliminary questions: 1) When does Hellenistic oratory start and end? 2) Who can be considered an orator of this period? 1) When does Hellenistic oratory start and end? As much as the conventional definition of the period from the death of Alexander in 323 BCE to the defeat of Cleopatra at Actium in 31 BCE can be useful for the analysis of many historical phenomena, I found it restrictive for a definition of Hellenistic oratory. The factors that produced changes in oratory in the Hellenistic age are many and diversified: they have to do with style but also and more importantly with the occasions for oratory and the audiences of oratory. The transition from an ‘Attic’ way of writing or delivering speeches to a ‘Hellenistic’ one is extremely slow: as a result, while Demochares (certainly a Hellenistic orator) is probably no less Attic than Demosthenes, conversely Pytheas12 (undoubtedly an Attic orator) was still politically active after the 323 BCE. In this rather complex situation, I decided to leave 31 BCE as the end-point, but chose to include fragments of speeches dateable to after 338 BCE, when the forerunners of Hellenism can begin to be traced. A broad chronological definition can help to better understand all the steps in this transition

 9 Among the vast bibliography on Theopompus, see esp. worthwhile discussions of his role as orator in Chávez Reino (2009), Flower (1994), and Pownall (2005). 10 See among others Brzoska (1894), Heckel (2006), Nitsche (1906), and Susemihl (1891) 140 n. 708, 511 n. 101, 655, Susemihl (1892) 452–453, 459 n. 17, 481 n. 93. 11 See the biographical entries in the Pauly/Wissowa: Brzoska (1895) and Weissenberger (2006). 12 Gärtner (1963).

  Roberta Berardi from Attic to Imperial oratory — to put it simplistically, from Demosthenes to Cicero. 2) Who can be considered an orator of this period? With regard to who can be considered an orator, the situation is even more complex. Fictitious speeches and rhetorical exercises already played a crucial role in the evolution and the practice of oratory during the Hellenistic period, and therefore the distinction between professional speeches and fictitious exercises is not always clear. Quotations from the Elder Seneca’s controversiae and suasoriae13 cannot be isolated or underestimated merely because of the nature of the work in which they are contained, and most importantly, rhetorical exercises, meletai, progymnasmata,14 tell us a lot about how the professional speeches we do not have anymore might have sounded. Papyrus fragments hastily categorised by their editors as ‘rhetorical exercises’ should be re-examined, as they may contain pieces of professional oratory.15 In the light of these preliminary remarks, an inclusive approach will be useful for a reconsideration of oratory throughout the whole period. Fragments of suasoriae and controversiae as well as some relevant rhetorical papyri should be included in a first edition of fragments of Hellenistic orators. On the contrary, fragments of rhetorical theorists, or even of rhetorical theory written by rhetors who were also professional orators, will be left aside for the moment and will constitute a subsidiary tool to the study of rhetorical education in the period.16

 A first survey on Hellenistic orators (with one example on papyrus) In the light of these premises, a first step is to understand the precise nature of the material that could be collected in a first edition. I have been able to count more than 80 names of Hellenistic orators. Such a high number derives from my very inclusive criterion of considering Hellenistic orators all writers for whom the activity of orator is attested in at least one single source. This leads to the inclusion of a number of personalities that only have one mention of this kind and for

 13 See more details below. 14 See Stramaglia (2010) and related bibliography (pp. 112–113). 15 Of this I will provide an example and discuss it below (pp. 87–94). 16 See for instance Anaximenes’ Rhetorica ad Alexandrum, Demetrius’ De Rhetorica, and the various papyri containing rhetorical technai (see Kremmydas 2013, 162–163).

Collecting Fragments for a Fragmentary Literary Genre  

whom only a brief biographical note will be given in the edition. From my selection are to be excluded late Attic or early Hellenistic orators, such as Dinarchus and Demades, who already have a recognised non-fragmentary corpus, or for whose extant fragments a context prior to 338 BCE is clearly inferable, such as is the case with Hegesippus17 and Python of Byzantium. Excluded from the selection also are rhetors who are attested as rhetorical instructors, but for whom there is no evidence of any active oratorical production.18 Some of the orators included in the list have only one or two testimonia,19 others have many testimonia but few fragments,20 yet others do not have any testimonia but have several fragments,21 and finally, some very fortunate examples have both a long list of testimonia and a number of fragments. The last group are orators who became famous for different reasons throughout the centuries, such as Demetrius of Phalerum, who was well-known for his political and philosophical activity (although the majority of his fragments are not oratorical), Theopompus of Chios, who was known as an historian, and to a lesser extent Demochares, who was a family relative of Demosthenes, and Hegesias,22 illustrious for his influence on later rhetorical theory and prose style. But most striking is the great variety of sources that transmit these fragments. A few are direct sources, among which is included one papyrus fragment,23 and some inscriptions containing decrees with the names of orators. However, the great majority are predictably quotations from literary or historical works from various periods and genres. There are among others Greek and Latin rhetoricians, both Classical and from Late Antiquity (precisely from Cicero to some medieval  17 The case of Hegesippus is quite controversial, as the only witnesses of his post Chaeronea activity are inscriptions (IG II2 23, IG II2 1623, 185 and 1629, 543, the latter mainly dealing with a decree from 325 about events prior to the battle, so not particularly relevant). The only actual piece of evidence of his oratory after Chaeronea is the speech On the Treaty with Alexander (333/332 BCE), attributed to Demosthenes but considered spurious with large consensus of scholars (as well as the speech On Halonnesus), but, being a fully preserved speech, this would not constitute a fragment stricto sensu, so it is safer not to include Hegesippus in the selection. 18 Given the ambiguity that lies in the Greek term ῥήτωρ, the benefit of the doubt should be given for those names for which sources witness the activity of ῥήτωρ but do not specify whether they worked as teachers. 19 For instance, Dionysius of Magnesia (Cic. Brut. 316, Plut. Cic. 4) and Diophanes of Mytilene (Cic. Brut. 104, Strab. 13.2.3, Plut. Ti. Gracch. 8, 20). 20 Such as the above-mentioned Anaximenes, whose fragments are mainly historical. 21 Such as Myron (Rut. Lup. 1.20, 2.1), otherwise unwitnessed, unless we identify him with Myron of Priene, but this hypothesis does not seem solid to me. 22 He was considered the ‘inventor’ of the Asian style. See Radermacher (1912) col. 2607. 23 The only fragment of an oration by Philiscus of Miletus is P. Oxy. XLI 2994, col. II.

  Roberta Berardi scholia to Hermogenes — the most relevant rhetorical sources are undoubtedly Dionysius of Halicarnassus and the treatise De figuris sententiarum et elocutionis by Rutilius Lupus, a Latin rhetorician),24 historians and biographers (especially Plutarch,25 and Diodorus Siculus),26 philosophers (a lot of information is derived from Diogenes Laertius), geographers (from Strabo27 to Stephanus of Byzantium,28 including the fragments of Agatharchides),29 lexica30 and encyclopaedias (from the second century CE Harpocration31 to the Byzantine Suda),32 anonymous gnomologies — namely the Gnomologium Vaticanum [Vat. gr. 744],33 and the Florilegium Baroccianum,34 in the version contained in the codex Monac. gr. 429.35 This tremendous variety of sources reveals a great deal about the circulation and the transmission of these texts and provides information about orators over a period of as many as sixteen centuries — if we are to count the extremely late witness of the Ps.-Eudocia, a forgery from the sixteenth century, an anthology (Violarium) falsely written under the name of the empress Aelia Eudocia Augusta.36  24 Rutilius Lupus was a Roman rhetorician whose activity can be placed during the age of Tiberius. His treatise De figuris sententiarum et elocutionis is a translation, adaptation and reduction of a Greek equivalent written by Gorgias of Athens (first century BCE, tutor of Cicero the younger). Among the others he quotes 11 Hellenistic orators. About Rutilius, see Barabino (1967), Brooks (1970), and Ruhnken (1768). 25 Mostly testimonia, but also a few fragments: Ant. 24 for Hybreas, Prae. ger. reip. 802e for Pytheas, and also Public. 9.11 for Anaximenes, Cons. ad Apoll. 104a for Demetrius, and Alex. 3.6 for Hegesias, for which the genre is not unequivocally oratory. 26 Exclusively testimonia. 27 Among the many testimonia (in some cases the only testimonium), one fragment: 14.2.24 for Hybreas. 28 S.v. Κορόπη for Demetrius (although the genre is not absolutely clear). 29 A fragment of Stratocles apud Phot. Bibl. cod. 250, 447a. 30 Interesting is the case of a fragment of Charisius in Phot. Lex. α 2091 Theodoridis from an otherwise non-attested speech Against Socrates. 31 Essentially testimonia except for Harp. ι 25 about Demochares but whose nature is not clear. 32 Among the various testimonia, there is a fragment of Democleides (ω 263, a quotation of a quotation by Timaeus of Tauromenion), see below. 33 253–260 for Demetrius and 307–308 for Hegesias. 34 One testimonium for Xenocles of Adramittium (second/first century BCE) in Florilegium Monacense (Monac. gr. 429), 220. 35 Worthy of a mention are also Athenaeus (main source for several orators), Eusebius of Caesarea, the so-called Alexander Romance by Ps.-Callisthenes, the Chronicum Lindium, and scholia to different authors. 36 Ps.-Eudocia Augusta, Violarium 712, where about Neanthes of Cyzicus we read he was a rhetor and pupil of Philiscus and that he wrote a treatise On Rhetorical Affectation and many panegyric speeches. This is a piece of information quite probably taken from the Suda (in the entry on Callinus), but here wrongly attributed to Neanthes.

Collecting Fragments for a Fragmentary Literary Genre  

How, then, should we proceed with the compilation of this list of names and testimonies? Part of this task was begun by Wooten in his doctoral thesis where he listed forty orators divided geographically (twelve from an Athenian mainland, twenty-four from Asia Minor, and four from Rhodes).37 On the one hand, this list is still extremely partial, but on the other, it includes rhetors such as the rather famous Apollodorus of Pergamon, for whom the activity of teacher is widely documented, but for whom no evidence of professional oratory exists. Some other names, at least for the earlier part of the Hellenistic age, can be found in David Ruhnken’s dated but still useful Historia Critica Oratorum Graecorum (1768), in which some later Attic or early Hellenistic orators are listed.38 An interesting case cited by Ruhnken is a certain Democleides,39 mentioned only in the lexicon Suda (ω 263 Adler) and Dionysius of Halicarnassus (Din. 11.73–78): in the former a fragment of this orator which appears to be from a speech against Demochares is quoted in the version given by Timaeus of Tauromenion, while the latter suggests the possibility of the existence of a speech in defence of Hermias of Atarneus attributed to Democleides40 or Menesaechmus, another almost unknown Hellenistic orator. Similarly, for the early part of the period, a scan of the fragments of the Attic Orators collected by Baiter and Sauppe (even the Adespota) is definitely useful.41 Furthermore, as noted above, the absence of an edition of Hellenistic orators is partly due to the overlap between the roles of historian and orator from the end of the fourth century BCE. This resulted in many Hellenistic orators being considered historians and ending up only in a collection of historical fragments. Therefore, a good starting point would be the systematic analysis of the main collections of fragmentary historians, with specific attention to the biographical  37 See Wooten (1972) 4–49, 143–145. 38 Apart from the names already mentioned above, we find a certain Coccus (a pupil of Isocrates quite difficult to date), Menesaechmus, Democleides (and/or Democles), Cephisodorus, Crates, Glaucippus, Hagnon (or Hagnonides), Callicrates, plus various disciples of Isocrates. 39 Sundwall (1906) 12 identified this orator with the homonymous archon of 316/315 BCE. Ruhnken (1768) xci–xcii does not exclude the possibility that Democleides is a corruption for the more widespread Athenian name Democles. An Attic orator called Democles is indeed witnessed by Ps.-Plut. X orat. 842e, where he is said to have been a pupil of Theophrastus, and was the object of a lost oration of Dinarchus (Antiatt. 113.7 Bk. = ρ 3 Valente: ῥαχίζει· ἀντὶ τοῦ μεγάλα ψεύδεται. Δείναρχος Κατὰ Δημοκλέους). The chronology would be consistent, as Dinarchus and Demochares were contemporaries. 40 See Ruhnken (1768) xci–xcii. 41 Additional yet indispensable tools are the works by Friedrich Blass: Die attische Beredsamkeit (1887–18982) and also Die griechische Beredsamkeit in dem Zeitraum von Alexander bis auf Augustus (1865), where a few of these orators find some space in the discussion.

  Roberta Berardi testimonia: for some of them some information about their activity as rhetors and orators is clearly given. This is the case not only with the above-mentioned Anaximenes, Theopompus, and other pupils of Isocrates (among whom Philiscus of Miletus has a fragment transmitted by direct tradition in a papyrus, see above n. 23), but also with other personalities such as Neanthes of Cyzicus,42 and Polyaenus of Sardis.43 It is worth looking, in addition, at all the names mentioned in the Greek and Latin rhetorical works. The main source is, as noted, the treatise of Rutilius Lupus, where eleven Hellenistic orators are identifiable, of which only eight are listed in Wooten’s thesis.44 The ninth is Demetrius of Phalerum, excluded by Wooten for no specific reason. The tenth is Pytheas, probably considered still ‘too Attic’ by Wooten, but at any rate in my opinion still worthy of being included in the edition, considering that all his fragments refer to orations that were surely pronounced after the Battle of Chaeronea. The eleventh is a certain Sosicrates,45 not mentioned elsewhere but dateable to the second century BCE, whose two fragments are particularly interesting for their Nachleben in Ciceronian oratory (see detailed analysis below). Among the others, fundamental works to be consulted are the rhetorical treatises of Cicero,46 and above all the whole production of Dionysius of Halicarnassus. It is in the latter’s works that we find the occasional mention of the names of orators active in the Hellenistic era, as is the case of the otherwise unknown Aristocles, Artamenes, and Sosigenes.47 As anticipated, this list needs to be complemented by a section of rhetorical papyri. In his contribution from 2013, Kremmydas listed twenty-one rhetorical papyri to be ascribed to the Hellenistic era. This list is a very inclusive one, and in my opinion not all of these pieces are worthy ‘fragments’ of Hellenistic orators. A selection leads me to exclude pieces of rhetorical treatises,48 later fragments  42 See above n. 8. 43 Peek/Stegemann (1952). 44 These are Charisius, Cleochares, Daphnis, Demochares, Hegesias, Isidorus, Myron, and Stratocles. 45 Wickert (1927). 46 For instance, Cicero (Brut. 316, 325) is the only witness for Aeschylus of Cnidus, defined an excellent example of Asianic style. 47 All three orators are mentioned at D.H. Din. 8. Only Aristocles has another witness (Strab. 14.2.13). 48 According to Kremmydas, P. Freib. 3. (inv. 9), PSI inv. 22012 (which, rather than rhetorical theory, seems actually a piece of a novel). While PSI inv. 22013, listed by Kremmydas as rhetorical techne (perhaps by mistake?) is definitely worth being edited as it is quite evidently an embedded speech in a historical work, or a rhetorical exercise. On the question, see Pintaudi/Canfora (2009) and Stramaglia (2015).

Collecting Fragments for a Fragmentary Literary Genre  

that cannot fairly be considered Hellenistic (some from the second century CE),49 and other fragments of non-Hellenistic content.50 Among others, there are some rather interesting pieces worthy of being considered fragments. In particular, P. Schub. 32 is both fascinating and understudied, and the possibility of its having belonged to a book containing professional orations cannot be ruled out completely. I will provide, in what follows, a new edition, translation, and some general observations on this piece. P. Schub. 32 = P. Berol. inv. 7445 (MP3 2507, LDAB 4265) Place and date Fayoum?, first century BCE/first century CE Editions Schubart (1950). Bibliography Merkelbach (1958) 126, no. 1153. Picture https://berlpap.smb.museum/00390/

. Diplomatic transcript ευδαιμονια[ τερουσαλλὰ[ ουςεἰληφοτας[ δικαιωσανα[ μισαμενοιτ[ ανδιατουτου[ τοισευεργ[ σωσαιτηνπ[ προθυμηθε[ λατουσεξαρ[ πρεσβευοντ[ καιτουτοισπ[ μενκαιπρ[ βαιοιπανδ[ βοηθηκασι[ τελευταιαμ[ πολινσυνοικ[ Λακεδαιμονι[  49 P. Schub. 33 (P. Berol. inv. 11492) + BKT 9.141 (P. Berol. inv. 21237), P. Hamb. 133 (inv. 16r), surely to be taken into account in a broader study on the evolution of rhetoric between Hellenism and Imperial age, but perhaps not perfectly fitting in an edition of Hellenistic material. 50 An alleged praise of Augustus, P. Bingen 3 (P. Med. 71.86f), and a funeral oration for M. Vipsanius Agrippa, P. Köln 1.10 (inv. 4701r) + 6.249 (inv. 4722r), probably a translation from an original Latin, for which the same considerations made about the examples in n. 55 below can be applied.

  Roberta Berardi κωλυκασιδο[ καὶμηνουδ[ λοισδυνατον[ γιαστυχεινθη[ ο̣ λιγωρησασι[ απαξουδεδισ [ πολλακισυπ[ ]ωνουτοιφαν[ ]πολεμησαν[ καιπολλαμεν[ ματαδαπαν[ τεσουκολιγ[ πολιτων [

. Critical text εὐδαιμονία[…] / τέρους ἀλλὰ[…τ-] / οὺς εἰληφότας[ καὶ] / δικαίως ἂν ἀ[ποκο-] / μισάμενοι τ[…] [5] / ἂν διὰ τούτου[…] / τοῖς εὐεργ[ετήμασι] / σῶσαι τὴν π[όλιν] / προθυμηθέ[ντες· ἀλ-] / λὰ τοὺς ἐξ Ἀρ[καδίας [10] / πρεσβεύοντ[ας]· / καὶ τούτοις π[…] / μὲν καὶ πρ[ῶτοι Θη-] / βαῖοι πανδ[ημεὶ βε-] / βοηθήκασι[ν. τὰ δὲ ] [15] / τελευταῖα Μ[εγαλό-] / πόλιν συνοικ[ίζειν] / Λακεδαιμόνι[οι κε-] / κωλύκασι δο[…] · / καὶ μὴν οὐδ[ὲ ἄλ-] [20] / λοις δυνατὸν[…] / γιας τυχεῖν Θη[βαίων] / ὀλιγωρήσασι[ν, οὐ γὰρ] / ἅπαξ οὐδὲ δὶς [ἀλλὰ] / πολλάκις ὑπ[ὲρ ἄλ- [25] / λ]ων οὗτοι φαν[ήσον- / ται] πολεμήσαν[τες] / καὶ πολλὰ μὲν [χρή] / ματα δαπαν[ήσαν-] / τες οὐκ ὀλίγ[ους δὲ τῶν] [30] / πολιτῶν [… … 5 τ[ὴν…] Schubart || 6 καὶ Schubart || 7 εὐεργ[ετήμασι] Schubart || 12 π[άλαι] Schubart, Merkelbach || 17 συνοικ[ίζειν] Berardi: συνοικ[ίσαι] Schubart: συνοικ[ίσαντες] Merkelbach || 19 δόλοις ci. Schubart || 20 οὐδ’ [ὧν παρ’ ἄλ-] / λοις Merkelbach || 21 fortasse εὐλο]γίας vel χορη]γίας: συνερ]γίας Schubart, Merkelbach || 22 Θη[βαῖοι] Merkelbach || 23 ὠλιγωρήκασι[ν Merkelbach || 31 ἀπωλέσαντες Schubart

. Translation Good fortune … but … those who have taken and rightly … carried back … therefore … to those eager to save the city with the services done … but what about the ambassadors from Arcadia? The Thebans first, at the time, came in mass to help them too and in the end the Spartans have prevented them from founding Megalopolis … and it is not possible for those others who esteem the Thebans little to obtain … not once or twice but often they will show they have fought for the others and to have spent a lot of money, not few of the citizens …

Collecting Fragments for a Fragmentary Literary Genre  

. Commentary This fragment appears to be a piece of an oration containing an encomium to the Thebans, and is, according to the editor princeps, W. Schubart, most likely a suasoria in defence of their courage. Location and date are rather uncertain, as it was most likely purchased, with the obvious consequence of a total loss of the archaeological context. Schubart dates it to the first century CE on palaeographical grounds (although he does not provide examples to support it), but this is a date that could be considered, seeing that since 1950 many other earlier fragments have been discovered and can constitute interesting parallels for the handwriting. The handwriting undoubtedly has some elements of the severe style (strenger Stil) with narrow ε, θ and σ, very small omicron, and pretty wide η, κ, μ, ν, and ω. Nevertheless, the curls at the upper ends of the υ and at the base of the υ and τ look earlier, and some parallels for this kind of handwriting can be found among the Herculaneum papyri: a similar large and stretched μ and ν are in P. Herc. 418 (third century BCE) and P. Herc. 330 (first century BCE), P. Herc. 154 (second century BCE) and P. Herc. 224 (first century BCE), of which the last two also present curled ends for υ. Moreover, significant similarities can be found in a much earlier papyrus, namely P. Hibeh. I. 15, dated to the first half of the third century BCE, another rhetorical piece containing an exhortation to be very likely attributed to a general (perhaps Leosthenes). Here we find a similarly narrow ε and σ (but not θ), a wide η, ν and κ (whose upper trait is quite high, as in our papyrus), but that υ is quite open (although without the base and the curled ends), ξ is written in two traits, and β has a similar small upper bow.51 In light of these observations, and in the absence of other elements, we can provisionally hypothesise a date for this papyrus between the first century BCE and the first century CE. The ideas evoked in the quoted speech are typical of panegyrical oratory: we hear about the eagerness of the Thebans to help other cities, and about their moral and material commitment not only to their own cause, but also to the interests of others. A famous archetype for an encomium of the Thebans is in Demosthenes’ oration De Corona (215), where the orator thanks the Thebans for three qualities: their courage, their loyalty and their wisdom, all virtues that they showed in 338 BCE when they welcomed an Athenian force into their city. If the reconstruction is correct, the episode used here as an example of Theban courage is their support for the Arcadians’ foundation of Megalopolis (Diod. Sic. 15.72.1–4),  51 Rather different are, on the contrary, α (more cursive in P. Schub. 32), ω (open in P. Hibeh. I. 15) and φ (sharper in P. Hibeh. I. 15).

  Roberta Berardi a city established, on the encouragement of the Theban general Epaminondas, between 371 and 368 BCE as a centre of power to oppose the Spartans. The text we have is not entirely clear as there is no historical source saying that the Spartans ever tried to prevent this foundation as the fragment seems to state — if the incomplete word Λακεδαιμονι[…] is indeed to be intended as a nominative. If this piece is really a rhetorical exercise, it is then possible that this unrealistic scenario was part of the theme of a suasoria. If Λακεδαιμονι[…] is, on the contrary, to be intended as an accusative with the Thebans as subject, we would need to think of a different meaning for συνοικίζω, maybe ‘join’ (in a non-pacific way) rather than ‘colonise’. This may be a reference to the Battle of Megalopolis in 331 BCE, when the Spartans were defeated by the Macedonians to whom Megalopolis was allied. Since Thebes was destroyed by order of Alexander in 335 BCE, and consequently could not play any role in the matters related to the Battle of Megalopolis, the first option is to be preferred. Examples of suasoria on Thebes are widely documented.52 If we exclude those dealing with the destruction of the city in 335 BCE, there are still many examples. Worthy of mention is one by Philostratus (VS 2.15.2), where Ptolemeus of Naucratis is said to have chosen a theme too disconnected from historical reality, namely ‘The Thebans accuse the Messenians of ingratitude because they did not welcome their refugees when Thebes was destroyed by Alexander’. Other remarkable examples of how the theme of Thebes remained of interest throughout the centuries are Aelius Aristides’ fictitious speeches to the Thebans (Or. 9–10), his Leuctran orations (11–15), and the suasoria about Leuctra at Cic. inv. 2.69–70. Particularly interesting is a parallel in Libanius’ Declamation 24,53 where we read about an embassy to the Spartans dealing with the question of the Messenians. However, in none of them is the episode of Megalopolis’ foundation mentioned. Another example of Thebes’ presence in declamations is also P. Oxy. LXXV 5025, a piece of oratorical prose dated to between the second and the third century CE, which we could imagine being delivered to the Amphictyonic Council or being included in a historical work (see P. Oxy. LXXV p. 18). Here a city is mentioned which the editor, R. Hatzilambrou, has conjectured may well be Thebes, due to a series of

 52 Hermog. Inv. 169.18–170.17, Philostr. VS 2.595–596, Rhet. Gr. 4.249–250 Walz, Sopater Rhet. Gr. 8.239.1–244.9 Walz, Rhet. Gr. 7.510–511 Walz, Apsines Rhet. Gr. 1.231 Spengel–Hammer. 53 νόμος ἐν Λακεδαίμονι τὸν εἴσω τριάκοντα ἐτῶν μὴ δημηγορεῖν. Θηβαῖοι τὴν ἐν Λεύκτροις μάχην νικήσαντες πρέσβεις πρὸς Λακεδαιμονίους πεπόμφασι πόλεμον ἀπειλοῦντες, ἢν μὴ Μεσήνην ἀφῶσιν αὐτόνομον. λεγόντων ἀφεῖναί τινων Ἀρχίδαμος νέος ὢν ἔτι παραινεῖ τὸν πόλεμον δέξασθαι. πέπεικε. τροπὴ γέγονε τῶν δυσμενῶν καὶ γράφεταί τις παρανόμων Ἀρχίδαμον.

Collecting Fragments for a Fragmentary Literary Genre  

possible references to the myth of Oedipus. If so, this would constitute another example of Thebes and its role in historical declamations. In order to place the fragment under discussion in its correct historical and rhetorical framework, the material conditions of the papyrus itself can be helpful. The text is written on the recto (the parallel side to the fibres) of a papyrus roll whose verso is blank, and the column is rather narrow, with a very wide upper margin and the remains of a left margin. The handwriting is very accurate and firm, the letters are well-spaced and clearly legible, no corrections are present, and the end of each sentence is marked by a paragraphos placed under the first letter of the line containing the last word of the clause (i.e. προθυμηθέ[ντες, βοηθήκασι[ν, καὶ μὴν, φαν[ήσον- / ται]). All these elements most definitely allow us to think of a refined, finely-produced book, rather than an extemporaneous exercise in a school manual (on the subject see Cribiore 1996, 97–118), a hypothesis which is consistent with the possibility of a longer circulation for this text. A parallel can be seen in P. Mil. Vogl. 3.123, a collection of progymnasmata of sophistic encomia of mythological characters: the handwriting (according to the editors, dateable to the third century BCE on palaeographical grounds) is rather accurate, overall correct, and there are coronides and paragraphoi. Furthermore, papyrologists commonly accept as a general rule the fact that very narrow columns — as the one in this case — are on average typical of oratory (see Turner/ Parsons 1987 2, 7), although Johnson (2004) 153–154 has shown that narrow columns are also widely employed for history and philosophy. At any rate, the hypothesis that this constitutes a piece of a roll containing speeches from professional, active orators is still possible, and considering the material conditions of the fragment, it seems more plausible for it to be a piece of oratory or history, than a draft of no value. Moreover, at the top of the very wide upper margin, traces of ink are clearly visible, the handwriting is different from that of the text (the few visible traits look bigger), and it is very difficult to decipher what is written. At a first glance, one may think of a marginal note or a title, but the ink looks different and the fibres of the papyrus appear somehow erased about halfway through each letter, as if the written word were neatly cut horizontally. What is more, corresponding with the traces of the second letter, just a few millimetres below, one can see another small trace of ink, perhaps part of the erased letter. A possibility is that the completely blank papyrus chosen as writing support for the book containing this oration had some sort of previously written text in that point, and the scribe tried to erase it manually (although unsuccessfully), in order to leave space for the upper margin. This would confirm the refined nature of this book, meant to be a good-looking copy and not just a scribbled manual. At any rate, it is fair

  Roberta Berardi to assume the text circulated for a certain period before it was copied, which would make it Hellenistic. From a stylistic point of view, a few observations can be made: (οὐκ) ἅπαξ οὐδὲ δὶς is never attested before the fourth century BCE, but is increasingly common from the second century CE onwards. It is remarkable to see the use of three forms of the perfect tense (εἰληφότας, βε-]βοηθήκασι[ν, κε-]κωλύκασι, καὶ μὴν οὐδ[ὲ, αι/πολεμήσαν[τες), a verbal characteristic that gradually disappears in Hellenistic prose, as explained in detail by Chantraine (1926) 214–252 [Le parfait à l’époque hellénistique]). For this reason, we would not expect to find it so prominent in a text dating to the first century BCE/CE. But at least two explanations are possible: (1) the composer of this rhetorical piece wanted to adopt a style as similar as possible to Attic; (2) again, as is the case with many papyrus fragments, this could easily be the copy of an earlier text, which might have circulated for a long time before it was copied. Schubart leaves the date of the text open to any possibility by placing the events related to the foundation of Megalopolis as terminus post quem. Here follows a lemmatic commentary: l. 1 εὐδαιμονία It is difficult to say what εὐδαιμονία referred to here, but it is worth noting the term is never linked to Thebes in the extant Greek literature. ll. 3–5 οὺς… ἀ[ποκο-] / μισάμενοι The positive connotation suggested by the adverb δικαίως may lead us to surmise that the subject is the Thebans, who are being praised. ll. 7–9 τοῖς… προθυμηθέ[ντες Again, the positive connotations of people who are ‘eager to save the city’, is most likely about the Thebans (see Dem. 18. 250). ll. 9–11 ἀλ-] / λὰ… πρεσβεύοντ[ας] What is not clear is the syntactical role of ἀλ-] / λὰ ] τούς ἐξ Ἀρ[καδίας / πρεσβεύοντ[ας]. If it were an accusative of respect, the construction would be particularly odd. Schubart’s hypothesis (p. 63) that it is a nominal clause, perhaps to be translated as a question ‘But what about the Arcadian ambassadors?’, is perhaps more sensible. There is no possibility of making πρεσβεύοντ[ας] agree with incomplete participle προθυμηθέ[…, because the paragraphos under the π of προθυμηθέ clearly marks the end of the sentence at that line, unless we consider that paragraphos misplaced. But we should rather adopt the cautiousness recommended by the lex Youtie: iuxta lacunam, ne mutaveris, especially considering the accuracy with which this text appears to be written and arranged on the column, and the exact correspondence of all the paragraphoi to actual strong pauses in the text.

Collecting Fragments for a Fragmentary Literary Genre  

ll. 12–15 καὶ τ… βοηθήκασι[ν The episode in which the Athenians helped the Arcadians was most likely during the war against the Spartans (370–368 BCE), when Epaminondas and the Thebans supported the Arcadians in the foundation of Megalopolis (see above). ll. 15–19 τὰ δὲ… κωλύκασι δο[ As I explained in the general commentary on the nature of this piece, the meaning of this sentence, in the light of our knowledge of Greek history, is unclear. Although the sentence works well syntactically, there is no attestation of the Spartans preventing the foundation of Megalopolis. See above for possible hypotheses to explain the incongruence, and for the reasons why the only possible supplement for Λακεδαιμονι[ is the nominative Λακεδαιμόνιοι. l. 17 συνοικ[ίζειν] I chose to restore the present infinitive, while Schubart restored the aorist συνοικ[ίσαι], which would imply the foundation of Megalopolis was a moment rather than a process. Merkelbach proposed a participle συνοικ[ίσαντες], which would make the sentence more meaningful if we also restored a dative for Λακεδαιμονι[…] (‘The Thebans, having founded Megalopolis, prevented the Spartans from …’), but unfortunately there is hardly enough space to restore seven letters, three of which (alpha, ny and tau) are quite large. ll. 19–20 οὐδ[ὲ ἄλ-] / λοις Merkelbach restored οὐδ[ὧν παρ’ ἄλ-] / λοις, again not acceptable due to the lack of space. Therefore, the simpler solution οὐδ[ὲ ἄλ-] / λοις is to be preferred. ll. 20–23 καὶ… ὀλιγωρήσασι[ν Although the object of the sentence is missing, the sense is easily understandable. The author blames those who unreasonably esteem the Thebans little, by saying that it will be difficult for them to obtain something, presumably a thing with a positive connotation. A possibility could be ‘cooperation’, if we consider Schubart’s supplement συνερ]γίας, also accepted by Merkelbach: it is indeed consistent with the idea of mutual assistance as main topic of the oration. However, considering the space occupied by the lacuna (four or five letters, depending on their size), at least two other sensible options can be taken into account: (1) ‘abundance/assistance/fortune’, if we restored χορη]γίας; (2) ‘plausibility/praise’, if we restored εὐλο]γίας. In all three cases the meaning restored is consistent with the sense that the sentence is supposed to have, as they all have positive connotations. Also, for all of them we find parallels of the construction with τυγχάνω: they are quite late for συνεργία (the earliest are Joseph. AJ 6.219.3, and Cassius Iatrosophista Quaestiones et Problemata 10.9, then the expression becomes very popular in Late Antique and Byzantine Christian literature) and εὐλογία (the earliest is Philo Quaestiones in Genesim 4 fr. 227.2, and consistently all the attestations are Christian, with the meaning of ‘blessing’). Much more interesting are the attestations of χορηγία with τυγχάνω: taken aside

  Roberta Berardi the earliest in Dem. Ep. 2.12.3, where χορηγία has its concrete original meaning of ‘equipping of choruses’, worthy of a mention are two attestations in Aristotle (Pol. 1.1255a14, about the way virtue can lead to obtaining resources, and in fr. 3, l. 2 of the Protreptic, where it is referred to estimable people), and the one in Speusippus’ Letter to Philip 12.12, referred to Theopompus and the assistance that Philip provided to him. With the meaning of ‘assistance’, χορηγία, not too differently from συνεργία, seems also to fit quite well the atmosphere of the oration and its focus on the ability of the Thebans to assist the others—but not those who esteem them little. Moreover, considering the financial nature of the original χορεγίαι, the presence of the word χρήματα at ll. 29–30 (see commentary and parallels below) seems to be particularly consistent with this context. Therefore, in the light of these considerations and of the relevant parallels, χορηγία seems to be preferable. ll. 22–23 Θη[βαίων] / ὀλιγωρήσασι[ν Merkelbach prefers ὠλιγωρήκασι[ν. This supplement is made perhaps on the basis of the presence of other perfects, and consequently he restores Θη[βαῖοι] as subject. However, although the first letter of the verb is uncertain, there is no space for both η and κ (whose shape is similar) after ρ, and σ is quite evident and not at all doubtful (therefore I do not mark it as littera incerta), although slightly erased. Moreover, if Merkelbach is correct here, the syntax of the sentence would be overly complicated. ll. 23–25 οὐ γὰρ… πολλάκις See P. Yale 2.105 (a rhetorical exercise on the theme of the Arginusae) col. 1, l. 19, for rhetorical commonplaces of the sort. The two papyri are probably to be dated back to the same century. ll. 24–31 The last legible sentence of the papyrus, consistent with the rest of the oration, focuses on the altruism of the Thebans, and underlines in an almost redundant way how much and how often they were ready to offer their help to other cities. ll. 27–28 χρήματα This point about the financial nature of the help provided by the Thebans corroborates the hypothesis of χορηγίας as supplement at ll. 21– 22, as the original meaning of the term is the liturgy (financial assistance) given from a chosen wealthy citizen to the city for the creation of a tragic or lyric chorus. Furthermore, interestingly, χορηγία and χρήματα appear associated twice in Demosthenes: Dem. 8.70.7, and Dem. Ep. 2.12.3, where, as seen above, the construction for χορηγία is the same as in this oration. ll. 29–30 οὐκ ὀλίγ[ους δὲ τῶν] / πολιτῶν It is safe to restore the correlation of μὲν (l. 28) and δὲ (in lacuna at l. 31), both because of space involved and for grammatical reasons. At this point, one might expect the speaker to praise Thebes not only as a state, but also for its citizens and the active participation of many of them (οὐκ ὀλίγ[ους) in the public matters and causes.

Collecting Fragments for a Fragmentary Literary Genre  

 Concentric Circles: a suggested model Going back to my selection of fragments, I have explained above why this is a list that includes all names, even those with only testimonies and no fragments, and those with paraphrased speeches. Consequently, in order to work on such a variegated selection of material, a precise method should be identified, so that it will be easier to proceed gradually with the edition and subsequently with the commentary. The paradigm I have found most effective is the one I call ʻConcentric Circlesʼ. This means we should start editing and commenting on the fragments where we can identify an exact quotation of an actual speech. The number of examples for this category is surprisingly high. For the innermost ‘Circle’, we have the twenty-four fragments transmitted in Latin translation by Rutilius,54 for Charisius, Cleochares,55 Daphnis,56 Demetrius of Phalerum, Demochares, Hegesias, Isidorus of Pergamon,57 Myron, Pytheas, Sosicrates, and Sratocles.58 For Charisius we also have one fragment contained in Photius’ lexicon,59 for Cleochares we have two more fragments in Greek, one

 54 During my presentation of a shorter version of this chapter at the 10th Celtic Conference in Montreal on July 10th 2017, a very interesting debate arose as to whether Latin translations should be considered actual quotations and proper fragments, despite what may have been lost in translation. The question was posed as to whether Rutilius is a particularly complicated case, given the likelihood that the rhetorician made his translations, as often happens, simply by recalling the original text by heart? As for the general question, I would safely consider translations to be real quotations, for the same principle according to which, for instance, we consider as actual Galen the Arabic translations of some of his works. In the specific case of Rutilius, this criterion is even more valid, as in his treatise there are quotations from extant speeches of Attic orators that can be verified, and the verification confirms that the translation is extremely faithful, rarely altering the original, and not at all made out of a text ‘simply remembered by heart’. Only occasionally the text is a paraphrase, and this only happens when Rutilius is creating a link between his observations and the actual quotation. Finally, some minor variants may depend on the different edition Rutilius was using and on the influence the Ciceronian theory of translation may have had on Rutilius’ method. It is not absolutely clear, at any rate, if the examples quoted by Rutilius were already present in Gorgias’ treatise, or were added later by the Latin rhetorician who surely had his own editions of the texts (quite unlikely is, on the other hand, the possibility that he was using Latin translations circulating in Rome). On these problems, see Barabino (1967) 75–133. 55 Aulitzky (1921). 56 Brzoska (1901). 57 Kroll (1916). 58 Fiehn (1945). 59 See above n. 30.

  Roberta Berardi transmitted by Photius,60 the other by Herodian.61 Demochares has two more fragments transmitted by Athenaeus,62 and for Hegesias there may be a further fragment transmitted by Plutarch.63 Pytheas has four more fragments, contained in Plutarch,64 Dionysius of Halicarnassus65 and Athenaeus.66 Finally, Stratocles has a fragment quoted by Agatharchides which is contained in Photius’ Bibliotheca.67 To these orators we can add Aeschines of Miletus, Damas, Dionysius of Pergamon or Atticus, Gorgias of Athens, Lesbocles, Nicetes, Plution, and Potamon, whose fragments are in the Elder Seneca’s Declamations,68 as well as Hybreas of Mylasa,69 who has eight fragments in the Elder Seneca (some in Greek, others in Latin),70 two from Strabo,71 and one from Plutarch.72 Then we have Democleides, with only one fragment in the Suda and perhaps a second one in Dionysius of Halicarnassus,73 Philiscus of Miletus, with one papyrus fragment,74 the late Attic orators Philinus and Polyeuctus of Sphettus, the panegyrist Matris, the abovementioned Apollonius Molon and his homonym Apollonius ὁ μαλακός, and finally the slightly more complex instance of Theopompus for whom, as well as for Demetrius of Phalerum, quotations from orations and historical works can be easily confused. For Theopompus, moreover, there is the case of some letters from exile, which are rhetorical products that appear to straddle the fence between oratory and private communication.75 The second ‘Concentric Circle’, a larger one, will be the one including paraphrased fragments,76 a typology of fragments that, ruling out any sort of stylistic

 60 Phot. Bibl. cod. 176, 121b. 61 Herodian. Rhet. Gr. 3.97 Spengel (= 40 Hajdú). 62 Ath. 5.187d and 5.215c. 63 Plut. Alex. 3.6 (history of Alexander or epideictic discourse?). 64 Plut. Dem. 27.5 and Prae. ger. reip. 802e–f. 65 D.H. Is. 4.4. 66 Ath. 2.44e–f. 67 Cod. 250, 447a. 68 See biographies in Bornecque (1902) 141–201. 69 Radermacher (1914). 70 Sen. contr. 1.2.23, 1.4.11, 2.5.20, 7.4.10, 9.1.12, 9.1.15, 9.6.16, suas. 4.5. 71 Strab. 14.2.24. 72 Plut. Ant. 24.7. 73 See above p. 85. 74 See above n. 23. 75 The case is perfectly explained and argued in Ottone (2005). 76 This is for instance the case of most of Theopompus’ fragments, and one fragment by Demochares.

Collecting Fragments for a Fragmentary Literary Genre  

considerations, requires a different approach for their commentary. Then, radiating out from the centre, a third ‘Concentric Circle’ could contain unclear fragments — for example, the entries about Demetrius of Phalerum in the Gnomologium Vaticanum, his sayings quoted by D.L. 5.82, that could at the same time be taken from speeches or just from anecdotes about his political activity but also life,77 and another unclear sentence attributed to him by Stob. Flor. 3.12.18. Also to be included here are some of Hegesias’ fragments quoted by Agatharchides De mari Erythraeo 5.21 which cannot definitively be attributed to history rather than oratory. Such a study may lead us to exclude from the selection most of these sententiae, due to their biographical rather than oratorical nature, but to include at least fragments that ancient sources attribute doubtfully to different orators: this is the case, for instance, of two fragments by Philinus, quoted by Harpocration (ε 100) and Athenaeus (10.425b), and all fragments by Menesaechmus (D.H. Din. 11.14–15; 73–78). A less detailed commentary is probably to be preferred for this type of fragment. Another issue requiring further investigation is the role of long speeches in historical works78 like those of Polybius,79 Diodorus of Sicily, and Livy,80 for which new complex questions arise. Are they worthy of being considered fragments or are they too long? Is there too much of the author’s reworking of the original text? Perhaps a final ‘Concentric Circle’ could be constituted by this kind of material, but it will be essential to consider these speeches case by case.

 A specimen case from Rutilius In order to illustrate some of the problems arising from the study of the fragments transmitted by Rutilius (and belonging therefore to the first ‘Concentric Circle’), I shall now provide the text, translation and commentary of one of the main cases

 77 For this third phase, the editor has to be very cautious and follow the fragment with a question mark. 78 Among others see Walbank (1965). 79 On Polybius’ speeches, see Wooten (1974). Moreover, an interesting reflection on the new genre of the ambassador’s speeches in the Hellenistic era is in Wooten (1973). At any rate, even if we accept to include Polybius’ speeches in our edition, it would be appropriate to leave those of the ambassadors aside, given their different role. 80 The long speech of Astymedes transmitted by Livy (15.22–25) is considered by Wooten as an actual fragment. There may be soon after hints to the fact that Livy was reading some sort of written text, but there is not enough evidence to support this idea.

  Roberta Berardi of indirect tradition (in translation), in which we are faced with an actual quotation whose nature is unclear. This is the case of Sosicrates, an orator attested only by Rutilius Lupus. There is only one useful element to date his life and work, which is the question of whether the ‘Philip’ mentioned in the fragment edited and analysed below is the one who was king of Macedonia between 221 and 179 BCE. If this were the case, we would be able to place Sosicrates’ life and work in the second century BCE. I should mention, though, that there is also a homonymous historian who lived in the same period and was born in Rhodes.81 If an identification of the two characters is possible, the Rhodian origin of the historian would be consistent with Sosicrates’ rhetorical abilities. Also, this would not contradict the possibility of a direct involvement of the orator in some of the events regarding Philip V of Macedon, since Philip (with the Aetolian league and other forces) waged war against Rhodes (and subsequently other allies) in the so-called Cretan War (205–200 BCE), at the end of which Rhodes acquired part of the territory of Crete. This hypothesis would also be supported by the fact that Sosicrates the historian is said to have written a History of Crete.82 But, according to the fragment below, the speaker, although blaming Philip, seems to have been, at least formerly, on the side of the Macedonian king, invalidating the previous argument. Nevertheless, this contradiction could be resolved if we imagine that this speech was embedded in a historical work (the History of Crete perhaps?), and that Sosicrates was simply quoting or paraphrasing the words of another orator.83 Or Sosicrates may be blaming Philip for causing a war that could have been avoided, the nos of the quotation referring not just to the Rhodians, but to all parties involved in the conflict. Given the lack of context for this fragment, this broad interpretation of the first person plural cannot be ruled out completely. Here follows the text and the commentary of what will be Sosicrates’ F1 in my edition: Sosicrates F1. Rutilius. Schemata Lexeos. 1.8 Sosicratis: ‘Non enim alius quis est, cuius opera in has difficultates inciderimus, sed initio ad bellum suscipiendum nos primum impulit Philippus, deinde in ipso belli labore ac periculo deseruit nos Philippus, novissime nunc calamitati nostrae proinde atque culpae succensuit idem Philippus.’

 81 Laqueur (1927). 82 Strab. 10.4.2, Ath. 6.263f. 83 Considering that we do not know much about Sosicrates’ historical work and its style, an inclusive approach is to be preferred in this dubious case.

Collecting Fragments for a Fragmentary Literary Genre  

inciderimus codd. Ald. : inciderim Basil. | initio codd. Ald. Basil. : invitos Ruhnken | nos B Basil. : vos A R V Ald. | nunc codd. Ald. Basil. : huic Ruhnken | proinde atque Basil. : atque codd. Ald. | succensuit codd. Ald. : successit Basil. Sosicrates: ‘We cannot have fallen into these difficulties because of anyone else but Philip, but the one who in the beginning first pushed us to start the war was Philip, then in the middle of that war’s fatigue and peril the one who deserted us was Philip, and finally now the person who is inflaming our disgrace and blame is again Philip.’

This fragment is quoted by Rutilius when he introduces the figure of ἐπιφορά, the repetition of a word at the end of a sentence, which is not very common in classical Attic oratory (see Barabino 1967, 26, nn. 89–92). The ἐπιφορά in this fragment is rhetorically particularly effective since it is constituted of the name of Philip. This was most probably a public oration pronounced against Philip V of Macedon, although the occasion for this is difficult to determine. As suggested, if we identify Sosicrates the orator with Sosicrates the historian, this could be part of an embedded speech in the History of Crete, pronounced by an understandably irate Cretan orator, who had been deluded by Philip — after all, at the end of the war, Crete lost part of its territory. Blass (1865) 35 believed this quotation was part of a historical work, but he did not establish any connection with Sosicrates the historian. On the other hand, although possible, it seems unlikely that this quotation belonged to a suasoria, as Philip V and events related to him do not seem to have been a very popular topic for declamations. We have only one example of a Roman suasoria where he is mentioned, which is at Cic. inv. 1.17: ‘Whether an army should be sent to Macedon to help allies against Philip, or retained in Italy to maximise forces against Hannibal’. In the absence of further evidence, the hypothesis that this came from a historical work seems plausible.84 A striking parallel for this passage can be found at Cic. Phil. 2.55, where the ἐπιφορά of the word Antonius — who is, like Philip, the one to blame — resembles the same figure in this fragment. This comparison would be very interesting because it could attest to the survival of Sosicrates’ orations in Cicero’s time (or at least to the survival of their traces), and that would be one of the proofs of the many non-Demosthenic influences on Cicero’s oratory. Thus, this fragment perfectly works as an example of how much certain almost forgotten sources can tell us about the survival — if not the influence — of Hellenistic oratory, especially in Roman oratory.

 84 utrum exercitus in Macedoniam contra Philippum mittatur, qui sociis sit auxilio, an teneatur in Italia, ut quam maximae contra Hannibalem copiae sint.

  Roberta Berardi

 Conclusions The aim of my research is to establish a critical text and to provide a first commentary on the extant fragments of Hellenistic orators. In this chapter I have discussed two relevant examples: I have investigated Sosicrates as emblematic of the issues posed by fragments transmitted by indirect sources, and specifically by the Latin translation of Rutilius, and I have also looked at P. Schub. 32, as a specimen case for understudied rhetorical papyri. From the first results of my research, it is clear that each fragment of this work-in-progress, if studied carefully and considered from every angle, could contribute to a more detailed overview of oratory in this period. We have seen how a methodology of ‘Concentric Circles’ can allow us to work with order and rigour on such an inclusive number of fragments, placed in a broad chronology. Certainly, a lot of work still needs to be done: further necessary steps will include a careful analysis of the prose rhythm in the longer Greek fragments and perhaps in some of the papyri, and I shall also undertake a re-examination of the epigraphical material. Furthermore, at the same time, the editor should never lose the connection with the historical context that allowed this kind of production to flourish; in particular, it is indispensable to make an in-depth study of the evolution of political institutions in the various areas of the Greek-speaking world, and to try to place each fragment in its most probable social contest. My hope is that all these aspects I have looked at in this chapter will contribute not only to the commentary on fragments on which I am currently working, but also to further studies on Hellenistic oratory.

References Aulitzky, K. (1921), ‘Kleochares (5)’, RE 11.1, cols 672–673. Baiter, J.G./Sauppe, H. (1893–1895), Oratores Attici, Zurich. Barabino, G. (1967), P. Rutilii Lupi Schemata dianoeas et lexeos. Saggio introduttivo, testo e traduzione, Genoa. Bayliss, A.J. (2011), After Demosthenes: The Politics of Early Hellenistic Athens, London. Blass, F. (1865), Die griechische Beredsamkeit in dem Zeitraum von Alexander bis auf Augustus, Berlin. Blass, F. (1868–1880, 1887–18982), Die attische Beredsamkeit, 4 vols, Leipzig. Bornecque, H. (1902), Les déclamations et les déclamateurs d’après Sénèque le père, Lille. Brooks, P. (1970), Rutilii Lupi De figuris sententiarum et elocutionis. Edited with Prolegomena and Commentary, Leiden. Brzoska, J. (1894), ‘Anaximenes (3)’, RE 1, cols 2086–2098. Brzoska, J. (1895), ‘Apollonios (85)’, RE 2, cols 141–144. Brzoska, J. (1901), ‘Daphnis (6)’, RE 4.2, col. 2146.

Collecting Fragments for a Fragmentary Literary Genre  

Chantraine, P. (1926), Histoire du parfait grec, Paris. Chávez Reino, A.L. (2009), ‘Teopompo orador: primeros apuntes’, in F. Gazzano/G. Ottone/ L. Santi Amantini (eds), Ingenia asiatica. Fortuna e tradizione di storici d’Asia minore. Atti della prima giornata di studio sulla storiografia greca frammentaria, Genova, 31 maggio 2007, Tivoli, 89–122. Cribiore, R. (1996), Writing, Teachers and Students in Greco-Roman Egypt, Atlanta. Fiehn, C. (1945), ‘Stratokles’, RE 4A.1, cols 269–271. Flower, M.A. (1994), Theopompus of Chios: History and Rhetoric in the Fourth Century BC, Oxford. Gärtner, H. (1963), ‘Pytheas (3)’, RE 24.1, cols 366–369. Habicht, C. (1995), Athen in hellenistischer Zeit, Munich. Heckel, W. (2006), Who’s Who in the Age of Alexander the Great, Malden, Ma. Johnson, W.A. (2004), Bookrolls and Scribes in Oxyrhynchus, Toronto. Kennedy, J. (1994), A New History of Classical Rhetoric, Princeton. Kremmydas, C. (2013), ‘Hellenistic Oratory and the Evidence of Rhetorical Exercises’, in C. Kremmydas/K. Tempest (eds), Hellenistic Oratory: Continuity and Change, Oxford. Kremmydas, C./Tempest, K. (eds) (2013), Hellenistic Oratory: Continuity and Change, Oxford. Kroll, W. (1916), ‘Isidoros (18)’, RE 9.2, col. 2064. Laqueur, R. (1927), ‘Sosikrates (3)’, RE 3A.1, cols 1160–1165. Laqueur, R. (1935), ‘Neanthes’, RE 16.2, cols 2108–2110. Marasco, G. (1984), Democare di Leuconoe: politica e cultura in Atene fra IV e III sec. a.C., Florence. Merkelbach, R. (1948), ‘Literarische Texte unter Ausschluß der christlichen’, APF 16, 82–129. Nitsche, W. (1906), Demosthenes und Anaximenes, Berlin. Norden, E. (1898), Die antike Kunstprosa vom VI. Jahrhundert v. Chr. bis in die Zeit der Renaissance, vol. 1, Leipzig. Ottone, G. (2005), ‘Alessandro, Teopompo e le ‘ἐπιστολαί πρὸς τοὺς Χίους’ ovvero autorità macedone e strumenti di interazione con la comunità poleica fra pubblico e privato’, in L.S. Amantini (ed.), Dalle parole ai fatti: relazioni interstatali e comunicazione politica nel mondo antico, Rome. Peek, W./Stegemann, W. (1952), ‘Polyainos (7)’, RE 21.2, cols 1431–1432. Pintaudi, R./Canfora, L. (2010), ‘PSI Laur. inv. 22013: retorica o romanzo?’, in G. Bastianini (ed.), I papiri del romanzo antico. Atti del Convegno internazionale di studi (Firenze, 11–12 giugno 2009), Florence, 81–93. Porter, S. (1997), Handbook of Classical Rhetoric in the Hellenistic Period, Leiden/New York/ Cologne. Pownall, F.S. (2005), ‘The Rhetoric of Theopompus’, CEA 42, 255–278. Radermacher, L. (1912), ‘Hegesias (13)’, RE 7.2, cols 2607–2608. Radermacher, L. (1914), ‘Hybreas (1)’, RE 9.1, cols 29–31. Ruhnken, D. (1768), P. Rutilii Lupi De figuris sententiarum et elocutionis libri duo, Leiden. Schubart, W. (1950), Griechische Literarische Papyri, Berlin. Stramaglia, A. (2010), ‘Come si insegnava a declamare? Riflessioni sulle ‘routines’ scolastiche nell’insegnamento retorico antico’, in L. Del Corso/O. Pecere (eds), Libri di scuola e pratiche didattiche. Dall’Antichità al Rinascimento. Atti del Convegno Internazionale di Studi, Cassino, 7–10 maggio 2008, Cassino, 111–151.

  Roberta Berardi Stramaglia, A. (2015), ‘Temi sommersi e trasmissione dei testi nella declamazione antica (con un regesto di papiri declamatori)’, in L. Del Corso/F. De Vivo/A. Stramaglia (eds), Nel segno del testo. Edizioni, materiali e studi per Oronzo Pecere, Florence, 147–178. Sundwall, J. (1906), Epigraphische Beiträge zur sozial-politischen Geschichte Athens im Zeitalter des Demosthenes, Leipzig. Susemihl, F. (1891), Geschichte der griechischen Litteratur in der Alexandrinerzeit, vol. 1, Leipzig. Susemihl, F. (1892), Geschichte der griechischen Litteratur in der Alexandrinerzeit, vol. 2, Leipzig. Turner, H./Parsons, P. (1971, 19872), Greek Manuscripts of the Ancient World, London. Walbank, F.W. (1965), Speeches in Greek Historians, Oxford. Weissenberger, M. (2006), ‘Molon [2]’, BNP 9, 130. Wickert, I. (1927), ‘Sosikrates (5)’, RE 3A.1, cols 1165–1166. Wooten, C.W. (1972), A Rhetorical and Historical Study of Hellenistic Oratory, diss. University of North Carolina at Chapel Hill. Wooten, C.W. (1973), ‘The Ambassador’s Speech. A Particularly Hellenistic Genre of Oratory’, QJS 69, 209–212. Wooten, C.W. (1974), ‘The Speeches in Polybius. An Insight into the Nature of Hellenistic Oratory’, AJPh 95, 235–251.

Francesco Ginelli

The New Nepos Prolegomena Toward a Renumbering of Cornelius Neposʼ Fragments Cornelius Nepos is commonly known for his biographical works. The Liber de excellentibus ducibus exterarum gentium (Book about excellent generals of foreign nations), consisting of twenty-two biographies of non-Roman generals, is the first collection of uitae still entirely preserved, not only in Latin literature, but in the entire classical canon. It was part of a larger De uiris illustribus (On famous men), now mostly lost. Only a few other uitae still exist, such as the short portrait of Cato the Censor, from a lost Liber de Latinis historicis (Book on Roman Historians), and that of Titus Pomponius Atticus, Neposʼ longest biography, partially written when its subject was still alive.1 Lastly, a few passages (mostly preserved by later authors) provide us with useful information about the structure and the contents of Neposʼ biographies. Two of these are particularly significant. The first is relayed by Flavius Sosipater Charisius, a Latin grammarian of the fourth century CE, who suggests that De uiris illustribus was arranged in at least sixteen books.2

 1 Cf. Nep. Att. 19.1 Haec hactenus Attico uiuo edita a nobis sunt. This is a disputed passage. Most scholars, beginning with Wissowa (1900) 1413‒1414, interpreted it as evidence of a second edition of Neposʼ uitae published after Atticusʼ death, but Rahn (1957) and, more recently, Toher (2002) and Holzberg (2013) denied this hypothesis, mainly on the assumption that edita … sunt should be translated as ʻwere writtenʼ or, at least, ʻwere divulgedʼ, not already ʻpublishedʼ. 2 Char. gramm. I p. 141.14 Keil = p. 178.20 Barwick = 33 Halm = V.5 Peter = 44 Malcovati = 43 Marshall = F13 Wirth = [fragment lacking] Briscoe/Drummond. Geiger (1985) 87‒88, however, supposed the existence of at least eighteen books. He assumed that the De uiris illustribus was arranged in pairs of books on Greek and Roman subjects and that the libri on the Greeks always preceded those on the Romans. Based on that, Geiger read fr. 55 Malcovati = 56 Marshall as evidence that books thirteen and fourteen were, respectively, the books on Greek and Roman historians. Moreover, Geiger asserted that Nepos in Dion 3.2 refers to the Book on Greek Historians. So he inferred that Liber de ducibus exterarum gentium was, at least, book fifteen. Therefore, the Liber de ducibus Romanorum should be the sixteenth. However Charisius in gramm. p. 179.5 Barwick = fr. 43 Malcovati = 42 Marshall argued about the particular use of partum for partium in book fifteen of the De uiris illustribus. Geiger, underlining the absence of this word in the Liber on foreign generals, maintained that Charisius was quoting from a different book. Therefore, the books on generals should be relegated to positions seventeen and eighteen. However, Geigerʼs hypothesis does not consider the (probably substantial) lacuna in Lys. 2.3, or even the one conjectured after Dion 5.6, where Nepos could have used the genitive partum recorded by Charisius. https://doi.org/10.1515/9783110712223-006

  Francesco Ginelli The second text, preserved at the end of many manuscripts and unanimously ascribed to the above-mentioned Book on Roman Historians, conveys part of a letter attributed to Cornelia, mother of the Gracchi.3 However, according to the indirect tradition, Nepos wrote not only biographies, but also erudite literary works, now largely lost, such as the Chronica, a chronological study which compared famous historical Greek and Roman events, and the Exempla, a collection of anecdotes on various topics. A small poetic corpus is also attested.4 Despite the significant number of critical editions of Nepos that appeared during the twentieth century, little attention has been paid to his fragmentary texts.5 Furthermore, the collections of his fragments, which are all preserved in the indirect tradition, have not always followed clear and common strategies. Moreover, the allocations of single texts proposed in the last century are sometimes controversial and problematic, since editors seem to have overestimated the potential information which could be deduced from the contents of several fragments. For this reason, the aim of this chapter is to propose some preliminary notes toward a new numbering of Cornelius Neposʼ fragments, based on an in-depth analysis of the literary sources still available. My preliminary discussion in what follows will be concerned both with the arrangements of Neposʼ fragments up until now, and with the methodological criteria I am proposing.

 Editions of Neposʼ fragments: preliminary and methodological remarks In 1946 E. Malcovati published a critical edition of Neposʼ texts in the series of the Corpus Paravianum,6 also including a collection of deperditorum librorum reliquiae (‘remains of lost books’). It was something new among Nepos studies. The Oxford edition by E.O. Winstedt, which appeared in 1904,7 comprised the Book on

 3 Fr. 28 Halm = V.15 Peter = 58 Malcovati = 59 Marshall = F64 Wirth = [fragment lacking] Briscoe/ Drummond. The text is also collected in the Epistolographi Latini Minores as fr. I.1.cxxiv.3‒4 Cugusi. It has been studied by Horsfall (1989) 41‒42, 125‒126. 4 Cf. Plin. epist. 5.3.6 = fr. 61 Malcovati = 63 Marshall. 5 On the Chronica, cf. the works collected by Spies (2017), entries: 956‒960, 971, 973‒974, 1115, 1160, 1400, 1615. On the Exempla and Neposʼ scientific-geographic interests, cf. entries: 1094, 1132‒1133, 1261, 1351, 1595, 1688, 1759, 1770, 1772. 6 Malcovati (19643). 7 Winstedt (1904).

The New Nepos  

Foreign Generals and the biographies of Cato and Atticus. The Belles Lettres edition by A.-M. Guillemin, published in 1923,8 only contained the surviving parts of the letter ascribed to Cornelia and three other fragments, two from the Book on Roman Historians, and one from Neposʼ letters to Cicero.9 Guilleminʼs odd selection was based on the fact that all these fragments were written in the direct style and, more significantly, they were the longest fragments. Short direct quotations, such as single words or very short clauses, were surprisingly not included.10 The criterion adopted by Malcovati was extremely different. Malcovati included a wide range of literary material in a single list of the remains of lost books, collecting not only fragments in the direct style, but also quotations in the indirect style and passages with information about the structure and the content of Neposʼ lost works. By so doing, Malcovati followed the criteria adopted both by Peter in the second volume of the Historicorum Romanorum reliquiae,11 and, prior to him, by Halm in his critical edition of Neposʼ works of 1871.12 The only differences between the three collections of Halm, Peter, and Malcovati, lay in the organisation or attribution of the narrative material and the number of the collected passages. On the basis of a few fragments conveying geographical information, Halm argued for the existence of a geographical treatise, although this was denied by Peter, Malcovati, and Marshall (quoted below). Excluding Peter, who was obviously not interested in Neposʼ poetic production, Halm listed 59 fragments and Malcovati 63. Marshall adopted Malcovati’s strategy thirty years later in his Teubner edition of Neposʼ works,13 and the number of the listed fragments increased from 63 to 65. The later edition by Wirth, which listed 64 texts with a German translation,14 is quite similar to these last two collections. The only innovation is a list of 6 passages concerning Neposʼ life.  8 Guillemin (1923). 9 Respectively nr. 3 and 4 Guillemin = Anon. Cod. Guelf. Gud. Lat. 278 (thirteenth century) = fr. 26 Halm = V.17 Peter = 57 Malcovati = 58 Marshall = F17 Wirth = [fragment lacking] Briscoe/Drummond, nr. 5 Guillemin = Lact. inst. 3.15.10 = Fr. 46 Halm = IV.1 Peter = 40 Malcovati = 39 Marshall = 9 Wirth = T16 Briscoe/Drummond. 10 Cf. Guillemin (1923) 169: “Les passages suivants ont été trouvés dans les manuscrits, à la suite di Traité des historiens latins, soit chez des auteurs qui les ont cités. Je n’ai pas recueilli une foule d’autres débris donnés par quelques éditeurs pour avoir été trouvés épars, sous forme de citations bien souvent rédigées en style indirect, chez divers écrivains: Pline l’Ancien, AuluGelle, Priscien, Charisius, Suétone, Lactance, Ausone, saint Jérome, etc. … On ne peut attribuer à C. N. des lambeaux de phrases sur la forme desquelles nous ne sommes pas exactement fixes”. 11 Peter (1967). 12 Halm (1871). 13 Marshall (19852). 14 Wirth (1994).

  Francesco Ginelli Something genuinely new appeared in 2013, in the Oxford Fragments of the Roman Historians (henceforth FRHist). Here, Briscoe and Drummond, following criteria common to their entire collection, divided the texts related to Nepos15 into two different categories. In the first, called Testimonia, they arranged information not only about Neposʼ life, but also concerning the general contents of his lost works. In the second group, entitled Fragments, Briscoe and Drummond put together only those texts which seemed to be both direct and indirect quotations from Neposʼ lost works.16 This is a criterion clearly explained in the introduction to the whole collection, cf. Cornell/Rich (2013) 14: “A text that purports to quote or paraphrase a particular passage of a lost original is a fragment, whereas a text that gives information about the author or about all or part of his work, but without reference to a particular passage is a testimonium”. For example, Text F3, which Briscoe and Drummond place in the category ʻUnassigned fragments, but probably from the Chronicaʼ, is an excerpt from the Collectanea rerum memorabilium by Gaius Julius Solinus, a Latin scholar and geographer who probably flourished in the middle of the third century BCE: Solin. 1.27 (fr. 5 Halm = I, 3 Peter = 5 Malcovati = 5 Marshall = F24 Wirth = F3 Briscoe/Drummond) Cincio Romam duodecima olympiade placet conditam: Pictori octaua: Nepoti et Lutatio opiniones Eratosthenis et Apollodori comprobantibus olympiadis septimae anno secundo. (Ed. Mommsen 18952) Cincius holds that Rome was founded in the twelfth Olympiad, Pictor in the eight, and Nepos and Lutatius, in conformity with the views of Eratosthenes and Apollodorus, place it in the second year of the seventh Olympiad. (Transl. Briscoe/Drummond 2013b)

Here Solinus is not directly quoting Neposʼ text, but summarises the original content. Even if the context and the references are extremely vague, we can surely assert that Nepos, somewhere in one of his lost works, probably in the Chronica (as suggested in comparison with the other quoted authors, such as the annalists Cincius Alimentus and Fabius Pictor, or the historian and poet Lutatius Catulus), maintained that Rome was founded in the second year of the seventh Olympiad (i.e. 751 BCE, contrary to the more famous Varronian date of 753 BCE).17 Despite

 15 Drummond and Briscoe excluded the texts related to the Exempla, De uiris illustribus, and a letter to Cicero. 16 This is an arrangement commonly used in collections of prose fragments. 17 Bernardi Perini (1992) 1272‒1273 n. 2.

The New Nepos  

the fact that it is a re-elaboration of a previous source, this text provides important information about the content of a lost work written by Nepos. From this example, it is clear that summaries or quotations in the indirect style are extremely helpful in deducing the general contents of fragmentary works. However, at the same time, we should be extremely cautious about relying on them for evidence, precisely because of their nature as ʻre-elaboratedʼ texts. If direct quotations, especially those based on memory, sometimes convey inaccurate texts, indirect quotations or short summaries could even alter or partially modify the original evidence of a text. This risk was clearly underlined, among others, by Grilli in a study on the general criteria for an edition of philosophical fragments. Grilli specified the problems and the doubts that an editor must deal with in editing fragmentary texts: It is a duty of the editor of philosophical fragments to inform his readers about the possibilities of distinguishing what is absolutely guaranteed and authentic from what could be a dilution of the original thought or a not-entirely-faithful paraphrase or a description that has been distorted by a polemical debate.18

These considerations, which concern philosophical fragments, can also be helpful in preparing an edition of any other kind of prose fragments.19 Grilli recognised the importance of differentiating the original words of a specific author from those texts transmitted by the indirect tradition or which convey only information on the contents of lost works. Similar issues were underlined also by Glenn W. Most in the well-known preface to the volume Collecting Fragments. Most summarised the differences between poetic fragments and prose fragments, and, among these, between philosophical fragments and historical fragments: Poetic fragments are usually cited because of their exact wording, and may be supposed to be transmitted fairly honestly (though of course their transmission is usually far more problematic than is that of complete works). Philosophical fragments are only rarely cited because of their exact wording, more usually because of their doctrine or argument; often they are cited by opponents, who

 18 Grilli (1981) 113: “È dovere dell’editore di frammenti filosofici ragguagliare il suo lettore delle possibilità di distinguere ciò che è assolutamente garantito e autentico da ciò che può rappresentare o una diluizione del pensiero originale o una parafrasi non sempre e del tutto fedele o una rappresentazione deformata attraverso la polemica”. The English translation is mine. 19 This is an issue that obviously concerns only those fragments that are specifically preserved in the indirect tradition. On a bibliographical update concerning the differences between the direct and the indirect traditions, and their peculiarities, cf. Curnis (2011) 71‒75, esp. n. 1.

  Francesco Ginelli disagree with the views expressed and may not be inclined to quote them accurately. Historical fragments are usually cited because of the historical or geographical information they contain, and are perhaps least likely of all to respect the exact wording of the original. When philosophical fragments are treated as though they were poetic fragments (e.g., in some research on the Presocratics) various kinds of confusion are likely to ensue.20 As concerns fragments conveyed in the indirect tradition, like those of Nepos, the main problem consists in telling a testimonium apart from a fragmentum, and, moreover, in distinguishing a quotation in the indirect style from a summary or a re-elaboration of a lost text, a problem which obviously does not affect collections of texts transmitted in the direct tradition. The lines which divide these categories are, unfortunately, extremely slight, sometimes even evanescent. From a strictly philological and etymological point of view, a section called Fragmenta should contain only those passages which can be clearly considered the original words of a lost author or work (directly transmitted fragments as well as direct quotations).21 On the contrary, under the heading Testimonia should be listed all those texts which convey information about the life of the author and the general features of his lost works, as well as those passages transmitted in the indirect manner (such as indirect quotations, summaries, re-elaborations), since it is not always discernible how closely the intermediary source has kept to the original version of the quoted text. However, herein lies the most insidious difficulty for the editor, since the distinction between a direct or indirect quotation, a summary and a re-elaboration is not always clear and indisputable when the original text is lacking. In such a context, an indirect quotation seems to be more accurate and closer to the original text than a general reference, and an editor should indicate these different levels of accuracy to his readers. One could argue that, in order to avoid such complications, a collection of fragments should be as close as possible to a ‘collection of textual references’, where all the texts concerning a specific author (general references to his life or literary works, direct and indirect quotations, summaries, allusions) are merely assembled in a single list, chronologically or alphabetically arranged as an index: it is up to the reader to use this collection in the way he finds most useful for his

 20 Most (1997) vii. Cf. the same issues discussed also in Most (2010). 21 On the ʻviolenceʼ that necessarily produces fragments, as suggested by the same Latin word fragmentum (as also the corresponding Greek words σπάσματα/ἀποσπάσματα and κλάσματα/ ἀποκλάσματα), cf. Most (2010) 371.

The New Nepos  

aims and analyses.22 After all, an edition of fragments is, at its core, a collection of texts and passages which can already be found in other works or sources. Its aim is to collect scattered texts into a single group in order to simplify consultation and analysis, and also to provide an overall idea of a work or of the literary production of a lost author.23 However, such a hypothetical edition would diminish the role of the editor, and it would also lack a ‘principle of arrangement’, which is the fundamental aim of the scholar who tries to sketch an overall portrait of an author by collecting literary fragments. On the other hand, a possible solution to this debate concerning the best allocation of testimonia and fragmenta in a hypothetical edition could consist in increasing the number of categories into which the fragments are divided, so as to be as accurate as possible in arranging the surviving textual material. The section Fragmenta could list only direct quotations, even if extremely irrelevant (as in the case of Nepos or prose authors whose works are mainly transmitted in the indirect manner), whereas the Testimonia could collect a great variety of texts, arranged into further sub-sets, including evidence about the author’s life, political or literary activity, as well as texts which convey information on the contents of lost works. This last category could therefore collect all those texts which are not direct quotations, such as summaries or indirect quotations. Such a methodology could be so systematic that it even reproduces the same text in multiple and different sets, if that text fits several categories. However, such an edition would eventually face the same problem of obscurity and lack of practicality as its opposite, the simple collection of sources mentioned above. From these premises, it is evident that editing prose texts that are mainly conveyed through the indirect tradition and in the reports of subsequent authors entails specific issues and features, which would differ from author to author (and from one literary genre to another). Many arrangements are, in fact, possible and equally plausible, and the duty of an editor is to select and justify the one which best fits the specific narrative material he or she is collecting.24 As concerns the preparations of an edition of Neposʼ fragmentary oeuvre, it will consequently be important to draw a plausible distinction between Testimonia and Fragmenta, giving also a clear account (where possible) of the ‘degree of reliability’ of the intermediary sources. Moreover, in doing so, the editor should also pay attention to two other important principles, namely ease of use and clarity. Bearing this in

 22 Balbo (20072) xv. 23 Most (1997) vii. 24 Cf. the bibliography collected by Fortenbaugh/Huby/Sharples/Gutas (1993) 4‒5.

  Francesco Ginelli mind, an edition of Neposʼ fragmentary texts should maintain a cautious but balanced approach. All those texts concerning the author’s life and general literary activity (but not specifically ascribable to specific works) should be collected in the Testimonia, whereas texts conveying information on the contents of lost but specific works, both directly and indirectly, should be arranged in the Fragmenta, as suggested by the editors of FRHist. In this last section, direct quotations should be highlighted and distinguished from the contextual material by using typographic markers — such as italics or underlining — in order to distinguish texts considered the real words of an author. This strategy would indicate clearly the different degrees of closeness to the original text and would also improve the intelligibility of short or obscure texts transmitted in the direct style. Even if such a distinction, one among others,25 does not strictly respect the philological difference between a fragmentum and a testimonium, it helps in setting out a clear and easy-to-analyse arrangement of texts through allocating (in the same set) those texts which convey ʻpartsʼ of the same work (both in the direct and the indirect manner). Moreover, the section entitled Testimonia will help the reader reconstruct the life and reception of Nepos. Lastly, a final and general remark will concern the order of the Testimonia and Fragmenta inside these two categories. Generally speaking, the arrangement of the texts should be in accordance with the issues and aims which characterise the study of specific genres and authors. For example, a thematic ordering of the Testimonia could help direct the reader towards those topics in which he is most

 25 Note the interesting distinction proposed by Balbo (20072) xvi in the collection of the fragments of Roman orators who lived during the Augustan and Tiberian era: “1) Frammento (F) è, in questa edizione, ogni testo che, con indicazione nominativa dell’autore o fornendo elementi tali da rendere evidente l’attribuzione, riporta non solo quello che possiamo presumere sia stato il testo originale dell’orazione, ma anche la sua eventuale parafrasi fatta dalla fonte, o — ancora — dà notizia dell’esistenza e dell’argomento dell’orazione stessa. Il testo originale (nei pochi luoghi in cui è stato possibile distinguerlo dalla parafrasi con sufficiente approssimazione) è stato stampato con il carattere spazieggiato. 2) Testimonianza (T) è, in questa edizione, ogni testo che riferisce elementi relativi all’attività letteraria (ed oratoria in particolare) dell’autore trattato, senza entrare nel merito del contenuto delle orazioni”. ‘1) A Fragment (F) is, in this edition, any text which, by referring to the author by name or by providing pieces of information which make the attribution evident, transmit not only that which we can presume is the original text of the speech, but also its own likely paraphrasing of the source, or that which — again — informs us of the existence and the argument of the speech itself. The original text (in the few places where it is possible to distinguish it from paraphrasing with sufficient approximation) has been printed with widely-spaced characters. 2) A Testimony (T) is, in this edition, any text which divulges pieces of information relating to the literary activity (and to the speeches in particular) of the author under discussion, without going into the merits of the subject matter of the speeches’.

The New Nepos  

interested, such as the private life of an author or his cultural activity.26 As concerns Nepos, a ‘conservative policy’, such as the one proposed by of FRHist, seems most reasonable. Testimonia should be arranged following the chronological order of citation, from the oldest to the most recent source, also giving the readers an idea of how the reputation or reception of a certain author has changed and developed over the centuries. Fragmenta will be conversely organised according to the order that they probably had in the original work. However, when this information is not available or easily inferable from the context, fragments will be arranged after the first group and according to the proven or most likely chronological order of publication.27

 Allocating fragments to Neposʼ lost works: some case studies Having determined the strategy we should adopt in collecting Neposʼ fragments, we have another issue to face. Sometimes editors have proposed ‘controversial’ attributions, and many fragments have been allocated according to contents or based on superficial affinities between texts. I will investigate a few fragments as case studies, in each case first examining the assignment proposed by Neposʼ editors, then suggesting a new allocation. The first case study is a passage from Plinyʼs Naturalis Historia concerning the tribe of the Caturiges: Plin. nat. 3.125 (fr. 53 Halm = II.9 Peter = 19 Malcovati = 18 Marshall = 32 Wirth = F10 Briscoe/ Drummond) Interiere et Caturiges, Insobrum exsules, et Spina supra dicta, item Melpum opulentia praecipuum, quod ab Insubribus et Bois et Senonibus deletum eo die, quo Camillus Veios ceperit, Nepos Cornelius tradidit. (Ed. Ian/Mayhoff 1906) There have also perished the Caturiges, Insubriam exiles, and the aforementioned Spina, and also Melpum, a town exceptional for its wealth, which Cornelius Nepos reports was destroyed by the Insubres, Boii, and Senones on the day on which Camillus captured Veii. (Transl. Briscoe/Drummond 2013b)

 26 It is the arrangement proposed by Fortenbaugh/Huby/Sharples/Gutas (1993) in their edition of the fragments of Theophrastus of Eresos. 27 Cornell/Rich (2013) 16‒18.

  Francesco Ginelli Peter, Malcovati, Marshall, and Wirth placed this text among the fragments from the Exempla, probably on the basis of its strongly episodic nature. Halm underlined the geographical content and proposed to read it as evidence of a geographical treatise, although this has never been attested in the indirect tradition. Briscoe and Drummond, more persuasively, placed the same text among the ‘Unassigned fragments, probably from the Chronica’, as suggested by the synchronism or chronological coincidence between the destruction of Melpum and Camillusʼ capture of Veii.28 The similarities with a passage from Gellius, clearly ascribed to the Chronica, are, in fact, very strong: Gell. 17.21.3 (fr. 4 Halm = 2 Peter = 4 Malcovati = 4 Marshall = 2 Wirth = F1 Briscoe/Drummond) … de Homero et Hesiodo inter omnes fere scriptores constitit aetatem eos egisse uel isdem fere temporibus uel Homerum aliquanto antiquiorem, utrumque tamen ante Romam conditam uixisse, Siluiis Albae regnantibus annis post bellum Troianum, ut Cassius in primo annalium de Homero atque Hesiodo scriptum reliquit, plus centum atque sexaginta, ante Romam autem conditam, ut Cornelius Nepos in primo chronicorum de Homero dixit, annis circiter centum et sexaginta. (Ed. Marshall 1968b) … for, as regards Homer or Hesiod, it is agreed by almost all writers, either that they lived at approximately the same period, or that Homer was somewhat the earlier; yet that both lived before the founding of Rome, when the Silvii were ruling in Alba, more than a hundred and sixty years after the Trojan war, as Cassius has written about Homer and Hesiod in the first book of his Annals, but about a hundred and sixty years before the founding of Rome, as Cornelius Nepos says of Homer in the first book of the Chronica. (Transl. Rolfe 1952)

In this last text, Nepos proposes a synchronism or chronological comparison between the life of Homer and the foundation of Rome. However, as Briscoe and Drummond clearly observed,29 the lack of the exact source in Plinyʼs passage does not allow the editor to assign the fragment with any certainty.30 In fact, Pliny did not specify which work of Nepos he used. For this reason, I would argue that the excerpt from the Naturalis historia should be assigned neither to the Exempla (following Peter, Malcovati, Marshall, and Wirth) nor to a lost geographical treatise. Moreover, it is important to note that Nepos uses synchronisms not only in the Chronica, but also in his biographies. It is indeed a common practice he follows to determine the relative dating of specific events, as in the last clause of Aristidesʼ biography:  28 Briscoe/Drummond (2013) 502. 29 Briscoe/Drummond (2013) 497‒498. 30 Similar doubts are also expressed by Ranucci (1982) 453 n. 3.

The New Nepos  

Nep. Arist. 3.3 Decessit [scil. Aristides] autem fere post annum quartum quam Themistocles Athenis erat expulsus. (Ed. Marshall 19852) Aristides died about four years after Themistocles had been banished from Athens. (Transl. Rolfe 1984)

Nepos did not record the exact year (or Olympiad) when Aristides died, but he made a connection between Aristides’ death and the ostracism of Themistocles. A similar passage is attested in Cimonʼs biography: Nep. Cim. 3.2 Itaque post annum quintum, quam expulsus erat, in patriam [scil. Cimon] reuocatus est. (Ed. Marshall 19852) Therefore, Cimon was recalled to his native land only four years after his banishment. (Transl. Rolfe 1984)

Another interesting case study concerns a passage from the grammarian Priscian: Prisc. gramm. II p. 383.4‒6 Hertz/Keil (fr. 13 Halm = II.26 Peter = 36 Malcovati = 35 Marshall = 49 Wirth = F11 Briscoe/Drummond) Nepos: Aedis Martis est in circo Flaminio architectata ab Hermodoro Salaminio: ‘architectata’ passive posuit, ἀρχιτεκτονευθεῖσα. (Ed. Hertz/Keil 1855) Nepos: The temple of Mars in the Flaminian Circus was designed by Hermodorus from Salamis: He used ‘architectata’ as a passive, as the Greek ἀρχιτεκτονευθεῖσα. (Transl. Briscoe/ Drummond 2013)

Here Priscian directly quoted Neposʼ text. The grammarian was specifically interested in the passive value of the participle architectata, which recalls the Greek ἀρχιτεκτονευθεῖσα. However, Priscian did not specify in which of Neposʼ works he found this specific clause. Halm, Peter, Malcovati, Marshall, and Wirth assigned the fragment to the Exempla; Joseph Geiger, in his study on Nepos and the political biography, assigned it to the De uiris illustribus.31 Briscoe and Drummond placed it among the ‘Unassigned fragments, but possibly from the Chronica’. The divergent opinions of these editors show that it is impossible to assign this fragment with any certainty. The text could have been used to establish a synchro-

 31 Geiger (1985) 106‒107.

  Francesco Ginelli nism, or to describe a building in a biography (as Nepos did in Milt. 6.3 in describing the Stoa Poikile),32 or even to provide an example of a famous building. A similar text could therefore be located among the Fragmenta ex libris incertis. Another passage from Plinyʼs Naturalis historia illustrates the same problems: Plin. nat. 6.31 (fr. 15 Peter = 51 Halm = 25 Malcovati = 24 Marshall = F38 Wirth = [fragment lacking] Briscoe/Drummond) Aliqui inter Pontum et Caspium mare CCCLXXV milia passum non amplius interesse tradiderunt, Cornelius Nepos CCL milia. (Ed. Ian/Mayhoff 1906) Some maintained that there are no more than 375 miles between Pontus and the Caspian Sea, Cornelius Nepos 250. (My translation)

This reference is very vague. Pliny, as usual, did not specify where Nepos wrote about the distance between Pontus and the Caspian Sea. Peter, Malcovati, and Marshall ascribed this text to the Exempla, whereas Halm placed it among the fragments that could support the existence of a geographical treatise. However, these hypotheses overlook the fact that Nepos also provided information about spatial distances in his biographies, cf. Milt. 4.2 Inde ad Atticam accesserunt ac suas copias in campum Marathona deduxerunt. Is est ab oppido circiter milia passuum decem,33 or Hann. 6.3 Post id factum paucis diebus apud Zamam cum eodem conflixit; pulsus (incredibile dictu) biduo et duabus noctibus Hadrumetum peruenit, quod abest ab Zama circiter milia passuum trecenta.34 From the examples analysed above, one could argue that, if an editor had to follow only what a fragmentum or a testimonium clearly and explicitly reports, his role would be reduced to that of a mere collector. This surely is not the role of an editor. An editor of fragments should restore the most trustworthy reconstruction of a lost text, resting his conclusions on the surviving material. For this reason, every kind of supposition must be built on solid bases, not on hypotheses, no matter how appealing. A clear example of this methodology is an excerpt from

 32 Note that Nepos does not abstain from using technical terminology in his biographies, esp. in military (cf. Milt. 7.2 uinea and testudo), legal (Them. 1.2 exheredo), and architectural (Paus. 1.4 exsculpo) contexts. 33 “Then they kept on to Attica and led their forces into the plain of Marathon, which is distant about ten miles from Athens” (transl. Rolfe 1952). 34 “A few days after the conferences he fought with Scipio at Zama. Defeated — incredible to relate — he succeeded in a day and two nights in reaching Hadrumentum, distant from Zama about three hundred miles” (transl. Rolfe 1952).

The New Nepos  

Gellius, that Briscoe and Drummond placed among ‘Unassigned fragments, but probably from the Chronica’. All the other editors, on the contrary, ascribed this text with certainty to the Chronica: Gell. 17.21.8 (fr. 6 Halm = I.4 Peter = 7 Malcovati = 7 Marshall = 25 Wirth = F4 Briscoe/Drummond) Archilochum autem Nepos Cornelius tradit Tullo Hostilio Romae regnante iam tunc fuisse poematis clarum et nobilem. (Ed. Marshall 1968b) And Cornelius Nepos adds that when Tullus Hostilius was king at Rome Archilochus was already illustrious and famous for his poems. (Transl. Rolfe 1952)

Gellius did not clearly specify his source. However, a few lines above, at the beginning of the chapter, Gellius wrote a long methodological note: Gell. 17.21.1 Ut conspectum quendam aetatum antiquissimarum, item uirorum inlustrium qui in his aetatibus nati fuissent, haberemus, ne in sermonibus forte inconspectum aliquid super aetate atque uita clarorum hominum temere diceremus … temporum aetatumque erroribus caueremus, excerpebamus ex libris, qui chronici appellantur quibus temporibus floruissent Graeci simul atque Romani uiri, qui uel ingenio uel imperio nobiles insignesque post conditam Romam fuissent ante secundum bellum Carthaginiensium, easque nunc excerptiones nostras uariis diuersisque in locis factas cursim digessimus. Neque enim id nobis negotium fuit, ut acri atque subtili cura excellentium in utraque gente hominum συγχρονισμούς componeremus, sed ut noctes istae quadamtenus his quoque historiae flosculis leuiter iniectis aspergerentur. (Ed. Marshall 1968b) I wished to have a kind of survey of ancient times, and also of the famous men who were born in those days, lest I might in conversation chance to make some careless remark about the date and life of celebrated men … in order, I say, to guard against such errors in dates and periods of time, I made notes from the books known as Chronicles of the times when those Greeks and Romans flourished who were famous and conspicuous either for talent or for political power, between the founding of Rome and the second Punic war. And these excerpts of mine, made in various and sundry places, I have now put hastily together. For it was not my endeavour with keen and subtle care to compile a catalogue of the eminent men of both nations who lived at the same time, but merely to strew these Nights of mine lightly here and there with a few of these flowers of history. (Transl. Rolfe 1952)

Gellius’ purpose here is to write a chapter on the dates and periods of the lives of famous and illustrious men. In doing so, he claims to have consulted those historical and erudite works that usually are titled Chronica: qui chronici appellantur. This is partially true, at least for Nepos. In fact, after a few lines, Gellius quoted a passage from the first book of Neposʼ Chronica:

  Francesco Ginelli Gell. 17.21.3 (fr. 4 Halm = 2 Peter = 4 Malcovati = 4 Marshall = F2 Wirth = F1 Briscoe/Drummond) … de Homero et Hesiodo inter omnes fere scriptores constitit aetatem eos egisse uel isdem fere temporibus uel Homerum aliquanto antiquiorem, utrumque tamen ante Romam conditam uixisse, Siluiis Albae regnantibus annis post bellum Troianum, ut Cassius in primo annalium de Homero atque Hesiodo scriptum reliquit, plus centum atque sexaginta, ante Romam autem conditam, ut Cornelius Nepos in primo chronicorum de Homero dixit, annis circiter centum et sexaginta. (Ed. Marshall 1968b) … for, as regards Homer or Hesiod, it is agreed by almost all writers, either that they lived at approximately the same period, or that Homer was somewhat the earlier; yet that both lived before the founding of Rome, when the Silvii were ruling in Alba, more than a hundred and sixty years after the Trojan war, as Cassius has written about Homer and Hesiod in the first book of his Annals, but about a hundred and sixty years before the founding of Rome, as Cornelius Nepos says of Homer in the first book of the Chronica. (Transl. Rolfe 1952)

The closeness of Gell. 17.21.3 to 17.21.8, the methodological introduction, and the comparisons between Greek and Roman chronologies (which characterise the whole chapter) ensure that the editor is able to assign, with a high degree of plausibility, the text to the Chronica, as well as the quotation from Gell. 17.21.23‒24: Gell. 17.21.23‒24 (fr. 7 Halm = I, 5 Peter = 8 Malcovati = 8 Marshall = F27 Wirth = F5 Briscoe/ Drummond) Neque multo postea Eudoxus astrologus in terra Graecia nobilitatus est Lacedaemoniique ab Atheniensibus apud Corinthum superati duce Phormione, et M. Manlius Romae, qui Gallos in obsidione Capitolii obrepentis per ardua depulerat, conuictus est consilium de regno occupando inisse, damnatusque capitis saxo Tarpeio, ut M. Varro ait, praeceps datus, ut Cornelius autem Nepos scriptum reliquit, uerberando necatus est. (Ed. Marshall 1968b) Not long after these events the astronomer Eudoxus was famed in the land of Greece, the Lacedaimonians were defeated by the Athenians at Corinth under the lead of Phormio, and at Rome Marcus Manlius, who during the siege of the Capitol had repulsed the Gauls as they were climbing up its steep cliffs, was convicted of having formed the design of making himself king. Marcus Varro says that he was condemned to death and hurled from the Tarpeian rock; but Cornelius Nepos has written that he was scourged to death. (Transl. Rolfe 1952)

However, in the absence of methodological passages or other cross-references, assignments based on contents cannot be persuasive. Let us see, for example, another chapter from Gellius: Gell. 4.18.1 (fr. [fragment lacking] Halm = [fragment lacking] Peter = 11 Malcovati = 11 Marshall = [fragment lacking] Wirth = [fragment lacking] Briscoe/Drummond)

The New Nepos  

Scipio Africanus antiquior quanta uirtutum gloria praestiterit et quam fuerit altus animi atque magnificus et qua sui conscientia subnixus, plurimis rebus, quae dixit quaeque fecit, declaratum est. Ex quibus sunt haec duo exempla eius fiduciae atque exuperantiae ingentis. (Marshall 1968a) How greatly the earlier Scipio Africanus excelled in the splendour of his merits, how lofty and noble of spirit he was, and to what an extent he was upheld by consciousness of his own rectitude, is evident from many of his words and acts. Among these are the following two instances of his extreme self-confidence and sense of superiority. (Transl. Rolfe 1927)

Two examples of Scipioʼs moral virtue follow. The first one is a defence speech against Marcus Naevius, a tribune of the plebs who accused Africanus of having accepted money from King Antiochus in order to promote a convenient peace. The second anecdote concerns another dispute between two tribunes of the plebs and Scipio on the destiny of the royal treasure brought to Rome after King Antiochusʼ defeat. Only two editors of Neposʼ fragments, Malcovati and Marshall, maintained that these two episodes should be ascribed to Neposʼ Exempla because of their ‘highly Nepotianʼ tone. However, Gellius did not explain which source he used in this chapter. Moreover, quite similar versions of these episodes are attested also in Livy (Liv. 37.55.10; 56.2) and Valerius Maximus (Val. Max. 3.7.1). It follows that Gellius could have found the two anecdotes concerning Scipio the Elder in the Ab Urbe condita or in the Factorum et dictorum memorabilium libri, or even in other collections of anecdotes used for rhetorical purposes. One last example of this kind of problematic attribution comes from Suetoniusʼ Life of Terence: Suet. uita Ter. 4 Wessner = p. 31.2 Reifferscheid (fr. 24 Halm = V, 12 Peter = 53 Malcovati = 54 Marshall = 56 Wirth = [fragment lacking] Briscoe/Drummond) Nepos auctore certo comperisse se ait C. Laelium quondam in Puteolano kalendis Martiis admonitum ab uxore, temperius ut discumbe, petisse ab ea ne interpellaretur seroque tandem ingressum triclinium dixisse non saepe in scribendo magis sibi successisse, deinde rogatum ut scripta illa proferret, pronuntiasse uersus qui sunt in Heautontimorumeno: “Satis pol proterue me Syri promissa huc induxerunt”. (Ed. Wessner 1902) Nepos says that he learned from a trustworthy source that once at his villa at Puteoli Gaius Laelius was urged by his wife to come to dinner at an earlier hour than common on the Kalends of March, but begged her not to interrupt him. When he at last entered the diningroom at a late hour, he said that he had seldom written more to his own satisfaction; and one being asked to read what he had written, he declaimed the lines of the Heautontimorumenos, beginning: “Impudently enough, by Heaven, has Syrus lured me here by promises”. (Transl. Rolfe 1914)

  Francesco Ginelli Suetonius records a bizarre anecdote preserved by Nepos on the controversial authorship of Terenceʼs comedies. The poet Laelius, blamed by his wife for being late for dinner, replied that he wrote lots of verses and he suddenly began to declaim a line (l. 723) from the Heautontimorumenos. The editors of Neposʼ fragments ascribed this excerpt to the De uiris illustribus. This passage was also used to speculate about the existence of a Life of Terence that could have been included in a hypothetical Book on Roman poets. Even if this hypothesis is highly attractive, it is not supported by solid evidence. The idea that Suetonius, as a biographer, used only biographical sources, is not convincing. In fact, the strongly anecdotal nature of the quoted text suggests that Nepos could have included it in the Exempla. So, in the absence of incontrovertible evidence, it is better to place this text among the Fragmenta ex libris incertiis.

 Final remarks The few examples discussed above have illustrated the problems that still afflict the arrangement of Neposʼ fragments. Problematic texts have been attributed to specific lost works without acknowledging that the allocation has been based more on optimistic probability than on concrete plausibility. As a consequence, some attributions, which are not supported by the evidence of the direct and the indirect tradition, compromise both the study and the interpretation of Neposʼ lost works. It is curious to note how this biographer, one of the most translated authors in Latin literature, has always received such scant attention. Even his most studied work, the De uiris illustribus, has been the object of optimistic parallelisms (specifically with Plutarchʼs Parallel Lives) instead of concrete analyses developed on the texts. I conclude by summarising the structure and arrangement that a future edition of Neposʼ fragments should follow: it should comprise two large sections, entitled Testimonia and Fragmenta respectively, the first containing information on Neposʼ biography and on his literary activity, the second collecting all those texts clearly identified as deriving from Neposʼ lost works. It should also distinguish, by means of typographical markers, direct from indirect quotations. Moreover, the Fragmenta should be divided into several sub-sections, each one devoted to a specific work of Nepos, for example, fragments from the Chronica, the Exempla, or lost biographic texts. Furthermore, every single attribution must be explained on a solid basis, so to avoid superficial attributions that could alter our interpretation of Neposʼ lost works. As Stem wrote in his monograph on Nepos: “Editing the fragments of Nepos poses particular challenges. Almost all of what

The New Nepos  

he wrote is lost, and an individual topic could have appeared in more than one work”.35 The accurate and careful approach which I have outlined above should therefore be applied to a proposed future collection of Neposʼ fragments.

References Balbo, A. (2004, 20072), I frammenti degli oratori romani dell’età augustea e tiberiana. Parte prima: Età augustea. Seconda edizione riveduta e corretta, Alessandria. Bernardi Perini, G. (1992), Aulo Gellio. Le notti attiche (Libri X‒XX), Turin. Briscoe, J./Drummond, A. (eds) (2013), ‘Cornelius Nepos. Texts and Translations’, in T.J. Cornell (ed.), The Fragments of the Roman Historians. Vol. II, Oxford, 798‒815. Cornell, T.J./Rich, J.E. (2013), ‘Introduction to FRHist’, in T.J. Cornell (ed.), The Fragments of the Roman Historians. In Three Volumes, Oxford, 3‒19. Curnis, M. (2011), ‘Plato Stobaeensis. Citazioni ed estratti platonici nellʼAnthologion’, in G. Reydams-Schils (ed.), Thinking Through Excerpts: Studies on Stobaeus, Turnhout, 71‒123. Fortenbaugh, W./Huby, P./Sharples, R./Gutas, D. (1993), Theophrastus of Eresus. Sources for His Life, Writings, Thought and Influence, 2 vols., Leiden/New York/Köln. FRHist = Cornell, T.J. (ed.) (2014), The Fragments of the Roman Historians. In Three Volumes, Oxford. Geiger, J. (1985), Cornelius Nepos and Ancient Political Biography, Stuttgart. Grilli, A. (1981), ‘Sui criteri per lʼedizione di frammenti filosofici’, in E. Flores (ed.), La critica testuale greco-latina, oggi. Metodi e Problemi. Atti del Convegno Internazionale (Napoli 29‒ 31 ottobre 1979), Naples, 111‒134. Guillemin, A.-M. (1923), Cornélius Népos. Oeuvres, Paris. Halm, C. (1871), C. Nepotis quae supersunt. Apparatu critico adiecto, Leipzig. Hertz, M./Keil, H. (1855), Grammatici Latini. Vol. II. Prisciani institutionum grammaticarum libri I‒XII ex recensione M. Hertzii, Leipzig. Holzberg, N. (2013), ‘Struttura encomiastica e riflessi di realtà tardo-repubblicana nella ‘Vita di Attico’ di Cornelio Nepote’, in G. Bernardi Perini/A. Cavarzere (eds), Orizzonti culturali di Cornelio Nepote dal Po a Roma. Atti del Convegno. Ostiglia, 27 aprile 2012 ‒ Mantova, 28 aprile 2012, Florence, 131‒145. Horsfall, N. (1989), Cornelius Nepos, A Selection Including the Lives of Cato and Atticus, Oxford. Ian, L./Mayhoff, C. (1906), C. Plini Secundi naturalis historiae libri XXXVII. Post Ludovici Iani obitum recognovit et scripturae discrepantia adiecta. Vol. I. Libri I‒VI, Leipzig. Malcovati, E. (1944, 19643), Cornelii Nepotis quae extant, Turin. Marshall, P.K. (1968a), A. Gellii noctes Atticae. Tomus I. Libri I‒X, Oxford. Marshall, P.K. (1968b), A. Gellii noctes Atticae. Tomus II. Libri XI‒XX, Oxford. Marshall, P.K. (1977, 19852), Cornelii Nepotis vitae cum fragmentis, Leipzig. Mommsen, Th. (1864, 18952), C. Iulii Solini collectanea rerum memorabilium, Berlin.

 35 Stem (2012) 86.

  Francesco Ginelli Most, G.W. (1997), ‘Some Reflections on Fragments’, in G.W. Most (ed.), Collecting Fragments. Fragmente sammeln, Göttingen, vi‒viii. Most, G.W. (2010), ‘Fragments’, in A. Grafton/G.W. Most/S. Settis (eds), The Classical Tradition, Cambridge (Mass.)/London, 371‒377. Peter, H. (1967), Historicorum Romanorum reliquiae. Volumen alterum. Ed. stereotypa editionis primae 1906 aucta conspectu librorum commentationum disputationum quae post annum 1900 conscripta sunt, Stuttgart. Rahn, H. (1957), ‘Die Atticus-Biographie und die Frage der zweiten Auflage der Biographiensammlung des Cornelius Nepos’, Hermes 85, 205‒215. Ranucci, G. (1982), ‘Plinio. Storia naturale. Libro III. Traduzione e note’, in G.B. Conte/A. Barchiesi/G. Ranucci (eds), Plinio. Storia naturale. Vol. I. Libri 1‒6, Turin. Rolfe, J.C. (1914), Lives of the Caesars. Vol. II. Claudius. Nero. Galba, Otho, and Vitellius. Vespasian, Titus, Domitian. Lives of Illustrious Men. Grammarians and Rhetoricians. Poets (Terence. Virgil. Horace. Tibullus. Persius. Lucan). Lives of Pliny the Elder and Passienus Crispus, Cambridge (Mass.). Rolfe, J.C. (1927), The Attic Nights of Aulus Gellius. Vol. I, Cambridge (Mass.). Rolfe, J.C. (1952), The Attic Nights of Aulus Gellius. Vol. III, Cambridge (Mass.). Rolfe, J.C. (1984), Cornelius Nepos. On Great Generals. On Historians, Cambridge (Mass.)/ London (first published [with Florus] 1929). Spies, Y. (2017), Kornelbibliographie. Die gesamte Literatur von und über Cornelius Nepos bis zum Ende des Jahres 2015, Heidelberg. Stem, R. (2012), The Political Biographies of Cornelius Nepos, Ann Arbor. Toher, M. (2002), ‘Nepos’ Second Edition’, Philologus 146.1, 139–149. Wessner, P. (1901), Aeli Donati quod fertur commentum Terenti. Accedunt Eugraphi commentum et scholia Bebina. Vol. I, Lepzig. Winstedt, E.O. (1904), Corneli Nepotis Vitae, Oxford. Wirth, G. (1994), Cornelius Nepos. Lateinisch-Deutsch, Amsterdam. Wissowa, G. (1900), ‘Cornelius Nepos (n. 275)’, RE 4.1, cols 1408‒1417.

Jarrett T. Welsh

The Fifth Glossary of Nonius Marcellus Open an edition of the De compendiosa doctrina of Nonius Marcellus, and pick a quotation.1 What next? Readers and editors interested less in Nonius for his own sake and more for the material that he preserves need to do a bit of sorting to determine how best to approach their chosen quotation. Perhaps the quotation is a bit of Lucretius, Plautus, Cicero, or Vergil that Nonius excerpted himself. Since most of those texts have survived independently, Nonius provides a useful and relatively early check on their direct tradition.2 Or perhaps it’s a bit of the satires of Lucilius or Varro, texts also known to have been excerpted by Nonius himself. In such a case, although Nonius is often our only witness to those words, their editor can avail herself of the information that has accumulated about Nonius’ working methods to guide her speculations.3 But if the fragment belonged originally to Titinius or Caecilius Statius or Laberius — and therefore came from a work of Republican literature or drama not excerpted by Nonius himself and not preserved independently — the situation is somewhat more hazardous, and the need for guidance in speculating is even more pressing, since for these quotations Nonius relies on works of earlier Imperial scholarship whose methods and whose priorities are, in important and diverse ways, different from his own. Those methods and priorities have left their  1 This chapter owes its existence to the courteousness extended to it in its various incarnations. Some preliminary remarks were presented at the 2012 meeting of the Classical Association of Canada in connection with the arguments that became Welsh (2012) and Welsh (2013). The panel organised by Francesco Ginelli and Francesco Lupi at the 2017 Celtic Conference in Classics at Montreal provided a perfect opportunity to take up again the ideas that had been patiently awaiting their turn while I sorted out other Nonian problems. I am grateful to all the participants at that conference for stimulating discussions on a range of fragmentary subjects. My argument has benefited likewise from discussion in connection with versions presented at the University of Colorado Boulder and the University of Toronto Mississauga. This project was supported by, and would have been impossible without, a grant from the Social Sciences and Humanities Research Council of Canada. All references to Nonius’ text are to the edition of Lindsay (1903), with Mercier’s pagination; unless otherwise specified, such references refer to the beginning of an entry, and not to the specific line at which a feature occurs. 2 See the studies listed at Barabino/Mazzacane (2000) 17–27, (2003) 150–151, adding now Butterfield (2013) 61–67. 3 The mechanics of Nonius’ methods have generally been well explained (see Lindsay 1901, White 1980, for revisions see Welsh forthcoming). The individual behind those methods has not received as much attention as he merits; Chahoud (2007) offers an excellent overview, but more remains to be said esp. about the thought and philology of Nonius. https://doi.org/10.1515/9783110712223-007

  Jarrett T. Welsh mark on the fragments that Nonius preserves, and failure to consider their impact can often lead to misguided inferences and speculations. A fragment of Pacuvius’ Periboea illustrates the potential pitfalls well. Among Nonius’ comments on case-usage are several remarks about the verb potior, which Republican Latinity deployed with either the accusative or ablative case; building on Lindsay’s foundations, we may assign these remarks to Nonius’ use of the so-called ‘first glossary’ (ʻGloss. iʼ).4 The fragment of the Periboea is quoted with its Nonian context (p. 501.22): Sic et POTIOR ILLAM REM. Pacuvius Periboea (289 Ribbeck3 = fr. 229 Schierl): regnum potior, coniugem macto inferis. One also says potior illam rem, with the accusative. Pacuvius in the Periboea: ‘I am in possession of (or I take possession of) the kingdom, I sacrifice the coniunx to the gods below’.

As Giovanni D’Anna noted fifty years ago, nearly everything that we would like to know about these five words of Pacuvius is unknown.5 The parts of speaker, addressee, and coniunx could be distributed to diverse characters in any number of ways. Since Nonius provides no guidance on the distribution of parts, students of Pacuvius have enjoyed their speculations: Oineus and Agrios are the obvious candidates to have spoken the line, while Periboea and even Oineus himself have been mooted as the threatened coniunx.6 There is, however, a better way to speculate about this fragment, even though I cannot offer a more conclusive distribution of parts, for some significant information about the history of this fragment — that is to say, its history after having been separated from a complete script of Pacuvius’ Periboea — has gone unremarked. Pacuvius’ words passed through the ‘first glossary’, and possibly other intermediaries, before they were set down by Nonius in this form. At some point on that path these words were handled by an individual who felt no great reverence for fragments — or, more precisely, felt that Pacuvius’ words could be modified to suit his own purposes. For that individual seems often to have altered the words he encountered in Republican drama to produce ‘standard’ examples of the sort one might encounter in an elementary work on grammar. Thus miseriam

 4 On this ‘glossary’, see Welsh (forthcoming), with arguments for the attribution of this material to ‘Gloss. I’. The following example elaborates on a problem that could be signalled only in passing there. 5 D’Anna (1967) 228: “Framm. oscurissimo: non si riesce a comprendere né chi parla, né chi sia il coniunx o la coniunx cui si allude”. 6 Schierl (2006) 465 gives a doxography.

The Fifth Glossary of Nonius Marcellus  

omnem ego capio, hic potitur gaudia (Ter. Ad. 876), a line spoken by Demea to describe the contrast between his own lifestyle and that of his brother Micio, is excerpted and rewritten as mille potior gaudia (“I am in possession of a thousand delights”, Non. p. 481.23), producing thereby the standard first-person form expected in Latin lists of idiomata casuum, and incidentally effacing entirely the differences that characterise Terence’s famous brothers.7 In similar fashion, Plautus’ maior potitus hostium est (Capt. 762), in which Hegio recounts the loss of his older son, is rewritten as ille alter potitus hostis est (‘The other brother has been captured by the enemy’, Non. p. 498.25). One change is meant to clarify, for readers without the script immediately in front of them, that Plautus’ maior refers to another brother (ille alter), but the other change (hostis) seems merely to replace the plural with the singular number, which overwhelmingly predominates in the Latin lists of idiomata. That habit of rewriting the words of the Republican dramatists in order to produce a simple and ‘standard’ example invites us to ask whether Nonius’ quotation regnum potior, coniugem macto inferis accurately replicates the syntax of Pacuvius’ own words.8 The next editor of Pacuvius should therefore regard the tendencies of Nonius’ source and the history of this fragment as a license to consider all the options for speaker, addressee, and coniunx when editing these words, whether or not those options can be made to accord precisely with the first-person verb forms potior and macto. In that process, perhaps the first item to consider afresh is a Paestan hydria plausibly connected with Euripides’ Oineus, a play thought also to lie behind Pacuvius’ Periboea.9 The vase, it seems, shows an aged Oineus, being persuaded, perhaps by Periboea and Diomedes, to take up the blade against Agrios, who sits bound upon an altar as a black Erinys rises from the earth below him.10 It is hard to imagine a scene better suited to the words coniugem macto inferis. This sketch of the problems of interpretation lurking in a fragment of Pacuvius, and of the neglected possible sources of information about it, is meant to  7 If the Adelphoe had been lost, speculations about the speaker of the ‘fragment’ mille potior gaudia would certainly have gone badly astray. 8 One could, of course, have raised such doubts even without supporting evidence from ‘Gloss. iʼ, although the secure iambo-trochaic rhythms of these five words seem to have forestalled such doubts. 9 See Taplin (2007) 198–199 for discussion and references to earlier bibliography on its connection with Euripides. On arguments connecting Pacuvius’ Periboea and Euripides’ Oineus, see Schierl (2006) 428–429. 10 The vase is London, British Museum, F155 (= 1772, 0320.37). Photographs are available on the British Museum’s website.

  Jarrett T. Welsh illustrate a general point about the problems attending Latin fragments in Nonius’ text. In his review of Friedrich Marx’s edition of the fragments of Lucilius’ satires, A.E. Housman took a pessimistic view of what could be achieved in editing book fragments: The truth is that the difficulties of the text of Lucilius are for the most part inexplicable and its corruptions for the most part irremediable. What more than anything else enables the critic and commentator of an ancient author to correct mistakes and to elucidate obscurities is their context; and a fragment has no context. An editor of Lucilius or Ennius or Nonius or the reliquiae scaenicae, unless he is grievously self-deluded, must know that the greater number of his corrections, and of his explanations also, are false.11

Although Housman was obviously right in stating that a fragment has nothing of its original literary or dramatic context, nevertheless the context of its survival can be no less significant for interpreters, and if we are prepared to pay careful attention to that context, we will often find something new to say about a fragment, or new information to provide a better guide to our speculations about it. In the case of fragments preserved by Nonius, that necessarily means ʻreverseengineeringʼ the De compendiosa doctrina, sorting the sources of individual fragments and understanding what guidance is offered by the fact that a fragment was taken from the ‘third glossary’ or ‘fifth glossary’, say, or by Nonius himself from a complete text. Each of the seven items of Imperial scholarship that Nonius used have left traces of their methods, interests, obsessions, and follies in the De compendiosa doctrina.12 Just as the methods and interests of Priscian differ from those of Gellius or Charisius or Festus, so too do the diverse and now-lost texts that Nonius used differ in their contents and purposes from one another and from Nonius’ own. Some of Nonius’ scholarly sources I have discussed elsewhere.13 This chapter  11 Housman (1907a) 53, reviewing Marx (1904–1905). That pessimism naturally did not stop Housman (1907b) from offering some selective conjectures of his own; they show little serious consideration of the particular circumstances of fragmentary transmission. 12 In Lindsay’s list of Nonius’ sources these are ‘Gloss. iʼ (1), ‘Gloss. iiiʼ (26), ‘Alph. Verbʼ (27), ‘Alph. adverbʼ (28), ‘Gelliusʼ (32), ‘Gloss. ivʼ (in two parts, 35A and 35B), and ‘Gloss. vʼ (in two parts, 38A and 38B). It is important always to remember that Nonius is not an impartial transmitter of that material; where his own priorities and interests align closely with those of the source (as, for example, seems to be the case with lists 26, 27, and 28), he may be treated as a reliable guide to its contents and shape, but where the alignments are not close Nonius’ excerpts cannot be assumed to be representative. Hertz’s (1862) assemblage illustrating Nonius’ use of Gellius makes the problem clear; about it Lindsay (1901, 104 n. f) wrote “they give us no real idea of what the work of Gellius was like.” 13 Welsh (2012), (2013), (forthcoming).

The Fifth Glossary of Nonius Marcellus  

aims to offer a description of the ‘fifth glossary’ (ʻGloss. vʼ), or, more accurately, a description of some particular features of its methods and an account of the implications of those features for the informed speculations of those who seek to edit or discuss a fragment that Nonius took from this source. It seems impossible to formulate laws on the basis of the observations I offer here, but some more general possibilities and probabilities do emerge that can form part of our calculus of speculation about these fragments. I conclude by examining three fragments where these probabilities seem to point the way down paths not preferred in recent discussions. Assembling the materials assigned to ‘Gloss. v’ yields a corpus of about 220 Latin quotations. Convenience and transparency have been the principal guides in that process, for I have chosen to follow attributions already signalled elsewhere,14 acknowledging that that decision necessarily includes a few quotations about which I myself would dissent, and omits a rather greater number that I believe belong to this source. Even that mechanical procedure makes the resulting corpus necessarily more tentative than others I have discussed, since Lindsay (and others) felt no confidence in distinguishing ‘Gloss. iiiʼ or ‘Gloss. ivʼ from ‘Gloss. vʼ, and so much that could have been assigned to one or more of these sources stands outside the confines of this study. A thorough-going revision and updating of Lindsay’s work on Nonius’ sources would therefore probably add considerable material to this corpus, but those additions would, I venture, augment rather than overturn the descriptions of ‘Gloss. vʼ that I offer in this chapter. Lindsay himself wrote very little about the individual glossaries as distinct works of scholarship. In 1901 he was naturally more concerned to prove their existence, and thus to support his theory about Nonius’ sources and procedures, than to delimit how to edit Ennius or Caecilius or Laberius on the basis of quotations from the glossaries. It would be unfair, to paraphrase Cicero, to expect simultaneous invention and perfection. However, Lindsay’s scattered comments on the fifth glossary do point the way to at least some of its significant features. In his list of Nonius’ sources he identified two parts of a source (numbered 38A and 38B) that he described as “a glossary in rough alphabetical order”.15 Later in his

 14 Most significantly, in Lindsay (1901), (1905), and Rychlewska (1954). 15 A rough alphabetical arrangement is suggested most clearly, but not unproblematically, in two series of entries assigned to it in Book 1; see Non. p. 62.6–65.8 and 66.18–67.25 alongside Lindsay’s (1901) 16–17 analysis of those entries. But there and elsewhere it is clear that the division between lists 38A and 38B, if an alphabetical arrangement is correctly perceived, will have fallen somewhere in the letter ‘P’. That division is odd; it seems even more suspect because of the fraught relationship between Nonius and the letter ‘P’ (see Welsh 2012, 839); and because

  Jarrett T. Welsh analysis he elaborated, remarking (1901, 101) that “[t]he ‘Gloss. v’ lemmas shew some striking resemblances to Festus’ epitome of Verrius Flaccus”, and describing its contents as follows (1901, 104 n. h): Thus it is the Epic of Ennius, not his dramas, which is drawn upon by ‘Gloss. v’, and the number of the book is as often omitted as not. In citations from Caecilius by ‘Gloss. v’ the name of the play is often not mentioned. Laberius furnishes a large number of quotations in this glossary. The quotations from Catullus seem to be peculiar to it.

These remarks, it will be clear, help to identify material that Nonius took from this source, but tell us little about the particular contours of what Nonius found in it, and nothing about the material that he passed over. Assembling the corpus of material that Nonius took from ‘Gloss. vʼ, however, reveals several significant features in that material. We cannot know to what extent these features were characteristic or typical of ‘Gloss. vʼ as a whole; certainly quite a lot of evidence suggests that Nonius found that source every bit as recalcitrant as he found ‘Gloss. iʼ to be, and that he had difficulty extracting from it material that met his standards.16 The safest claim, therefore, is that the following features represent significant patterns in the material that Nonius extracted from ‘Gloss. vʼ, making no positive statement about their prominence or distribution in that text as a whole.17 I discuss three features observable in the Latin quotations taken from this source, before turning to four features observable in its scholarly treatment and presentation of them. At least some of the quotations taken over from ‘Gloss. vʼ show signs of having been rewritten, when compared to versions of the quotations that survive elsewhere. Moreover, that rewriting seems to be more invasive than the kind of rewriting of quotations discussed earlier in the quotations illustrating the verb potior from ‘Gloss. iʼ. Thus Nonius quotes some words from Plautus’ Casina as follows (Non. p. 135.7): LVSTRATVS, luxuriosus: a lustris,

abditis locis. Plautus Casina (245): unde es, nihili? ubi lustretur? ubi bibatur?

 entries illustrating differentiae show no significant alphabetical arrangement and occur in a position most suitable to list 38B (see Non. p. 445.2–447.26 with Lindsay 1901, 20). I am not convinced that Nonius’ excerpts from this source give an accurate impression of its nature, and therefore I refrain from speculating about the shape, arrangement, or identity of the entire work. I suspect, however, that only a part of it was arranged alphabetically. 16 I collect some of this material in section iv of Welsh (forthcoming). 17 Cf. above n. 12.

The Fifth Glossary of Nonius Marcellus  

Lustratus means immoderate, given to luxury, deriving from lustra, concealed places. Plautus in the Casina: ‘Where are you from, worthless man? Where there is (?) whoring and drinking?’ Cf. Plaut. Cas. 245 unde is, nihili? ubi fuisti? ubi lustratu’s? ubi bibisti?

The changes evident in Nonius’ version of the quotation are not particularly sensible: in the place of Cleostrata’s angry questions to her husband stand now some awkward and almost intelligible Latinity, in which the colourless question ubi fuisti has been omitted and the questions ubi lustratu’s and ubi bibisti now stand, perhaps, as impersonal verbs; the reason for their change to subjunctives I cannot explain. Likewise, a line which this source assigns to Catullus,18 but which better evidence assigns to Cinna, has evidently undergone some substantial alterations (Non. p. 546.15): CARCHESIA,

genera poculorum … Alias summa pars mali, id est foramina quae summo mali funes recipiunt … Catullus Veronensis: lucida qua splendet19 carchesia mali. Carchesia, types of cups. … Elsewhere the top part of a mast, that is, the holes which receive ropes at the top of the mast. … Catullus of Verona: ‘where the gleaming top of the mast shines’. Cf. Cinna fr. 2 apud Isid. orig. 19.2.9 lucida confulgent alti carchesia mali.20

Quite apart from the question of the attribution of the quotation, Nonius’ version has lost its descriptor of the height of the ship’s mast and has qua splendet in the place of cum fulgent. Something like alti in this verse would have been useful to illustrate the definition of carchesia that Nonius records, and so it is difficult to explain why it would have been omitted.21 The words qua splendet are obviously similar to the idea expressed elsewhere by cum fulgent, but are equally obviously less persuasive than that version. These kinds of changes should at least caution us against accepting anything from this source as the ipsissima verba of an author.

 18 Supported by Zaffagno (1978). But at Non. p. 517.3 this source also evidently assigned to Catullus a fragment attributed instead to Serenus by Diomedes (gramm. I p. 513.10 = fr. 17 Courtney). 19 I do not adopt Giocondo’s splendet, because I cannot shake a suspicion that Nonius copied from his source a quotation in which carchesia had been interpreted, incorrectly, as a nominative singular form. Admittedly I may be badly mistaken in that suspicion. 20 On this fragment, see Courtney (2003) 216 and Hollis (2007) 24. 21 I am, however, not inclined to restore any such descriptor here, in view of the points I make below about the gaps between lexicographical definitions and quoted examples.

  Jarrett T. Welsh There are, it must be said, very few instances in which the quotations deriving from ʻGloss. vʼ can be checked against an independent source. More readily identifiable, because they require no external source of corroboration, are the numerous literary quotations set down in a form best described as abrupt or syntactically incomplete. The original compiler of this material seems content to have allowed two or three words to illustrate his lexicographical point, no matter how compressed or obscure the original quotation and its context thereby became. A selection of examples illustrates their own obscurity: Non. p. 63.3 degrumare forum (Ennius, Ann. 441 Skutsch) Non. p. 67.25 ueteres prosapia (Cato, Orig. fr. 27 Chassignet) Non. p. 80.22 bipedem bliteam beluam (Laberius, fr. 62 Panayotakis) Non. p. 137.19 cuius amicam mammeatam (Plautus, Poen. 393)22

Quotations like these share a certain opacity and a certain recalcitrance. Readers who feel hard-pressed to explain the meaning of Ennius’ degrumare forum on the basis of that quotation alone, or who want to know more about Laberius’ ‘twofooted spinachy beast’ than that quotation reveals, or who feel no confidence that any of these quotations in fact corroborate the interpretations of the words that Nonius reports would, I imagine, not find themselves alone. If we feel a sense of uncertainty or even aporia when faced with such quotations, then, we should be as ready to accept that that uncertainty is correctly felt, a necessary correlate of the history of these fragments, as we are to suppose that some error or misinterpretation can be diagnosed or corrected. There is sometimes a substantial gap between what Nonius reports about the meaning or usage of a word and what a quotation taken from this source actually demonstrates. The entries delibratum and exterebrare in the first book of the De compendiosa doctrina illustrate well the clash of lemma and example (Non. p. 62.23 and 62.26): DELIBRATVM,

decorticatum; ut DEARTVATVM, per artus discissum. Vergilius Georgicorum lib. II (77): udoque docent inolescere libro. EXTEREBRARE est ui aliquid extorquere et scrutari aut curiosius quaerere. Plautus in Astraba, cum in curiosum iocaretur (fr. VI Monda): terebra tu quidem pertundis. Delibratum, stripped of bark, like deartuatum, cut apart at the joints. Vergil in Book 2 of the Georgics: ‘they teach (them) to grow in the moist bark.’

 22 Cf. the Palatine text of this verse, huius amica mammeata, mea inimica et maleuola.

The Fifth Glossary of Nonius Marcellus  

Exterebrare is to wrest something away by force and to examine or investigate inquisitively. Plautus in the Astraba, when a joke is made at the expense of a busybody: ‘you’re beating me with a drill.’

Vergil’s description of grafted shoots growing in another tree’s moist inner bark illustrates well the meaning not of delibratum,23 but of its root word liber. The Plautine quotation goes a step further still, illustrating neither exterebrare in its literal sense (as would have been appropriate in the first book of the De compendiosa doctrina)24 nor the same verb in any of its metaphorical senses ui aliquid extorquere et scrutari aut curiosius quaerere,25 but a joking riposte that approximates the sense curiosius quaerere with the phrase terebra … pertundis. That riposte, moreover, bears only a tenuous connection to the proprietas of the word exterebrare. Those gaps are particularly significant because they are most subversive to the goals of those who are interested in Nonius’ fragments for their own sake: although the nature of the gaps between the stated subject of these two entries and what their quotations in fact demonstrate is readily recognisable, we cannot assume that every gap will be comparably visible. Instead, there is a good chance that what Nonius tells us about a word bears only a tenuous relationship to the quotation that seems to illustrate that point. Turning from the contents and treatments of the quotations to their history and presentation as a component of the verbal scholarship presented by this text, four significant features emerge. The first of these are the indications that the compiler of this ‘glossary’ sometimes abbreviated the titles of literary works he quoted. This expedient practice was common, but far from ubiquitous, in ancient Latin verbal scholarship. We can trace it, most immediately, in ‘Gloss. ivʼ, a work evidently based on Varro’s writings, but we cannot trace it in any of the other sources Nonius used.26 The clearest evidence for this practice in the fifth glossary is in its quotations of the play we know as the Plautine (or pseudo-Plautine) Cor-

 23 Cf. Masurius Sabinus apud Serv. Auct. Verg. Aen. 1.225 delubrum, effigies, a delibratione corticis; nam antiqui felicium arborum ramos cortice detracto in effigies deorum formabant, unde Graeci ξόανον dicunt (cf. Paulus p. 64.6, Ps.-Ascon. p. 188.2–3 Stangl), Caes. Gall. 7.73.2, Colum. 4.24.6, 5.11.1 (et saepius), 12.58.1, Ps.-Colum. arb. 26.1, 26.8. 24 For which something like Cic. div. 1.48 ex eo auro, quod exterebratum esset would have been appropriate. 25 For the first of which Plaut. Persa 237 numquam hercle istuc exterebrabis, tu ut sis peior quam ego siem would have sufficed. 26 I discuss some Varronian examples in section II of Welsh (forthcoming). Ritschl (1845) 155 was right to feel doubts about an abbreviation in the citation at Nonius p. 147.22; the quotation comes from list 28, and so the best explanation is that Cornicula was corrupted to Corniquia.

  Jarrett T. Welsh nicula. Nonius evidently encountered in his source quotations of this play recorded against the abbreviated title Cornic, which has survived in connection with one of Nonius’ fragments of the play from this source (Non. p. 220.11): generis masculini … Neutri. Plautus Cornic (fr. VI Monda = fr. 65–66 Lindsay): em te obsecro, Lyde, pilleum meum, mi sodalis, mea salubritas.

PILLEVS

Pilleus, used in the masculine gender. … In the neuter. Plautus in the Cornic: ‘Now please, Lydus, my freedom-cap, my comrade, my healthfulness.’

There, it seems, Nonius transcribed that abbreviation unaltered. On other occasions he seems to have expanded the abbreviations in different ways: once as Cornicula (Non. p. 63.11), but twice as Cornicularia (Non. p. 134.28, 134.35),27 on the model of the Plautine Aulularia, Cistellaria, and Mostellaria. That Nonius would have expanded abbreviations inconsistently is hardly surprising, and finds parallels elsewhere in the De compendiosa doctrina.28 But the mere habit of expanding abbreviations necessarily effaces most traces of the original citation on which Nonius relied, and therefore we can normally only identify this feature where obvious mistakes or divergences present themselves, or where a title has been transcribed in its unexpanded form. Second, a recurring pattern in entries taken from ‘Gloss. vʼ has multiple words sharing similar patterns of formation or derivation grouped together in the same entry, in the manner of the entry delibratum (ut deartuatum) already quoted. The habit appears often in ‘Gloss. vʼ entries throughout the De compendiosa doctrina, with particularly fine examples in the first book. Thus the entry consedo (p. 62.18) relates that word to adsedo, the entry moletrina (p. 63.22) to furatrina aut fetutina, the entry mustulentum to uinulentum et faeculentum.29 Comparable groupings occur for words deriving from the same root.30 The first compiler of this material occasionally imposes relationships not corroborated by the comparison.31 A productive point of comparison is Nonius’ use of Gellius’ Noctes

 27 I am not persuaded that Lindsay’s deletion (Cornicula[ria]) in each passage was well motivated in an edition of Nonius. 28 Cf. n. 26 above. 29 Cf. Non. p. 93.26, 101.23. 30 Cf. Non. p. 64.29, 65.4, 66.18, 108.3, 118.22, 134.24. 31 E.g. Non. p. 63.19 fulguratores ut extispices et aruspices, ita hi fulgurum inspectores. The entry prodigiatores harispices, prodigiorum interpretes at Fest. p. 254.29 Lindsay perhaps reveals what has been garbled.

The Fifth Glossary of Nonius Marcellus  

Atticae. Nonius handles material from Gellius in similar fashion, sometimes allowing grouped words to stand in the same entry32 and sometimes (often for the sake of the alphabetical arrangement of Books 2‒4) distributing such groups across diverse entries,33 but not obviously grouping scattered but related material together. It therefore seems reasonable to conclude from at least the ‘Gloss. v’ entries collected here that the compiler of that text employed this methodology, and possibly that other entries now dispersed in the De compendiosa doctrina once stood together.34 Third, occasionally quotations deriving from ‘Gloss. vʼ are equipped with a paraphrase or with an explanatory comment intended to clarify or restate the feature of Latinity under discussion. The entry for latrocinari quotes a fragment of Plautus against the title Cornicularia and follows the quotation with a restatement of the same idea (Non. p. 134.29): LATROCINARI, militare mercede. Plautus Cornicularia (fr. II Monda = fr. 61 Lindsay):

… latrocinatus annos decem Demetrio.35 qui aput regem in latrocinio fuisti, stipendium acceptitasti.

Latrocinari means to serve as a soldier for pay. Plautus in the Cornicularia: ‘served as a soldier to Demetrius for ten years’, you who were in military service for the king, received payment.

Readers who compare the abbreviated quotation here with its counterpart at Varro ling. 7.52 (qui regi latrocinatu’s decem annos Demetrio) will see that Nonius’ words qui aput regem in latrocinio fuisti restate and clarify the meaning of the Plautine words qui regi latrocinatu’s, even though the beginning of the verse has been omitted or lost in Nonius’ quotation; stipendium acceptitasti clarifies still further the meaning expressed by latrocinatus.36 Similar paraphrases occur in the  32 E.g. Non. p. 140.20 = Gell. 6.9. 33 E.g. Non. p. 133.23, 145.9, 148.15, 188.5 = Gell. 18.11, Non. p. 100.19, 129.24 = Gell. 19.8.6. 34 Cf. Non. p. 63.27 mustulentum ut uinulentum et faeculentum and p. 114.15 floces faex uini. The entries at Non. p. 101.23 and 142.23 quote the same verse of Pacuvius (109–110 Ribbeck3 = fr. 83 Schierl), and the definition of memoria in the latter entry should make any reader suspect Nonius’ own handiwork. 35 Lindsay’s restoration of decem Demetrio (building on a suggestion of Onions 1895, 166) seems a more plausible correction of the corrupted text X mercedem in tiberio than does Müller’s decem [mercedem] in Tiberium, which presupposes the unlikely coincidence of a quotation of Cato’s In Tiberium occuring alongside a Plautine verse referring originally to Demetrius, a name easily extricated from mercedem in tiberio. 36 Cf. Plaut. Mil. 499, Poen. 704–705, Trin. 599. In treating this paraphrase as a ‘fragmentum incertum’ TLL 1017.45 s.v. latrocinium seems to me to err.

  Jarrett T. Welsh extraordinary discussion of protinus and related words (Non. p. 375.31), in which Plautus’ in protenam dedi (Plaut. Curc. 363) is explained as meaning quasi in longam fugam; Naevius’ protinam me dedi (Naev. com. 35) as id est pedibus longius fugi;37 and, somewhat later in the discussion, Vergil’s protinus aeger ago is glossed with alternative explanations (ut sit animo et corpore ualde aeger; aut si aliud enuntiat, refertur ad illud ut sit longe, porro ago). At least three other instances of comparable paraphrase, although textually uncertain, have sometimes been identified by editors.38 This kind of paraphrase of Latin literary quotations, otherwise quite uncommon in the De compendiosa doctrina, thus occurs often enough in connection specifically with material taken from ‘Gloss. vʼ that we may safely regard it as a feature of its scholarly explanation. Finally, I return to a feature pointed out already by Lindsay, namely that Nonius’ material from the fifth glossary shows many resemblances and connections with the descendants of the enormous Augustan-era treasure trove of lexicography known to us as Verrius Flaccus’ De uerborum significatu, now lost but incompletely digested by much imperial Latin scholarship, including especially the first- and second-generation epitomes compiled by Festus and by Paul the Deacon.39 The following examples of connections between ‘Gloss. v’ entries and material attributable to Verrius Flaccus focus on correspondences beyond mere identical definitions of words, and give an intimation of the relationship between the two texts: (c)

Non. p. 62.12 CALONVM quoque proprietas haec habetur, quod ligna militibus subministrent: κᾶλα enim Graeci ligna dicunt.

 37 Although these paraphrases of Plautus and Naevius seem, at first glance, to describe distance, they probably refer instead to time. Varro’s continuitatem significans (ling. 7.107) has been clarified in the expansion continuationem uel praeteritae uel futurae longitudinis dicit; for this meaning cf. TLL 1630.26 s.v. longitudo. The paraphrases probably owe something to ancient scholarship on Ter. Phorm. 190–191; cf. Paulus p. 253.17 Lindsay and Don. Ter. Phorm. 190. 38 At Non. p. 181.5 the last four words of the entry (ut quaestus sit magnus) should therefore be accepted as an explanatory paraphrase of Varro’s adjective uber, as Lindsay printed them, and not bracketed (“f. secludenda” Müller) or assigned to Varro himself (so Mazzacane 2014 and, tentatively, Lindsay 1903 in app.). At Non. p. 212.7, the words hic latrinum uentri † finem locum dixit plainly once explained Laberius’ latrinum as being something different from the la(ua)trina quod nunc balneum dicitur. (Cf. the definition of foriolus at Non. p. 114.9 and of mustulentus at Non. p. 63.27.) At Non. p. 446.34 the text is uncertain per se and rendered even more so by the presence of a (different) interpolation at the same point in the quotation of Plautus, Persa 408 at Non. p. 10.10, but the words hoc est uiuens sine lege, sine iure seem difficult to explain as anything but a similar paraphrase. 39 Some of this material was identified already by Froehde (1890).

The Fifth Glossary of Nonius Marcellus  

Paulus p. 54.19 CALONES militum serui dicti, quia ligneas clauas gerebant, quae Graeci κᾶλα uocant.40 Nonius: This too is regarded as the precise meaning (proprietas) of calones, in that they supply sticks to soldiers: for the Greeks call logs κᾶλα. Paulus: Slaves of soldiers are called calones, because they carried wooden sticks, which the Greeks call κᾶλα (‘logs’). (b) Non. p. 63.19 FVLGORATORES ut extispices et aruspices, ita hi fulgurum inspectores. Fest. p. 254.29 PRODIGIATORES harispices, prodigiorum interpretes.41 Nonius: Fulgoratores (‘lightning-interpreters’), as extispices and aruspices (‘entrail-diviners’) (scil. are inspectors of entrails), so these are inspectors of lightning. Festus: Prodigiatores are entrail-diviners, interpreters of prodigies. Non. p. 66.18 POLITIONES, agrorum cultus diligentes, ut POLITA omnia dicimus exculta et ad nitorem deducta. Fest. p. 266.19 . . . dicebant testiculos porcorum, cum eos castrabant, a politione segetum aut uestimentorum, quod similiter atque illa curentur. Nonius: Politiones, diligent cultivations of fields, just as we call polita all things that are well cultivated and polished to a shine. Festus: They used to call the testicles of pigs, when they castrated them, , from the cultivation (politio) of crops or clothes, because they are cared for similarly.

Some correspondences between the texts are fairly substantial, such as their shared etymology connecting calones with the Greek κᾶλα; that etymology is especially significant because of the general sparseness of Greek material across these texts.42 Others are more slight, though similar enough to suggest that we are simply missing a crucial piece of corroborative evidence, as in the case of the entries fulguratores and prodigiatores. Still others are more complicated, as in Nonius’ connection of politiones with cultus echoing, perhaps, the logical connection drawn by Verrius Flaccus between the treatment of pig testicles at castration and the similar care taken over crops and clothing. Some significant implications can be seen from even that limited sample of correspondences. Nonius’ ʻfifth glossaryʼ was not any of the extant descendants

 40 Cf. Serv. Verg. Aen. 1.39, Serv. Auct. Verg. Aen. 6.1. 41 Cf. n. 32 above. 42 Convenient measures of that sparsity can be had from the indices uerborum Latinorum et Graecorum in Lindsay’s edition of Festus and Paulus (40 pages for the former, just over 2 pages for the latter), and the index entry for sermo graecus in Lindsay’s edition of Nonius; only a handful of the latter pertain to ‘Gloss. vʼ.

  Jarrett T. Welsh of Verrius Flaccus’ massive and encyclopedic work of lexicography. It was, however, related in some indirect way to them, or had at least absorbed quite a lot of Verrius’ learning. About forty per cent of the corpus securely assigned to ‘Gloss. vʼ shows some connection with Verrius’ intellectual legacy.43 That proportion is too large to be a result of accident or chance; apart from the epitomes of Festus and Paul, no other work of Latin scholarship known to me comes close to rivalling this proportion of Verrian doctrina. But the other part of that equation is equally significant, for some sixty per cent of the material therefore shows no connection, or at least no identifiable connection, with Verrius. Even allowing for some ‘missing links’, it seems an inevitable conclusion that ‘Gloss. v’ included some material that did not originate in Verrius’ own work. The most obvious corroboration of that fact is the point that this source provided Nonius with very nearly all of his quotations of Laberius, an author evidently neglected by Verrius, and likewise those of Catullus, Horace, and others whose Latinity was, although not in doubt, certainly not in keeping with the standards that Verrius employed when selecting his exempla. A source not dependent exclusively on Verrius is, implicitly, also not constrained by his methods; many of the features I have discussed diverge in obvious ways from the methods employed by Verrius and still observable in the works of Festus and Paul.44 There are, therefore, indications that Nonius found in the fifth glossary a source whose priorities, methods, interests, and contents did not mirror his own. Faced with such material, he excerpted from that source what he could add to the De compendiosa doctrina without overthrowing entirely his own criteria. That necessarily means that we have an imperfect impression of the fifth glossary, and those imperfections mean that it is impossible to formulate ‘laws’ or ‘principles’ for fragments deriving from this source. Even so, a number of considerations present themselves: problems worth considering, questions worth asking, and possibilities worth worrying over when speculating about a fragment. I signal four considerations that have occurred to me in examining this material. First, there seems to be a greater likelihood — as compared to the quotations that Nonius excerpted at first hand from complete texts, and as compared to quotations taken from other ‘glossaries’ — that the fragments preserved from ‘Gloss.

 43 That is not to say that quotations are necessarily shared by the two; rather, that the interpretations of individual words and expressions are identical or similar enough to suggest a plausible connection between the two. 44 The most obvious divergence is that the compiler of ‘Gloss. v’ quoted Latin verse in units of complete sense without regard for metrical completeness, in contrast to Verrius’ principle of quoting metrical units without regard for the completeness of sense, for which see Welsh (2015).

The Fifth Glossary of Nonius Marcellus  

vʼ have been corrupted or rewritten in ways that are not readily transparent, not readily intelligible, and even, when recognised, not readily justifiable. It is commonly thought that the text of the De compendiosa doctrina is badly corrupted, and that view has much in its favor, since scribes confronted with quotations out of context often had difficulty following the logic of what they were writing. However, on many occasions it seems that the text of Nonius is in fact less corrupt than the editions of Müller, Onions, and Lindsay imply; more precisely, many of the alleged errors and omissions are as likely to have been transcribed or even committed by Nonius himself.45 Nonius’ quotations of Plautus and Cicero, for example, are generally good because he took them from complete texts. Likewise the quotations taken from lists 26, 27, and 28 are generally well preserved because the quality of those sources was very high. Quotations from sources like ʻGloss. iʼ (which shows a bewildering patchwork of accurate and inaccurate material) and ‘Gloss. v’ (which shows frequent inaccuracies of several types), however, are rather more often inaccurate and slipshod. Second, since the compiler of ‘Gloss. v’ seems to have regarded a few words as sufficient to illustrate his scholarly or lexicographical point, we should never trust that something like cuius amicam mammeatam or degrumare forum or any of the other egregiously curtailed quotations preserved from this source adequately represent their original context. The problems that that observation poses to readers and editors interested in the fragments for their own sake will be obvious, since these kinds of quotations pose great obstacles to any attempt to understand or recover that original, now-lost context. Third, we should not put too much trust in the titles attached to quotations from ‘Gloss. v’ where we cannot verify them from an independent source; Nonius seems often to have expanded abbreviations he found in that source, and at least sometimes he erred or was inconsistent in what he wrote. That pattern has implications both for the text of the De compendiosa doctrina, where editors ought not normalise variations in titles attached to quotations from this source,46 and for the fragments themselves, where problems of attribution are even more fraught than usual. Thus, for example, since the manuscripts of Nonius preserve two quotations of Laberius that evidently were taken from ‘Gloss. v’ and that are recorded against the titles †there (Non. p. 108.5 = Laber. fr. 26 Panayotakis) and Cythera

 45 On the difficulties of editing a text whose author transcribes much of his material from earlier sources, see esp. Kaster (2010) 65–84. 46 Lindsay’s habit of regularising in the De compendiosa doctrina titles of works Nonius excerpted himself otherwise has value, although he applied it more extensively than is appropriate; some examples are discussed in the appendix to Welsh (forthcoming).

  Jarrett T. Welsh (Non. p. 151.19 = Laber. fr. 27 Panayotakis), the next editor of Laberius ought to consider whether the fragments, linked long ago by emendation of their titles, instead properly belong to two different scripts, one being graced with a title that could have been abbreviated as (something like) †there.47 Fourth, where there is agreement between a ‘Gloss. v’ entry and material deriving ultimately from Verrius Flaccus — whether that connection presents itself in scholarly interpretations or in the quotations adduced to support them — it is at least methodologically unsound, perhaps even unwise, to reject that agreement in favor of some other interpretation, unless that interpretation is supported by cogent arguments sufficient to override the impression of agreement between ‘Gloss. v’ and Verrius Flaccus. In this case, as always, we will be weighing probabilities against one another. It would be foolish to lay down a ‘law’ about the fifth glossary and Verrius, not least since Nonius’ source only derived some materials from the Augustan scholar, and since it was compiled by someone considerably less careful and methodical than Verrius himself. Nevertheless, probability does not favor arguments that discard evidence from Festus or Paulus arbitrarily. None of these points rises to the level of a generally applicable ‘law’, and accordingly someone confronted with a fragment illustrating one or another of these phenomena will still, at best, be speculating about the original context. But informed speculation is better than wild speculation, and looking at this material en masse has suggested some ways that we might guide our speculations away from some possibilities and towards others. Three examples illustrate what could, and even should, be done with this sort of material. In illustration of the ʻregularʼ masculine gender of pilleus (‘freedom-hat’) Nonius quotes from the fifth glossary eight words that allegedly belong to Plautus’ Captiui, followed by a fragment of another Plautine play to illustrate the irregular neuter gender (Non. p. 220.11, with modifications to Lindsay’s presentation of the text): PILLEVS generis masculini. Plautus Captiuis (fr. dubium Lindsay) :

pilleum,

 47 Panayotakis (2011) 241 states fully the implicit assumptions of previous centuries: “It is not unreasonable to conclude, given that there is some similarity between these words, that cythera and there were variant readings of the title of the same mime”. But in this case ‘not unreasonable’ is not the same as ‘more plausible than other speculations in view of the characteristics of this source’. I leave it to the next editor of Laberius to weigh these speculations against one another. It troubles me not at all that the only known title of Laberius beginning with the letter ‘T’ (viz. Tusca) is rather different from †there.

The Fifth Glossary of Nonius Marcellus  

quem habuit, deripuit eumque ad caelum tollit … Neutri. Plautus Cornic (fr. VI Monda = fr. 65–66 Lindsay) : em, te obsecro, Lyde, pilleum meum, mi sodalis, mea salubritas. Pilleus, used in the masculine gender. Plautus in the Captiui: ‘he snatched off the pilleus, which he had, and raises it to the sky.’ In the neuter. Plautus in the Cornic: ‘now please, Lydus, my freedom-cap, my comrade, my healthfulness.’

The Captiui has, of course, survived, and in that script one finds nothing resembling this quotation, nor indeed any instance of the word pilleus in any gender. Unsurprisingly, then, when students of Plautus have taken note of this text, they have given free reign to their speculations. The attribution of this quotation to the Captiui has been explained by appeal to a lacunose text of Plautus,48 to inexact quotation from memory,49 and to simple error over the title.50 Each of those general explanations is possible, and any of their specific permutations could be correct: the evidence is insufficient to permit a certain solution. The safest editorial solution therefore would be to set the fragment among the incerta, as recent editors have (almost) preferred to do.51 But attention to the source that provided Nonius with this quotation perhaps will point the way to the source of the original error, and thus towards a more reasonable manner of speculating about the fragment. In illustration of the lesscommon neuter gender of pilleum, Nonius took from the fifth glossary the quotation of Plautus’ Cornicula that I have already mentioned for its abbreviated title ‘Cornic’. The presence of that abbreviation, in turn, suggests a plausible speculation to explain this problem. I submit that the easiest way that Nonius could produce the attribution ‘Plautus Captiuis’ attaching to a quotation not occurring in that script is by encountering an abbreviated title attached to this quotation in the fifth glossary, which he simply expanded incorrectly as Captiuis. Although this explanation (whatever support it draws from the present argument) is necessarily as speculative as any other, I would venture the guess that Carbonaria is

 48 Pascal (1901) 13–15, Havet (1910). 49 Lindsay (1901) 106 n. m. 50 E.g., Schoell (1887) 95, Leo (1895) 221, Aragosti (2009) 296–297. 51 More precisely, Monda (2004) 103, Torino (2013) 88, and De Melo (2013) 490 each put the fragment among the dubia. I leave it as a puzzle to readers to explain how ‘Plautus Captiuis’ could have been attached to a non-Plautine quotation in the entry pilleum; certainly I can produce no explanation that merits consideration, and so I am inclined to think that this quotation is ‘dubiously Plautine’ only in the sense that every quotation attributed to Plautus in antiquity and taken from a play that has not survived independently is ‘dubiously Plautine’.

  Jarrett T. Welsh the correct attribution, and that Nonius’ Captiuis represents a mistaken expansion of something like CAR or CARB (if the latter, perhaps written CARV). My second example considers a fragment which Lindsay attributed to ‘Gloss. iʼ, for reasons that no longer seem cogent, and which I would assign instead to ‘Gloss. v’.52 At p. 84.5 Nonius illustrates the innovative word conspicillum, allegedly meaning ʻa place for spyingʼ (the quotation itself is presented after Monda’s scansion): CONSPICILLVM, unde conspicere possis. Plautus Medico (= Parasitus Medicus fr. I Monda = fr. 99–100 Lindsay): in conspicillo | adseruabam pallium, obseruabam

Conspicillum, a place from which you could observe. Plautus in the Medicus: ‘I watched over the pallium in a place for watching, I observed’

Onions deserves the credit for first (insofar as I have found) suggesting that obseruabam is a gloss on adseruabam;53 L. Müller subsequently reached the same conclusion,54 but more recent editors have preserved obseruabam as authentic to the fragment.55 The arguments adduced in defense of obseruabam — namely that the verb is sometimes used absolutely56 — are not conclusive, since in conspicillo adseruabam pallium is a complete thought, and we might instead assume that obseruabam was imperfectly excerpted to corroborate the alleged meaning of conspicillum, even though it properly introduced unquoted material that followed. The insufficiency of those arguments does not, of course, prove that obseruabam is an interpolation, and editors who choose to preserve it as part of the Plautine quotation would not be unjustified in that decision. But readers who harbour suspicions about obseruabam would be equally justified in treating it as non-

 52 See Lindsay (1901) 40, for this attribution, and Welsh (forthcoming) for arguments about the nature of the ‘Gloss. iʼ sequences, with n. 88 there for suggestions about this entry. 53 Onions (1883) 88, not superseded by Onions (1888) 173. His third thoughts, in my view no better than his first, are printed at Onions (1895) 103. 54 Müller (1888) 115. 55 Lindsay (1903), Monda (2004) 79, Mazzacane (2014) 145. 56 Aragosti (2009) 187.

The Fifth Glossary of Nonius Marcellus  

Plautine, regarding it now not as an interpolation, but as an explanatory paraphrase or gloss on adseruabam (‘keep watch over, guard’) of the sort found in other ‘Gloss. vʼ. entries.57 The third example illustrates an argument made from several of the features of the fifth glossary. From that source Nonius preserves a fragment of Laberius’ mime entitled Imago, given in illustration of the phrase genius generis (Nonius p. 119.26): GENIVS GENERIS. Laberius in Imagine (54 Ribbeck3 = fr. 34 Panayotakis):

genius generis nostri parens.

The entry occurs in the second book of the De compendiosa doctrina, entitled De honestis et noue ueterum dictis (‘On respectable and innovative usages of the ancients’). Nonius therefore thought that he was recording an expression that was innovative but respectable and, more importantly, sanctioned by the usage of ueteres deemed authoritative.58 In his edition of Laberius, Panayotakis (2011) 273 forbears; following Carilli’s (1982) 70 claim that Nonius mistakenly believed that genius generis was a formulaic expression, he asserts: Nonius cannot be right in his interpretation of the syntax in L.’s fragment, because if we assume that genius forms an unparalleled expression when coupled with generis, and if we take nostri to be the genitive singular of the adjective noster qualifying generis, then parens is left in a very odd position in the line (?‘the parental guardian spirit of our clan’?).

That is all well and good; it is speculation, but on the evidence put on record before 2011, it was reasonable speculation. Whether it is justifiable speculation is another question, and quite a lot of evidence suggests that it is less justifiable than other explanations. Nonius, after all, was not interpreting Laberius’ syntax off his own bat, nor, given the frequency of abrupt or syntactically incomplete quotations from ‘Gloss. v’, can the objection about the odd position of parens be considered cogent. That objection, as formulated, is rather vague, but it especially has no force given that the compiler of this material sometimes felt no concern over metrical completeness, clarity, or even intelligibility. I can see no reason to assume that parens stood as a one-word apposition to genius in Laberius’ text, but I can see many reasons why the compiler of ‘Gloss. v’ would stop excerpting after the word parens. Nor can there be any force in the objection that  57 I pass over the problems posed by assuming that this title refers to the Parasitus Medicus (defended by Aragosti 2009, 186 with n. 263), but would point out that the entry lopades at p. 551.3 comes from ‘Gloss. vʼ. 58 On the implications of the title see Welsh (forthcoming) n. 52.

  Jarrett T. Welsh genius generis would be an unparalleled expression: the lighter dramatic genres have more than their fair share of singular diction and hapax legomena,59 and few books of the De compendiosa doctrina are more accommodating of unparalleled expressions than the second. In supporting the claim that Nonius mistakenly interpreted genius generis as a formulaic expression modeled after genius loci and other similar expressions, Carilli pointed to a definition of genius recorded against the name of Aufustius and preserved in Paul’s epitome of Festus’ epitome of Verrius Flaccus’ De significatu uerborum (‘On the meaning of words’). The full entry bears repeating here (Paulus p. 84.3 Lindsay): appellabant deum, qui uim optineret rerum omnium gerundarum. Aufustius: ‘genius’ inquit ‘est deorum filius et parens hominum, ex quo homines gignuntur. Et propterea genius meus nominatur, quia me genuit.’ Alii genium esse putarunt uniuscuiusque loci deum. GENIVM

They called a god who had the power of doing all things a genius. Aufustius says ‘a genius is a son of gods and a parent of humans, from which humans are produced. And my genius is so named for that reason, because it produced me (me genuit).’ Others thought that a genius was the god of each individual place.

The connection between this entry and Nonius’ interpretation is perhaps more significant than Carilli thought.60 We have seen that the fifth glossary shows connections with the work of Verrius Flaccus that cannot be dismissed as accidental. Aufustius’ definition accords tidily with Nonius’ interpretation of genius generis, which would mean not the ‘guardian-spirit of our clan’, but something like its ‘productive-spiritʼ or even ‘ancestor’. That connection should not be abandoned or overridden without cogent reasons. For abandoning it requires the assumption that someone who could read Laberius’ play, who evidently thought that these words illustrated something like Aufustius’ explanation of genius, and who excerpted enough of Laberius’ words to corroborate that explanation, nevertheless did so wrongly, against the proper interpretation of what Laberius wrote. Such a source, I submit, is at least incompetent and untrustworthy, very possibly even subversive, and if that explanation is even partially right, the obelus surely ought to be applied to this fragment.  59 Cf. Carilli (1980) and Panayotakis (2011) 63. Carilli (1982) 70 n. 86 was categorically wrong to assert that the lack of a parallel “conferma che si tratta di un errore d’interpretazione di Nonio”. 60 Inter alia, the last sentence of Paulus’ entry suggests that at least one branch of ancient lexicographical thought would have regarded a usage like genius generis as fundamentally distinct from that of genius loci.

The Fifth Glossary of Nonius Marcellus  

But a different explanation is both easier to accept and better supported by what we can say about the fifth glossary. My own argument would run as follows. Nonius preserves four words of Laberius’ Imago in illustration of the phrase genius generis. Their first excerpter understood those words to illustrate satisfactorily the meaning of genius as explained by Aufustius in material descending ultimately from Verrius Flaccus. That compiler regarded semantically and syntactically incomplete quotations as adequate illustrations of lexical phenomena; accordingly he quoted genius generis nostri and the word parens, which is slight material, but sufficient to corroborate Aufustius’ definition. Laberius’ words need mean no more than ‘the genius (the ‘productive spirit’ or ‘ancestor’) of our clan, the parent’ — of some individuals, not mentioned in the excerpted quotation, whose identity was not needed to prove the compiler’s point, and could therefore be left to the side. Sometimes the simpler explanation is also the one most supported by the available evidence.61 Attention to this kind of argumentation is essential because it provides the best kind of ‘context’ (to revisit Housman’s grim warning) for a literary fragment that we can reasonably hope to explore, and because it provides a firmer foothold for our speculations about such material, where firm footholds are few and far between. The specific arguments that I have made are, of course, not transferable to Nonius’ other sources, nor indeed to the texts that he excerpted himself, but they do tell us about this source and about the kinds of questions we should be asking about other ancient scholarship that gives us our book fragments. It also points to an assumption that is, it seems to me, in need of more interrogation than it has received, namely the assumption that editing a corpus like the fragmentary tragedies of Pacuvius or the fragmentary mimes of Laberius is a task identical to and coterminous with analysing and editing the fragments of those authors. For quite a lot of information seems to have escaped the attention of editors because it has been assumed, not always rightly, that editing those fragments can be done without taking account also of, for example, some rewritten quotations of Terence and Plautus or of a pattern of abrupt quotations sometimes attached to abbreviated play-titles. I freely admit that that is something of a horrifying prospect, because it does not make it easier to put fragments back into our general conversations about extant texts, but its horror does not diminish the fact that it needs doing, and in some cases desperately needs doing; otherwise a lot of plausible

 61 Readers may at this point wish to revisit Laber. fr. 67 Panayotakis (a quotation that Lindsay admittedly assigned to ‘Gloss. iʼ for reasons no longer defensible), recalling that ablative mea regularly scans as a single heavy syllable in scenic verse.

  Jarrett T. Welsh but inaccurate speculation about fragments will go undetected, and our reconstructions will necessarily remain more uncertain than is commonly recognised.

References Aragosti, A. (2009), Frammenti plautini dalle commedie extravarroniane, Bologna. Barabino, G./Mazzacane, R. (2000), ‘Bibliografia Noniana’, in F. Bertini (ed.), Prolegomena Noniana I, Genoa, 7–77. Barabino, G./Mazzacane, R. (2003), ‘Aggiornamento’, in F. Bertini (ed.), Prolegomena Noniana II, Genoa, 149–152. Butterfield, D. (2013), The Early Textual History of Lucretius’ De rerum natura, Cambridge. Carilli, M. (1980), ‘Artificiosità ed espressività negli «hapax» di Laberio’, Studi e ricerche dell’Istituto di Latino 3, 19–33. Carilli, M. (1982), ‘Note ai frammenti di Laberio tramandati da Nonio’, Studi Noniani 7, 33–88. Chahoud, A. (2007), ‘Antiquity and Authority in Nonius Marcellus’, in J.H.D. Scourfield (ed.), Texts and Culture in Late Antiquity: Inheritance, Authority, and Change, Swansea, 69–96. Courtney, E. (2003), The fragmentary Latin poets, Oxford. D’Anna, G. (1967), M. Pacuvii Fragmenta, Rome. De Melo, W. (2013), Plautus. Vol. 5, Cambridge (Mass.). Froehde, O. (1890), De Nonio Marcello et Verrio Flacco, Berlin. Havet, L. (1910), ‘La lacune des Captifs’, in Mélanges offerts a M. Émile Chatelain, Paris, 26–32. Hertz, M. (1862), ‘A. Gellius und Nonius Marcellus’, Jahrbücher für classische Philologie 8 (= NJPhP 85), 706‒726, 779‒797 (= 1886, Opuscula Gelliana, Berlin, 85–146). Hollis, A.S. (2007), Fragments of Roman Poetry c. 60 BC–AD 20, Oxford. Housman, A.E. (1907a), ‘Luciliana’, CQ 1, 53–74 (= Classical Papers ii, 662–684). Housman, A.E. (1907b), ‘Luciliana’, CQ 1, 148–159 (= Classical Papers ii, 685–697). Kaster, R.A. (2010), Studies on the Text of Macrobius’ Saturnalia, Oxford. Leo, F. (1895), Plautus: Comoediae, Berlin. Lindsay, W.M. (1901), Nonius Marcellus’ Dictionary of Republican Latin, Oxford. Lindsay, W.M. (1903), Nonii Marcelli De compendiosa doctrina, 3 vols, Leipzig. Lindsay, W.M. (1905), ‘De citationibus apud Nonium Marcellum’, Philologus 64, 438–464. Marx, F. (1904–1905), C. Lucilii carminum reliquiae, Leipzig. Mazzacane, R. (2014), Nonio Marcello: De conpendiosa doctrina I, libri I-III. With E. Magioncalda and P. Gatti, Florence. Monda, S. (2004), Titus Maccius Plautus: Vidularia et deperditarum fabularum fragmenta, Sarsina/Urbino. Müller, L. (1888), Noni Marcelli Compendiosa Doctrina, Leipzig. Onions, J.H. (1883), ‘Notes on Nonius’, JPh 12, 79–98. Onions, J.H. (1888), ‘Noniana Quaedam’, JPh 16, 161–182. Onions, J.H. (1895), Nonius Marcellus: De conpendiosa doctrina I–III, Oxford. Panayotakis, C. (2011), Decimus Laberius: The Fragments, Cambridge. Pascal, C. (1901), ‘Osservazioni critiche sui Captivi di Plauto’, RFIC 29, 1–15. Ritschl, F. (1845), Parerga zu Plautus und Terenz, Leipzig.

The Fifth Glossary of Nonius Marcellus  

Rychlewska, L. (1954), ‘Quaestiones Nonianae: De librorum XI–XX compositione et fontibus’, in Tragica II, Wroclaw, 117–141. Schierl, P. (2006), Die Tragödien des Pacuvius, Berlin. Schoell, F. (1887), T. Macci Plauti Captivi, Leipzig. Taplin, O. (2007), Pots and Plays: Interactions between Tragedy and Greek Vase-painting of the Fourth Century B.C., Los Angeles. Torino, A. (2013), Titus Maccius Plautus: Captivi, Sarsina/Urbino. Welsh, J.T. (2012), ‘The Methods of Nonius Marcellus’ Sources 26, 27, and 28’, CQ 62, 827– 845. Welsh, J.T. (2013), ‘Some Fragments of Republican Drama from Nonius Marcellus’ Sources 26, 27, and 28’, CQ 63, 253–276. Welsh, J.T. (2015), ‘Verse Quotations from Festus’, HSCP 108, 403–465. Welsh, J.T. (forthcoming), ‘Nonius Marcellus and the Source Called ‘Gloss. i.’’, HSCP 111. White, D.C. (1980), ‘The Method of Composition and Sources of Nonius Marcellus’, Studi Noniani 8, 111–211. Zaffagno, E. (1978), ‘Carchesia, summa pars mali’, Studi Noniani 5, 313–321.

Nereida Villagra

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers A Case Study on Phineus and the Argonauts In this chapter,1 I shall examine closely one ἱστορία (historia), which must have belonged to the Mythographus Homericus to the Odyssey, and I shall use it as a case study in order to try to throw some light on the methodological issues the editor of such a textual artefact has to face when ‘reconstructing’ the Mythographus Homericus — in particular, the nature of the work and the state of its text. My analysis of this ἱστορία will also give me an opportunity to address some problems related to the fragmentary mythographers quoted in the subscriptiones.

 From Panzerʼs hypothesis to the papyri discoveries Mythographus Homericus (henceforth MH) is the designation created by Johannes Panzer in the nineteenth century to name the author of a hypothetical book on mythography.2 He argued that the ἱστορίαι in the D-scholia to Homer came from this book.3 In the twentieth century, the edition of several subliterary papyri proved the existence of such a mythographical commentary.4 Indeed, some papyri contain ἱστορίαι with a parallel redaction to the ἱστορίαι in the D-scholia, arranged by Homeric lemmata and followed by a subscription to an authority in  1 This paper has been written with the support of the FCT post-doctoral scholarship SFRH/BPD/90803/2012. The funded project has been developed at the host institution, the Centre for Classical Studies at the School of Arts and Humanities at the University of Lisbon. 2 Panzer (1892). His dissertation was forgotten until 1937, when Pfeiffer (1937) 1‒18 published a paper on the P. Lond. 142 and identified it as a fragment of the Mythographus Homericus, recovering Panzerʼs theory. 3 By ἱστορία I refer to scholia containing narratives on mythical subjects with subscriptiones with the formula ἡ ἱστορία παρὰ X, ἱστορεῖ X, or equivalent expressions. On ἱστορίαι, see Cameron (2004) 91‒93. See also Vassilaki (2015) and Delattre (2016). 4 Pfeiffer (1938) 1‒18. For a comprehensive edition and description of the papyri of the MH, see van Rossum-Steenbeek (1998) 85‒118, 278‒309. See also Pagès (2007) 26‒35, Montanari (1995) 229‒242, Haslam (1990) 31‒36, and Montanari (2002) 129‒144. https://doi.org/10.1515/9783110712223-008

  Nereida Villagra agreement with the scholia. Thus, the papyri proved three important points: first, that the ἱστορίαι derive from a strictly mythographical work and not from a commentary of miscellanea; secondly, that this work had the format of an ὑπόμνημα and was not a continuous book such as Apollodorusʼ Library;5 and, third, that the subscriptiones were original and not added by a Byzantine scholiast. The date of the oldest surviving papyrus containing a ἱστορία and the terminus post quem indicated by a subscription which quotes Didymus suggest that this mythographical work must have circulated already by the end of the first century CE.6 This fits in well with the cultural panorama of the Augustan era, for this period saw the development of many mythographical works.7 The MH is not only fragmentary, but in addition, while it is a work whose existence can be deduced with considerable certainty thanks to the papyri, it is also a work for which we do not have any independent testimonia. Neither the scholia nor the papyri provide any information regarding its authorship; nor is there any testimony which refers to the nature or the title of this work.8 Scholars have been confronted by the complexity of the transmission of the ἱστορίαι and by the heterogeneity of the fragments themselves, both in the papyri and in the Byzantine manuscript tradition. Indeed, the edition and study of such a work poses several methodological questions, since the analysis of its nature and of the state of the text will depend on which ἱστορίαι are considered to belong to this work. In general, when a ἱστορία in the D-scholia, usually with a subscription, is also attested in a papyrus, it is considered to belong to the MH. However, not all the ἱστορίαι in the scholia have a corresponding papyrus. On the other hand, in order to classify a papyrus fragment as belonging to the MH, it is required that it has, in turn, a Homeric lemma and a mythographical narrative paralleled by the

 5 Panzer (1892) 61‒63, who could not know the papyri yet, proposed the MH to be similar to the Library. 6 On the date, see van Rossum-Steenbeek (1998) 115‒116, Montanari (2002) 132. 7 This coincides with the Second Sophistic. Mythographical works around this period are: Parthenius (Lightfoot 1999) and Conon (Brown 2002) in the Augustan period, the narrative hypotheseis to Euripides (Meccariello 2014) and Apollodorusʼ Library (Cuartero 2010, 2012) around the second century CE, Antoninus Liberalis in the third century CE (Papathomopoulos 1968). A description of mythography can be found in Pellizer (1993), Higbie (2007), Meliadò (2015). Discussions in Henrichs (1987), Fowler (2000) xxvii‒xxxviii, Alganza Roldán (2006), Trzaskoma/Smith (2013). Specifically on the Imperial times, see Cameron (2004). 8 Since the MH is a hypothetical author, it is common practice to use the designation MH also to refer to the work.

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

D-scholia to Homer. But, again, there are papyri with several ἱστορίαι, some of which have no parallel in the scholia. Regarding the mythographical scholia with no parallel in papyri, scholars consider only the material which can be traced back to the D-scholia to have belonged to the MH.9 However, some cases cannot be clearly ascribed only to the minor tradition. As a matter of fact, a number of mythographical narratives are transmitted also by Homeric scholia from other manuscript traditions, with similar redaction but with differences which prevent us from ascribing both ἱστορίαι to the same source.10 And there are even some cases of ἱστορίαι which have been transmitted only outside the D-scholia.11 Another problem in the study of the MH is that opinions are divided as to whether it is the papyri or the scholia which preserve a state of the text closer to the original. Whereas van Rossum-Steenbeek gives priority to the papyri because of their earlier date,12 Montanari and Pagès argue that the scholia can sometimes preserve a better state of the text, which was simplified in the version of the papyri. This opinion rests on the fact that the papyri and the scholia represent different stages in the transmission, yet neither the papyri nor the scholia preserve the original wording of the MH.13 Regarding the nature of the original work and its purpose, it has been stressed that the relationship of the ἱστορίαι to the lemmata is sometimes puzzling, because some fragments show a closer relationship to the Homeric poems, whereas others take the poem as an excuse to insert a mythical narrative extremely loosely related to the context.14 Different opinions have arisen. Montanari thinks that the work of the MH was originally a sophisticated commentary which incorporated learned citations, but which lost this component in the process of transmission.15 Van der Valk believes that the process of simplification — in terms of the workʼs sophistication — occurred in the redaction of the commentary itself

 9 Montanari (1995) 135‒138, (2002) 132. 10 Montanari (1995) 4, Pontani (2005) 190 n. 421. 11 Panzerʼs theory on the textual history of the ἱστορίαι in the scholia to the Iliad is summarised by Pagès (2007) 36‒39. Panzer considers that the ἱστορίαι preserved only in bT derive also from the MH and had been omitted in the D-scholia. The specific cases are analysed in Pagès (2007) 16‒18. 12 Van Rossum-Steenbeek (1998) 86, 103‒108. 13 Montanari (1995) 148‒149, (2002) 135, Pagès (2007) 43‒44. 14 Pagès (2007) 48 confesses being perplexed at the lack of uniformity of the relationship to the lemmata in ἱστορίαι that are preserved in the same papyrus. He concludes that variety is in fact one of the characteristics of the MH, shared with Alexandrian and Imperial aesthetics (2007, 84). 15 Montanari (1995) 165‒168.

  Nereida Villagra and not in the transmission of the text.16 Van Rossum-Steenbeek cautiously accepts the learned origin of the MH, but warns against the evidence derived from ἱστορίαι preserved in scholia for they, she points out, could have been altered by copyists. She also acknowledges the similarities of the MH to mythographical works.17 Cameron proposes that, whatever the somewhat learned origins of the MH might have been, the MH as we have it is not a learned commentary but a mythographical work with “simpler goals”.18 Pagès proposes a double tradition with the transmission, on the one hand, of a learned and exegetical commentary (MH-A), and, on the other, of a relatively superficial commentary, a compilation of narratives of myths alluded to by Homer (MH-B).19

 Schol. Hom. Od. 12.69‒70: Phineus and the Argonauts That the narrative transmitted by the D-scholium to Homer Odyssey 12.69 did belong to the corpus of the MH is indisputable. It presents all the formal features considered to be typical of the MH: a lemma, a narrative text, and a subscription. It is preserved both in the D-scholia to the Odyssey and in some fragments of the famous papyrus PSI 10.1173, a third-century CE papyrus codex, of which several fragments preserve ἱστορίαι which comment on verses of the Odyssey comprised between Books 3 and 14.20 The medieval manuscripts which contain the ἱστορία are the Oxoniensis Bodleianus auct. V 1.51, from the tenth century, which preserves the D-scholia independently from the poetic text (Z in Ernstʼs edition), the Marcianus 613, a manuscript from the thirteenth century containing the text of the Odyssey with interlinear and marginal scholia (M), the Vindobonensis phil. gr. 56, from the fifteenth century, which contains also the text of the Odyssey with D-scholia mostly in agreement with M (Y), and a fifteenth-century manuscript, the Cantabrigensis Corp. Christ. Coll. 8, which contains the Homeric text and  16 Van der Valk (1963) 310‒312. 17 Van Rossum-Steenbeek’s arguments in favour of a learned origin are the format of hypomnema, the subscriptions, the cases of references to authorities, and the presence of obscure material. See (1998) 105‒108, 116‒117. 18 Cameron (2004) 104‒106. He accepts Haslamʼs opinion that MHʼs objective was to offer “its users more distraction than elucidation” (Haslam 1990, 30). 19 Pagès (2007) 117‒128. 20 See van Rossum-Steenbeek (1998) 99‒100. The first edition is Coppola (1932) 131‒132 (MP3 01209.000 / TM 61611 / LDAB 2760). See also Pfeiffer (1937) 1‒18.

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

scholia written by different hands. The scholia were copied by Demetrius Chalcondyles, who had consulted Eustathius.21 (1) Schol. Hom. Od. 12.69

5

10

15

20

25

οἵη δὴ κείνη γε παρέπλω ποντόπορος νηῦς / Ἀργὼ πᾶσι μέλουσα: Τυρὼ ἡ Σαλμωνέως ἔχουσα δύο παῖδας ἐκ Ποσειδῶνος, Νηλέα τε καὶ Πελίαν, ἔγημε Κρηθέα. καὶ ἴσχει παῖδας ἐξ αὐτοῦ τρεῖς, Αἴσονα καὶ Φέρητα καὶ Ἀμυθάονα. Αἴσονος δὲ καὶ Πολυμήλας καθʼ Ἡσίοδον γίνεται Ἰάσων, κατὰ δὲ Φερεκύδην ἐξ Ἀλκιμέδης. τελευτῶν δὲ οὗτος καταλείπει τοῦ παιδὸς ἐπίτροπον τὸν ἀδελφὸν Πελίαν, ἐγχειρίσας αὐτῷ καὶ τὴν βασιλείαν, ἵνα αὐξηθέντι τῷ υἱῷ παράσχῃ. ἡ δὲ τοῦ Ἰάσονος μήτηρ Ἀλκιμέδη δείσασα δίδωσιν αὐτὸν τρέφεσθαι Χείρωνι τῷ Κενταύρῳ. τραφεὶς δὲ καὶ ἡβήσας ἔρχεται εἰς Ἰωλκὸν ἀπαιτῶν τὴν πατρῴαν ἀρχὴν τὸν Πελίαν. ὁ δὲ ἔφασκε χρῆναι αὐτὸν πρότερον διακομίσαι τὸ χρυσοῦν δέρος ἀπὸ Κόλχων καὶ τοὺς πυριπνόους ἀνελεῖν ταύρους. ἀκούσας δὲ ταῦτα ὁ Ἰάσων λέγει τῷ Χείρωνι. καὶ αὐτῷ ὁ Χείρων συνεκπέμπει τοὺς ἠϊθέους. κατασκευάζει δὲ ἡ Ἀθηνᾶ τὴν Ἀργώ. πλέοντες δὲ ἀφικνοῦνται ἐπὶ τὴν τῶν Βιθυνῶν χώραν, καὶ ὁρῶσι τὸν Φινέα πηρὸν διὰ ταύτην τὴν αἰτίαν. ἔχων γὰρ παῖδας ἐκ Κλεοπάτρας τῆς Βορέου καὶ ἐπιγήμας Εὐρυτίαν δίδωσιν αὐτοὺς τῇ μητρυιᾷ διαβληθέντας πρὸς ἀναίρεσιν. Ζεὺς δὲ χαλεπήνας λέγει αὐτῷ πότερον βούλεται ἀποθανεῖν ἢ τυφλὸς γενέσθαι. ὁ δὲ αἱρεῖται μὴ ὁρᾶν τὸν ἥλιον. ἀγανακτήσας δὲ ὁ Ἥλιος Ἁρπυίας ἐπιπέμπει αὐτῷ, αἵτινες, εἴ ποτε μέλλοι αὐτοῦ διέπρασσον ἐμβάλλουσαι φθοράν τινα. καὶ οὕτω Φινεὺς ἐτιμωρεῖτο. τοῦτον οὖν τὸν Φινέα θεωρήσαντες οἱ περὶ τὸν Ἰάσονα παρεκάλουν ὑποθέσθαι πῶς δεῖ πλεῦσαι τὰς Πλαγκτὰς λεγομένας πέτρας. ὁ δὲ εἶπεν, ἐὰν τὰς Ἁρπυίας ἀπαλλάξωσι τῆς πρὸς αὐτὸν ὁρμῆς. θέντες δὲ συνθήκας ἐπαγγέλλονται αὐτῷ τοῦτο δρᾶσαι. ὁ δὲ λέγει αὐτοῖς, πόσον δύναται ἔχειν τάχος ἡ Ἀργώ; φάντων δὲ πελειάδος, ἐκέλευσεν ἀφεῖναι περιστερὰν κατὰ τὴν συμβολὴν τῶν πετρῶν, κἂν μὲν μεσολαβηθῇ, μὴ πλεῖν, ἐὰν δὲ σωθῇ, τότε περαίνειν τὸν πλοῦν. οἱ δὲ ταῦτα ἀκούσαντες ποιοῦσι. κατασχεθείσης δὲ τῆς περιστερᾶς διὰ τῆς οὐρᾶς, προσβάλλουσι τῇ Ἀργοῖ δυοῖν πληγάδες πέτραι συνελθοῦσαι τῆς νεὼς συμμύουσιν, αὐτοὶ δὲ σώζονται. οἱ δὲ Βορεάδαι Ζήτης καὶ Κάλαϊς ἀποδιώκουσι τὰς Ἁρπυίας ἀπὸ τῶν τοῦ Φινέως δείπνων. καὶ οὕτω παραγίνονται εἰς Κόλχους. ἡ δὲ ἱστορία παρὰ Ἀσκληπιάδῃ.

Fontes: ZMCY Testimonia: PSI 10.1173 frr. 5, 6r van Rossum-Steenbeek; Eust. 1712.17; Pherecyd. fr. 104c Fowler; Hes. fr. 38 M.–W.; schol. A.R. proleg. Ba Wendel. lemma οἴηι (sic) δὴ κείνη γʼ ἐπέπλω ποντόπορος ναῦς ἀργῶ πᾶσι μέλλουσα Z : nihil nisi ἀργὼ πασιμέλλουσα C : πασιμέλουσα Μ : οm. Y || 1‒11 Τυρὼ–χώραν : ἰάσονος πλέοντος μετὰ τῶν αὐτοῦ συντρόφων σὺν τῇ ἀργῶ νηὶ πρὸς τὸ κομῖσαι τὸ χρυσοῦν δέρας ἀπὸ κόλχων. καὶ φθασάντων ἐπὶ τὴν τῶν ἀθηνῶν χώραν hab. Y || 1 ἡ om. M || 2 ἔχουσα ZM : δύο παῖδας ἐκ Ποσειδῶνος τεκοῦσα C; Κρηθέα : κρηθέα τὸν καὶ ὅλου (τὸν Αἰόλου prop. Fowler) M; παῖδας post αὐτοῦ C || 3 καὶ (1) om. C; Πολυμήλας Barnes : πολυμίλας ZMC : Πολυμήδας uel -μήδης  21 Ernst (2006) viii‒xii. See also Pontani (2005), who uses the siglum Vo for the Oxoniensis, and who provides a fuller description of the history of the manuscripts. I follow Ernst’s edition of the text as basis. I choose a divergent reading in line 10 (ἠϊθέους). I add Fowler’s (2000) readings to Ernst’s apparatus.

  Nereida Villagra Sturz (1789) ex Apollod. 1.9.16 [107] || 3 καθʼ Ἡσίοδον : καθʼ Ἠρόδωρον Ruhnken (1823) 580 || 4 ἰάσων ZrecMC : ἰάσονι Ζ || 6 παράσχῃ : παράσχει C; τοῦ‒μήτηρ om. C || 6–7 τῷ Κενταύρῳ om. C || 8 χρῆναι αὐτὸν C, Barnes : ἐχρῆν αὐτὸν ZM : αὐτὸν δεῖν PSI 10.1173, fr. 5v, l. 127; δέρος : δέρας C; ἀπὸ ZM : ἐκ C, PSI 10.1173 fr. 5v, l. 129 || 8–9 καὶ‒ταύρους om. C || 9 ταῦτα om. M || 10 ἠϊθέους MC et PSI 10.1173 fr. 5v, l. 132 : ἡϊθέους Ζ : ἡμιθέους Ernst, corr. Barnes ab schol. Od. 14.533a || 11 τῶν om. C; Βιθυνῶν : Βυθινῶν M; Φινέα CYpc, Barnes : φηνέα ZMYac || 12 τῆς : τοῦ Μ; Βορέου MC : βορσοῦ Ζ : βόρσου Y; καὶ : εἶτα C; Εὐρυτίαν : εὐρυτέαν Y || 13 Ζεὺς δὲ χαλεπήνας : ἀγανακτήσαντος δὲ τοῦ διὸς ἐν τούτ ut. uid. Y || 15 Ἁρπυίας : ἁρπύας Y || 15‒23 αἵτινες‒συμμύουσι : aliter hab. C, coll. Eust. 1712.17 αἳ λιμώττειν ἐποίουν τὸν ἄνθρωπον ἀφαιροῦσαι δαινυμένου τὰ βρώματα· τοῦτον ἰδόντες οἱ περὶ τὸν Ἰάσονα καὶ σοφὸν εἶναι μαθόντες παρεκάλουν ὑποθέσθαι πῶς δεῖ πλεῦσαι τὰς πλαγκτὰς πέτρας· ὁ δὲ ἔφη ὑποθέσθαι εἰ τὰς ἁρπυίας ἀπελάσουσιν· ἐπαγγελλομένων δὲ τοῦτο ποιήσειν ἠρώτα ὁ φινεὺς ὅσον ἂν εἴη τὸ τάχος τῆς ἀργοῦς· φάντων δὲ ὅσον πελειάδος ἐκέλευσεν ἀφεῖναι περιστερὰν μέσον τῶν πετρῶν· κἂν μὲν μεσολαβηθῆ μὴ πλεῖν· ἐὰν δὲ σωθῆ τότε περαίνειν τὸν πλοῦν. οἱ δὲ τοῦτα (sic Ernst) ἀκούσαντες ποιοῦσι· κατασχεθείσης δὲ τῆς περιστερᾶς προσβάλλουσι τὴν ἀργῶ κατὰ μέσον τῶν συμπληγάδων· αἱ δὲ συμμύσασαι τὴν ἄκραν πρύμνην τῆς ἀργοῦς ἔκλασαν || 15‒17 αἵτινες‒Ἰάσονα : aliter Y τὴν καθημερινὴν διφθείρουσαι (-ειν s.l.) τροφήν· θεωρήσαντες οὖν οἱ περὶ τὸν Ἰάσονα τὸν φινέα || 15 post μέλλοι spatio relicto ca. 20 litt. Z et ἐσθίειν suppl. Zrec : μέλλοι κτείνασθαι spatio relicto M : μέλλει ἐσθίειν Asulanus : μέλλοι ἐσθίειν τὰ βρώματα Buttmann coll. Eust.; διέπρασσον ZM : διήρπαζον Buttmann : τὰ μὲν διήρπαζον τὰ δὲ καὶ διέφθειρον Jacoby; ἐμβάλλουσαι Ζ : ἐμβαλοῦσαι M || 17 δεῖ πλεῦσαι ZMCY : δεῖ πλεῦσαι Buttmann, Ernst : διαπλεύσωσι PSI 10.1173 fr. 6r, l. 139 : διεκπλευσοῦνται Eust. || 18–19 θέντες‒αὐτῷ : ὑπέσχοντο δὲ αὐτοὶ Y || 19‒20 λέγει‒πελειάδος om. Y || 22 lac. ind. Buttmann coll. Eust. : ὀργυιῶν ἄφλαστον καὶ τούτου τὰ ἄκρα περικόψασθαι καὶ Συμ‒ suppl. Jacoby || 23 αὐτοὶ δὲ σώζονται om. C; Βορεάδαι : βορίαδαι M || 24 καὶ hab. C, Barnes; post Κάλαϊς hab. C πτερωτοὶ ὄντες cf. Eust. || 24‒25 ἀπὸ‒Ἀσκληπιάδῃ om. Y || 25 ἀσκληπιάδῃ : ἀσκληπιάδην M

Translation by S. Asirvatham (2014) slightly adapted: ‘Only one seafaring ship has made it through, the Argo renown by all’: Tyro the daughter of Salmoneus, having two children from Poseidon — Neleus and Pelias — married Cretheus and had three children from him, Aeson, Pheres and Amythaon. According to Hesiod, Jason was the son of Aeson and Polymele, but according to Pherecydes he was the son of Aeson and Alcimede. When Aison died he left his brother Pelias as the guardian of the child, also entrusting the kingdom to him, in order that he give it to his son when he was grown. But Jason’s mother Alcimede, out of fear gave him to Chiron the Centaur to raise. When he had grown to be a young man, Jason went to Iolkos demanding his ancestral kingdom from Pelias. But Pelias said that he first had to bring the golden fleece from Colchis and to capture the fire-breathing bulls. Upon hearing this, Jason told Chiron, so Chiron sent some youths along with him, and Athena prepared the Argo. Sailing, they came to the land of the Bithynians, and saw Phineus, who had been blinded for this reason: having children by Cleopatra, the daughter of Boreas, and next marrying Eurytia, he gave them to their stepmother to be killed, because they had been traduced. Zeus was angry at this and asked him whether he wanted to die or be blind. And he chose not to see Helios, but Helios was offended and sent

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

the Harpies against him, who whenever he was about to eat anything would ruin it by depositing some kind of corruption upon it; and in this way Phineus was punished. So Jason’s crew, catching sight of this Phineus, asked for advice on how to sail through the so-called Planctae [‘Wandering’] Rocks. He said he would help if they put an end to the Harpies’ attacks. Making a pact with him, they promised to do this, and he said to him: ‘How much speed can the Argo attain?’ When they said, as much as a pigeon, he commanded them to release a pigeon as the rocks were coming together, and if it was caught mid-way, not to sail, but if it got safely through, then they could accomplish the passage. They did what they were told, and when the pigeon was caught by the tail, they attached a sternpiece two fathoms long to the Argo. The Symplegades clipped the edge of this as they came together; the ship passed through, the crew was saved, and the rocks closed for good. The Boreades Zetes and Cales chased the Harpies away from Phineus’ dinners. And so they came to Colchis. The story is told by Asclepiades.

(2) PSI 10.1173 van Rossum-Steenbeek Fr. 5r ll. 116‒120: … π]αρὰ Ἀσκληπιάδηι. ʼΟδ. μ. Ἀργὼ πασιμέλουσα παρʼ Ἀ[ιήταο. Ṭυρὼ ἡ Σαλμωνέως ἔχ̣[ουσα δύο παῖδ]ας ἐκ Ποσειδῶνο[ς, Νηλέα τε 120 καὶ Πελία]ν, ἐπέγημε Κ[ρηθέα ........ 116 ο\δ/ μ : Montanari (1995) 146 supplements Ὀδ(ύσσείας) μ || 117 παρʼ || 119 ποσιFr. 5v ........ ]ια̣δ[ πέμπ]ει παρὰ Χ[είρων]α[ τραφ]ε̣ὶς δὲ παρὰ τούτωι καὶ ἠ[ΐθε]ος γενόμενος ἦλθεν εἰς 125 Ἰω]λκὸν ἀπαιτῶν τὴν πατρώι]αν ἀρχὴν τὸν Πελίαν. ὁ [δ]ὲ[ ἔφασ]κεν αὐτὸν δεῖν πρότερον] κ̣ομίσαι τὸ χρυσοῦν δέρος ἐ]κ Κόλχων. ἀκούσας δὲ τα[ῦ130 τα ὁ] Ἰάσων λέγει τῷ Χείρωνι[˙ καὶ α]ὐτῷ ὁ Χείρων συνεξέ[πεμ]ψεν τοὺς ἠϊθέους. κατα[σκευάζ]ει δʼ ἡ Ἀθηνᾶ τὴν Ἀ[ργώ. πλέ]οντες δὲ ἀφικ̣[νοῦν135 ται πρὸς τὰς Πλα]γκτὰ[ς ........

  Nereida Villagra 122 Χ[είρων]α ⟦παρ[.]⟧|| 126 πελ\ι/ανo[ || 128 after ρον there is no space for δια] as supplemented in ed. pr. || 130 χειρο- || 131 ὁ add. ed. pr. || 132 ηϊθε- || 133 δʼ || 135 ]γʼκτα[ Fr. 6r ........ Φι]ν̣ εύς. [τοῦτον δὲ τὸν Φιν]έα ἰδόντε[ς οἱ περὶ τὸν Ἰάσο]να παρεκάλο[υν ὑποθέσθαι π]ῶς διαπλεύσωσι τὰ[ς Πλαγ140 κτ]ὰς πέτρας. ὁ δʼεἶπεν δ̣[ιηγήσασθ]αι, ἐὰν ἀποστήσωσι τὰ[ς Ἀρπ]υίας. θέντες δὲ [σ]υνθ[ήκας ἐπαγγέλλονται αὐτῶι το̣ ῦ[τ]ο δρᾶν ἐφʼ ὅσον δύνανται. [ὁ δὲ Φι145 νε]ὺς ἐν τῇ Ἀργοῖ κελεύε[ι φέρ]ειν πελειάδα καὶ α[ ........ 137 ϊδ- || 140 δʼ | δ̣[ιαπλεύσε- ed. pr. : δ̣[ιηγήσα- suppl. van Rossum-Steenbeek || 142 ]υϊας || 143 α̣τεπετελλονται : ἐπαγγέλλονται schol. || 144 letter crossed between first α and ν || 146 πελι-

Translation: fr. 5r in Asclepiades. Od. 12 ‘Argo, renowned by all, (sailing on her way) from Eetes’: Tyro, Salmoneus’ daughter, after having two children from Poseidon, Neleus and Pelias, was married to Cretheus … fr. 5v Sent to Cheiron … was raised by him and when he became a young man he went to Iolchus and demanded Pelias for the kingdom inherited from his father, but he said that he had to fetch the golden fleece from Colchis. After hearing that, Jason told Cheiron and Cheiron sent the youth with him. Athena prepared the ship Argo. After navigating they got to Colchis … fr. 6r Phineus, this … When Jason’s crew saw Phineus, they asked him how to navigate through the Clashing Rocks. He said he would explain it if they got rid of the Harpies. After they came to an agreement, they were told to do … as much as they could. [Phineus] told them to take a dove in the Argo …

The ἱστορία summarises the Argonautic expedition and pays special attention to the encounter of the Argonauts with Phineus and the causes of Phineusʼ punishment. The lemma which introduces the narrative text has been transmitted with some divergences. Manuscript Z gives verse 69 and the first half of verse 70 of

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

Book 12 of the Odyssey.22 Other manuscripts give only Ἀργὼ πασιμέλουσα or just πασιμέλουσα. Manuscript Y omits the lemma. The papyrus gives a greater part of verse 70.23 I shall summarise the contents of the ἱστορία, noting the divergences between the Byzantine manuscripts and the papyrus. Nearly the entire manuscript tradition (ZMC) begins with Jasonʼs genealogy. Hesiod and Pherecydes are quoted as authorities on variants of the name of Jasonʼs mother, who is Polymela according to Hesiod, and Alcymede according to Pherecydes. Then Aesonʼs death is mentioned, and the transfer of the control of the city to his brother Pelias until Jason comes of age. Then we hear about Alcymedeʼs fear of Pelias and her decision to send the baby Jason to be raised by Cheiron. Jason grows up and goes back to Iolcus to claim power. Pelias insists on the condition that he fetch the Golden Fleece and kill the Fire-Breathing Bulls. Jason relates this to Cheiron, who gathers the Argonauts, and the goddess Athena prepares the ship. The voyage begins, and they arrive at Bithynia, where they see Phineus being punished. Manuscript Y diverges in the first part of the narrative: it completely omits the genealogy and the antecedents of the expedition and starts the account with the voyage of the Argonauts and the fact that they were the first to arrive to the land of the Athenians (sic), where they found Phineus being punished. The papyrus breaks off in the middle of the genealogy. It is impossible to know whether it went on to quote the authorities Hesiod and Pherecydes. The text of the papyrus restarts again referring to the upbringing and education of Jason by Cheiron and lasts until the beginning of the navigation and the first stop of the Argonauts. The redaction is very similar to the one in the scholium with minor variations. Nevertheless, there are some divergences. Specifically, the papyrus omits the demand by Pelias that Jason kill the Fire-Breathing Bulls. The part of the papyrus corresponding to the first stop of the Argonauts is very badly preserved. The editors restore Πλαγκτὰς from γκτα. If we accept the restitution, the Argonauts got directly to the Symplegades. When the papyrus restarts, we read about the encounter with Phineus. All the Byzantine manuscripts include a story explaining the causes of Phineusʼ punishment. Ιt is introduced by διὰ ταύτην τὴν αἰτίαν and finishes with

 22 The Homeric verses are οἴη δὴ κείνηι (schol. κείνη) γε παρέπλω ποντοπόρος νηῦς/Ἀργὼ πᾶσι μέλουσα. Note that the scholium preserves different readings of the Homeric text. Cf. Westʼs app. ad 69‒70 (2017, 256): οἴηι for οἶη; κείνη appears also in Homer’s manuscripts and is corrected to κείνηι by Matthiae; γʼ ἐπέπλω for γε παρέπλω; πασιμέλουσα for πᾶσι μέλουσα. 23 Ἀργὼ πασιμέλουσα, παρʼ Αἰήταο, which is almost the complete verse. It lacks the last word, πλέουσα.

  Nereida Villagra καὶ οὕτω Φινεὺς ἐτιμωρεῖτο. The wording of lines 17 and 20‒22 is different in Manuscript Y, but there are no differences in the content. After the digression, the medieval tradition resumes the narrative with τοῦτον οὖν τὸν Φινέα θεωρήσαντες. The papyrus, after the lacuna, continues with Φι]ν̣ εύς ̣[τοῦτον δὲ τὸν Φιν]έα ἰδόντε[ς. According to both the medieval tradition and the papyrus, the Argonauts asked Phineus how to cross the Symplegades and he demanded the killing of the Harpies in return. They accept his proposal, and he tells them how to navigate using a dove. There is a difference between the papyrus and the scholia with regard to Phineusʼ answer: in the scholia he replies by asking how fast a ship Argo is. In the papyrus, this question is omitted and he directly advises them to take a dove with them. After that, the papyrus breaks off. The end of the ἱστορία and the subscriptio are omitted by Y.

 Reconstructing the MH: purpose and text The relationship of the story to the lemma is ambiguous. The narrative takes as its starting point a verse of the Odyssey which mentions the ship Argo (12.69) to recount an episode of the trip to Colchis, together with the antecedents of the Argonautic expedition — Jasonʼs education, childhood and origins. Thus the narrative starts by referring to Tyro, Jasonʼs grandmother.24 Surprisingly, the sole adventure in the entire Argonautic voyage mentioned here is the stop at Bithynia and the encounter with Phineus. The inclusion of this episode seems arbitrary in relation to the lemma, which focuses on the Argo itself in all the manuscript versions. The mythographer could have chosen any other episode of the rich Argonautic myth. However, if we take into account the wider context of the Homeric poem, we can see that several verses above, the Clashing Rocks (Symplegades) are mentioned. After the katabasis, Circe tells Odysseus about his future trip. She informs him that one of the two possible routes is to sail through the Clashing Rocks, which were very difficult to navigate through, and that the only ship which succeeded in doing so was ‘the Argo known to all’ (Od. 12.70 Ἀργὼ πᾶσι μέλουσα) on its way back from Aeetesʼ land. It is appropriate, then, that Phineusʼ episode is told, for he is the one who reveals to the Argonauts how to navigate through the Symplegades. Therefore, the ἱστορία is completing the reference to

 24 This allows it to supply in addition the genealogy of Pelias, a son of Poseidon, who plays an important role in the myth.

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

the navigation of the Argo through the Clashing Rocks, and not explaining the epithet πασιμέλουσα or the name of the ship. The relationship to the poetic text shows that, at least in this case, the function of the lemma, both in the papyrus and in the medieval tradition, was to serve as a guide to locate the passage of the poem which was commented on, and not necessarily to point to a specific term.25 The variations in the transmission of the lemma, thus, should be interpreted as a consequence of this relationship: in a codex with marginal scholia it is no longer necessary to copy the entire verse, but the reader still needs some indication of the passage under discussion. Since manuscript Z contains only scholia, the complete verse of the poem is copied. The Argonauts had to liberate Phineus from the Harpies in exchange for the information he divulges. The story on the causes of Phineusʼ punishment bears no direct connection to the Homeric poem, but is a development of the ἱστορία itself. This is why I indicated above that its relationship to the main text is ambiguous, for the general narrative is well connected to the broad passage of the poem, but the digression on Phineus bears no direct relation to it. Regarding the purpose of the commentary, in the Odyssey it is specified that the navigation through the Clashing Rocks occurred on the way back from Colchis. The lemma in the papyrus even quotes that detail from verse 70 (παρʼ Αἰήταο). However, the ἱστορία places the episode on the way to Colchis (line 30). Nevertheless, the MH does not point out that this is an alternative or conflicting variant. This suggests that the objective of the narrative is to complete the information of the poetic allusion and not to contrast its version with other treatments. The papyrus and the scholium preserve very close redactions to the ἱστορία. The differences between the two recensiones affect mainly the syntactic or lexical level but the final meaning of the texts is the same. However, there are differences which deserve some attention. There are two omissions in the papyrus: first, the fact that Pelias demanded that Jason kill the Fire-Breathing Bulls, and second, Phineusʼ question about how fast a ship the Argo was. This agrees with the general panorama of the MH, for the papyri tend to present a redaction shorter than the version in the D-scholia.26 It is possible, then, that the papyrus abridged the original text and deleted these parts. However, it is also possible that the scholium added the information and, therefore, the reference to the Fire-Breathing Bulls and Phineus’ inquiry about the shipʼs speed did not go back to the MH.  25 Cf. Schironi (2012) 409‒410. 26 Pagès (2007) 46‒49 considers that the papyri are abridged versions; contra cf. van RossumSteenbeek (1998) 103‒108.

  Nereida Villagra Regarding Peliasʼ demands, all other sources mention only the fetching of the Golden Fleece.27 The labour of ploughing with the Fire-Breathing Bulls is well known, but is usually described as a condition made by Aeetes, not by Pelias.28 The fact that it is a popular episode suggests that the detail could be a later addition by a commentator or scholiast, inserted in a sloppy way in the wrong sentence. One can even imagine that the addition could come from an interlinear note. However, we cannot exclude the possibility that Peliasʼ demand for Jason to kill the Fire-Breathing Bulls was already in the MH. The fact that the killing of the bulls is a demand by Pelias could be simply a consequence of condensing the Argonautic expedition to a short summary, precisely the work that the MH would have done. In fact, the narrative of the ἱστορία stops when the Argonauts get to Colchis and, therefore, it does not describe the labours which Jason has to confront there, which are only mentioned prospectively. In other words, the author of the summary — the MH — would have known that Jason killed the Fire-Breathing Bulls and fetched the Golden Fleece at Colchis. When he condensed all the information, he would have mentioned both deeds together at the beginning of the narrative, because he would not relate the labours as taking place at Colchis, but would finish the narrative when the Argonauts reach Colchis. Phineusʼ question about how fast a ship the Argo was, is not preserved in the papyrus. Apollodorusʼ version explains that Phineus advises them to use a dove but omits the question about the shipʼs speed. So does Apollonius of Rhodes.29 The question about the speed of the Argo seems a rationalising version of the use of the dove as an omen: if the Argo is as fast as a dove, then the Argonauts should use a dove to calculate if they will be able to cross the rocks. Does the MH show a tendency to rationalisation? It has been pointed out that, taking into account all the preserved material attributable to the MH to the Iliad, one can see some hints of rationalisation but no clear traces of it. Pagès considers that sometimes the rationalising tendency in the ἱστορίαι transmitted by scholia would fit

 27 Pi. P. 4.156‒167, Apollod. 1.8.12 [109]. A.R. 1.15‒18 omits the exact demand. 28 P. 4.224‒239. Schol. Pi. P. 4.414a describes the indications which Medea gave Jason to accomplish that proof. A.R. 3. 409‒421 Aeetesʼ request, 728‒739 Medea accepts the proposal to help Jason, 1026‒1062 Medeaʼs advice to Jason, 1247‒1407 Jason accomplishes the deed, Val. Fl. 7.58‒ 77 Aeetesʼ petition, 539‒653 Jason accomplishes the deed with Medea’s help, Apollod. 1.9.23 [126‒132], but Arg. Orph. 870‒873 does not refer the deed as Peliasʼ demand nor does it mention Medea’s help. 29 Apollod. 1.9.21‒22 [120‒125], A.R. 2.178‒341.

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

better the activity of a medieval commentator such as Eustathius who was re-telling an ancient myth, rather than the MH itself.30 Eustathiusʼ commentary on verse 70 of the Odyssey provides, conflated with other material, a text that parallels part of our ἱστορία, starting with the arrival of the Argonauts at Bithynia. His text also contains the question about Argoʼs speed. Its redaction presents some syntactic differences with the text in the scholium and in the papyrus, but it is clearly the same narrative, which suggests that Eustathius knew a version of the D-scholia similar to that in the Bodleian manuscript (Z). Therefore, Eustathiusʼ recensio should be regarded as an independent witness.31 As a consequence, the presence of the question about Argoʼs speed in two independent testimonies of the ἱστορία suggests, but does not prove, that it was originally found in the MH. In turn, if we accept that this element goes back to the MH, but that the rationalising tendency is a priori not a characteristic of MHʼs way of recounting myths, one wonders if the rationalistic twist could go back to its source. This is a complex discussion which boils down to the problem as to whether we consider the MH the work of an author or of a compiler.32 Indeed, in most cases it is impossible to compare the MHʼs ἱστορίαι to the version of the author quoted in the subscriptiones, because these are also non-preserved texts.33 Therefore, the boundaries between the features of the MH and its sources are extremely permeable. Two other elements in the text of the scholium relate to this discussion: the quoting of authorities on genealogical variants and the digression on Phineus. The papyrus does not preserve the corresponding part of the text for any of them. Therefore, it is uncertain whether these two elements appeared in the lost parts of the papyrus. Regarding the variants, they deal with Jasonʼs mother.34 Variants

 30 Pagès (2017) 213, 303. On Eustathius, see the volume by Pontani/Katsaros/Sarris (2017), esp. Paganiʼs contribution. 31 Other ἱστορίαι of the MH are also partially paralleled by Eustathius. See Pagès (2007) 24, 214. 32 Montanari synthesised in the following terms “il nocciolo del problema sta nel significato che diamo al fatto che i contenuti di queste historiae siano basati su opere precedenti” (Montanari 2002, 133). 33 On the subscriptions which could be comparable to some preserved material, see Lunstedt (1961). 34 The fact that the tradition gives different possibilities for the mother of a hero is a common feature in heroic genealogies. Genealogies are often given at the beginning of a ἱστορία. However, they are normally about the mythical figure under discussion, which is the departing point of the narrative and they do not include variants with their corresponding authorities. In our text, the lemma does not correspond to a mythical figure, but to a reference to the ship Argo. The genealogy starts with Tyro and advances until Jason, for he is the main figure of the expedition and Tyro is his grandmother.

  Nereida Villagra and the quotation of authorities inside a ἱστορία are rare in the MH and unparalleled in the ἱστορίαι found in the scholia to the Odyssey which have a correspondence with the PSI 1173. Only one scholium to the Odyssey considered to have belonged to the MH — but not paralleled in papyrus — preserves a similar feature.35 This is schol. Od. 15.16, which quotes Pherecydes as authority for a variant of the name of Icariusʼ wife. Pherecydes is also mentioned in the subscription.36 In our text, two authorities — Hesiod and Pherecydes — are quoted for alternative variants, but neither of them is mentioned by the subscription, which cites Asclepiades (most probably) of Tragilos. This is in turn puzzling because Asclepiades is the author of a work called Tragoidoumena, which must have dealt with tragedy, but none of the quoted authorities are tragedians. It has traditionally been considered that Asclepiades of Tragilos compared tragic myths to epic elaborations and former mythographical treatments.37 However, the evidence for this assumption is very scanty and open to debate.38 Thus, it is not only hard to determine whether the quotation belongs to the original MH or is a later addition, but also whether it could ultimately derive from the authority quoted in the subscription.39 As for the excursus on Phineus, it deals with the antecedents of the hero’s punishment. From a structural point of view, its content does not affect the relationship of the ἱστορία to the main poetic text, but such a long digression has no parallel in the other ἱστορίαι transmitted by papyri or by scholia which can be classified as MH.40 However, Eustathius does preserve the excursus on Phineusʼ punishment. It has a different redaction, although the contents and structure are the same.41 This does not prove per se that the digression goes back to the MH, but if we accept Eustathius as an independent testimony, it would tilt the scales in favour of its originality.  35 Some cases of the MH to the Iliad refer to authorities or even quote verses from poets (schol. Il. 1.5, 1.39, 1.59, 1.264, 2.157, 2.339, 5.392, 6.396, 10.435, 21.448). Only one case cites authorities in a genealogy: schol. Il. 1.7. 36 Van Rossum-Steenbeek (1998) 88 n. 11 gives also examples of references to anonymous variants in schol. Od. 7.324, 12.39, 19.518. 37 Robert (1873) 72‒76, Wilamowitz-Moellendorff (1875) 181‒183, Wendel (1935) 1353‒1354, Jacoby (1957) 484, Graf (1993) 193. On Asclepiades see Villagra (2009), Asirvatham (2014). 38 Villagra (2012) 289‒92, (2014) 27‒41. 39 Lunstedt (1961) concluded that the subscriptions were valid in the sense that they pointed to sources which contained part of the material in the ἱστορία or shared the subject matter. Van der Valk (1963‒1964) 303‒413 also reached a positive conclusion on the subscriptions. 40 Some longer narratives refer to the causes of a punishment or previous events, but they do so in a very brief and straightforward way: schol. Od. 11.271 (Ernst), on Oedipus schol. Od. 11.287 (Ernst), schol. Il. 6.155 (van Thiel) for digression, and schol. Il. 23.144 (van Thiel). 41 Eustathius adds some details.

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

Also, it must be stressed that in the redaction of the scholium there is an effort to identify and integrate the excursus on Phineus’ crime. Indeed, the text is introduced with the expression διὰ ταύτην τὴν αἰτίαν, which is reinforced by the γὰρ in the following sentence (lines 13‒14: καὶ ὁρῶσι τὸν Φινέα πηρὸν διὰ ταύτην τὴν αἰτίαν. ἔχων γὰρ …). More importantly, when the account of the advice of Phineus is resumed, the reader is reminded of the story, which was left unfinished, through repetition of the last sentence of the main story in an equivalent expression: τὸν Φινέα θεωρήσαντες. This is not a mechanical repetition, but a reformulation using a synonym. The connection to the main narrative line is also reinforced by the illative and deictic particles οὕτω, τοῦτον, οὖν (lines 19‒20: καὶ οὕτω Φινεὺς ἐτιμωρεῖτο. τοῦτον οὖν τὸν Φινέα θεωρήσαντες). This strongly suggests that one hand tried to harmonise the different stories. Does that hand correspond to the writer of the MH or to a later copyist? The conjunctive particle οὖν in line 20 of the scholium, which marks the fact that the narrative is picking up the interrupted story of the Argonauts, is missing in the papyrusʼ fragment. In this part the wording of the papyrus is heavily reconstructed by the editors, who do not include the particle in their reconstructions. If they are right, this may suggest that the papyrus version omitted the digression and this, in turn, can be interpreted as a sign that this part did not belong to the MH. Two lost plays by Sophocles and one more by Aeschylus were called after Phineus.42 This fits in well with the fact that the subscription of our ἱστορία cites Asclepiades of Tragilos.43 In turn, this speaks in favour of the possibility that the digression does belong to the MH, for he could have found all the information on Phineus in his source, Asclepiades, or in its re-elaboration.44 Aeschylus’ Phineus and the subsequent Phineus by Sophocles are believed to have represented the liberation of Phineus by the Boreads. This is the subject matter of the last part of our ἱστορία. We could then postulate that Asclepiades referred to one of these plays, or to both. We may imagine that the digression on Phineus could correspond to the tragic prologue of one of the pieces where the causes of Phineus’ punishment would have been presented.45 However, our knowledge about these fragmentary plays is very scanty and the plot of the pieces  42 Phineus A, B: TrGF 4 FF 645, 704‒705, TrGF 3 FF 258‒260. 43 See above n. 37. Even though Asclepiadesʼ Tragoidoumena is lost, indirect references to his work in other corpora of scholia provide positive arguments for the fact that Phineusʼ story was treated in this work: schol. A.R. 2.178‒182a (BNJ 12 F 22), 328, 562 (BNJ 12 F 2a–b). 44 It has been argued that the MH probably derived his ἱστορίαι from Homeric scholarship (hypomnemata or syngrammata), i.e. Mittelquelle, rather than the original quoted works, see Montanari (2002) 134‒135. 45 The accumulation of causes is typical in tragedy.

  Nereida Villagra is conjectural. There are also several other problems for this proposal: scholars think that Aeschylus did not set Phineusʼ story in the context of the Argonautic expedition, which is considered to be an innovation by Sophocles.46 But our text does not agree in some details with what we know about Sophocles’ tragedy.47 At the same time, we do not know if Asclepiades’ work summarised any of these tragedies, or what sort of information the MH would have found in his work. Be that as it may, the main problem we are confronted with is the fact that none of these arguments have a direct impact on the MH’s text or nature. We infer that the content is likely to come from the MH because the subscription points to Asclepiades, whose work dealt with tragedy, and we know that were tragedies on the same topic. However, this does not provide any solid evidence regarding the text of the MH itself. Moreover, as already noted, the subscription itself is most probably a vague way to point to a source which dealt with the same story, rather than a rigorous quotation. Lastly, Manuscript Y, which omits, as noted, the first part of the ἱστορία,48 offers an unusual variant in the route of the Argonauts, who here arrive first at the ‘land of the Athenians’, where they find Phineus. There are slight divergences even in the other sources on Phineus’ location, although he is always placed on one of the shores of the Black Sea: Apollodorus places him at Salmydessus in Thrace, on the opposite side of Bithynia, while Apollonius places the episode in Thynis, which is also in Thrace, and Eustathius situates him in in Bithynia. The surprising variant of Y is otherwise unattested. It is hard to explain it as a palaeographical mistake, but it could be a mistake introduced by the copyist of Y.49 From a critical point of view, the reading can be considered a banalisation, for

 46 Asclepiades would be the first testimony of the combination of both traditions. On Aeschylusʼ Phineus, see Deforge (1987) 30‒44, Stoessl (1937) 162ff. 47 According to a scholium on Apollonius Rhodius (schol. A.R. 2.178‒182b), Sophocles presented Phineus as being punished because, convinced by the lies of his second wife, he blinded the sons born to him by his first wife Cleopatra. This agrees with our text in the fact that there is violence against Phineusʼ children. But our text says that the stepmother plotted ‘their destruction’ generally, and does not refer specifically to the blinding of the children by their father. Also, the name of the second wife in the scholium on Apollonius is Dia, whereas in our text it is Eurytia. Therefore, our text cannot be a straightforward summary of Sophocles’ tragedy, but it could be the summary of Aeschylus, who would be responsible for the innovation of combining Phineus’ story with that of the Argonauts. Alternately, our text may be a re-elaboration of Asclepiades. 48 Eustathius also starts the narrative at this point. 49 Βιθυνῶν probably sounds similar to ἀθηνῶν due to the itacism; moreover ‘Athensʼ is a more common toponym than Bithynia.

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

Athens is a more common toponym than Bithynia. However, if we take into account the context of the Argonautic navigation, it would be the lectio difficilior, for the Argonauts do not travel south in any version. One wonders if this variant may be related to the fact that the first wife of Phineus, Cleopatra, was the granddaughter of Erechtheus and sister of the Boreads and, therefore, ‘the Athenian land’ in Y could be an old variant.50 The papyrus, as already mentioned, is mutilated in the part corresponding to the arrival of the Argonauts at Bithynia. According to the restitution — which, in my opinion, is highly probable — the expedition would first go through the Symplegades. Therefore, the papyrus and the version in the Bodleian manuscript of the scholia would also differ on the first stop of the Argonauts. Which was the original version of the MH? Could the Argonauts have tried unsuccessfully at first to pass through the Clashing Rocks, and thereafter have stopped at Bithynia? We cannot know if the papyrus subsequently made mention of Bithynia, Thynis, Salmydessus, or even Athens. However, given the proximity of the papyrus to the text of Z, and the fact that Eustathius also mentions Bithynia, one may expect that toponym as well.

 Conclusions Given the state and the nature of the preserved testimonies of the MH, reconstructing its ‘original’ text is a very uncertain operation. Indeed, this analysis has aroused several questions which are very hard to answer, since the various traditions preserve different stages of the text. Moreover, even though the ἱστορία of the Argonauts offers no doubt in its attribution, and even though the text of the papyrus and the text of the D-scholium are very close, it is not possible to reduce both versions to an archetype and establish which were the ‘original features’ of the text. For instance, the scholium quotes authorities and demonstrates a rationalising tendency, which the papyrus does not. However, we cannot assume that either of them is ‘original’, for both texts — the papyrus and the scholium — have been adapted to specific needs and audiences, all of which are irretrievable from the current state of the material.

 50 Orythia, Erechtheus’ daughter, is the mother of Cleopatra, Phineus’ first wife, and of the Boreads: schol. A.R. 1.211‒215c. In Valerius Flaccusʼ version, Phineus invokes this connection to get the help of the Boreads (4.460‒464). In other versions the Boreads are the ones who punish Phineus for blinding the sons of their sister Cleopatra (Arg. Orph. 671‒676).

  Nereida Villagra It is just this situation which is the most interesting aspect of such a text. We are not dealing with a work which was perceived as an authorial work in antiquity, as the text of an authority. As a matter of fact, the subscriptions show the need to provide each ἱστορία with a valid authority.51 On the other hand, the impulse to quote reliable, ancient authorities loomed large in Imperial times, and there are many cases where the author of a commentary or a mythographical text is known. For instance, we know that Dicaearchus wrote summaries to tragedies, and that Didymus wrote commentaries on Homer, Pindar, and Euripides. We also know that Apollodorus — whoever he was — wrote the Library. Therefore, it is no insignificant fact that no ancient work mentions the author of the compilation of ἱστορίαι which we call MH. This is an argumento ex silentio, but it still reveals that already in antiquity this text was perceived as an open text, and not as the work of an authority. As a consequence, such a work must have been subjected to a great deal of alteration in its transmission from the very beginning. We are not able to establish whether the changes to MH’s text operated when the text was copied in the preserved papyri, or if the ἱστορίαι were augmented in the medieval tradition. However, it is precisely the variations across the different traditions and testimonies that speak about the nature of the MH itself.

References Alganza Roldán, M. (2006), ‘La mitografía como género de la prosa helenística’, FlorIlib. 17, 9‒ 37. Asirvatham, S.R. (2014), ‘Asklepiades of Tragilos (12)’, in I. Worthington (ed.), Brill’s New Jacoby (Brill Online Reference Works). BNJ = Worthington, I. (ed.), Brillʼs New Jacoby, Brill Online Reference Works. (https://referenceworks.brillonline.com/browse/brill‒s‒new‒jacoby). Brown, M.K. (2002), The Narratives of Konon. Text, Translation and Commentary on the Diegeseis, Munich/Leipzig. Cameron, A. (2004), Greek Mythography in the Roman World, Oxford/New York. Coppola, G. (1932), ‘1173. Schol. Hom. Odyss.’, Papiri greci e latini (Pubblicazioni della Società Italiana per la ricerca dei Papiri greci e latini in Egitto) 10.2 (‘ni. 1163–1181’), 131–140. Cuartero, F.J. (2010), Pseudo-Apollodor. Biblioteca, vol. 1, Barcelona. Cuartero, F.J. (2012), Pseudo-Apollodor. Biblioteca, vol. 2, Barcelona. Deforge, B. (1987), ‘Eschyle et la légende des Argonautes’, REG 100, 30‒44. Delattre, Ch. (2016), ‘Référence et corpus dans les pratiques de commentaire. Les emplois de historia’, Revue de Philologie, de Littérature et d’Histoire Anciennes 90.2, 89–110.

 51 A different interpretation of the subscriptions has recently been proposed by Delattre (2016).

Mythographus Homericus, Ἱστορίαι and Fragmentary Mythographers  

Ernst, N. (2006), Die D-Scholien zur Odyssee. Kritische Ausgabe, Köln (https://kups.ub.uni‒ koeln.de/id/eprint/1831). Fowler, R.L. (2000), Early Greek Mythography. I. Texts, Oxford/New York. Graf, F. (1993), Greek Mythology. An Introduction, Baltimore. Henrichs, A. (1987), ‘Three Approaches to Greek Mythography’, in J. Bremmer (ed.), Interpretations of Greek Mythology, London/Sidney, 242–277. Haslam, M.W. (1990), ‘A New Papyrus of the Mythographus Homericus’, BASP 27.1, 31‒36. Higbie, C. (2007), ‘Hellenistic Mythographers’, in R.D. Woodard et al. (eds), The Cambridge Companion to Greek Mythology, Buffalo, 237‒254. Jacoby, F. (1957), Die Fragmente der griechischen Historiker. Erster Teil. Genealogie und Mythographie. Kommentar, Leiden. Lightfoot, J.L. (1999), Parthenius of Nicaea: The poetical fragments and the Erotica Pathemata, Oxford. Lunstedt, P. (1961), Untersuchungen zu den mythologischen Abschnitten der D-Scholien, diss. Hamburg. Meccariello, C. (2014), Le hypotheseis narrative dei drammi euripidei. Testo, contesto, fortuna, Rome. Meliadò, C. (2015), ‘Mythography’, in F. Montanari/S. Matthaios/A. Rengakos (eds), Brillʼs Companion to Ancient Scholarship, 2 vols, Leiden, 1057‒1089. Montanari, F. (1995), ‘The Mythographus Homericus’, in J.G.J. Abbenes/S.R. Slings/I. Sluiter (eds), Greek literary theory after Aristotle: A collection of papers in honour of D.M. Schenkeveld, Amsterdam, 135–172. Montanari, F. (2002), ‘Ancora sul Mythographus Homericus (e l’Odissea)’, in A. Hurst/F. Létoublon (eds), La mythologie et l’Odyssée. Hommage à Gabriel Germain. Actes du colloque international de Grenoble 20–22 mai 1999, Geneva, 129–144. Pagani, L. (2017), ‘Eustathius’ Use of Ancient Scholarship in his Commentary on the Iliad: Some Remarks’, in F. Pontani/V. Katsaros/V. Sarris (eds), Reading Eustathios of Thessalonike, Berlin/Boston, 79‒110. Pagès, J. (2007), Mythographus Homericus. Estudi i edició comentada, diss. Universitat Autònoma de Barcelona. Panzer, J. (1892), De Mythographo Homerico restituendo, diss. Greifswald. Papathomopoulos, M. (1968), Antoninus Liberalis. Les Métamorphoses, Paris. Pellizer, E. (1993), ‘La mitografía’, in G. Cambiano et al. (eds), Lo spazio letterario della Grecia antica, volume I. La produzione e la circolazione del testo, tomo II. LʼEllenismo, Rome, 283‒303. Pfeiffer, R. (1937), ‘Hesiodisches und Homerisches’, Philologus 92, 1–18. Pontani, F. (2005), Sguardi su Ulisse. La tradizione esegetica greca all’Odissea, Rome. Pontani, F./Katsaros, V./Sarris, V. (eds) (2017), Reading Eustathios of Thessalonike, Berlin/Boston. Robert, C. (1873), De Apollodori Bibliotheca, Berlin. Ruhnken, D. (1823), Opuscula. Tomus secundus, Leiden. Schironi, F. (2012), ‘Greek Commentaries’, Dead Sea Discoveries 19.3, 399‒441. Stoessl, F. (1937), Die Trilogie des Aischylos. Formgesetze und Wege der Rekonstruktion, Baden bei Wien. Sturz, F.W. (1789), Pherecydis fragmenta, Thüringen. TrGF 3 = Radt, S. (1985), Tragicorum Graecorum Fragmenta, vol. 3, Aeschylus, Göttingen.

  Nereida Villagra TrGF 4 = Radt, S. (1977, 19992), Tragicorum Graecorum Fragmenta, vol. 4, Sophocles, Göttingen. Trzaskoma, S.M./Smith, R.S. (eds) (2013), Writing Myth. Mythography in the Ancient World, Leuven. Van der Valk, M. (1963), Researches on the Text and Scholia of the Iliad. Part 1, Leiden. Van Rossum-Steenbeek, M. (1998), Greek Readersʼ Digests? Studies on a Selection of Subliterary Papyri, Leiden. Vassilaki, E. (2015), ‘Entre histoire et légende : recherche sur les emplois des mots ἱστορία, ἱστορεῖν, ἱστοριογράφος et ἱστορικός dans les scholies aux Olympiques de Pindare’, in C. Poulle-Muckenstrum/S. David-Guignard (eds), Traduire les scholies de Pindare II. Interprétation, histoire, spectacle, Dialogues d’Histoire Ancienne. Supplément 13, 93–117. Villagra, N. (2009 [2008]), ‘Los Τραγῳδούμενα de Asclepíades de Tragilo: una obra mitográfica’, Faventia 30.1–2, 285–295. Villagra, N. (2012), ‘Commenting on Asclepiades of Tragilos: Methological Considerations on a Fragmentary Mythographer’, in A. Castro et al. (eds), Estudiar el pasado: aspectos metodológicos de la investigación en Ciencias de la Antigüedad y de la Edad Media. Proceedings of the First Postgraduate Conference on Studies of Antiquity and Middle Ages, Universitat Autònoma de Barcelona 26‒28th October 2010, BAR Series 2412, 289‒292. Villagra, N. (2014), ‘Fragmentary Mythography as a Source: Neoptolemos at Delphi in the Tragodumena’, in M. Carrive/M.-A. Le Guennec/L. Rossi (eds), Aux sources de la Méditerranée antique, Aix-en-Provence, 27‒42. Wendel, C. (1935), ‘Mythographie’, RE 16.2, cols 1353‒1374. West, M.L. (2017), Odyssea, Berlin/Boston (Mass.). Wilamowitz-Moellendorff, U. von (1875), Analecta Euripidea, Berlin.

Giulio Iovine

The Unruly Fragments Old Problems and New Perspectives in Latin Military Papyri from Dura-Europos (P. Dura 56, 64, 72, 74, 76, 89, 113) This paper presents some of the preliminary results derived from a papyrological reassessment, through direct inspection, of the so-called P. Dura 55 to 145:1 namely, the Latin military papyri from the archive of the cohors XX Palmyrenorum. It is the only paper in this volume not dealing with literary texts; instead, it enters the realm of documents, populated by lists, private letters, contracts and testaments, notices, subpoenae, loans, permits and records, and a thousand further typologies. These two sets of texts (literary and documentary) are usually kept well separated by papyrologists, who generally prefer focusing on the latter ones, leaving the former to scholars working in the fields of Greek and Latin literature.2 This separation mirrors a strong difference in aim which exists between, for instance, a tragedy of Aeschylus and a list of auxiliary soldiers from a Roman cohort drafted in the same day and in the same Egyptian town. The former text is copied to be perused, annotated, and enjoyed by the next generations of readers and scholars; the latter will be discarded in the moment it becomes obsolete, and  The research leading to these results has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement no. 636983); ERC-PLATINUM project, University of Naples ‘Federico II’. I wish to thank Francesco Lupi and Francesco Ginelli of the University of Verona and Milan for admitting a preliminary version of this chapter to their panel at the 10th Celtic Conference in Classics (Montreal, July 2017) and for including the full version in this book; I also thank R. Matera of Beloit College for a vital check on my English, and finally, for their constant encouragement and support, the whole staff of the Beinecke Rare Book and Manuscript Library at Yale University. I wish to credit the Yale Papyrus Collection, Beinecke Rare Book and Manuscript Library, as the source of the images in this paper. All images are accessible via the search engine of the Library’s website: https://brbl-dl.library.yale.edu/vufind/Search.  1 I prefer not to dwell in this paper on P. Dura 54, the Feriale Duranum (TM 44772), as there is no real reason to attribute it to the military milieu of the town, and its complexity would require a whole paper, rather than a section of it. Cf. Reeves (2005) for a detailed analysis. 2 See e.g. Vandoni (1966) 1: “[dalla papirologia] sono esclusi (dopo la trascrizione e la prima pubblicazione, compito del papirologo) i papiri letterari, campo di studio più specificamente filologico, e i papiri scritti in lingue orientali (demotico, copto, aramaico, ebraico, siriaco, persiano, arabo) … Il compito principale del papirologo si può definire come lo studio e la pubblicazione dei testi papiracei”. Exceptions, of course, do occur. https://doi.org/10.1515/9783110712223-009

  Giulio Iovine military lists become outdated at a very quick pace. The works dealt with in the other papers of this volume were meant to outlast their times; the documents referred to throughout this chapter were discarded as soon as their original context ceased to exist. That said, something can be pointed out with regard to what these documents share with literary fragments. First, they are in Latin: if they do not enter the history of Latin literature, they heavily contribute to the history of Latin language, which in turn is a priceless tool for literary history. The language of a document teaches us something whether it is very distant or very close to the language of literature.3 Secondly, they are on papyrus. Papyri are the main — or the only — source for some of the most important collections of fragments from Greek literature; all the problems typically connected with poetry or prose fragments (erasures, damage to the fibres, cuts and lacunae, sequences of wormholes and recurring damages, dirt to be removed) are the same problems one faces when dealing with any documentary papyrus, let alone the Dura papyri. This is the obvious consequence of the fact that texts on papyrus were discarded in the same place when no longer of use.4 Since the methodology is the same, sharing a single individual’s experience on some specific fragments with other scholars potentially means that further knowledge can be produced and new frontiers opened, regardless of the differences in content between the scrutinised texts. The cohors XX Palmyrenorum I mentioned at the beginning of this paragraph was an auxiliary cohort of Roman soldiers, of Palmyrene first, then probably Durene extraction, who were garrisoned in the north portion of the town of Dura, a settlement on the right bank of the Euphrates re-founded as a Seleucid colony and heavily Hellenised. Romans came there with Lucius Verus (ca. 165 CE) and were forced to go after Shapur I’s campaigns in the East: the final siege happened in 256/257 CE.5 This long occupation made Dura a pivotal element in the Roman-

 3 Linguistics and sociolinguistics can draw (and have drawn) several important conclusions both from the letter of Terentianus and Tiberianus and their substandard second-century-AD Latin (e.g. P. Mich. VIII 468, TM 27081 and 471, TM 27084, 100–125 CE) and from the flourishing sentences of Latin documents from Late Antique Egypt (e.g. P. Ryl. IV 609, 505 CE, TM 17309), which mirror the products of coeval prose writers (e.g. Ambrosius, Augustinus) and the laws collected in the Digestum and the Theodosian Code. On the weakening distinction between ‘documentary’ and ‘literary’ in texts from Late Antique Egypt, see Fournet (2017). 4 One of the most famous dumping places for papyri in the history of papyrology is Oxyrhynchus: see the monumental book by Bowman/Coles/Gonis/Obbink/Parsons (2007). 5 A good summary, with extensive bibliography, is in Austin (2010) esp. 17–32. See also Welles/ Fink/Gilliam (1959) 24, Sommer (2005) 311–312, and the (now indispensable) James (2019). More on the final chronology of Dura in James (1985), MacDonald (1986), James (2011).

The Unruly Fragments  

Parthian (later Sasanian) border, and after the siege, the settlement was abandoned. Like all other units in Roman army, the Palmyrene cohort had an office and an archive, where documents referring to its everyday life and operations were received, produced, and if necessary, stored. During the final siege, Romans resorted to reinforcing the north wall of the town with ramparts in order to face the incoming Sasanian army. They vacated all the rooms adjacent to that wall, and threw in the fill every available object, including dismissed documents which they did not feel the need of keeping. It was in one of those rooms, room W13 in the Temple of Artemis Azzanathkona, that fragments from more than a hundred papyri resurfaced in the 1920–30s, when the buried town was excavated by a joint archaeological mission involving Yale University and the Académie Française des Inscriptions et Belles-Lettres.6 A considerable quantity of the papyri and parchments found in room W13 (which constitute the greatest majority of papyri and parchments found in Dura-Europos) were in Latin language, and referred to the cohors XX Palmyrenorum, proofs of whose existence are to be found only in the manuscripts themselves, and in the inscriptions excavated in the city. The Latin subset of this unique collection has benefited from the editorial attention of J.F. Gilliam, R.O. Fink, C.B. Welles,7 R. Marichal8 and, partially, P. Cugusi.9 The last full edition (Marichal’s) dates back to the 1970s, but the papyri themselves were inspected by Marichal in the 1960s, and he subsequently worked with photographs.10 The time has come for an overhaul, especially from the textual point of view. I have been privileged to be able to inspect the manuscripts

 6 Hopkins (1979) esp. 75–105 for the papyri findings. 7 Welles/Fink/Gilliam (1959), Fink (1971). The former book collects all Durene papyri, not just the Latin ones; Fink’s book is devoted to documents coming from Roman army, therefore not just from Dura and also in other languages than Latin. 8 Bruckner/Marichal et al. (1954–). The Latin Dura papyri (and a small bunch of Greek ones from the same milieu) were published by Marichal in volumes 6, 7, 8, and 9. 9 In his CEL, P. Cugusi has re-published, often with personal supplements and readings, all the official correspondence in the Durene collection (P. Dura 55–80). 10 “En 1966, lorsque j’ai collationné les P. Dura à l’Université de Yale, ceux qui étaient déjà sous verre se trouvaient à la Beinecke Rare Book and Manuscript Library, ceux qui étaient encore sous dossier se trouvaient à Winchester Hall dans le bureau de C. B. Welles. Ils doivent tout trouver maintenant à la Beinecke et sont probablement montés sous verre. Mes descriptions donnent donc l’état de la collection tel qu’il était à la fin de 1966. Certaines photographies avaient été prises antérieurement à un moment où la disposition des fragments était différente; la plupart d’entre elles ont été refaites; dans certaines cas je les ai conservées telles quelles, à titre documentaire, en indiquant dans le commentaire ou les notes les modifications apportées et le plus souvent retenues dans l’édition” (R. Marichal in ChLA VI, p. v).

  Giulio Iovine directly over a six-month period (April to September 2017) at the Beinecke Library in New Haven; I was provided with two brand new tools, an electronic microscope and high-quality digital images.11 While addressing a papyrological archive which has been left untouched for forty years, the chief and fondest hope of the papyrologist is to find unpublished pieces. Be it a scrap resurrected from the stacks of the library or reassigned to an already known papyrus, or a larger piece simply unheeded by former publishers, one always welcomes fresh additions to one’s editorial toil, bearing in mind that in most cases, when previous editors leave something behind, there usually is a good reason for this. Secondly, the papyrologist will take pains to fix any mistake or confusion created by previous publications, such as different numberings of the fragments and consequent confusion for the reader, and will be on the lookout for fragments which have vanished from one edition, and which have reappeared in another, or which have not reappeared at all. The final and probably most feasible aim is to improve the given text with a fresh look, direct inspection, and new tools. Within this framework, I shall present some of my results.

 P. Dura 5612 contains fragments from at least three official letters issued by the chancery of Marius Maximus, governor of Syria.13 They are dated by editors around 208 CE.14 The letters were sent from Maximus to the tribune of the Palmyrene cohort, an otherwise unknown Ulpius Valentinus. After the addressee read them, they were pasted together by a clerk in a probably pre-existing τόμος συγκολλήσιμος, a liber epistularum (‘register of correspondence’) in which all letters concerning horses or cavalrymen of cohort were probably preserved. Traces of this inclusion in the cohort’s liber epistularum are in the commonly attested formula acc(epta) ‘received’, and then the dating formula, septimum decimum [K]al(endas) A[p]ril[es d(omino) n(ostro) Imp(eratore) A]ntonino Aug(usto) I[II et

 11 For the images my gratitude must go to M. Custer and E. Doon, both working at the Beinecke Library, who have been most helpful and accommodating. 12 The first edition is Gilliam (1950) 171–189; then the others follow: Welles/Fink/Gilliam (1959) (P. Dura 56), Fink (RMRP 99), Marichal (ChLA VI 311), and Cugusi (CEL I 179). TM 44776. 13 For Marius Maximus and his rich cursus honorum, see Miltner (1930). 14 Dates are seen in the texts of the three letters; the most complete one (208 CE) is in letter A, ll. 7–9 ex die [qu]ar[to ̣ ̣ ̣ ] | Iuniạṣ [d n Imp Antonino A]ụg· III [e]ṭ ⟦et[a] | Caesare it[e]r[um⟧ cos].

The Unruly Fragments  

⟦Geta Caesare II15⟧ co(n)s(ulibus)], which can be found in the blank space between the sender’s and addressee’s names and salutem (fr. A, ll. 1–2).16 In fr. A, above l. 1 and in the right section of the upper margin, a few small letters can be read, which Marichal identified as a docket added later to the letter, but which he was unable to read.17

Fig. 1: P. Dura 56: Docket in fr. A, above l. 1.

Parallels for a docket inserted in the upper margin of an official letter can be seen elsewhere in the Dura papyri: 1. P. Dura 63,18 two letters addressed to a Saturninus: an otherwise unknown Agathonius, again in the dative case, is added in the upper margin (] Agathonio). He is, in fact, the ultimate addressee of the letters themselves, which were forwarded to him (his name appears in the verso), after having been sent to Saturninus; 2. P. Dura 66,19 letter f — the letter is addressed to somebody we cannot know for certain, but a docket is inserted above, with the name of the cohort’s commanding officer, Postumius Aurelianus (]tumio), in the dative case; 3. Eventually, P. Dura 71,20 addressed to some Artemidorus, has a small name in the dative case drawn in the upper margin, ]elio.

 15 The erasure of Geta’s name from this inserted formula, which cannot have taken place before 212 CE, is conjectural: one infers it from a comparable erasure in ll. 7–9. 16 The archival annotations ‘accepta + dating formula’ in Latin documents on papyrus and ostrakon are further scrutinised in Iovine (2019). 17 “Traces très incertaines … peut-être ajouté lors de l’enregistrement et du collage des lettres” (ChLA VI, p. 15). 18 211 CE, TM 44791. 19 216 CE, TM 44795. 20 219 CE, TM 44800.

  Giulio Iovine

Fig. 2: P. Dura 71: Docket in the upper margin.

Fig. 3: P. Dura 66 f: Docket in the upper margin.

Fig. 4: P. Dura 63: Docket in the upper margin.

Given that in two out of three cases (P. Dura 63, 66), the added name is the name of the addressee (either the original or a later one), we are allowed to guess that the annotation in the upper margin of P. Dura 56 is in fact, a name, that it is in the dative case, and that it could also be the name of one of the letter’s addressees. Before i, which is quite certain, one might see a rather squarely drawn p, and after i an o with (perhaps) a protruding acute accent.21 The first visible letter, on

 21 Or so it seems. Evidence of acute accents in Latin literary texts and letters on papyrus is quite attested; not so for Latin documentary papyri, where acute accents are rarely seen (for a full overview, cf. Scappaticcio 2012). It might as well be the case that the protruding stroke of this o

The Unruly Fragments  

the other hand, consists of the left edge of a circular stroke; only weak spots of ink are to be seen after it. Perhaps one is here witnessing the annotation in the dative case Ụḷpió. Needless to say, this might be a remark concerning the addressee, Ulpius Valentinus. Little do we know of such customs and it seems impossible to clarify the purpose of these annotations; it is perhaps safer to assume they were done by the clerk inserting the letter in the liber, for internal purposes. This papyrus also offers occasions to rectify previous editions. As often happens with Dura papyri, the exact number and disposition of the fragments in P. Dura 56 has undergone many fluctuations in subsequent editions. According to the editor princeps Gilliam, this particular papyrus included about forty fragments.22 Most of them had probably been rejoined before the princeps, as in the 1950 contribution only twelve fragments appear, three of which (a, b and c) offer some articulated text. Gilliam’s further edition in 1959 and Marichal’s in 1976 have altered the order and names of the fragments due to rejoinings and removals. A further reorganisation, which can be seen from the online images, has been done by the German papyrologist R. Duttenhöfer in the early 1990s.23 The table given below provides a record of the alterations. Tab. 3: P. Dura 56: Ordering the fragments. Gilliam 

Gilliam 

Marichal

Online photos

fr. a

fr. a

fr. a

fr. A

fr. b

fr. b (= b + d + e)

fr. b

fr. B (= b + d + e)

fr. c

fr. c (= c + f + k + l)

fr. c

fr. C (= c + f + k + l)

fr. d (e]x ui Kal [)

joins fr. b

removed from fr. b, becomes fr. h

joins fr. B

fr. e (]ṃo cos)

joins fr. b (Apro et Maxi]ṃo cos)

removed from fr. b, becomes fr. i

joins fr. B

fr. f (] ̣ ̣ [)

joins fr. c (iṇ [)

idem

joins fr. C

fr. g (]Ṃạriuṣ[)

becomes fr. d

idem

idem

 is nothing but the artificial continuation of the single stroke which was necessary to draw o itself. A similar phenomenon occurs in P. Dura 63, where the o’s of Agathonio, Saturnino and suo (letter a ll. 1–2), as well as the d’s in letter a l. 4 (ad) and 7 (uendidisse) all feature an ornamental continuation of the stroke employed to draw the circular portion of the letter. 22 Gilliam (1950) 173. 23 I have obtained this piece of information from Dr Duttenhöfer herself, to whom I am very grateful.

  Giulio Iovine

Gilliam 

Gilliam 

Marichal

Online photos

(M]ạrius [Maximus) fr. h (] ̣ [)

becomes fr. e (]ị[ ̣ ]ọ[)

idem

idem

fr. i (] ̣ d ̣ [)

becomes fr. g

idem

idem

fr. j (]ẹta ̣ ̣ [)

becomes fr. f

idem

idem

fr. k (] ṃo ̣ [)

joins fr. c (] mọs ̣ [)

idem

joins fr. C

fr. l (]al[)

joins fr. c (K]ạl ̣ [)

idem

joins fr. C

-

-

fr. m (blank)

fr. h (blank)

As far as the disposition of the fragments in the original roll is concerned, their shape gives us very strong clues, which did not escape scholars. Fr. A (here featuring with its right portion) shows a large triangular scrap with two vertical, narrow holes—and below it a smaller, rounder triangle with two round tops. The same shape can be seen in frr. B and C, smaller and smaller. As Gilliam thought, frr. B and C followed A in the original roll (it is unlikely that they preceded it, as the left margin is clearly visible in letter A).

Fig. 5: Fr. A.

The Unruly Fragments  

Fig. 6: Fr. B.

Fig. 7: Fr. C.

In his edition, Marichal has removed the original fr. d and e from B and renamed them h and i, but there is no need for that, as their shapes allow us to connect them to the larger fragment; the current disposition of the pieces in the glass frames at the Beinecke Library and Gilliam and Fink’s editions clearly take this into account. By keeping a close association between the main fragments and the smaller scraps, we can see in fr. B = letter B, ll. 7–8 — together with Gilliam and Fink — the dating formula of the relatio in acta, [± 11 e]x VI Kal(endas) [ ̣ ̣ ̣ ̣ Apro et Maxi]ṃo co(n)s(ulibus), i.e. 207 CE. The fact that this formula precedes the one given in fr. A = letter A, ll. 7–9 ex die [qu]ar[to ̣ ̣ ̣ ] | Iuniạṣ [ domino nostro Imperatore Antonino A]ụg· III [e]ṭ ⟦et[a] | Caesare it[e]r[um⟧ co(n)s(ulibus)], i.e. 208 CE, and the very

  Giulio Iovine likely chronological order observed by libri epistularum,24 create a chronological issue, as one supposes the letters in the liber to have been pasted together in chronological order. This is what convinced Marichal to separate the scraps from letter B.25 However, the shapes of the scraps are a very strong clue. One might follow Gilliam,26 and accept that in this particular circumstance the letter B, despite following letter A in the liber, predates it in composition; that the scribe, in short, failed to respect the expected chronological order in the roll he was patching together. Marichal also separated (or kept apart) what is now fr. h in the online images, and m in his own edition, from the main fr. B: its shape is clue enough for us to reconnect the two fragments, filling the only missing portion of the picture.

Fig. 8: Fr. B with smaller scraps. Marichal removes them from fr. B, but they can keep their position

Fig. 9: Fr. h (m in Marichal’s edition).

 24 Welles/Fink/Gilliam (1959) 219. 25 Cf. ChLA VI, p. 19, particularly “mais rien, semble-t-il, n’oblige à considérer h et i comme appartenant au frag. B”. 26 “Unless this letter is in fact earlier than A and was written in 207, the effective date for the records is earlier than that of the letter by at least several months. I have assumed that the letters were filed in chronological order and were added to the right edge of the roll” (Welles/ Fink/ Gilliam 1959, 219).

The Unruly Fragments  

To conclude, it is perhaps safer not to turn upside down, as Marichal does, fr. l from Gilliam’s first edition, now joining fr. C as its smallest scrap. Firstly, the shapes of the preceding scraps clearly show the top of the scrap to be tip-like and sharp, not blunt; secondly, if we accept Marichal’s alteration, the resulting kappa appears in fact to be oddly written, as if it were dislodged.

Fig. 10a–b: a) Current status: ] ḳal ̣ [.

b) Marichal’s edition (upside down): e]x̣ Kạl[.

 J.F. Gilliam and R.O. Fink, the first editors of P. Dura 64,27 which contains, as well as P. Dura 56, official correspondence of the cohort, counted in it two fragments: a (containing letter A) and b (containing letter B). Gilliam in 1957 and 1959 and Fink in 1971 only have frr. a, b. But if one looks at the frames today, there are actually six fragments.

 27 First edition in Gilliam (1957) 49–62, then Welles/Fink/Gilliam (1959) (P. Dura 64), Fink (RMRP 99), Marichal (ChLA VI 319), and Cugusi (CEL I 199). TM 44792.

  Giulio Iovine

Fig. 11: Letter A: The smaller scraps.

Fig. 12: Letter B: The smaller scrap.

In the frame containing letter A three smaller scraps have apparently been added, b, c and d; and in the frame containing letter B, another tiny scrap can now be found (b). These are probably the very same scraps that Gilliam thought of no consequence enough to be discussed in his first editions, and which have been ignored since.28 The most recent edition, produced by Marichal for the Chartae

 28 “Some small fragments, containing nothing of consequence, may belong to these letters or at any rate to the same roll. They will not be discussed here” (Gilliam 1957, 49).

The Unruly Fragments  

Latinae Antiquiores, features only fr. d. Frr. b, c and the other b remained unpublished. I have taken the liberty of publishing them here: recto, letter A, fr. b --[- - - dumta-]|1-xat se ̣ ̣ ̣ [- - -] |2 ueṣẹ ̣ [- - -] --recto, letter A, fr. b: 1 bottom of an oblique stroke, pointing upwards | bottom of P, B, less likely D, and dot-like trace at the bottom of the writing line; otherwise, the two traces are the basis of a Q || 2 S or X recto, letter A, fr. b: 1 fortasse -ue sex̣[ vel seṣ ̣ [quiplicar-

recto, letter A, fr. c --|1 [- - -] ̣ ̣ [- - - |2 - - -] ̣ mar[ ̣ ] ̣ [- - -] --recto, letter A, fr. c: 1 scattered traces, the last letter either a I or a L || 2 dot-like traces | lower portion of an oblique stroke, pointing upwards, trespassing the writing line from below

recto, letter B, fr. b --|1 [- - -] ̣ [- - - |2 - - -]s[ --Moreover, in the verso of letter A, the nomen of Iustillus, so far undeciphered, is perhaps Aetius (A]ẹṭịo): before o, one can clearly see e, t and i:

Fig. 13: Letter A, verso: The addressee’s name.

  Giulio Iovine One last remark can be made concerning the hands at work on P. Dura 64. In both letters, the sender forwards to the addressee a second letter, which was sent by the uir consularis and provincial governor to the addressee himself. Reasons for the provincial governor not to send the letter directly to the addressee may include the fact that the original letter contained only a section concerning orders for the addressee, the rest of it being intended for the initial recipient and present sender. This second letter apparently begins right after the first one closes, at the bottom of the first column, and continues in column II. It is generally agreed that the hand working on both letters A and B is the same; it is likewise the same hand who copies down the two epistles forwarded by the sender of A and B. One may however remark that in letter A, frr. c–d the hand, although very similar to the one in frr. a–b, is smaller in magnitude. This may suggest that frr. c–d come from yet another letter, now missing from the extant portions of the manuscript.

 It often happens in the Dura papyri that mention is made of the 20th cohort of the Palmyrenes; in fact, the Dura papyri, together with some Durene inscriptions, are the only source for the existence of such an auxiliary unit.29 When the cohort is mentioned, normally after the name of a soldier and his rank (miles, milites, militibus, eques, equitibus30), we find coh, abbreviation for cohortis;31 the figure XX;

 29 No list of units or coeval document to the Dura papyri, unless it comes directly from the Durene region, offers any information on the cohort of the Palmyrenes; instead, Greek papyri from Dura-Europos mention the σπείρη εἰκοστὴ Παλμυρηνῶν (P. Dura 126, 235 CE, TM 44860, P. Dura 129, 225 CE, TM 17235, P. Euphr. 14, 241 CE, TM 44672). For inscriptions, see for instance AE 1923, 23; 1940, 240; all from third-century CE Dura-Europos. See also James (2019) 245–250. 30 See P. Dura 58, l. 3 eq(uitibus) [eqq.] coh(ortis) XX [Palmyrenorum (225–235 CE, TM 44780), 66 fr. d recto, ll. 3–4 [Milit]es numero uiginṭi octo, d[omine, cohortis | XX Palmyrenorum] (216 CE, TM 44795), 76 fr. c, l. 2 mị[l(-)] cohortis suprascriptae [ (222–235 CE, TM 44806), 82, col. I, ll. 1–2 numerus purus militum ca]ligatorum DCCCCXXII[I]I … | cohortis XX [Palmyrenorum (passim; 223–233 CE, TM 44813). Often in the verso of official correspondence one finds cohortis after tribuno in address formulae, when the tribune of the cohort is the recipient of the letter. 31 Or cohorte? In the whole Durene corpus one only finds coh abbreviated, and editors resolve it as cohortis. One must bear in mind that there is no absolute certainty in this respect. Latin papyri sometimes feature the ablative instead of the genitive to point the unit to which the mentioned soldier belongs: see e.g. ChLA XLV 1340 (= P. Vindob. inv. L 135), ll. 1–3 Lucius Caẹcilius Secundus eques ala Paullini | turma Dicaci, Caio Pompeio militi coh(orte) | Ạ ̣ ̣ ̣ l [ ̣] Habeti centuria Betiti salutem (27 CE, TM 16273).

The Unruly Fragments  

and Palmyrenorum, sometimes unabridged, often abbreviated in Palm or Palmyr.32 This is also what one finds in P. Dura 72,33 a small scrap probably coming from official correspondence. One sees the last stroke of the second X, then Palmyr and a small medial dot, or interpunctum:34

Fig. 14: P. Dura 72.

Editors have so far read and reconstructed the letters after Palmyrenorum as ạṇ[. This naturally led them to think that what is to be seen here is the honorific title Ạṇ[toniniana, normally attached to names of legions, cohorts, alae and numeri during the reign of Caracalla and Elagabalus, who were both Antonini, as far as their names are concerned. Other titles are to be found in several papyri and inscriptions;

 32 Unabridged, e.g., in P. Dura 82, col. II, l. 15 (223–233 CE, TM 44813), Palm in P. Dura 82, col. I, l. 10, II, l. 1, Palmyr in P. Dura 56, letter B, and here in P. Dura 72. 33 Welles/Fink/Gilliam 1959 (P. Dura 72), Fink (RMRP 110), Marichal in (ChLA VI 327), and Cugusi (CEL I 188). TM 44801. 34 The interpunctum is the small medial dot that Romans used from presumably the earliest stages of their written culture, to separate words from one another. They allegedly felt this custom to be peculiarly Roman, in contrast with the Greeks, which did not use such a tool for the diuisio verborum: see Ammirati (2015) 32–33. In papyri, one begins to see that the interpuncta fade at the end of the first century CE, surviving no later than the fourth as markers inconsistently used for abbreviated words.

  Giulio Iovine they normally changed when a new Emperor rose to the throne.35 The cohors XX Palmyrenorum is Antoniniana in several Dura papyri,36 and Seueriana Alexandriana in others,37 or even Gordiana38 in manuscripts written later in the third century CE, during the reign of Gordian III.

Fig. 15: P. Dura 72: ạṇ[, or ṃạ[?

However, if we look again at the manuscript, we would rather read a pretty clear m in ligature with an oblique stroke, pointing upwards: this suggests Mạ[ximiniana, another imperial title attested for cohortes.39 This particular papyrus might perhaps be dated later — not to Caracalla’s or Elagabalus’ reign, but to the shortlived monarchy of the Thracian Maximinus (235–238 CE).

 35 “Gli auxilia, come le legioni, le coorti pretorie, urbane, dei vigili, le flotte, il quartiere degli equites singulares se non gli stessi equites, nel III secolo d.C. aggiunsero, com’è noto, alle altre parti della loro denominazione un soprannome, non però obbligatorio, derivato dal nome dell’imperatore regnante e che cambiava quindi, eccettuato quello di Antoniniana, ad ogni cambiamento di imperatore: tali furono, come risulterà pienamente dall’elenco che darò in seguito, i soprannomi di Antoniniana, Alexandriana o Seueriana, o Seueriana Alexandriana, Maximiana etc.” (Bersanetti 1940, 105–106, and see also, for an updating by the same author, Bersanetti 1943). Gilliam likewise noted that “epithets such as Antoniniana, formed by the reigning Emperor’s cognomen, were regularly added to the names of units in the third century … They changed at the accession of each new Emperor” (Gilliam 1957, 57). 36 P. Dura 66 a recto, l. 4, verso, l. 2, 66 t, l. 2, 66 mm, l. 5. 37 P. Dura 82, col. I, l. 2. 38 P. Dura 89, col. I, l. 5 and passim. 39 See Bersanetti (1940) 127.

The Unruly Fragments  

 Honorific titles also come to relevance in P. Dura 74.40 This papyrus consists of two fragments, a and b, which couldn’t be either aligned or rejoined by former editors. We know, however, that the original document contained official correspondence. In fr. a, l. 4, Fink and Marichal believed that ante should be read; Marichal, as a palaeographer, was not particularly happy with that t, which he described as “très mauvais”.41

Fig. 16: P. Dura 74: -anae, not -ante.

I agree with Marichal and suggest that the supposed t is rather an a. We could therefore be viewing the last two syllables of the already seen imperial honorific title of the cohors: -anae, and then cp̣ [, normally an abbreviation for c(ui) p̣(raees), ‘which you command’. If one considers that the preceding line, l. 3, was occupied by the name of the cohort itself, ll. 3–4 become instantly clearer: 3 4

coh(ortis) XX P[almyrenorum - - - i-] -anae, c(ui) p̣ (raees) [

 40 Welles/Fink/Gilliam 1959 (P. Dura 74), Fink (RMRP 97), Marichal (ChLA VII 329) and Cugusi (CEL I 186). TM 44803. 41 “Le t est très mauvais, mais peut-être acceptable” (ChLA VII, 16).

  Giulio Iovine That is, ‘of the 20th cohort of the Palmyrenes’, then the honorific title, and finally ‘which you command’, a formulaic expression normally used when you are giving orders to the tribune of an auxiliary unit. Unfortunately, this papyrus has no certain dating, and since we only see the end of the title, this could have been any title between Caracalla (Antoniniana) and Decius (Deciana). There is no way to determine the exact length of the lines in fr. a. One can, however, realise that if Palmyrenorum was abbreviated Palm or Palmyr, as it most likely was, the remaining lacuna would have room only for half of the imperial title (e.g. P[almyr Antonini]|anae, or Seueri]|anae or Gordi]|anae and the like), thus determining a very short line if we compare it with other instances of official correspondence in the Dura papyri.42 As I previously stated, P. Dura 74, which goes by the catalogue number P. CtYBR inv. DP 21, includes two fragments, a and b. As far as I know, they have been considered part of the same document, probably on the account of archive records or excavation documents: no absolute certainty in this respect can be gathered from any recurring damage or textual evidence. Fr. a features the beginning of a number of lines, is brown and dark in colour, and there are faint remnants of letters on the verso; fr. b contains middle sections of a number of lines and is lighter in colour; its ink is more vivid and its verso is blank. This may mean, however, that they were distant enough in the original papyrus sheet for a change of colour to take place. The reconstruction of fr. a, ll. 3–4 proposed above does not match any line in fr. b; one must infer that, if fr. b is to be aligned to fr. a, its first line ] Ka[lendas must be located lower than l. 4. On the other hand, traces of a dating formula in fr. a, l. 5 and fr. b, ll. 1–2 might be of use in proposing a possible alignment of the two portions of the manuscripts: fr. a

5

fr. b

̣[ s ̣u ̣ ̣[ coh(ortis) XX Pạ[lmyrenorum - - -i-] -anae c(ui) p(raees) [ - - - ] Ḳạ[lendas Decembris [ - - - ] co(n)s(ulibus) s ̣ ̣ ̣ [ -ri ̣ [ ̣ ]um suor[um - - - ] Ịdus Ian[u]ạ[rias [ ̣ ̣ ̣ ̣ ̣ ]n mil(-) [ - - - ] ̣no co(n)s(ulibus). Opto [ ]ṃ[

 42 For instance, the aforementioned P. Dura 56, or 60, or 64, have apparently longer lines than the perspective line one would have in P. Dura 74. Perhaps this might strengthen the possibility of a longer supplement, such as P[almyrenorum unabridged, and then Seuerianae Alexandri]|anae.

The Unruly Fragments  

I am not fully satisfied, however, by the fact that at ll. 6–7 of the reconstructed fragment one sees Idus Ianuarias and then the consular couple, not together, but split by another sentence where a miles is involved. And the lines, we have remarked before, were probably very short. Unless this difficulty is resolved (perhaps Idus Ianuarias and the consular couple were part of two different, shorter dating formulae within the text?), this realignment must remain only an attractive hypothesis.

 By contrast, a case in which a realignment is verified but does not solve anything either, is P. Dura 76.43 This papyrus too preserves at least one, but probably more official letters. It consists of four fragments, three of which (a, b, d) have most probably been written by the same hand. Given the remarkable similarity between frr. a–b (colour, shape, ink, hand), one would be tempted to align them in some way; this is further suggested by similar wormholes on the surface of the two fragments. There is no clue concerning their disposition in the original roll; however, if they really belong to the same letter, fr. b, which contains the closing greeting formula normally located in the right portion of the letter, might be placed to the right of fr. a, exactly as it is in current editions.

 43 Welles/Fink/Gilliam 1959 (P. Dura 76), Fink (RMRP 93), Marichal (ChLA VII 331) and Cugusi (CEL I 182). TM 44806.

  Giulio Iovine

Fig. 17: P.Dura 76: Similar damages.

] ualer[e] | [- - - feli]cissimum [ (opto te domine felicissimum multis annis bene ualere)144 In fr. c, two details deserve a remark. At l. 2, before coh s·s[, we can clearly see ] ̣s mị[l]; the second word may be either mị[l(es)] or mị[l(ites)]. The sequence has not been detected by former editors. The following abbreviation might therefore be construed as coh(ortis) s(upra) s(criptae).45 At l. 3, one can read an almost illegible docket, added by another hand, and written with a high degree of carelessness and informality. The most certain letters to be read form the sequence obt. The sequence obt normally matches obtuli or obtulit, a verb usually employed in Late Antiquity Latin documents to mark the action of ‘bringing’ — a letter, a libellus, any specific thing — to a higher authority.46 After obt, ụl might well be  44 For the salutation formulae in Roman tradition with a focus on how they appear in papyri and ostraka, see some discussion – with extensive bibliography – in CEL II, pp. 60-64 and in P. Rain Cent., pp. 488–489 (P.J. Parsons). 45 For suprascriptus in papyri earlier than the fourth century CE, see P. Mich. VII 438, ll. 9–10 in vexil]|lati[o]ne c̣ọḥ(ortis) [·] ṣ(upra) · s(criptae), and ChLA X 445 recto, l. 3 co]h(or-) · s(upra) s(cripta) (centuria) (225–227 CE, TM 69943). 46 See ChLA X 463, l. 28 Fl(auius) · Antirus exsceptor obtuli (ca. 350 CE, TM 17284), ChLA XLIII 1245, l. 10 Fl(auius) Asclepius exceptor obtuli. Edantur (late fourth century CE, TM 70037), P. Iand. IV 68 b ☧ Fl(auius) Symeonius cornicularius obtul(it) (sixth century CE, TM 78417).

The Unruly Fragments  

read, with -i or -it omitted in the abbreviation, as customary. Before the verb, a personal name might occur, ending with a very quickly and informally drawn ligature us. This name must for the moment remain unclear: one can see perhaps a n (in ligature with a preceding letter), and two oddly drawn letters, an a and a p (less likely an l): ] ̣ ṇạp̣us or ] ̣ṇạḷus. This formula, apparently distinct from the greeting formula opto te domine, might point out the physical bearer of this letter (or of these letters), and function as a subscription: …us obtul(i) or obtul(it), ‘I …us (or just …us) brought this’.

] ̣ṇ ạ

̣ u s |o b

t

ụ

l-

Fig. 18: P. Dura 74: A possible subscription.

 P. Dura 8947 preserves the fragments of some large acta diurna, a typical document employed in the Roman army to note every single daily activity of the unit to which it referred.48 When first published by Gilliam in 1950, this amounted to 8 fragments. Seven more were added in his second edition in 1959, and on the first eight he performed two relevant re-joinings, which altered the disposition and names of the fragments. The table below offers a clear account of the current editorial situation.

 47 The first edition is in the aforementioned Gilliam (1950), then come Welles/Fink/Gilliam (1959) (P. Dura 89), Fink (RMRP 50), R. Marichal (ChLA VII 344). TM 44820. 48 The term acta diurna (or cottidiana) is used for the first time by M.T. Rostovtzeff (1934) 367 on account of a passage of Vegetius: totius enim legionis ratio, siue obsequiorum siue militarium munerum siue pecuniae, cotidie adscribitur actis. ‘For an account of the whole legion, whether of indulgences or of military gifts or of money is written up in the acta’ (2.11). See Stauner (2004) 74–90 and Phang (2007) 293.

  Giulio Iovine Tab. 4: P. Dura 89: Ordering the fragments. Gilliam 

Gilliam 

Fink 

Marichal 

frr. a + b + c

becomes ‘col. I’

idem

idem

fr. d (A[urelius and IIII joins fr. f and beKal Iun) comes ‘col. II’

idem

idem (fragments are called a b)

fr. e (quod imp)

joins fr. g and becomes fr. b

idem

fr. d

fr. f (Palmyrenorum Gordianae, Avito centurio etc.)

joins fr. d and becomes ‘col. II’

idem

idem (fragments are called a b)

fr. g (P]riscus)

joins fr. g and becomes fr. b

idem

fr. c

idem

fr. e

fr. c ( illegible lines)

idem

fr. m

fr. d (]n ̣[)

idem

fr. h

fr. e ( illegible lines)

idem

fr. k

fr. f (] ̣ ̣ on ̣ ̣ ̣ ̣ ̣ ̣ ̣ [)

fr. f (]tiron ̣ ̣ ̣ [)

fr. l

fr. g (blank)

idem

fr. g

fr. h (]ẹṣ[)

idem

fr. f

fr. i (]uṣ[

idem

fr. i

fr. h (A[u]rel Gerfr. a man[us, ordinatus Aurelius etc.)

This is not consistent, however, with the current conservation status of this papyrus. The frame of P. Dura 89 (P. CtYBR. inv. DP 9) only contains col. I, col. II and fr. e in Marichal’s reckoning. Fr. e is connected to col. II through a small scrap, which Marichal did not see, and must therefore have been put there later;49 this was the right thing to do, as the reconstructed texts both on the recto and on the verso provide very strong clues for this joining.

 49 This may also have been done by R. Duttenhöfer.

The Unruly Fragments  

sunt ịṇ [hibernis co]h(ortis) X̣ X P̣alm[yrenorum - - -] | coh(ortis) XX̣ Palmyrenorum Goṛd[ianae

Fig. 19: P. Dura 89 (= DP 9 recto), col. II, ll. 5–6 (the scrap provides coh and traces of the figure XX, directly to be linked to Pal of Palmyrenorum).

 Heliodori Mocimus Salmes, Ŧ Paulini is now entirely legible

Fig. 20: P. Dura 107 (= DP 9 verso), col. II, l. 12.

Where are the other fragments? As for f, g, h, i, k, l and m, they are currently under conservation process and have been so since many years; nevertheless, I have

  Giulio Iovine been granted permission to see them. As for c and d, they were not to be found in the Beinecke Library, nor anywhere else to my knowledge. A final remark. When publishing fr. l, which should be f in Gilliam’s reckoning, Marichal was very uncertain, as the fragment he had before his eyes, and which he therefore published, clearly was not what Gilliam and Fink believed it to be.50 Marichal saw two lines, while Gilliam saw only one, and very differently written. To solve the riddle, it is necessary simply to turn the fragment around. Marichal was actually looking by mistake at the verso of the scrap; if one turns the scrap and regains the proper recto, it perfectly matches with Gilliam’s earlier description.

 P. Dura 11351 was only described by Fink in Welles/Fink/Gilliam (1959); the actual editio princeps appeared in 1971, in Fink’s Roman Military Records on Papyrus (RMRP). The papyrus preserves portions from a guard roster; the dating is on palaeographical grounds and also relies on the fact that the recto (P. Dura 85) is dated to about 230 CE, and is probably earlier than the verso. Only a handful of names survive. In the frame now visible at the Beinecke Library, one can see seven fragments: a, b, c, d, f, g, h. Editors Fink and Marichal regard frr. a, b, c as forming a single fragment, as well as d and h; and in their editions, they treat them as such. Why they thought so is not evident from the shape of the fragments, nor explicitly stated by them. In the edition I am currently preparing, the layout of the papyrus as it appears in the frame is preserved, so that only frr. b and c are united in a single fragment. Something is missing from the given list of scraps. Fr. e, mentioned by Fink, has been removed from the frame before Marichal worked on his edition, and neither he nor I have been able to inspect it. On the other hand, in fr. b+c I have been able to improve the readings in ll. 7–15.

 50 “G. et F. … ne transcrivent qu’une ligne … Peut-être s’agit-il d’un autre fragment que je n’ai pas retrouvé” (ChLA VII, p. 49). 51 Welles/Fink/Gilliam (1959) (P. Dura 113, only described), Fink (RMRP 12), Marichal (ChLA IX 368), TM 44845.

The Unruly Fragments  

Tab. 5: Fr. b+c: A reassessment. fr. b+c (Marichal ) ] ̣ au ̣ [ ]us Zaḅdib[ol ]ṣ Ṃalch[ ] ̣ rus ̣ ̣ o ̣ [ ] ̣ Theṃ[ ] (centuria) `ạ ̣ ̣ ̣ ´ Nạṣọṇ[i]ṣ Mo ̣ ̣ ̣ ̣ [ ] au ̣ [ ]s l ̣ ̣ [ ̣ ] ̣ [ ] ̣ ̣ ̣ ̣ ̣ ̣ ̣ ̣[ ] Zebid[a ] ̣ ̣ sam ̣ ̣ [ ]e ̣ a[ ] ̣ ̣ au ̣ [  ]us ma ̣ [

Tab. 6: Fr. b+c: A reassessment. fr. b+c (Iovine )





] ̣ Aur[eli]us Zabdib[olus T]h[e]mes Malch[i I]arh[a]boḷẹs Thẹm[arsa ]  Nạsoṇ[i]s Mocim[us ]nor[ ̣ ̣ ̣ ̣ ]s Lol ̣ ̣ [ ] ̣ ̣ ̣ [ ̣ ̣ ̣ ̣ ̣ ] Zebid[a The]ṃ[a]rsa⟦m⟧s ̣ [ ̣ ̣ ̣ ̣ ̣ ̣ ]ei ̣ [ ] ̣ Aur[ ] ̣ Diom[ed-

The names which appear in the improved text are not surprising: we already know these people from other Dura papyri, and names such as Zabdibolus, Themarsas, Iarhaboles or Themes are no strangers to the Palmyrene cohort. Nevertheless, to find them once again in a roster is potentially useful, as a new prosopography of the soldiers in the cohort will have to take advantage of any source of names in the papyri, no matter how often those names are found. One last remark: if one looks closely at the last two lines of fr. b+c, one finds letters very hard to construe at first sight. Fink’s and especially Marichal’s solu-

  Giulio Iovine tions, though the most precise possible, fail, I believe, to solve the riddle elegantly. But if we resort to Debernardi’s principle,52 and turn the papyrus upside down, things start to work.

] ̣[ ]aum[ (Fink, Marichal)

Fig. 21: The papyrus in its current position.

] ̣ al ̣ [ (l. 20) ]us[ (l. 19)

Fig. 22: The papyrus turned 180o.

 52 The principle states that an apparently unsolvable sequence of characters in a manuscript can be sometimes understood if one turns the manuscript itself 180°. Written sequences in papyri and ostraka are not always and not necessarily all drafted in the same direction and side within the writing frame, and many documents stored in libraries which are not yet deciphered have been mistakenly placed upside down in the glass frames. The principle helps us bearing in mind these difficulties. I am indebted to my colleague D. Debernardi (Genoa) for sharing this insight with me.

The Unruly Fragments  

One could read ] Ṃal[hi, ] P̣alṃ[yrenorum or ] al(igati) ̣[: the little remains of this text discourage further supplements. Ll. 19–20 of fr. b+c most likely belong to another document or to a portion of the original document which was drawn by turning the sheet upside down, perhaps to take advantage of the larger lower margin. More than that, nothing can be said.

References Ammirati, S. (2015), Sul libro latino antico. Ricerche bibliografiche e paleografiche, Pisa/Rome. Austin, J. (2010), Writers and Writing in the Roman Army at Dura-Europos, diss. University of Birmingham. Bersanetti, G.M. (1940), ‘I soprannomi imperiali variabili negli auxilia dell’esercito romano’, Athenaeum 18, 105–135. Bersanetti, G.M. (1943), ‘Sui soprannomi imperiali variabili nelle legioni’, Athenaeum 21, 79– 91. Bowman, A.K./Coles, R.A./Gonis, N./Obbink, D./Parsons, P.J. (2007), Oxyrhynchus. A City and Its Texts, London. CEL = Corpus Epistularum Latinarum Papyris Tabulis Ostracis servatarum. Collegit, commentario instruxit P. Cugusi, 2 vols, I. Textus – II. Commentarius (Florence 1992); III. Addenda, Corrigenda, Indices rerum, Indices verborum omnium (Florence 2002). ChLA = Bruckner, A./Marichal R. et al. (1954–), Chartae Latinae Antiquiores I–, Dietikon and Zürich. Fink, R.O. (1971), Roman Military Records on Papyrus, Cleveland (Ohio). Fournet, J.-L. (2017), ‘Culture grecque et document dans l’Égypte de l’Antiquité tardive’, in T. Derda/A. Lajtar/J. Urbanik (eds), Proceedings of the 27th International Congress of Papyrology, Warsaw, 29 July – 3 August, 2013, 3 vols, Warsaw, 135–162. Gilliam, J.F. (1957), ‘Two Latin Letters from Dura-Europos of A.D. 221’, Études de Papyrologie 8, 49–62. Gilliam, J.F. (1959), ‘Some Latin Military Papyri from Dura’, YClS 11, 171–189. Hopkins, C. (1979), The Discovery of Dura-Europos, New Haven. Iovine, G. (2019), ‘Data epistula: Later Additions of Roman Dating Formulae in Latin and Greek Papyri and Ostraka from the First to the Sixth Centuries AD’, Manuscripta 63.2, 157–230. James, S. (1985), ‘Dura-Europos and the Chronology of Syria in the 250s AD’, Chiron 15, 115– 124. James, S. (2011), ‘Stratagems, Combat, and ‘Chemical Warfare’ in the Siege Mines of Dura-Europos’, AJA 115, 69–101. James, S. (2019), The Roman Military Base at Dura-Europos, Syria: An Archaeological Visualisation, Oxford. MacDonald, D. (1986), ‘Dating the Fall of Dura-Europos’, Historia 35, 45–68. Miltner, R. (1930), ‘L. Marius Maximus’, RE 14.2, cols 1828–1831. Phang, S.E. (2007), ‘Military Documents, Languages, and Literacy’, in P. Erdkamp (ed.), A Companion to the Roman Army, Main Street (MA), 286–338.

  Giulio Iovine P. Rain. Cent. = Festschrift zum 100-jährigen Bestehen der Papyrussammlung der Österreichischen Nationalbibliothek, Papyrus Erzherzog Rainer, Vienna 1983. Reeves, M.B. (2005), The ‘Feriale Duranum’, Roman Military Religion, and Dura-Europos. A Reassessment, Buffalo. RMRP = Fink (1971) Rostovtzeff, M.T. (1934), ‘Der Militärarchiv von Dura’, in W.F. Otto/L. Wenger (eds), Papyri und Altertumswissenschaft. Vorträge des 3. Internationalen Papyrologentages in München vom 4. bis 7. vom 4. bis 7. September 1933 [= Münchener Beiträge zur Papyrusforschung und antiken Rechtsgeschichte, 19], 351–378. Scappaticcio, M.C. (2012), Accentus, apex e distinctio. L’accentazione grafica tra Grammatici Latini e papiri virgiliani, Turnhout. Sommer, M. (2005), Roms orientalische Steppengrenze. Palmyra – Edessa – Dura-Europos – Hatra. Eine Kulturgeschichte von Pompeius bis Diokletian, Stuttgart. Stauner, K. (2004), Das offizielle Schriftwesen des römischen Heeres von Augustus bis Gallienus (27 v.Chr. – 268 n.Chr.). Eine Untersuchung zur Struktur, Funktion und Bedeutung der offiziellen militärischen Verwaltungsdokumentation und zu deren Schreibern, Bonn. Vandoni, M. (1966), Appunti dalle lezioni di papirologia, Milan. Welles, C.B./Fink, R.O./Gilliam, J.F. (1959), The Parchments and Papyri. With an Account of Three Iranian Fragments by W. B. Henning, New Haven.

List of Contributors Roberta Berardi is Research Associate in Classics at the University of Oxford ([email protected]) Francesco Ginelli is Postdoc in Roman History at the University of Milan ([email protected]) Giulio Iovine is Marie Curie Global Fellow at the University of Bologna ‘Alma Mater Studiorum’ ([email protected]) Francesco Lupi is Postdoc in Greek Language and Literature at the University of Verona ([email protected]) Chiara Meccariello is Research and Teaching Associate at the Georg-August University of Göttingen ([email protected]) Stefano Vecchiato is PhD student in Classics at the Scuola Normale Superiore in Pisa ([email protected]) Nereida Villagra is Assistant Professor in Classics at the University of Lisbon ([email protected]) Jarrett T. Welsh is Associate Professor in Classics at the University of Toronto ([email protected])

Index of Names Acusilaus 25 Aeschines of Miletus 96 Aeschylus 159, 160 and nn. 46‒47, 165 Aeschylus of Cnidus 86 n. 46 Agatharchides of Cnidus 66 n. 35, 84, 97 Agrippa, Marcus Vipsanius 87 n. 50 Alexander III (‘the Great’) 81, 90, 96 n. 63 Alganza Roldán, M. 146 n. 7 Alpers, K. 31 n. 42 Ambrose 166 n.3 Ammirati, S. 179 n. 34 Anaximenes of Lampsacus 81, 82 n. 16, 83 n. 20, 84 n. 25, 86 Antiochus III (‘the Great’) 117 Antoninus Liberalis 146 n. 7 Antonius, Marcus 99 Apollodorus (mythographer) 70 n. 51, 146 and n. 7, 156, 160 Apollodorus of Athens 106 Apollodorus of Pergamon 85 Apollonius Molon 81, 96 Apollonius ὁ μαλακός 96 Apollonius Rhodius 156, 160 and n. 47 Aragosti, A. 137 n. 50, 138 n. 56, 139 n. 57 Archelaus (king of Macedon) 68 and n. 41 Archilochus 28, 34, 115 Aristarchus (grammarian) 64, 65, 66 and n. 35, 67, 68, 69 Aristides (Athenian general) 112, 113 Aristides, Publius Aelius 59 n. 7, 90 Aristocles (orator) 86 Aristophanes 6 n. 17, 13, 58 and n. 4, 59 n. 7, 65, 66, 67, 70, 71 and n. 55, 71, 73, 75 n. 65 Aristophanes of Byzantium 59 and n. 8, 67 n. 37 Aristotle 60 and n. 9, 68 n. 41, 69 n. 49 Armstrong, A.H. 64 n. 26 Artamenes 86 Artemidorus 169

Asclepiades of Tragilos 151, 152, 158 and n. 37, 159 and n. 43, 160 and n. 46, 160 n. 47 Asirvatham, S. 150, 158 n. 37 Astymedes 97 n. 80 Asulanus, Andreas 150 Athenaeus of Naucratis 63, 84 n. 35, 96, 97 Atticus, Titus Pomponius 103, 105 Audax 24 Aufustius (grammarian) 140, 141 Augustinus, Aurelius 166 n. 3 Augustus 87 n. 50 Aulitzky, K. 95 n. 55 Ausonius, Decimus Magnus 105 n. 10 Austin, C. 49 n. 44 Austin, J. 166 n. 5 Avezzù, G. 37, 41 n. 15 Baiter, J.G. 85 Balbo, A. 7 n. 21, 109 n. 22, 110 n. 25 Baldi, D. 30 n. 40, 31 nn. 42–43 Baltussen, H. 5 n. 14, 8 n. 23, 10 n. 28 Barabino, G. 84 n. 24, 95 n. 54, 99, 121 n. 2 Barnes, J. 149, 150 Bassino, P. 63 n. 23 Berardi, R. 7 n. 19, 13, 14, 88 Bergk, T. 38 n. 5, 39, 41, 42 n. 20, 43, 46 Bernardi Perini, G. 106 n. 17 Bernsdorff, H. 61 n. 13 Bersanetti, G.M. 180 nn. 35+39 Bethe, E. 28 Blass, F. 85 n. 41, 99 Blaydes, F.H.M. 39 Blum, R. 64 n. 25 Blumenthal, A. von 39 Borges, C. 60 n. 12, 61 n. 13 Bornecque, H. 96 n. 68 Borrelli, B. 60 n. 12, 61 n. 13

 Names referring to mythical entities are excluded from the index. https://doi.org/10.1515/9783110712223-011

  Index of Names Bouchard, E. V Bowman, A.K. 166 n. 4 Briscoe, J. 106 and n. 15, 111, 112 and nn. 28‒29, 113, 115 Brooks, P. 84 n. 24 Brown, M.K. 146 n. 7 Bruckner, A. 167 n. 8 Brzoska, J. 81 nn. 10–11, 95 n. 56 Buchwald, W. 39 Burges, G. 39 Butterfield, D. 121 n. 2 Buttmann, Ph. 150 Caecilius Statius 121 Callicrates (orator) 85 n. 38 Callimachus 61, 63 Callinus 84 n. 36 Cameron, A. 145 n. 3, 146 n. 7, 148 and n. 18 Canfora, L. 5 n. 10, 86 n. 48 Capra, R.L. V Caracalla 179, 180, 182 Carilli, M. 139, 140 and n. 59 Carrara, P. 44 n. 26, 49 n. 43, 72 n. 57 Casadio, V. 30 n. 40 Casanova, A. 26 and nn. 19+21, 37 nn. 1–2 Cato, Marcus Porcius (the Censor) 105, 131 n. 35 Catullus, Caius Valerius 126, 127 and n. 18, 134 Ceceidas 59 n. 8 Cephisodorus (orator) 85 n. 38 Chahoud, A. 121 n. 3 Chamaeleon 59 nn. 6–7, 62 Chantraine, P. 92 Chantry, M. 66 n. 34 Charisius (orator) 84 n. 30, 86 n. 44, 95 Charisius, Flavius Sosipater 103 and n. 2, 105 n. 10, 124 Chávez Reino, A.L. 81 n. 9 Cicero, Marcus Tullius 81, 82, 83, 86 and n. 46, 99, 105, 106 n. 15, 121, 125, 135 Cicero, Marcus Tullius Minor 84 n. 24 Cimon Atheniensis 113 Cincius Alimentus, Lucius 106 Cingano, E. 26 n. 23, 65 n. 29 Cinna, Caius Helvius 127

Clement of Alexandria 27, 40 n. 8 Cleochares 86 n. 44, 95 Cleopatra 81 Coccus 85 n. 38 Coles, R.A. 166 n. 4 Collard, C. 69 n. 44 Conon (mythographer) 146 n. 7 Conte, G.B. 11 n. 31 Coppola, G. 148 n. 20 Cornelia (mother of the Gracchi) 104, 105 Cornell, T.J. 106, 111 n. 27 Costanza, S. 32 n. 45 Courtney, E. 127 n. 20 Cozzoli, A.T. 26 n. 23, 27 nn. 24–25 Crates (orator) 85 n. 38 Crates of Mallus 64 Cribiore, R. 61 n. 14, 91 Cropp, M. 69 n. 44 Cuartero, F.J. 146 n. 7 Cugusi, P. 167 and n. 9, 168 n. 12, 175 n. 27, 179 n. 33, 181 n. 40, 183 n. 43 Curnis, M. 53 and n. 55, 107 n. 19 Custer, M. 168 n. 11 Cydias 59 n. 8 Cydidas of Hermione 59 and n. 8 D’Alessio, G.B. 26 n. 22 D’Anna, G. 122 and n. 5 Dalby, A. 57 nn. 1–2 Damas 96 Danbeck, D. 28 and nn. 27–28 Daphnis (orator) 86 n. 44, 95 Davies, M. 59 n. 7, 62 n. 21 De Melo, W. 137 n. 51 De Stefani, C. 37 n. 1 Debernardi, D. 190 and n. 52 Deforge, B. 160 n. 46 Del Corno, D. 59 n. 8 Delattre, Ch. 145 n. 3, 162 n. 51 Demades 83 Demetrius Chalcondyles 149 Demetrius of Phalerum 82 n. 16, 83, 84 nn. 25+28+32, 86, 95, 96, 97 Demochares of Leuconoe 79, 81, 83, 84 n. 31, 85 and n. 39, 86 n. 44, 95, 96 and n. 76

Index of Names  

Democleides 84 n. 32, 85 and nn. 38–39, 96 Democles 85 nn. 38–39 Demosthenes 80 and n. 5, 81, 82, 83 and n. 17, 88, 94 Denniston, J.D. 46 Derda, T. 1 n. 2 Di Gregorio, L. 45 n. 30 Di Lello-Finuoli, A.L. 43 n. 23 Dicaearchus of Messana 65, 72 and n. 61, 73, 74 Dickey, E. 20 nn. 1+3 Didymus Chalcenterus 146 Diggle, J. 38 n. 3, 41 nn. 16–17, 43 n. 21, 47 n. 35, 69 n. 44 Dinarchus 83, 85 Dindorf, W. 39, 43, 48 n. 41 Diodorus Siculus 60, 84, 97 Diogenes Laertius 60 n. 9, 84 Diomedes Grammaticus 127 n. 18 Dionisotti, A.C. 4 n. 6 Dionysius of Halicarnassus 75, 84, 85, 86, 96 Dionysius of Magnesia 83 n. 19 Dionysius of Pergamon 96 Diophanes of Mytilene 83 n. 19 Dobrov, G. 51 n. 49 Doon, E. V, 168 n. 11 Dover, K. 59 nn. 7–8, 69 n. 47, 74 n. 63 Dräger, P. 23 n. 13 Drummond, A. 106 and n. 15, 111, 112 and nn. 28–29, 113, 115 Duttenhöfer, R. 171 and n. 23, 186 n. 49 Dyck, A.R. 20 nn. 1+3

Eusebius of Caesarea 84 n. 35 Eustathius of Thessalonica 150, 156, 157 nn. 30–31, 158 and n. 41, 160 and n. 48

Elagabalus 179, 180 Ennius, Quintus 124, 125, 126, 128 Epaminondas 90, 93 Eratosthenes 59 n. 60, 106 Ernst, N. 149, 150, 158 n. 40 Eudocia Augusta, Aelia 84 Eudoxus astrologus 116 Euripides 6 n. 17, 24, 44, 45, 46, 49 n. 43, 53 n. 55, 60, 62 and n. 19, 66, 67 and n. 36, 68 and nn. 40–41, 69 and n. 47, 123 and n. 9, 146 n. 7

Galen 95 n. 54 Gambetti, S. 66 n. 35 Gärtner, H. 81 n. 12 Geiger, J. 103 n. 2, 113 and n. 31 Geißler, C.J. 61 n. 13 Gellius, Aulus 105 n. 10, 112, 115, 116, 117, 124 and n. 12, 130, 131 Gentili, B. 40 nn. 9+12, 47 nn. 34+36 Giannachi, F.G. 47 n. 36 Giannini, P. 21 n. 8 Gibert, J. 69 n. 44

Fabius Pictor 106 Falaschi, E. V Falivene, M.R. 61 n. 17 Fassino, M. 49 n. 43 Favorinus 40 and n. 8 Ferreri, L. 43 n. 23, 48 n. 39 Festus, Sextus Pompeius 124, 126, 132, 133 n. 42, 134, 136, 140 Fick, A. 34 n. 57 Fiehn, C. 95 n. 58 Finglass, P.J. V, 19, 37 n. 1, 47 n. 33, 52 n. 52, 59 n. 7, 62 n. 21 Fink, R.O. 166 n. 5, 167 and n. 7, 168 n. 12, 173, 174 nn. 24+26, 175 and n. 27, 179 n. 33, 181 and n. 40, 183 n. 43, 185 n. 47, 186, 188 and n. 51, 189, 190 Fitzpatrick, D. 37 n. 2, 38 n. 3, 40 and n. 14, 41 and nn. 16+19, 42 n. 19, 43 n. 21, 46 nn. 31+33, 51, 52 and nn. 51+53, 53 and n. 54 Flower, M.A. 81 n. 9 Fortenbaugh, W. 109 n. 24, 111 n. 26 Fournet, J.-L. 166 n. 3 Fowler, R.L. 146 n. 7, 149 and n. 21 Friedländer, P. 39 Fritzsche, F.V. 71 n. 55 Froehde, O. 132 n. 39 Fronterotta, F. 9 n. 27 Furley, W. 31, 32 and nn. 45+47–48, 33 and nn. 49–51

  Index of Names Gilliam, J.F. 166 n. 5, 167 and n. 7, 168 n. 12, 171 and n. 22, 172, 173, 174 nn. 24+26, 175 and n. 27, 176 and n. 28, 179 n. 33, 180 n. 35, 181 n. 40, 183 n. 43, 185 and n. 47, 186, 188 and n. 51 Ginelli, F. 1, 5 n. 13, 7 n. 22, 8 n. 25, 14, 19, 121 n. 1, 165 Giocondo, G. 127 n. 19 Gladhill, W. V Glaucippus (orator) 85 n. 38 Gleditsch, H. 39, 42 n. 20 Goetz, G. 24 n. 15 Gonis, N. 166 n. 4 Gordianus III, Marcus Antonius 180 Gorgias of Athens 84 n. 24, 95 n. 54, 96 Gracchi (brothers) 104 Graf, F. 158 n. 37 Grilli, A. 107 and n. 18 Guidorizzi, G. 59 n. 8 Guillemin, A.-M. 105 and nn. 8+10 Gutas, D. 109 n. 24, 111 n. 26 Gysembergh, V. 31, 32 and nn. 45+47– 48, 33 and nn. 49–51 Habicht, C. 79 n. 3 Hagnon (or Hagnonides) (orator) 85 n. 38 Hallo, W.W. 57 n. 2 Halm, K. 105 and n. 12, 112, 113, 114 Hannibal 99 and n. 84 Harder, A. 69 nn. 45–46 Harmodius 58 n. 4 Harpocration, Valerius 84, 97 Haslam, M.W. 65 n. 31, 69 n. 47, 71 n. 55, 145 n. 4, 148 n. 18 Hatzilambrou, R. 90 Havet, L. 137 n. 48 Heckel, W. 81 n. 10 Hegesias of Magnesia 83, 84 nn. 25+33, 86 n. 44, 95, 96, 97 Hegesippus (orator) 83 and n. 17 Heliodorus (grammarian) 24 n. 14 Hemina, Lucius Cassius 112, 116 Henrichs, A. 146 n. 7 Hense, O. 43 n. 23, 44, 47 n. 37, 48 and n. 38, 49 n. 44 Heraclitus 9 n. 27 Hermann, G. 32

Hermias of Atarneus 85 Hermodorus of Salamis 113 Hermogenes 84 Herodian, Aelius 20 and n. 3 Herodotus 29 Hertz, M. 113, 124 n. 12 Hesiod 12, 19, 20, 21, 22, 23 n. 12, 24, 25, 26, 27 n. 26, 28, 29, 30 and n. 40, 32 and n. 45, 33 and n. 50, 44, 64, 112, 116, 149, 150, 153, 158 Higbie, C. 146 n. 7 Hilder, J. 1 n. 2 Hilton, J. 37 Hirschberger, M. 19, 22, 26 n. 22 Hollis, A.S. 127 n. 20 Holzberg, N. 103 n. 1 Homer 20, 23 n. 12, 28, 53 n. 55, 63, 75, 112, 116, 145, 146, 147, 148, 153 n. 22, 154, 155, 159 n. 44 Hopkins, C. 167 n. 6 Horace 134 Horsfall, N. 104 n. 3 Hourmouziades, N.C. 51 n. 49 Housman, A.E. 124 and n. 11, 141 Huby, P. 109 n. 24, 111 n. 26 Hybreas 84 nn. 25+27 Hyperides 61 and n. 15 Ian, L. 111, 114 Ioannidou, K. V Iovine, G. V, 6 n. 16, 15, 169 n. 16 Isidorus (tragedian) 29 Isidorus of Pergamon 86 n. 44, 95 Isocrates 53 n. 55, 81, 85 n. 38, 86 Jacoby, F. 80, 150, 158 n. 37 James, S. 166 n. 5, 178 n. 29 Janko, R. 22 n. 11 Jerome 105 n. 10 Johnson, W.A. 91 Kannicht, R. 49 n. 44, 50 n. 44, 67 n. 38, 70 n. 53, 71 n. 55, 72 n. 59 Kassel, R. 4 n. 6, 31 n. 41 Kaster, R.A. 135 n. 45 Katsaros, V. 157 n. 30 Katz, J.T. 4 n. 5

Index of Names  

Kenney, E.J. 75 Kiso, A. 51 n. 49 Kremmydas, C. 80 and n. 6, 82 n. 16, 86 and n. 48 Kroll, W. 95 n. 57 Kwapisz, J. 1 n. 2 Laberius, Decimus 121, 125, 126, 132 n. 38, 134, 135, 136 and n. 47, 139, 140, 141 Lactantius, Lucius Cae(ci)lius Firmianus 105 n. 10 Laelius, Caius 117, 118 Laks, A. 7 and n. 21, 8 n. 24 Lamprocles 59 and nn. 6–7 Laqueur, R. 80 n. 8, 98 n. 81 Larcher, P.H. 30 n. 40 Lasserre, F. 31 n. 43 Lefkowitz, M. 68 n. 41 Lehrs, K. 39 Leo, F. 137 n. 50 Leosthenes 89 Lesbocles (orator) 96 Libanius 90 Librán Moreno, M. 37 n. 1 Lightfoot, J.L. 146 n. 7 Lindsay, W.M. 121 nn. 1+3, 122, 124 n. 12, 125 and nn. 14–15, 126 n. 15, 130 n. 27, 131 n. 35, 132 and n. 38, 133 n. 42, 135 and n. 46, 136, 137 n. 49, 138 and nn. 52+55 Livadaras, N. 31 n. 43 Livy 97 and n. 80, 117 Lloyd-Jones, H. 38 n. 3, 39, 41 n. 17, 42, 43 n. 21, 46 n. 32 Löffler, I. 26 n. 23 Lomiento, L. 40 nn. 9+12, 47 nn. 34, 36 Lucilius, Caius 121, 123, 124 Lucretius Carus, Titus 121 Lunstedt, P. 157 n. 33, 158 n. 39 Lupi, F. V, 1, 5 nn. 11–12, 13 and n. 33, 19, 50 nn. 45+47–48, 121 n. 1, 165 Luppe, W. 72 and n. 61 Lutatius Catulus, Quintus 106 Maass, E. 28 MacDonald, D. 166 n. 5

Maehler, H. 60 n. 11 Maganuco, A. 37, 48 n. 41 Malcovati, E. 104 and n. 6, 105, 112, 113, 114, 117 Maltomini, F. 60 n. 11, 61 n. 15 Manlius Capitolinus, Marcus 116 Marasco, G. 79 and 4 Marichal, R. 167 and nn. 8+10, 168 n. 12, 169, 171, 172, 173, 174, 175 and n. 27, 176, 179 n. 33, 181 and nn. 40–41, 183 n. 43, 185 n. 47, 186, 188 and nn. 50– 51, 189, 190 Marshall, P.K. 105 and n. 13, 112, 113, 114, 115, 116, 117 Martano, A. 62 n. 21 Martinelli, M.C. 49 n. 44 Marx, F. 124 and n. 11 Masurius Sabinus 129 n. 23 Matelli, E. 65 n. 28 Matera, R. 165 Matijašić, I. 28 and nn. 29–30 Matris 96 Matthiae, A.H. 153 n. 22 Maximinus Thrax 180 Mayhoff, C. 111, 114 Mazzacane, R. 121 n. 2, 132 n. 38, 138 n. 55 Meccariello, C. 13, 62 n. 20, 67 n. 38, 68 n. 40, 72 n. 60, 146 n. 7 Meliadò, C. 146 n. 7 Menesaechmus 85 and n. 38, 97 Mercier, J. 121 n. 1 Meridor, R. 50 nn. 45–46+48 Merkelbach, R. 19, 20, 21 and n. 9, 22, 30 nn. 37–38, 87, 88, 93, 94 Merro, G. 72 n. 59 Milo, D. 37 n. 2, 38 n. 3, 40 nn. 8+13, 41 nn. 16–17, 43 n. 21, 46 n. 31, 47 n. 35 Miltner, R. 168 n. 13 Momigliano, A. 27 n. 24 Mommsen, T. 106 Monaco, C. V Monda, S. 137 n. 51, 138 and n. 55 Monella, P. 37 n. 2, 52 n. 51, 53 n. 54 Montanari, F. 62 n. 21, 65 n. 29, 145 n. 4, 146 n. 6, 147 and nn. 9–10+13+15, 151, 157 n. 32, 159 n. 44

  Index of Names Most, G.W. 1 n. 1, 2 n. 3, 5 n. 14, 7 n. 18, 11 nn. 31–32, 19, 22 and nn. 10–11, 24, 26 n. 20, 28 nn. 28+31, 29, 30 n. 37, 107, 108 nn. 20–21, 109 n. 23 Müller, K. 80 Müller, L. 131 nn. 35+38, 135, 138 and n. 54 Musurus, M. 48 and n. 40 Myron (orator) 83 n. 21, 86 n. 44, 95 Myron of Priene 83 n. 21 Nachmanson, E. 60 n. 9, 63 n. 24 Naevius, Cnaeus 132 and n. 37 Naevius, Marcus (tribune) 117 Nauck, A. 72 Neanthes of Cyzicus 80 n. 8, 84 n. 36, 86 Nepos, Cornelius 14, 103 and n. 2, 104 and n. 5, 105 and n. 10, 106, 107, 108, 109, 110, 111, 112, 113, 114 and n. 32, 115, 116, 117, 118 Nicetes 96 Nitsche, W. 81 n. 10 Nonius Marcellus 14, 121 and nn. 1+3, 122, 123, 124 and n. 12, 125 and n. 15, 126, 127 and n. 19, 128, 129, 130 and n. 27, 131 and n. 34, 132, 133 and n. 42, 134, 135 and n. 46, 136, 137, 138, 139, 140 and n. 59, 141 Norden, E. 79 n. 2 Obbink, D. 166 n. 4 Olson, S.D. 5 n. 14, 10 n. 28, 29 n. 36, 58 n. 4 Onions, J.H. 131 n. 35, 135, 138 n. 53 Orion (lexicographer) 30 and n. 49 Otranto, R. 61 n. 15 Ottone, G. 96 n. 75 Pacuvius, Marcus 122, 123 and n. 9, 131 n. 34, 141 Pagani, L. 157 n. 30 Pagès, J. 145 n. 4, 147 and nn. 11+13–14, 148 and n. 19, 155 n. 26, 156, 157 nn. 30–31 Panayotakis, C. 136 n. 47, 139, 140 n. 59 Panzer, J. 145 and n. 2, 146 n. 5, 147 n. 11 Papathomopoulos, M. 146 n. 7

Parsons, P.J. 60 n. 11, 91, 166 n. 4, 184 n. 44 Parthenius of Nicaea 146 n. 7 Pascal, C. 137 n. 48 Paulus Diaconus 132, 133 n. 42, 134, 136, 140 and n. 60 Pausanias (geographer) 65 Pearson, A.C. 38 n. 5 Peek, W. 86 n. 43 Pellizer, E. 146 n. 7 Peter, H. 105 and n. 11, 112, 113, 114 Pfeiffer, R. 64 n. 25, 145 nn. 2+4, 148 n. 20 Phang, S.E. 185 n. 48 Pherecydes of Athens 149, 150, 153, 158 Philinus (orator) 96, 97 Philip II (king of Macedon) 94 Philip V (king of Macedon) 98, 99 and n. 84 Philiscus of Miletus 83 n. 23, 84 n. 36, 86, 96 Philostratus, Flavius 90 Phormio Atheniensis 116 Photius 95, 96 Phrynichus (comedian) 59 nn. 6–7 Phrynichus (grammarian) 59 n. 6 Piccione, R.M. 43 nn. 23–24, 44 and nn. 25–26+28, 45 n. 30 Pindar 28, 34 and n. 57 Pintaudi, R. 86 n. 48 Pisander 28 Plato 53 n. 55, 65, 74, 75 Plautus, Titus Maccius 121, 123, 126, 127, 128, 129, 130, 131, 132 and n. 37, 135, 136, 137 and n. 51, 138, 139, 141 Pliny the Elder 105 n. 10, 111, 112, 114 Plotinus 64 Plutarch 9, 28, 29, 39 n. 6, 40 and n. 13, 60, 63 n. 24, 84, 96, 118 Plution 96 Pollux, Julius 20, 28 Polyaenus of Sardis 86 Polybius 49 n. 44, 50 n. 44, 97 and n. 79 Polyeuctus of Sphettus 96 Pontani, F. 147 n. 10, 149, 157 n. 30 Pordomingo, F. 47 n. 37, 49 n. 43 Porphyry 64 and n. 26

Index of Names  

Porson, R. 48 n. 41 Potamon (orator) 96 Pownall, F.S. 81 n. 9 Praxiphanes 64 Priscianus 105 n. 10, 113, 124 Prodi, E.E. V, 19, 61 n. 15 Ps.-Ammonius 29 Ps.-Callisthenes 84 n. 35 Ps.-Eudocia 84 and n. 36 Ps.-Herodian 21, 22, 23 n. 12, 96 Ps.-Zonaras 31 n. 42 Ptolemaeus of Naucratis 90 Puglia, E. 61 n. 15 Pytheas (orator) 81, 84 n. 25, 86, 95, 96 Python of Byzantium 83 Radermacher, L. 83 n. 22, 96 n. 69 Radt, S. 38 and nn. 3+5, 39, 40 n. 9, 41, 42 n. 20, 43 n. 21, 50 n. 47 Rahn, H. 103 n. 1 Ranocchia, G. 43 n. 23 Ranucci, G. 112 n. 30 Reeves, M.B. 165 n. 1 Reifferscheid, A. 117 Reiske, J.J. 39 Reitzenstein, R. 20 and n. 2 Rich, J.E. 106, 111 n. 27 Ritschl, F. 129 n. 26 Robert, C. 158 n. 37 Rodighiero, A. 37 Roemer, A. 66 Rolfe, J.C. 112, 113, 114 nn. 33–34, 115, 116, 117 Rossbach, A. 42 n. 20 Rostovtzeff, M.T. 185 n. 48 Ruggeri, L. 19 Ruhnken, D. 84 n. 24, 85 and nn. 39–40, 99, 150 Rutilius Lupus 84 and n. 24, 86, 95 and n. 54, 97, 98, 99, 100 Rychlewska, L. 125 n. 14 Rzach, A. 33 and n. 54 Sampson, M. 60 n. 12, 61 n. 13 Sanzo, J.E. 58 n. 3 Sappho 61 n. 15 Sarris, V. 157 n. 30

Sauppe, H. 85 Scappaticcio, M.C. 170 n. 21 Scattolin, P. 53 n. 54 Schierl, P. 122 n. 6, 123 n. 9, 131 n. 34 Schironi, F. 155 n. 25 Schoell, F. 137 n. 50 Schubart, W. 87, 88, 89, 92, 93 Schwartz, E. 72 Schwartz, J. 21 and n. 6, 24 n. 14, 26 n. 23 Scipio Africanus, Publius Cornelius 114 n. 34, 117 Scullion, S. 67 n. 38, 68 nn. 41–42, 69 n. 47, 71 n. 55 Secundus, Lucius Caecilius 178 n. 31 Seneca the Elder 82, 96 Serenus, Septimius 127 n. 18 Shapur I 166 Sharland, S. V, 37 Sharples, R. 109 n. 24, 111 n. 26 Sittl, K. 24 n. 14, 30 n. 40, 31 n. 42 Slattery, S. 37 Smith, R.S. 146 n. 7 Solinus, Caius Iulius 106 Solmsen, F. 33 Solon 60, 62 n. 24 Sommer, M. 166 n. 5 Sommerstein, A.H. 37 n. 2, 38 n. 3, 40 and nn. 8+14, 41 and nn. 16+19, 42 n. 19, 43 n. 21, 46 nn. 31+33, 51, 52 and nn. 51+53, 53 and n. 54, 58 n. 4, 63 n. 22, 69 n. 47 Sophocles 4, 13, 37 and n. 2, 39 n. 6, 40 and nn. 8+12, 41, 46, 47 n. 36, 49, 50, 52, 69 n. 49, 159, 160 and n. 47 Sosicrates (historian) 98 and n. 83, 99 Sosicrates (orator) 86, 95, 98, 99, 100 Sosigenes (orator) 86 Speranzi, D. 43 n. 23, 48 nn. 39–40 Speusippus 94 Spies, Y. 104 n. 5 Stauner, K. 185 n. 48 Stegemann, W. 86 Stem, R. 119 n. 35 Stephanus of Byzantium 84 Stesichorus 59 and n. 6, 62

  Index of Names Stobaeus, Johannes 13, 37, 39, 40 and n. 9, 42, 43 and n. 21, 44 and n. 28, 45 and n. 30, 47, 48, 49 and n. 44, 50 and n. 44, 51, 53 and n. 55 Stoessl, F. 160 n. 46 Strabo 84, 96 Stramaglia, A. 82 n. 16, 86 n. 48 Stratocles (orator) 84 n. 29, 86 n. 44, 95, 96 Strecker, K. 66 Sturz, F.W. 25, 30 n. 40, 150 Suda 84 and n. 36, 85, 96 Sundwall, J. 85 n. 39 Susemihl, F. 81 n. 10 Sutton, D.F. 51 n. 49 Svetonius Tranquillus, Caius 105 n. 10, 117, 118 Talboy, T. 37 n. 2, 40 n. 8 Tammaro, V. 37 n. 1 Taplin, O. 123 n. 9 Tempest, K. 80 and n. 6 Terence 118, 123 Themistocles 113 Theocritus 68 n. 43 Theodosius II 166 n. 3 Theophrastus 85 n. 39 Theophrastus of Eresos 111 n. 26 Theopompus of Chios 81 and n. 9, 83, 86, 94, 96 and n. 76 Tiberius 84 n. 24 Timaeus of Tauromenion 84 n. 32, 85 Titinius 121 Toher, M. 103 n. 1 Torino, A. 137 n. 51 Trajan 182 Trzaskoma, S.M. 146 n. 7 Tsagalis, C. 19 Tucker, T.G. 39 Tullus Hostilius 115 Turner, H. 91 Tzetzes, J. 59 n. 7, 70 Valerius Flaccus Setinus Balbus, Caius 161 n. 50 Valerius Maximus 117 Van der Valk, M. 147, 148 n. 16, 158 n. 39

Van Herwerden, H. 39 Van Rossum-Steenbeek, M. 61 n. 16, 62 nn. 18–19, 145 n. 4, 146 n. 6, 147 and n. 12, 148 and nn. 17+20, 152, 155 n. 26, 158 n. 36 Van Thiel, H. 158 n. 40 Vandoni, M. 165 n. 2 Varro Reatinus, Marcus Terentius 106, 116, 121, 129 and n. 26, 132 nn. 37–38 Vassilaki, E. 145 n. 3 Vecchiato, S. 5 nn. 12–13, 12, 29 nn. 32+34–36 Vegetius Renatus, Publius Flavius 185 n. 48 Vergados, A. 33 and nn. 50–51, 34 Verhasselt, G. 72 n. 58 Verrius Flaccus, Marcus 15, 126, 132, 133, 134 and n. 44, 136, 140, 141 Verus, Lucius 166 Villagra, N. 15, 158 nn. 37–38 Virgil 121, 128, 129, 132 Vos, G. 52 Vulcanius, B. 29 Walbank, F.W. 97 n. 78 Waser, O. 25 n. 17 Webster, T.B.L 51 n. 49 Weissenberger, M. 81 n. 11 Welles, C.B. 166 n. 5, 167 and nn. 7+10, 168 n. 12, 174 nn. 24+26, 175 n. 27, 179 n. 33, 181 n. 40, 183 n. 43, 185 n. 47, 188 and n. 51 Welsh, J.T. 10 and n. 29, 11, 14, 15, 121 nn. 1+3, 122 n. 4, 124 n. 13, 125 n. 15, 126 n. 16, 129 n. 26, 134 n. 44, 135 n. 46, 138 n. 52, 139 n. 58 Wendel, C. 158 n. 37 Wessner, P. 117 West, M.L. 8 n. 23, 9 n. 26, 19, 20, 21 and nn. 7+9, 22, 23 and n. 12, 24, 25 and n. 18, 26 n. 19, 30 nn. 37–38+41, 31 n. 41, 33 and nn. 52–54, 34 nn. 56–57, 39, 61 n. 13, 153 n. 22 Westphal, R. 42 n. 20 White, D.C. 121 n. 3 Wickert, I. 86

Index of Names  

Wilamowitz-Moellendorff, U. von 39, 48 n. 41, 72, 158 n. 37 Winstedt, E.O. 104 and n. 7 Wirth, G. 105 and n. 14, 112, 113 Wissowa, G.O.A. 103 n. 1 Wooten, C.W. 79 and n. 5, 85 and n. 37, 86, 97 nn. 79–80 Wright, M. 4 n. 6, 6 n. 15, 8 n. 25 Xenocles of Adramittium 84 n. 33 Xenophon (historian) 60

Zaffagno, E. 127 n. 18 Zagari, E. V Zenodotus 75

Index Locorum Acusilaus Genealogiae fr. 34 Fowler

25

Aeschylus Prometheus Vinctus 666 30 Supplices 890 48 n. 41 Fragmenta TrGF 3 FF 258‒260 159 n. 42 Agatharchides De mari Erythraeo 5.21 Fragmenta FGrHist 86 F 19

97 66 n. 35

[Ammonius Grammaticus] De adfinium vocabulorum differentia 340, p. 88.11–13 Nickau 29 Antiatticista ρ 3 Valente

85 n. 39

Apollodorus Mythographus Bibliotheca 1.8.2 70 n. 51 1.8.12 156 1.9.16 150 1.9.21‒22 156 n. 29 1.9.23 156 n. 28 Apollonius Rhodius Argonautica 1.15‒18 2.178‒341 3.409‒421 3.728‒739

156 n. 27 156 n. 29 156 n. 28 156 n. 28

3.1026‒1062 156 n. 28 3.1247‒1407 156 n. 28 4.1717 30 Scholia proleg. Ba Wendel 149 schol. 1.211‒215c Wendel 161 n. 50 schol. 1.1289 Wendel 29–30 schol. 2.178‒182a Wendel 159 n. 43 schol. 2.178‒182b Wendel 160 n. 47 schol. 2.328a Wendel 159 n. 43 schol. 2.562 Wendel 159 n. 43 Apsines, Valerius Rhet. Gr. 1.231 Spengel–Hammer 90 n. 52 Archilochus Fragmenta fr. 182.1 W.2

34 n. 57

Aristides, Publius Aelius Ἱεροὶ λόγοι 1, p. 280.20–21 Jebb Scholia schol. Or. 46.162 Dindorf Aristophanes Acharnenses 1093 Aves 452–453 Nubes 961–968 Pax 1270 Ranae 860–864 1119–1121 1119–1248 1198

60 n. 9 59 n. 7

58 n. 4 47 n. 35 58 n. 4 58 n. 4 67 n. 39 67 n. 39 65 65

 For the sake of convenience, quotations both from Hesiod’s ‘genuine’ works and from other works traditionally ascribed to him are all arranged under the heading ‘Hesiod’. https://doi.org/10.1515/9783110712223-012

  Index Locorum 1206–1208 66 n. 33, 69 n. 47 1238–1242 71 n. 54 1240–1241 69 n. 48 Scholia schol. Nub. 967 Holwerda 59 n. 5 schol. Nub. 967a.α Holwerda 59 nn. 7–8 schol. Nub. 967a.β Holwerda 59 n. 8 schol. Nub. 967b.α Holwerda 59 nn. 6–8 schol. Nub. 967b.β Holwerda 59 n. 7 schol. Ran. 1206 Chantry 66 n. 34, 74 n. 63 schol. Ran. 1206a Chantry 66 schol. Ran. 1206b Chantry 66 schol. Ran. 1206c Chantry 66 schol. Ran. 1213b Chantry 70 n. 52 schol. Ran. 1219a Chantry 70 n. 52 schol. Ran. 1233 Chantry 70 n. 52 schol. Ran. 1238 Chantry 69 n. 49 schol. Tzetz. Ran. 1225 Koster 70 n. 53 Aristophanes Byzantinus Fragmenta incertae sedis fr. 379 Slater 59 n. 8 Aristoteles Atheniensium respublica 5.2.4–7 Kenyon 60 n. 9 Historia animalium 601b 29 Politica 1.1255a14 94 5.1311b30–34 68 n. 41 Protrepticus fr. 3 94 Rhetorica 3.1409b8 69 n. 49 3.1418b29–30 60 n. 9 Asclepiades Tragilensis Tragoidoumena BNJ 12 F2 159 n. 43 BNJ 12 F22 159 n. 43 [Asconius Pedianus] Commentarii vel scholia Ciceronis orationum p. 188.2–3 Stangl 129 n. 23

Athenaeus Grammaticus Deipnosophistae 2.44e–f 96 n. 66 5.187d 96 n. 62 5.215c 96 n. 62 6.244a 63 n. 25 6.263f 98 n. 82 10.425b 97 13.585b–c 63 n. 25 Audax De Scauri et Palladii libris excerpta Gramm. VII p. 332.5–7 Keil 24 Caesar, Caius Iulius De bello Gallico 7.73.2 129 n. 23 Callimachus Fragmenta fr. 433 Pfeiffer fr. 434 Pfeiffer

63 n. 25 63 n. 25

Cassius Iatrosophista Quaestiones et problemata 10.9 93 Cato, Marcus Porcius Origines fr. 27 Chassignet 128 Chamaeleon Heracleotes Fragmenta fr. 31A Martano 59 n. 6 fr. 31B Martano 59 n. 6 fr. 32 Martano 62 n. 21 Charisius, Flavius Sosipater Artis grammaticae quae exstant p. 178.20 Barwick 103 n. 2 p. 179.5 Barwick 103 n. 2 Cicero, Marcus Tullius Brutus 104 83 n. 19 316 83 n. 19, 86 n. 46 325 86 n. 46

Index Locorum  

De inventione 1.17 99 1.48 129 n. 24 2.69–70 90 In M. Antonium orationes Philippicae 2.55 99 Cinna, Caius Helvius Carmina 127 fr. 2 Blänsdorf2 Clemens Alexandrinus Stromateis 1.12.134 27 1.21.107 27 Columella, Lucius Iunius Moderatus Res rustica 4.24.6 129 n. 23 5.11.1 129 n. 23 12.58.1 129 n. 23 [Columella] De arboribus 26.1 26.8

129 n. 23 129 n. 23

Crates Mallota Fragmenta fr. 78 Broggiato

65 n. 27

Demosthenes Orationes 8.70 18.215 18.250 Epistulae 2.12.3

94 89 92 94

Dicaearchus Messanius Fragmenta fr. 114 Mirhady 71 n. 56 Diodorus Siculus Bibliotheca historica 1.38.4 66 n. 35 4.85.4–5 28 n. 30

10.9.8 15.72.1–4

60 n. 10 89

Diogenes Laertius Vitae philosophorum 1.119 60 n. 9 2.42 60 n. 9 3.37 75 n. 66 5.27 60 n. 9 5.60 60 n. 9 5.82 97 Diomedes Ars Grammatica Gramm. I p. 513.10 = fr. 17 Courtney 127 n. 18 Dionysius Halicarnassensis Ad Ammaeum 2 81 De compositione verborum 22.58 60 n. 9 25.209–218 75 n. 66 De Dinarcho 8 86 n. 47 11 85, 97 De Isaeo 4 96 n. 65 7 75 n. 65 De Lysia 17 75 n. 65 Dionysius Periegeta Orbis descriptio 537 29 Donatus, Aelius Commentum in Terentii Phormionem 190 132 n. 37 Ennius, Quintus Annales fr. 441 Skutsch

128

Eratosthenes Cyrenaeus Fragmenta fr. 101 Strecker 59 n. 6

  Index Locorum Etymologicum Genuinum AB α 926 Lasserre–Livadaras 31 AIAIIB s.v. εἰλυφάζων 30 n. 40 AB s.v. ἱμερτῆς 29 Etymologicum Gudianum s.v. ἀφωσιωμένος De Stefani 29 Etymologicum Magnum s.v. τῶ p. 773.22 Gaisford

30 n. 40

Etymologicum Symeonis α 1074/75 Lasserre–Livadaras 31 ε 153 Baldi 30 n. 40 [Eudocia] Violarium 712 Euripides Alcestis 780–789 782 782–786 783–786 908–910 932–934 Bacchae 389–394 Electra 167–168 369–370 550–551 Helena 726–727 728–733 Hercules Furens 10 64 664 Hippolytus 24 203–207 403 404 406 407

84 n. 36

46 and n. 31 47 n. 33 46 46 n. 31 47 n. 36 47 n. 36 50 n. 48 60 n. 10 45 n. 29 45 n. 29 44 n. 25 44 n. 25 70 n. 50 70 n. 50 48 n. 41 70 n. 50 50 n. 48 44 n. 26 44 n. 26 44 n. 26 44 n. 26

407a 44 n. 26 408 44 n. 26 409 44 n. 26 410 44 n. 26 413–414 44 n. 26 415–423 44 n. 26 Iphigenia Aulidensis 28–33 50 n. 48 Medea 408–409 45 n. 29 446–447 45 n. 29 520–521 45 n. 29 568–570 45 n. 29 hyp. l. 43 Diggle 67 n. 37 Orestes 211–212 44 n. 25, 45 n. 29 229–230 44 n. 25, 45 n. 29 229–236 44 231–232 44 n. 25, 45 n. 29 233–236 44 n. 25, 45 n. 29 234 44 n. 25 1034 45 n. 29 1509 45 n. 29 1523 45 n. 29 Phoenissae 49 71 50 71 1364 60 n. 10 1375 60 n. 10 Troades 101–102 50 n. 48 Fragmenta TrGF 5 F 103 47 n. 37 TrGF 5 F 119 50 n. 48 TrGF 5 F 228 66 n. 35 TrGF 5 F 228.3 69 n. 45 TrGF 5 F 228.3–5 69 n. 44 TrGF 5 F 228a 68 50 n. 47 TrGF 5 F 261 TrGF 5 F 453 49 n. 44 TrGF 5 F 453.1 49 n. 44 TrGF 5 F 453.1–8 49 n. 44 TrGF 5 F 453.1–9 49 n. 44 TrGF 5 F 453.4 49 n. 44, 50 n. 44 TrGF 5 F 453.6 49 n. 44, 50 n. 44 TrGF 5 F 453.9 49 n. 44 TrGF 5 F 453.10 49 n. 44

Index Locorum  

TrGF 5 F 453.10–12 49 n. 44 TrGF 5 F 515 71 n. 55 TrGF 5 F 515.1 69 n. 49 TrGF 5 F 516 69 n. 48, 71 n. 55 TrGF 5 F 516.2 70 n. 53 TrGF 5 F 539a 70 TrGF 5 F 819.1 70 TrGF 5 F 846 66 n. 33 TrGF 5 F **953m Kn. 49 n. 43 Scholia schol. Or. 1094 Schwartz 24 Testimonia TrGF 5 T 1 IA.9, 28–29 67 n. 36 TrGF 5 T 1 IB.5, 57–59 67 n. 36 [Euripides] Rhesus hyp. (b) ll. 26–44 Diggle hyp. (c) ll. 52–53 Diggle

71 n. 56 73 n. 62

Eustathius Thessalonicensis Commentarii ad Homeri Odysseam 1712.17 149, 150 Favorinus De exilio 11.1 (col. 9, ll. 25–26)

40 n. 8

Festus, Sextus Pompeius Epitomae operis de verborum significatu Verrii Flacci quae extant p. 254.29 Lindsay 130 n. 31, 133 p. 266.19 Lindsay 133 Florilegium Monacense 220 84 n. 34 Gellius, Aulus Noctes Atticae 4.18.1 6.9 17.21.1 17.21.3 17.21.8 17.21.23–24 18.11 19.8.6

116 131 n. 32 115 112, 116 115, 116 116 131 n. 33 131 n. 33

Gnomologium Vaticanum 253–260 84 n. 33 307–308 84 n. 33 Harpocratio Grammaticus Lexicon in decem oratores ε 100 Keaney = p. 125.10–16 Dindorf ι 25 Keaney = pp. 163.14–164.2 Dindorf

97 84 n. 31

Hephaestio Grammaticus De poematibus 7.2 34 n. 57 Enchiridion de metris 15.10, p. 50.21–22 Consbruch 60 n. 9 [Hermogenes] Περὶ εὑρέσεως 169.18–170.17 Rabe

90 n. 52

[Herodianus Grammaticus] De figuris 40 Hajdú 96 n. 61 Philetaerus 242, p. 66 Dain 19 Herodotus Historiae 1.52 1.106.2 2.10.3 5.82.2

29 29 n. 36 29 n. 36 29 n. 36

[Herodotus] Vita Homeri 16, ll. 203–205 Allen

63 n. 24

Hesiodus Catalogus mulierum fr. 4 M.–W. = 5 M. 25 fr. 7 M.–W. = 7 M. = 3 H. 25 fr. 7–8 M.–W. = 7–8 M. = 3, 42 H. 26 fr. 8 M.–W. = 8 M. = 42 H. 25 fr. 9 M.–W. = 9 M. = 4 H. 25 fr. 23a M.–W. = 19 M. = 15 H. 28 fr. 26 M.–W. = 23 M. = 17 H. 28

  Index Locorum fr. 38 M.–W. = 37 M. fr. 40 M.–W. = 36 M. = 28 H. fr. 41 M.–W. = 183 M. = *23 H.

149 21 19, 21, 22 and n. 11, 23 n. 13 fr. 42 M.–W. = 163 M. 22 n. 11 fr. 43a.41–43 M.–W. = 69 M. = 37 H. 21 n. 7 fr. 75.13–25 M.–W. = 48 M. = *3 H. 21 n. 7 fr. 76.9–23 M.–W. = 48.34–48 M. = *4 H. 21 n. 7 fr. 85 M.–W. = 120 H. 28 fr. 117 M.–W. = 8 H. 28 fr. 149 M.–W. = 245 M. 28 n. 30 fr. 165.3 M.–W. = 117 M. = 72 H. 21 n. 7 fr. 165.4 M.–W. = 117 M. = 72 H. 21 n. 7 fr. 188A M.–W. = 130 M. 28 n. 30 fr. 211.7–13 M.–W. = 152.14–20 M. = 100 H. 21 n. 7 frr. 2–6 M. 26 n. 20 fr. 162 M. = 103 H. 22 n. 11 fr. 162.2 M. = 103 H. 22 n. 11 fr. 162.2–3 M. = 103 H. 22 n. 11 fr. 307 M.2 = frr. 117 + 85 M.–W. = 8 + 120 H. 28 n. 28 Ceycis nuptiae fr. 263 M.–W. = 202 M. 29, 30 n. 37 Melampodia fr. 278 M.–W. = 214 M. 26 fr. 279 M.–W. = 215 M. 27 Opera et dies 21–24 45 n. 29 40–41 45 n. 29 56–57 45 n. 29 67–68 45 n. 29 210 31 211 31 270–272 45 n. 29 308–313 45 n. 29 317–320 45 n. 29 346–348 45 n. 29 352 45 n. 29 656 33, 34 Theogonia 435 33 800 33, 34 875 30

Fragmenta incertae sedis frr. 303–342 M.–W. fr. 305 M.–W. = 255 M. = *11 H. fr. 306 M.–W. = 256 M. = *12 H. fr. 327 M.–W. = 277 M. fr. 308 M.2 fr. 309 M.2 Fragmenta dubia fr. 347 M.–W. 29 n. 36 Fragmenta spuria fr. 364 M.–W. = 303 M. 29 n. 36 fr. 368 M.–W. 29 n. 36 fr. 403 M.–W. 30 n. 40 fr. 406 M.–W. 30 n. 40 Testimonia T 42 M. 65 n. 29 T 49 M. 65 n. 28 T 50 M. 64 n. 27 TT 152–157 M. 29

23 27 n. 26 27 n. 26 23 29 28 n. 31

Homerus Ilias 1.25 30 n. 39 1.379 30 n. 39 1.589 31 5.478 20 n. 5, 21 5.526 30 and n. 41 8.19 19 8.21 19 17.548 19 20.220–222 28 24.492 23 n. 12 Odyssea 7.94 32 n. 46 9.38 23 n. 12 12.69 154 152 12.69‒70 12.70 154, 155, 157 12.184 61 n. 10 13.267–268 20 15.428 20 Scholia D schol. Il. 1.5 van Thiel 158 n. 35 D schol. Il. 1.7 van Thiel 158 n. 35 D schol. Il. 1.39 van Thiel 158 n. 35 D schol. Il. 1.59 van Thiel 158 n. 35 D schol. Il. 1.264 van Thiel 158 n. 35

Index Locorum  

D schol. Il. 2.157 van Thiel D schol. Il. 2.339 van Thiel D schol. Il. 5.392 van Thiel D schol. Il. 6.155 van Thiel D schol. Il. 6.396 van Thiel D schol. Il. 10.435 van Thiel D schol. Il. 21.448 van Thiel D schol. Il. 23.144 van Thiel schol. Il. 1.1a Erbse schol. Il. 1.1b Erbse schol. Il. 1.589 Erbse D schol. Od. 7.324 Ernst D schol. Od. 11.271 Ernst D schol. Od. 11.287 Ernst D schol. Od. 12.39 Ernst D schol. Od. 12.69 Ernst

D schol. Od. 12.69‒70 Ernst D schol. Od. 14.533a Ernst D schol. Od. 15.16 Ernst D schol. Od. 19.518 Ernst schol. Od. 10.2 Dindorf Isidorus Hispalensis Origines 19.2.9 127 Isocrates Orationes 10.29 17.29

62 n. 21 30

Josephus, Flavius Antiquitates Judaicae 6.219.3 93 Laberius, Decimus Mimorum fragmenta fr. 26 Panayotakis 135 fr. 27 Panayotakis 135, 136 fr. 34 Panayotakis 139, 140 fr. 62 Panayotakis 128 fr. 67 Panayotakis 141 n. 61

158 n. 35 158 n. 35 158 n. 35 158 n. 40 158 n. 35 158 n. 35 158 n. 35 158 n. 40 75 n. 64 75 n. 64 31 158 n. 36 158 n. 40 158 n. 40 158 n. 36 148, 149, 150, 151, 159 15, 148 150 158 158 n. 36 25

Lactantius, Lucius Cae(ci)lius Firmianus Divinae institutiones 3.15.10 105 n. 9 Livius, Titus Ab Urbe condita 15.22–25 37.55.10 37.56.2

97 n. 80 117 117

Lucianus Verae historiae 2.20

75 n. 64

Naevius, Cnaeus Comoediarum fragmenta 132 fr. 35 Ribbeck3 Nepos, Cornelius Aristides 3.3 Atticus 19.1 Carmina fr. 63 Marshall Chronica fr. 4 Marshall fr. 5 Marshall fr. 7 Marshall fr. 8 Marshall Cimon 3.2 De viris illustribus fr. 42 Marshall fr. 43 Marshall fr. 54 Marshall fr. 56 Marshall fr. 58 Marshall fr. 59 Marshall Dion 3.2 5.6 Epistulae fr. 39 Marshall Exempla fr. 11 Marshall fr. 18 Marshall

113 103 n. 1 104 n. 4 112 106 115 116 113 103 n. 2 103 n. 2 117 103 n. 2 105 n. 9 104 n. 3 103 n. 2 103 n. 2 105 n. 9 116 111

  Index Locorum fr. 24 Marshall fr. 35 Marshall Hannibal 6.3 Lysander 2.3 Miltiades 4.2 6.3 7.2 Pausanias 1.4 Themistocles 1.2

114 113 114 103 n. 2 114 114 114 n. 32 114 114 n. 32

Nonius Marcellus De compendiosa doctrina p. 10.10 132 n. 38 pp. 62.6–65.8 125 n. 15 p. 62.12 132, 133 p. 62.18 130 p. 62.23 128 p. 62.26 128 p. 63.3 128 p. 63.11 130 p. 63.19 130 n. 31, 133 p. 63.22 130 p. 63.27 130 n. 34, 132 n. 38 p. 64.29 130 n. 30 p. 65.4 130 n. 30 p. 66.18 130 n. 30, 133 pp. 66.18–67.25 125 n. 15 p. 67.25 128 p. 80.22 128 p. 84.5 138 p. 93.26 130 n. 29 p. 100.19 130 n. 33 p. 101.23 130 n. 29, 131 n. 34 p. 108.3 130 n. 30 p. 108.5 135 p. 114.9 132 n. 38 p. 114.15 131 n. 34 p. 118.22 130 n. 30 p. 119.26 139, 140 p. 129.24 130 n. 33 p. 133.23 131 n. 33 p. 134.24 130 n. 30

p. 134.28 p. 134.29 p. 134.35 p. 135.7 p. 137.19 p. 140.20 p. 142.23 p. 145.9 p. 147.22 p. 148.15 p. 151.19 p. 181.5 p. 188.5 p. 212.7 p. 220.11 p. 375.31 p. 445.2–447.26 p. 446.34 p. 481.23 p. 498.25 p. 501.22 p. 517.3 p. 546.15 p. 551.3

130 131 130 126 128 130 n. 32 131 n. 34 131 n. 33 129 n. 26 131 n. 33 135, 136 132 n. 38 131 n. 33 132 n. 38 130, 136 132 126 n. 15 132 n. 38 123 123 122 127 127 139 n. 57

Orion Etymologicum s.v. ἀήτης, p. 19 Sturz Orphica Argonautica 671‒676 870‒873

30

161 n. 50 156 n. 28

Pacuvius, Marcus Tragoediarum fragmenta fr. 83 Schierl 131 n. 34 fr. 229 Schierl 122 Paulus Diaconus Excerpta ex libris Pompeii Festi de significatione verborum p. 54.19 Lindsay 133 p. 64.6 Lindsay 129 n. 23 p. 84.3 Lindsay 140 p. 253.17 Lindsay 132 n. 37

Index Locorum  

Pausanias Periegeta Graeciae descriptio 9.31.4 65 n. 29 10.5.7 24 n. 16 10.6.7 24 n. 16 10.12.10 24 n. 16

Scholia schol. P. 4.414a Drachmann

Pherecydes Atheniensis Historiae fr. 104c Fowler 149

Plautus, Titus Macc(i)us Astraba fr. VI Monda 128, 129 Captivi 762 123 fr. dub. Lindsay 136, 137 Casina 245 126, 127 Cornicula fr. II Monda 131 fr. VI Monda 130, 137 Curculio 363 132 Miles gloriosus 499 131 n. 36 Parasitus Medicus fr. I Monda 138 Persa 237 129 n. 25 408 132 n. 38 Poenulus 393 128 704–705 131 n. 36 Trinummus 599 131 n. 36

Philo Quaestiones in Genesim 4 fr. 227.2 93 Philostratus Vitae sophistarum 2.15.2 90 2.595–596 90 n. 52 Photius Bibliotheca cod. 176, 121b 96 n. 60 cod. 250, 447a 84 n. 29, 96 n. 67 Lexicon α 2091 Theodoridis 84 n. 30 Phrynichus Comicus Fragmenta PCG 78 59 n. 6 Pindarus Isthmia 3.9 6.48 Nemea 9.9 10.32 11.23 Olympia 3.15 9.108 Pythia 1.99 4.102 4.156‒167 4.224‒239

34 n. 57 34 n. 57 34 n. 57 34 n. 57 34 n. 57 34 n. 57 34 n. 57 34 n. 57 21 156 n. 27 156 n. 28

Plato Phaedrus 243b

62 n. 21

Plinius Secundus, Caius Naturalis historia 3.125 111 6.31 114 15.3 29 n. 36 Plinius Caecilius, Caius Epistulae 5.3.6 104 n. 4 Plutarchus Amatorius 756b–c

68 n. 40

156 n. 28

  Index Locorum De audiendis poetis 21b–c 39 De Herodoti malignitate 857f 28 Praecepta gerendae reipublicae 802e 84 n. 25 802e–f 96 n. 64 Alexander 3.6 84 n. 25, 96 n. 63 Antonius 24 84 n. 25 24.7 96 n. 72 Cicero 4 83 n. 19 Demosthenes 27.5 96 n. 64 Lysander 15.3 60 n. 10 Publicola 9.11 84 n. 25 Solon 8.2–3 63 n. 24 8.3 63 n. 24 Tiberius Gracchus 8 83 n. 19 20 83 n. 19 [Plutarchus] Consolatio ad Apollonium 104a 84 n. 25 Vitae decem oratorum 842e 85 n. 39 Poetae melici Graeci (PMG) fr. 960 40 n. 8 Pollux, Julius Onomasticon 9.12.3 Bethe 9.18 Bethe Polybius Historiae 12.26.5 Porphyrius Vita Plotini 4

20 n. 4 28

49 n. 44

64 n. 26

Praxiphanes fr. 28A Matelli fr. 28B Matelli

65 n. 28 65 n. 28

Priscianus Institutiones grammaticae Gramm. II p. 383.4–6 Hertz/Keil Rhetores Graeci 4.249–250 Walz 7.510–511 Walz

113

90 n. 52 90 n. 52

Rutilius Lupus Schemata dianoeas et lexeos 1.8 98 1.20 83 n. 21 2.1 83 n. 21 Seneca, Lucius Annaeus (rhetor) Controversiae 1.2.23 96 n. 70 1.4.11 96 n. 70 2.5.20 96 n. 70 7.4.10 96 n. 70 9.1.12 96 n. 70 9.1.15 96 n. 70 9.6.16 96 n. 70 Suasoriarum liber 4.5 96 n. 70 Servius (grammaticus) Aeneidos additamenta 1.225 129 n. 23 6.1 133 n. 40 Commentarius in Vergilii Aen. 1.39 133 n. 40 Solinus, Caius Iulius Collectanea rerum memorabilium 1.27 106 Sopater Rhetor Διαίρεσις ζητημάτων Rhet. Gr. 8.239.1–244.9 Walz

90 n. 52

Index Locorum  

Sophocles Ajax 125–130 181 191 692 758–761 815 Antigone 614–614a 625–625a Electra 1082–1083 Oedipus Tyrannus 380–382 873–874 977–979 979 Trachiniae 1023 Fragmenta TrGF 4 F 61 TrGF 4 F 88 TrGF 4 F 556 TrGF 4 FF 581–595b TrGF 4 F 591

45 n. 29 47 n. 36 47 n. 36 52 45 n. 29 52 47 n. 36 47 n. 36 50 n. 48 45 n. 29 45 n. 29 46 and n. 32, 51 47 n. 33 48 n. 41

50 n. 47 39 and n. 6 48 n. 41 38 n. 3 38 n. 5, 40 and nn. 8+12 TrGF 4 FF 591–593 51 n. 49 TrGF 4 F 591.1–2 40 and n. 8 TrGF 4 F 592 38, 40 and n. 12, 42 n. 20, 43 n. 21, 51 n. 49 TrGF 4 FF 592.1–3 38 and n. 5, 39, 40 and n. 13, 41 and nn. 18+20 TrGF 4 F 592.2–3 40 TrGF 4 F 592.4–6 37, 38 and n. 5, 40 and n. 14, 41 and n. 19, 42 and n. 20, 43, 44, 45, 46 n. 33, 47, 48, 49, 50, 51, 52, 53 TrGF 4 F 592.5 40 TrGF 4 FF 592–593 51 n. 49 TrGF 4 F 593 37, 38, 39, 41 and n. 16, 42 and nn. 19– 20, 43 and n. 21, 44, 45, 46 n. 33, 47, 48

and n. 38, 49, 50, 51 and n. 49, 52, 53 TrGF 4 F 593.1 48 n. 38 TrGF 4 F 593.1–2 46 n. 33 TrGF 4 F 593.2 48 n. 38 TrGF 4 F 593.2–3 46 n. 31 TrGF 4 F 645 159 n. 42 TrGF 4 FF 704‒705 159 n. 42 TrGF 4 F 835 39 TrGF 4 F 836 39 TrGF 4 F 879a 41 and n. 16, 42 n. 20 Speusippus Epistula ad Philippum regem 12.12 94 Stephanus Byzantius Ethnica s.v. Κορόπη 84 n. 28 Stesichorus Fragmenta fr. 90 Finglass fr. 322a Finglass

62 n. 21 59 n. 6

Stobaeus Anthologium 3.2.11–12 3.8.2 3.10.11–12 3.12.18 3.20.34–35 3.22.20–21 3.29.2–4 3.37.4 4.8.10–11 4.14.1 4.16.11 4.19.1 4.19.2 4.19.13 4.22g.169–170 4.22g.194–195 4.26.6 4.29a.12 4.29b.36–37 4.34.39

45 n. 29 50 n. 47 45 n. 29 97 45 n. 29 45 n. 29 45 n. 29 50 n. 48 45 n. 29 49 n. 44 50 n. 48 44 n. 25 44 n. 25 50 n. 47 45 n. 29 45 n. 29 47 n. 38 40 n. 8 45 n. 29 39

  Index Locorum 4.34.40 4.36.1 4.36.2 4.36.3 4.36.4 4.36.1–4 4.41.6 4.44.15 4.44.34 4.48b.17 4.50.63 4.52a.6–7 4.52a.7–8

39 44 n. 25 44 n. 25 44 n. 25 44 n. 25 45 n. 29 50 n. 48 50 n. 48 50 n. 48 50 n. 48 48 n. 41 45 n. 29 45 n. 29

Strabo Geographica 1.3.18 9.3.5 10.3.13 10.4.2 13.2.3 14.2.13 14.2.24

29 n. 36 24 n. 16 60 n. 9 98 n. 82 23 n. 19 86 n. 48 85 n. 27, 96 n. 71

Suda ω 263 Adler

84 n. 31, 85

Svetonius Tranquillus, Caius Vita Terentii p. 31.2 Reifferscheid 117 Terentius Afer, Publius Adelphoe 876 123 Heautontimorumenos 723 118 Phormio 190–191 132 n. 37 Theocritus Idyllia 17.13–33

68 n. 43

Tiberius Rhetor De figuris Demosthenicis 48 69 n. 44 Tzetzes, Johannes Chiliades 1.25, l. 686 Leone 59 n. 7 Valerius Flaccus Setinus Balbus, Caius Argonautica 7.58‒77 156 n. 28 7.539‒653 156 n. 28 Valerius Maximus Facta et dicta memorabilia 3.7.1 117 Varro, Marcus Terentius De lingua Latina 7.52 131 7.107 132 Vergilius Maro, Publius Eclogae sive bucolica 1.13 132 n. 37 Georgica 2.77 128 Xenophon Memorabilia 2.6.11 Oeconomicus 11.15

60 n. 10 29

[Zonaras] Lexicon s.v. ἀντιφερίζω p. 216 Tittmann 31 n. 42